Extending Opportunities: How Active Social Policy Can Benefit Us All

Extending Opportunities HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL Social policy is often disparaged as being a burden ...

Author: Donald J. Johnston

9 downloads 1051 Views 1MB Size Report

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

DOWNLOAD PDF

Extending Opportunities HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL Social policy is often disparaged as being a burden on society, undermining the entrepreneurial spirit, discouraging work and savings, and fostering dependency on the state. Bad social policy can indeed have negative effects, but social policy is not inevitably bad. An increasing body of evidence has been gathered in recent years that shows which programmes are effective in improving social outcomes, and which are not. Well-designed social protection is not a cumbersome inheritance from the past, but an asset that is critical for sustaining social development. Indeed, its role will inevitably grow in the future as technological change, lifelong learning and globalisation increase opportunities for those who are well-equipped in skills and resources and further disadvantage those who are not. To fulfil its potential, however, social protection needs to be forward-looking to deal with the challenges emerging in the 21st century, rather than reacting to the problems of the past. It needs to recognise new needs of individuals and families, and new constraints on their functioning. Successful programmes require new means to attain their goals, to leverage the initiatives of a broad range of actors, and to involve clients at every stage in the design and delivery of programmes.

This book is available to subscribers to the following SourceOECD theme: Social Issues/Migration/Health Ask your librarian for more details of how to access OECD books on line, or write to us at

[email protected]

w w w. oe c d . org

-:HSTCQE=UU\^Y^:

ISBN 92-64-00794-6 81 2005 05 1 P

HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL

OECD’s books, periodicals and statistical databases are now available via www.SourceOECD.org, our online library.

HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL

Extending Opportunities

In exploring responses to these challenges, this report promotes the concept of active social policies. This concept stresses the importance of shifting the focus of social programmes from insuring individuals against a few, well-defined contingencies towards investing in their capabilities and making use of them to the best of their potential at every stage of the life course. It also underscores the importance of broadening the roles played by individuals, employers and trade unions, as well as profit and not-for-profit providers of social services within a more ambitious social protection system.

Extending Opportunities

Extending Opportunities HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

ORGANISATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT The OECD is a unique forum where the governments of 30 democracies work together to address the economic, social and environmental challenges of globalisation. The OECD is also at the forefront of efforts to understand and to help governments respond to new developments and concerns, such as corporate governance, the information economy and the challenges of an ageing population. The Organisation provides a setting where governments can compare policy experiences, seek answers to common problems, identify good practice and work to co-ordinate domestic and international policies. The OECD member countries are: Australia, Austria, Belgium, Canada, the Czech Republic, Denmark, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Japan, Korea, Luxembourg, Mexico, the Netherlands, New Zealand, Norway, Poland, Portugal, the Slovak Republic, Spain, Sweden, Switzerland, Turkey, the United Kingdom and the United States. The Commission of the European Communities takes part in the work of the OECD. OECD Publishing disseminates widely the results of the Organisation’s statistics gathering and research on economic, social and environmental issues, as well as the conventions, guidelines and standards agreed by its members.

This work is published on the responsibility of the Secretary-General of the OECD. The opinions expressed and arguments employed herein do not necessarily reflect the official views of the Organisation or of the governments of its member countries.

Also available in French under the title: Accroître les chances de chacun Pour une politique sociale active au bénéfice de tous

© OECD 2005 No reproduction, copy, transmission or translation of this publication may be made without written permission. Applications should be sent to OECD Publishing: [email protected] or by fax (33 1) 45 24 13 91. Permission to photocopy a portion of this work should be addressed to the Centre français d'exploitation du droit de copie, 20, rue des Grands-Augustins, 75006 Paris, France ([email protected]).

FOREWORD

Foreword

E

conomic growth has resulted in huge improvements in the social conditions of millions of OECD citizens over the past 50 years. Sustaining the pace of economic growth will remain critical for these improvements in social conditions to continue in the future. But economic growth alone is not enough. Also important is the existence of well-functioning institutions, in particular those devoted to providing social protection to families and individuals. This document builds on the experiences of OECD countries over the past two decades. It argues that social protection is as important as ever for attaining a broad range of social goals. Recent policy experimentation – together with better research and evaluation – has started to pay off. More is known today about effective policies to address problems that seemed intractable only a few years ago. In particular, we have learned how to better integrate social and labour market interventions aimed at reducing poverty, exclusion and dependency among some groups dependent on welfare benefits. However, while many policies have been effective in achieving desired outcomes, others have not, and difficult challenges lie ahead. Indeed, the recent optimism that we can improve the lives of those most exposed to disadvantage poses a challenge in itself: citizens – knowing that social protection can do more to help people on welfare or with disabilities, or to enhance the life chances of children and to help parents – expect quality interventions in each of these fields. But society has already committed vast resources to meeting other social policy challenges – in particular, to provide old-age pensions to an increasing number of elderly persons. In this context, is an expansion of public responsibilities affordable or desirable? If not, how can society rebalance its interventions to meet the needs of individuals at different stages of their life-cycle? Are there better means of achieving social goals? These are some of the questions that will be confronting Social Policy Ministers as they meet in Paris on 31 March and 1 April 2005 (www.oecd.org/socialmin2005). This report has been prepared by the OECD Secretariat to inform those discussions. The heart of this report puts the case for active social policies, a concept that stresses the importance of shifting the focus of social programmes from insuring individuals against a few, welldefined, risks towards investing in their capabilities and making the best use of them. This report – which complements the evidence provided in the recently released bi-annual compendium of social indicators, Society at a Glance – brings together results from a range of analyses carried out by the OECD in recent years, in particular work undertaken in the Directorate for Employment, Labour and Social Affairs, the Centre for Tax Policy and Administration, the Directorate for Education and the Economics Department. This report, which has been prepared by Marco Mira d’Ercole of the OECD Social Policy Division, draws on contributions by various members of the OECD Social Policy Division: Peter Whiteford

EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

3

FOREWORD

(Part I); Willem Adema, Anna D’Addio and Peter Whiteford (Part II); Manfred Huber and Monika Queisser (Part IV). David Dowey assured the production of most of the policy scenarios presented in the report, as well as of the tables and charts. Patrick Hamm contributed to the final editing of the report. Mark Pearson, Head of the OECD Social Policy Division, supervised the preparation of this report and provided useful comments on various versions. Donald J. Johnston OECD Secretary-General

4


TABLE OF CONTENTS

Table of Contents Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

9

Part I The Social Policy Context Chapter 1. Social Development in OECD Countries: Mapping the Progress. . . . . . . . . . . . . . .

17

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Monitoring social development through indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

18 18 25 25

Chapter 2. The Need for More Active Social Policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

27

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. The traditional model of social protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Changes in the context of social policies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. More ambitious goals for social protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

28 28 31 38 39 40

Chapter 3. Broadening Roles for Social Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

41

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. The theoretical case for and against private roles for social protection . . . . . . . . . . . . . . 3. Differences in social protection arrangements across OECD countries . . . . . . . . . . . . . . . 4. The experience of reform in OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Towards a better mesh of public and private roles in social protection . . . . . . . . . . . . . . 6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

42 42 45 47 48 52 54

Part II Giving Children and Parents a Brighter Future Chapter 4. Families with Children: Achievements and Challenges . . . . . . . . . . . . . . . . . . . . .

57

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Policies and achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. The challenges today . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

58 58 61 70 70

Chapter 5. More Effective Policies for Families with Children . . . . . . . . . . . . . . . . . . . . . . . . . .

73

1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Public transfers and tax advantages can reduce child poverty . . . . . . . . . . . . . . . . . . . . . . 3. Investing in children has long-term pay-offs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Increasing maternal employment reduces childhood poverty . . . . . . . . . . . . . . . . . . . . . . 5. Reconciling work and family life brings benefits to parents and children . . . . . . . . . . . . 6. Raising fertility: satisfying couples’ preferences, boosting the labour force . . . . . . . . . . . 7. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

74 74 75 79 84 91 97 98


5

TABLE OF CONTENTS

Part III Combating Poverty and Exclusion among Prime-aged Persons Chapter 6. Poverty and Exclusion in Prime-age: Achievements and Challenges . . . . . . . . . . 103 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Policies and achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. The challenges today . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

104 104 108 113

Chapter 7. More Effective Policies to Reduce Poverty and Exclusion . . . . . . . . . . . . . . . . . . . . 115 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Labour market integration is the key . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. Increasing the rewards of work makes labour market integration more sustainable. . . . . . 4. More effective social assistance makes a difference to the lives of the poor . . . . . . . . . . 5. Improving policy coherence can reduce unintended adverse effects . . . . . . . . . . . . . . . . 6. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

116 116 128 132 136 138 139

Part IV Responding to Shifts in the Risks Confronting Older Persons Chapter 8. The Social Conditions of Older People: Achievements and Challenges . . . . . . . . 143 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Policies and achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. The challenges today . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

144 144 146 162 162

Chapter 9. More Effective Policies for Elderly People . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2. Reducing the burden of public pensions systems while diversifying retirement income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3. A longer working life reduces both old-age poverty and pension costs. . . . . . . . . . . . . . . 4. Better care for increasing numbers of the frail elderly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Notes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

166 166 174 180 187 188

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 191 Boxes 1.1. OECD social indicators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. New partnerships across government, business and community: the Australian experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Conditional benefits to poor families: the experience of Oportunidades in Mexico . . . . . . 5.2. Investment policies targeting child poverty and disadvantage: Job Corps and Head Start in the United States . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Reconciling work and family life in Sweden: a continuum of public policy supports . . . . . . 5.4. Government policies to make business more family-friendly . . . . . . . . . . . . . . . . . . . . . . . 7.1. Increasing employment among lone parents: the United Kingdom experience . . . . . . . . 7.2. Policies to promote the rehabilitation of disabled persons: a new reform strategy in Luxembourg. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3. Reforms of invalidity benefits: Pathways to Work pilots in the United Kingdom . . . . . . . . 7.4. Ireland’s National Action Plan against poverty and social exclusion. . . . . . . . . . . . . . . . . . . 9.1. Linking pensions to life expectancy: notionally defined pension systems in Italy, Sweden and Poland. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2. The notions of actuarial fairness and actuarial neutrality applied to pension systems . . . . 9.3. Improving quality through better co-ordination of long-term care policies: Aged Care Assessment Teams in Australia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.4. Changes in the financing of long-term care: the experience of France . . . . . . . . . . . . . . .

6

19 53 77 79 89 92 121 124 125 137 168 177 185 186


TABLE OF CONTENTS

Tables 1.1. Share of the population who did not have enough money to buy family necessities during the past year, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1. Different arrangements of social protection. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Employment and part-time work among women in 2003, by presence and age of children . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.1. Assistance for families with children, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.2. Effective tax rates on second earners at different income levels, and differences relative to singles and primary earners in couple families, 2002. . . . . . . . . . . . . . . . . . . . . 5.3. Maternity and parental leave, duration and benefits in 2003-2004 . . . . . . . . . . . . . . . . . . . 5.4. Effective tax rates on couples with two children at different income levels, and differences relative to singles and childless couples, 2002 . . . . . . . . . . . . . . . . . . . . . . 7.1. Changes in disability policies in the period 1985-2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2. Legislative framework and extent of employer responsibility towards disabled persons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.3. Examples of employment-conditional benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4. Proportions of disabled and unemployed persons receiving benefits, and proportion of benefit recipients who are not disabled or unemployed . . . . . . . . . . . 8.1. Benefit entitlement from safety-net arrangements for older people . . . . . . . . . . . . . . . . . 8.2. Relationship between care recipient and informal care-giver . . . . . . . . . . . . . . . . . . . . . . . 8.3. Policy concerns about the quality of long-term care . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1. Examples of pension reform measures, 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2. Persons aged 65 and over receiving institutional and home care services . . . . . . . . . . . . . . . .

20 49 60 76 81 87 94 122 127 130 133 154 157 161 169 182

Figures 1.1. 1.2. 1.3. 1.4.

Trends in absolute poverty rates in selected OECD countries . . . . . . . . . . . . . . . . . . . . . . Relative poverty rates among the entire population . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Social isolation in OECD countries, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cross-country correlations between per capita income and different social indicators in OECD countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.1. Gross public social spending in OECD countries, as a share of GDP, 1960-2001. . . . . . . . 2.2. Net public social benefits per capita by age of recipients, 1998 . . . . . . . . . . . . . . . . . . . . . 2.3. Recipients of cash public social benefits, persons of working age . . . . . . . . . . . . . . . . . . . 2.4. Trends in the distribution of equivalised household income, OECD average. . . . . . . . . . 2.5. Relative poverty rates by age of individuals, OECD average . . . . . . . . . . . . . . . . . . . . . . . . 2.6. Trends in fertility rates in selected OECD countries. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.7. Percentage of foreign-born persons and of non-citizens in OECD countries, around 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.8. Ratio of annual divorces to marriages in selected OECD countries . . . . . . . . . . . . . . . . . . 2.9. Distribution of the population by household types, mid-1980s and 2000 . . . . . . . . . . . . . 2.10. Persons in households with a head of working age and no adult working, 2000 . . . . . . 3.1. Public and private social spending, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.2. Employment in the non-profit sector in 1995 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Mortality and immunisation rates among children, OECD average. . . . . . . . . . . . . . . . . . 4.2. Relative poverty among children and employment rates among mothers, 2000 . . . . . . 4.3. Gross public spending on families as a percentage of GDP . . . . . . . . . . . . . . . . . . . . . . . . . 4.4. Relative poverty rates among children. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.5. Different forms of child disadvantage in OECD countries, late 1990s . . . . . . . . . . . . . . . . 4.6. Cumulative years spent by children in different family types . . . . . . . . . . . . . . . . . . . . . . 4.7. Desired and observed fertility in different years . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.8. Desired and observed fertility among women of different ages. . . . . . . . . . . . . . . . . . . . . 4.9. Gender gaps in employment of persons aged 25 to 54 by educational attainment, 2000 . . . . 4.10. Gender gaps in earnings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.11. Time spent daily by women and men on childcare, paid and unpaid work. . . . . . . . . . . 5.1. Effect of taxes and transfers in reducing relative poverty in families with children, 2000 . . .


21 22 23 24 29 30 30 31 32 33 34 35 36 37 47 51 58 59 61 62 63 64 66 67 68 69 70 75

7

TABLE OF CONTENTS

5.2. Potential increase in female labour force participation resulting from various policy reforms. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.3. Potential impact of higher employment among mothers on child poverty . . . . . . . . . . . 5.4. Non-equivalised and equivalised household income for different family types and gross earnings, OECD average 2002 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.5. Cross-country relations between total fertility and female employment rates . . . . . . . . 5.6. Potential impact of various policy reforms on total fertility rates . . . . . . . . . . . . . . . . . . . 5.7. Potential impact of a recovery in fertility rates on population size and structure . . . . . 6.1. Relative poverty rates among the population of working age . . . . . . . . . . . . . . . . . . . . . . 6.2. Relative poverty rates among the working-age population, non-employment rates and joblessness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3. Relative poverty among the working-age population and social spending in 2000 . . . . . . . 6.4. Effects of taxes and transfers in reducing relative poverty among the population of working age . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.5. Gross public spending on disability benefits, as a percentage of GDP . . . . . . . . . . . . . . . . 6.6. Gross public spending on social assistance, as a percentage of GDP. . . . . . . . . . . . . . . . . 6.7. Relative income of non-employed disabled persons and single parents, late 1990s. . . . 6.8. Full-time recipients of various income-replacing benefits among the population of working age, 1980-1999. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.9. Outflow rates from disability benefits . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.1. Changes in welfare caseloads following welfare reforms . . . . . . . . . . . . . . . . . . . . . . . . . . 7.2. Single mothers receiving welfare benefits and in employment, United States . . . . . . . . 7.3. Potential impact of higher employment among lone parents on relative poverty rates of single-parent households . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.4. Potential impact of higher employment rates among disabled persons on their relative income . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.5. Relative poverty rates among households with a working-age head around 2000, by number of adults working . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7.6. Impact of full take-up of benefits on poverty in the United States . . . . . . . . . . . . . . . . . . 7.7. Net income of social assistance recipients, 2001 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.1. Mean disposable income of the elderly relative to persons aged 55 to 64, 2000 . . . . . . . 8.2. Gross public spending on old-age pensions, as a percentage of GDP . . . . . . . . . . . . . . . . 8.3. Trends in life expectancy in old age, OECD average. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.4. Considerable country differences in the projected pace of population ageing . . . . . . . . 8.5. Projected effects of ageing on public spending, 2000-2050 . . . . . . . . . . . . . . . . . . . . . . . . . 8.6. Population ageing could lead to a sharp slowdown in the growth of the OECD labour force . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.7. Employment rates by age and gender, 2003 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.8. Large country differences in the effective age of retirement . . . . . . . . . . . . . . . . . . . . . . . 8.9. Percentage of employees by age who received job-related training over the previous year . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.10. Relative poverty rates among persons aged 65 and above, 2000 . . . . . . . . . . . . . . . . . . . . 8.11. Older persons living alone, 1990 to 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.12. Sensitivity of long-term care spending to different factors in four European countries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8.13. Expenditure on long-term care, 2000 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.1. Net replacement rates of OECD pension systems for workers at different earnings levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9.2. Defined-contribution and defined-benefit plans in the United States, 1985-2002. . . . . . 9.3. Net tax cost per dollar of contributions to tax-favoured pension schemes . . . . . . . . . . . 9.4. Potential impact of various reforms on labour-force participation of the elderly . . . . . . 9.5. Potential impact of higher employment rates for elderly people on their relative poverty rates. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

8

82 83 95 96 97 98 104 105 106 106 107 108 109 111 112 118 119 120 126 128 134 135 145 145 146 147 148 149 151 152 153 153 158 159 160 160 172 174 179 180


ISBN 92-64-00794-6 Extending Opportunities: How Active Social Policy Can Benefit Us All © OECD 2005

Summary


9

SUMMARY

T

he lives of hundreds of millions of OECD citizens have improved radically in recent decades. People today live longer, healthier and more prosperous lives. Millions have entered or re-entered the world of work during the economic recovery of the 1990s, thus allowing major reductions in unemployment and benefit rolls. New initiatives are giving children and their families effective support in getting the best possible start in life. Much of this improvement has been the fruit of the economic growth that has taken place throughout the OECD. But, however essential economic growth is to improving people’s lives, it has not been sufficient to solve all social problems. Indeed, despite greater prosperity, a substantial portion of the population in every OECD country continues to face great risks: risks of disadvantage in childhood, of exclusion from work in prime age, of isolation and limited self-sufficiency in old age. That traditional social policies have not helped people address all these risks adequately is highlighted by one crucial fact: from the mid-1970s to the mid-1990s, labour and capital incomes have become more unequally distributed among the population in every OECD country. Opinion is, and will always be, divided as to how much we should care about income inequality. The fact that inequality in market income had been widening everywhere in OECD countries until recently is, however, a source of concern that is shared among policy makers. First, although people differ in how much they care about inequality as opposed to other public goals, such as economic growth and rewarding enterprise, few are entirely indifferent to the distributional outcomes of a market economy. Second, the fact that people are poor is a sign that they have not successfully participated in the labour force or in society: in other words, poverty and inequality are evidence of an inefficient society, which wastes human resources, opportunities and life chances. Third, children of poor parents have less chance of succeeding in life than children of rich parents: a widening inequality of income risks leading to a widening inequality of opportunity. Because of these factors, a failure to tackle the poverty facing millions of families and their children is not only socially reprehensible, it will also weigh heavily on our capacity to sustain economic growth for years to come. Widening inequality in the distribution of market income stems from a variety of sources, including advances in technology and globalisation, changes in demography and family patterns, and many others. What this report addresses, however, is not the cause of the change, but what social policies can do about it. The way social policies were set up in the past provides one possible approach to dealing with the more unequal distribution of market income: tax those who have benefited most from economic growth, those with higher incomes, to compensate those who are unable to find well-paid work, be it because of disability, sickness, lack of skills, or whatever. All countries continue to use such an approach as a central way of alleviating poverty and other forms of disadvantage. An intrinsic problem with this approach, however, is that if the widening of market income continues, sooner or later it will be harder to redistribute yet more cash, as better-off voters may reject continuing tax increases and climbing tax rates may deter investment and work effort. Furthermore, the obstacles to higher public social spending are already being

10


SUMMARY

exacerbated by the greater burden being placed on workers by the need to support an ageing population. Instead of relying on taxes and public transfers alone, OECD countries need to look for other ways to deal with the social challenges of today. In this volume, the policies that aim to do this are called active social policies because they try to change the conditions in which individuals develop, rather than limiting themselves to ameliorating the distress these conditions cause. This shift away from the reactive, compensatory approach of the past gives greater emphasis to investing in people and giving them incentives to participate actively in the labour market, so as to maximise their potential to become self-sufficient, autonomous members of society. It also emphasises the importance of basing social policies on a life-course perspective, so as to recognise how conditions in a given phase of an individual’s life inevitably influence those in the next, and to anticipate problems that may rise tomorrow. This report uses the life course to link experiences in childhood, primeage and old age. Social policy is often disparaged as being a burden on society, damaging the entrepreneurial spirit, discouraging work and savings, and creating dependency on the state. Bad social policies can indeed result in all these negative effects, but there is nothing inevitable about bad policies. Many examples of successful policy are presented in this report. This report also presents an agenda for active social policies and discusses how they can benefit us all. Using the life-course perspective, the report is structured around three key objectives of social policy: ●

Giving children the best possible start in life. Many social problems have their roots in childhood. Ending childhood poverty is as important for social policy as it is possible to imagine: children who grow up in disadvantaged households are more likely to do poorly at school, to struggle to find a job, and to be unemployed, sick and disabled when they become adults, precipitating an inter-generational cycle of disadvantage and deprivation. More generally, family instability can damage life chances, as can inadequate care – be it by parents who lack the time and skills to raise children or by paid carers with inadequate qualifications. What is needed is a mixture of reforms to the tax and transfer system, childcare support, and help for working parents – particularly mothers – to help parents reconcile their work and family responsibilities. Achieving this will also create more favourable circumstances for raising the level of fertility rates, which are currently below replacement levels in most OECD countries. Belowreplacement fertility rates impose large social and economic costs on society: the cost of pension payments falling on each worker rises; investment is discouraged due to a lack of consumers and savers; and the capacity of extended families to attend to the needs of their members is reduced. Governments recoil from pursuing avowedly natalist policies for obvious historical and cultural reasons. However, given that women themselves say that they want more children than they actually have, it is reasonable to consider whether there is anything governments could do to help them realise their goals. Policy priorities in this area thus include: ❖ Investing in children – including early childhood intervention programmes, particularly well-designed ones that closely involve the family. ❖ Boosting maternal employment – including by adjusting tax and benefit systems so as not to discourage second earners.


11

SUMMARY

❖ Reconciling family and work responsibilities – through co-ordination across a range of areas such as childcare provision, parental leave and family-friendly workplaces. ❖ Creating a framework favourable for raising fertility rates – through policies that share the costs of childrearing more broadly and that allow young couples to get a secure footing in labour markets. ●

Helping prime-aged individuals overcome barriers to quality jobs. Getting the jobless work and assisting the disadvantaged to acquire the skills to get higher pay are the best ways in which economic and social progress can be reconciled. The vicious cycle that goes from joblessness to loss of self-confidence, deteriorating skills, isolation, and exclusion needs to be broken – and the recent success achieved by many OECD countries in slashing some benefit rolls shows that it can. The progress achieved in integrating the unemployed into the labour force now needs to be extended to other groups, such as lone parents and those with disabilities, through tailored interventions that give greater emphasis to labour-market integration. But the results of these welfare-to-work policies, while positive, also point to the need for complementary welfare-in-work policies. Welfare-in-work is essential if benefit recipients are not only to get jobs but also to keep them and avoid poverty. For more socially disadvantaged individuals, more effective social assistance remains essential to provide adequate living standards, while alternatives to paid work in the market economy need to be found to encourage participation in social life, and to ensure that a culture of dependency does not take hold of children who grow up in families that are dependent on welfare benefits for their daily living. Policy priorities in this area thus include: ❖ Completing the welfare-to-work agenda – in particular with regard to lone parents and, in most countries, people with disability. ❖ Making progress with welfare-in-work – including through policies to “make work pay” and to increase job retention and the career prospects of low-paid workers. ❖ Strengthening the effectiveness of social programmes targeted to persons for whom paid work in the market economy is less feasible – including through extending coverage and take-up of existing programmes to all persons in need, assuring the adequacy of the benefit provided, and moving beyond “work” as the only focus for social policies. ❖ Promoting the coherence of different policies affecting poverty and exclusion and making long-term commitments to achieve poverty-reduction goals.

●

12

Protecting the well-being of the elderly by enhancing their participation in economic and social life. Pay-as-you-go pension systems have been crucial to improvements in the well-being of the elderly, but their increasingly high costs as the population ages are threatening their financial sustainability as well as the capacity to make social investments in younger generations. The solution cannot be only to reduce the level of old-age pensions (although such cuts have been made in some countries and further cuts are needed in others): breaking pension commitments that people have based their retirement plans on would undermine trust in government and put living standards in old age at risk. Moreover, old-age poverty – though hugely reduced from its peak – has not gone away. To put pension promises on a sustainable footing, pension systems need a more realistic link between benefits, on one side, and the life expectancy and income of the workingage population, on the other: it is not reasonable – nor does it correspond to the wishes of many older persons – for successive generations to spend ever-decreasing proportions EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

SUMMARY

of their lives in work. The need to induce larger numbers of older people to remain in the workforce does not, however, imply a blanket increase in retirement ages: sickness and disability rates indicate that some older people have barriers to work other than age that have to be addressed as well. In addition, increasing numbers of older people will need quality and affordable long-term care. This calls for policies that judiciously mesh formal and informal care, so that the frail elderly receive the care best suited to their circumstances and so that care responsibilities, most of which are assumed by women, do not unduly burden them or push them out of the labour force. Policy priorities in this area thus include: ❖ Limiting the costs of old-age pensions on public budgets – including by promoting greater diversification of retirement income among middle and higher-income retirees and better targeting public pensions on lower-income retirees. ❖ Favouring a longer working life – through a combination of steps to close pathways to retire early, increase the standard age for entitlement to a public pension, correct disincentives to retire early that are embedded in pension and other benefit systems and encourage employers to hire and retain older workers. ❖ Improving quality and access to long-term care – through policies to support informal carers, increase the availability of formal care services to frail elderly at home, increase users’ choice among alternative types of provision and better monitor care quality. Successful implementation of this agenda of active social policies would increase both the level and the quality of employment; it would eventually reduce reliance on transfer payments; and it would narrow the distribution of market income. Because of these effects, active social policies hold the promise of reducing the negative effects of social protection systems on economic growth that have long dominated public discussions about the welfare state. Implementing this agenda would not only better serve the “public good”, but would also reduce social isolation and exclusion, help individuals realise their personal goals, and provide a more secure life in old age. The challenge to public policy that this agenda implies is weighty: it is all well and good to talk of such active policies in the abstract, far less easy to make them work in practice. To a significant extent, social problems reflect actions that are outside the immediate remit of social policies. The complex links between different social problems, and between economic and social processes, point to the need to move away from the typical “one problem/one instrument” approach to many policy decisions, towards one based on explicit assessments of the social consequences of policies in different fields, identifying trade-offs between competing goals, and finding ways of shifting these tradeoffs over time. More than that, investing in active social policies is expensive. The returns are in the future – often the distant future, when today’s children enter the labour market, for example. This implies that reforms of social policies confront a double burden: paying for the failures of the past and investing now to ensure that such failures are not reproduced in the future. As pressures on public expenditures are already heavy, OECD societies are increasingly looking beyond public programmes to achieve social goals. The importance of supplementing government action is already evident, not only in the increasing role of privately-financed pensions, but also with regard to long-term care, most of which is provided by informal carers. What is the basis for further progress? Employers will benefit


13

SUMMARY

from a larger, more productive workforce, so they should bear some responsibility for ensuring that parents can juggle work and care for children, and that older workers or those with poor health can find a “niche” in the labour market. Individuals, too, should have greater responsibility to take advantage of those opportunities that society pays to make available to them, thereby increasing the effectiveness of interventions. Private financing and delivery of social protection can sometimes result in more efficient and responsive forms of social protection, just as non-governmental organisations are often able to mobilise resources and enthusiasm to an extent unavailable to publicly-run organisations. But shifting the financing and delivery of social protection outside government raises difficult issues about the extent and fairness of coverage, and it does not always deliver the efficiency gains and cost savings expected. It also requires that government move from being the direct provider of protection to assuming new and more complex governance functions. The present report is organised as follows. Part I provides a general context for a discussion of active social policies, linking the different issues and perspectives, reviewing past achievements, identifying drivers of future challenges and analysing the need and basis for mobilising non-governmental forces. Part II focuses on children and families and considers how best to balance the competing demands on families to work and care, and how to invest in children to give them the best possible start in life and secure equal opportunities for all. Part III focuses on people of working-age and considers how to combat poverty and exclusion among them, with particular attention to social assistance and disability benefits. Part IV focuses on the elderly and looks at the challenge of adapting social protection to shifts in the risks they confront.

14


PART I

The Social Policy Context

Significant improvements in social conditions have been achieved in recent years due to the combination of economic growth and effective social protection. Nevertheless, social distress remains only too common. The main way traditional social policy has addressed this is through publicly-provided benefits – to protect individuals against the risks of life and to ensure that economic and social development move in tandem. However, this traditional model is under increasing pressure today, because of concerns about its affordability and because of a widelyheld view that many clients are badly served by existing programmes. At the root of social distress are structural factors that are common to all OECD countries: the more unequal distribution of market income, shifts in disadvantage among stages in the individual’s life-course, rapid ageing of the population, more diverse family patterns and changing labour-market conditions. More and more countries are introducing active social policies to tackle the challenges posed by these trends. This move to a more active approach aims to encourage individuals to participate as fully as possible in economic and social life. It is based on recognition of the rights and responsibilities of individuals and organisations as part of the broader community where they live and function. These reforms typically combine efforts to give all children a better start in life, to help parents reconcile work and family responsibilities, to help individuals overcome barriers to work, to make work pay and to respond to the needs of people in the latter part of their life. While the specifics of active social policies vary according to the life-course of individuals and specific country circumstances, their common thrust is the need to go beyond insuring individuals against the risks of life, towards a greater focus on investing in peoples’ capabilities and enabling them to realise their full potential throughout their lives.



PART I

Chapter 1

Social Development in OECD Countries: Mapping the Progress


17

I.1.

SOCIAL DEVELOPMENT IN OECD COUNTRIES: MAPPING THE PROGRESS

1. Introduction This chapter sets the stage for the more specific discussion of how active social policies can better achieve the goal of improving social conditions for individuals and communities. It describes the progress that has been achieved over the past few decades in improving social conditions and reducing social distress in OECD countries, and the relationship between economic growth and social development.

2. Monitoring social development through indicators Social policy seeks to influence not only the distribution of resources in society, but also their levels. Its goals encompass more tangible life conditions, like health, but also less tangible conditions such as autonomy, feelings of social isolation, connectedness, and the ability to rely on the support of others when the need arises. As a convenient shorthand, we could refer to progress in achieving these objectives as social development. Because of the breadth of objectives, a broad range of indicators has to be used to chart progress in social development, to quantify the scope for further improvement, and to analyse the relationship between its different dimensions (Box 1.1).

2.1. The positives Social conditions in OECD societies have improved dramatically in many ways. Changes in economic well-being have translated into significant gains in living conditions, leisure time, and material ownership. Of special salience for an assessment of social conditions in OECD countries are indicators of health, education and labour-market performance.

18

●

Improvements in economic well-being over the past 50 years qualify as extraordinary when compared to historical precedents. Average real growth in per capita GDP in the OECD countries from 1950 to 1973 implied a doubling in living standards every 20 years. Despite significant deceleration since the early 1970s, growth in real per capita GDP since 1973 was still sufficient to double living standards every 35 years (Maddison, 2001).

●

Gains in real income of this size have been accompanied by an accumulation of assets that households can use in the event of adversity. The net worth of households in the (seven) major OECD economies in 2003 (or in the most recent year available) ranged between 5 and 7½ times their annual disposable income and – despite reductions since the levels recorded at the peak of asset prices in 1999 – remained higher relative to the early 1990s in several of them.1

●

Life expectancy at birth has increased by nearly 2.5 years per decade, with no deceleration in this trend, while infant mortality has dropped by three-quarters in the 30 years to 2002.2

●

The average amount of time spent in education across OECD countries has increased by close to one year per decade, to a level of 11½ years by 2000. Education not only has a significant impact on economic growth, but also delivers benefits both to individuals – in EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

I.1.


Box 1.1. OECD social indicators Social indicators are measures covering a variety of fields that reflect people’s objective conditions of life in a given country and period (Diener and Suh, 1997). Their ambition is to provide information beyond that conveyed by conventional measures of material wellbeing such as per capita income. Indeed, historically, much of the drive to develop social indicators originates in the belief that conventional measuring systems underestimate the value of production outside the market economy and thus give more prominence in decision-making to economic considerations rather than to the social conditions in which people live. Strengths of social indicators are the widespread agreement about the value of what is being measured, a presumption that these social characteristics can be measured reliably, and a belief that they are relatively independent of subjective perception. Social indicators also have weaknesses, which are usually related to differences in methodology and measurement systems, the ad hoc manner of selecting indicators, the difficulty of providing a parsimonious representation of the quality of life in a given society, and lack of agreement on how to aggregate indicators in the presence of trade-offs between different dimensions of well-being. The OECD has a long association with social indicators. A Working Group on Social Indicators, set up in 1971, led to the release of a first progress report in 1976 – Measuring Social Well-being – and to the publication in 1986 of the first OECD collection of social indicators in the report Living Conditions in OECD Countries. In 1999, this endeavour was resumed, giving rise to the bi-annual publication Society at a Glance. The social indicators presented in Society at a Glance may be represented along a two-dimensional classification. The first dimension corresponds to three main goals of social policy, i.e. self-sufficiency, equity and social cohesion. The second dimension corresponds to the nature of the indicators, i.e. social context, social status, and societal responses. Around 35 different indicators have been published in each issue of Society at a Glance, most recently in OECD (2005a). Evidence based on these indicators is reviewed below.

terms of higher earnings, better health, better parenting practices – and to society – lower crime and poverty, a lower likelihood of receiving social assistance, and greater participation in community and political life. ●

The position of women has improved in almost every area of social and economic life. The educational attainment and performance of girls is today as high as that of boys, and the extent of women’s participation in the labour market – while remaining below that of men – has increased substantially from the levels prevailing just a generation ago.

At the root of many of these improvements is the functioning of labour markets. Work gives individuals not only financial security, but also a sense of identity, belonging and selfrespect.3 The better functioning of labour markets over the second half of the 1990s in many countries has contributed to dispel much of the pessimism prevailing in previous years about the possibility of improving the life chances of the unemployed and of others with weak labour-market attachment. Overall, the employment-to-population ratio rose by around 1 percentage point from 1991 to 2001, and by much more in several OECD countries (OECD, 2003d).


19

I.1.


2.2. The negatives Despite these positive developments, much avoidable distress (inadequate income, weak labour market attachment, social isolation) remains and new social problems are emerging. Distress and disadvantage are functions of a broad range of factors, not all of which are within the immediate reach of social policy; in all cases, however, social policies can be called upon to remedy their consequences. Hunger and food insecurity persist amidst affluence in many OECD countries, and significant numbers of survey respondents (especially in Turkey, Mexico and Poland) report having been unable to afford basic food items at some point during the year (Table 1.1, 1st column),4 and even higher proportions report having been unable to buy health care and clothing the family needed (2nd and 3rd columns). Statistics on homelessness that are available in a few countries suggest that it continues to affect a non-negligible proportion of OECD citizens.5

Table 1.1. Share of the population who did not have enough money to buy family necessities during the past year, 2001 Percentage of respondents reporting that during the last year there have been times when they did not have enough money to: Buy food the family needed

Pay for health care the family needed

Buy clothing the family needed

Canada

9.5

12.5

16.4

Czech Republic

8.5

5.8

18.0 11.6

France

8.1

5.3

Germany

4.8

7.8

9.9

11.2

12.1

16.4

Italy Japan

3.9

4.4

5.0

Korea

18.3

14.9

20.9

Mexico

44.0

45.1

42.7

Poland

35.1

41.6

42.5

Slovak Republic

16.2

13.6

29.7

Turkey

44.7

47.5

47.5

United Kingdom

11.3

10.8

20.1

United States

15.4

26.5

18.8

Source: Pew Research Center (2002), What the World Thinks in 2002, The Pew Global Attitudes Project, Washington DC.

Lack of income remains the most basic determinant of social deprivation. There are two basic approaches to the notion of income poverty, which are reflected in two standards of measurement: “absolute” thresholds and “relative” thresholds. Both of these provide important information to policy makers, and they are ultimately complementary. For instance, while economic growth (i.e. a shift to the right of the entire distribution of household income) tends to pull people out of absolute poverty, its impact will decrease if it is accompanied by a widening of the lower tail of the income distribution.6 “Absolute” thresholds, as used by several OECD countries, typically relate to the costs of a given basket of necessities in a given year, updated annually in line with inflation. While absolute thresholds pose difficult methodological problems for cross-country comparisons, one way to show changes in absolute poverty over time is to use a relative income threshold in a base year for each country, and keep it unchanged in real terms. On average, among the 15 OECD countries for which these data are available over the entire

20


I.1.


Figure 1.1. Trends in absolute poverty rates in selected OECD countries Mid-1980s

Mid-1990s

Around 2000

Mid-1990s = 1.00 7.0 6.0 5.0 4.0 3.0 2.0 1.0

Ita ly

Ja

bl ic nm ar k Fin lan d Fr an c Ge e rm an y Gr ee ce Hu ng ar y Ire lan d

pu

De

Re

pa n M e Ne xic o th er lan ds No rw ay Po lan Po d rtu ga l Sp ain Un Sw ed ite en d Ki ng Un d ite om d St at OE es CD -1 5

Cz

ec h

Be

Au

st ra

lia lg iu m

0

Note: Levels of absolute poverty rates in the mid-1990s are set equal to 1. Thresholds for absolute poverty, as here defined, refer to thresholds set at 50% of median equivalised disposable income in the “base year”, kept constant in real terms in the following years. The “base year” differs across countries (mid-1980s for most countries, except the Czech Republic, Hungary and Poland – where it refers to the mid-1990s – and Australia and the United States – where it refers to the mid-1970s). The OECD average value is the unweighted average of 15 OECD countries for which information is available both in the mid-1980s and in 2000. Source: OECD questionnaire on income distribution and poverty.

period covered by Figure 1.1, absolute poverty has declined by over a third in the decade since the mid-1980s, and by close to 25% in the five years since the mid-1990s. This decline reflected gains in real household disposable income, which more than offset changes in the shape of the distribution.7 “Relative” income poverty refers to the proportion of the population falling below a particular proportion of median income. Relative poverty is a factor in social exclusion, which has a central role in the social policy agendas of several OECD countries, as it stresses the deprivation of individuals and families who are denied the goods and services that are customary in a given society.8 Indeed, when family budgets are directly examined to determine the amount of resources needed to afford decent living conditions, the range of expenditure items broadens: child-care costs, for example, hardly qualify as an essential item for meeting the basic needs of a person living on benefits, but become important when the same person is expected to work. One of the key implications of increasing relative poverty is precisely that larger numbers of people risk being excluded from economic and social life. For thresholds set between 50% and 60% of median disposable income, the assessment of progress made in reducing poverty in OECD countries is more muted.9 On the basis of a 50% threshold, the proportion of the population experiencing poverty ranges between 4% in Denmark and the Czech Republic to levels close to four times higher in Ireland, Japan, Mexico, Turkey and the United States. On average, across the 21 countries for which data are available in each of the years shown in Figure 1.2, close to 10% of the population had a disposable income less than half of the median around the year 2000, with an increase since the mid-1990s comparable to that recorded in the preceding decade. When a higher threshold is used, a significant share of the population – 8% or more in some countries – appears to be clustered between the 50% and 60% thresholds, and


21

I.1.


Figure 1.2. Relative poverty rates among the entire population %

Poverty at 50%

Increase in poverty when threshold moves from 50% to 60%

30 25 20

Mid-1990s Mid-1980s

2000

15 10 5

Au

st ra l Au ia st Be ria lg iu m Cz ec Can h a Re da pu De blic nm a Fin rk lan Fr d a Ge nce rm an Gr y ee Hu ce ng ar Ire y lan d Ita ly Lu Ja xe pa m n bo u M rg Ne ex th ico Ne erla w nd Ze s ala No nd rw a Po y la Po nd rtu ga Sp l S ain Sw wed itz en er lan Un ite Tu d d rk K e Un ing y ite dom d St OE ates CD -2 1

0

Note: Poverty rates are measured as the share of individuals with equivalised disposable income less than 50% and 60% of the median for the entire population. Data for Canada and Sweden for the mid-1980s are adjusted to take into account breaks in series in the mid-1990s. The OECD average refers to countries for which data in each of the three periods is available. “2000” data refer to the year 2000 in all countries except 1999 for Australia, Austria and Greece; 2001 for Germany, Luxembourg, New Zealand and Switzerland; and 2002 for the Czech Republic, Mexico and Turkey. “Mid-1990s” data refer to the year 1995 in all countries except 1993 for Austria; 1994 for Australia, Denmark, France, Germany, Greece, Ireland, Japan, Mexico and Turkey; and 1996 for the Czech Republic and New Zealand. “Mid-1980s” data refer to the year 1983 for Austria, Belgium, Denmark and Sweden; 1984 for Australia, France, Italy and Mexico; 1985 for Canada, Japan, the Netherlands, Spain and the United Kingdom; 1986 data for Finland, Luxembourg, New Zealand and Norway; 1987 for Ireland and Turkey; 1988 for Greece; and 1989 for the United States. Source: OECD questionnaire on income distribution and poverty.

experience difficulties in making ends meet. Poverty gaps – a measure of the extent to which the income of the poor falls below the poverty line – narrowed only marginally over the 1990s. The evidence that there has been little overall progress in reducing the extent and severity of relative poverty is disappointing in the light of the significant improvement in labour-market conditions over this period.10 Beyond income, subjective perceptions of social deprivation also relate to a lack of support or help from family, friends or the community where individuals live. In the United States, a large majority (86%) of all households declared that they expected to receive help in case of need, and this proportion did not vary much across households with different characteristics; however, when households did experience difficulties in meeting basic needs, less than 20% actually received the help they expected (Bauman, 2003). For European countries, an index of potential support from others in situations of difficulty shows higher levels of potential support among the population as a whole than among those in the lowest quartile of the income distribution. Potential support increased from 1996 to 2001 for the first group and declined for the second (Gallie and Paugam, 2002). A measure of the frequency of social contacts outside the family indicates that, on average, around 8% of survey respondents rarely or never meet with friends, colleagues or others outside their family, with much higher proportions in Mexico and Japan (Figure 1.3); the risk of social isolation is around 10% higher among persons with low income than among the population at large. With regard to health, there are significant inequalities in outcomes by race, gender and socio-economic status. For example, in the United States life expectancy at birth is 6½ years shorter for Afro-American men than for white men, while the equivalent gap for

22


I.1.


Figure 1.3. Social isolation in OECD countries, 2001 Rarely

%

Never

16 14 12 10 8

Low income people

6

Total population

4 2

n

ico

pa Ja

M ex

l

bl ic

ce

ga ec

h

Re

pu

rtu Cz

Ita ly

an

Po

ite d

Fr

d St at e De s nm ar Ge k rm an y Gr e ec Gr e ea tB rit ain Be lg iu m Ic ela nd Ca na da Sp ain Fin lan d Ko re a Au st ria

lan Ire

Un

Ne

th e

rla

nd

s

0

Note: This measure of subjective isolation refers to the proportion of respondents who indicate that they “never” or “only rarely” meet friends, colleagues and other non-family members for socialising. Countries are ranked in increasing order of social isolation. Low-income persons are those in the lower third of the distribution, as reported by respondents. Source: World Values Survey, 2001.

the two respective groups of women is slightly over 5 years. In the United Kingdom, it was around 7½ years shorter for unskilled men than for professionals (and around 5½ years shorter for women) at the end of the 1990s. In France, life expectancy at 60 is 5½ years shorter for blue-collar men than for professionals (and 3 years shorter between the two groups of women). The life expectancies of indigenous populations in Australia, Canada and New Zealand are also well below those of other groups.11 Data on health inequalities by education, income and occupation of individuals in a broad sample of EU countries suggest no consistent narrowing over time (Institute of Public Health, 2003).

2.3. The relation between social and economic variables Economic growth has been a driving force in improving social conditions in the OECD countries. But economic growth alone will not solve all social problems and in fact – by changing social and economic conditions – inevitably generates new social problems as well. Figure 1.4, which shows cross-country correlations between per capita income and different social indicators, highlights several patterns. ●

Richer OECD countries tend, on average, to enjoy better health and education. There is, however, much variation in the pattern of associations. For instance, the correlations for both health and education outcomes are about 70% but significantly lower in the case of child mortality and even lower for reading underachievement among students aged 15.

●

Subjective measures of satisfaction with own life, happiness and social isolation tend also to be positively and significantly correlated with per capita income.

●

Indicators of income distribution and poverty, measured using a relative income threshold in each country, display a weaker but positive association with levels of per capita income. The association is significant in the case of inequality (i.e. richer countries display lower income inequality than poorer ones), but not so in the case of relative poverty rates.


23

I.1.


●

Richer countries display higher prevalence of crime victimisation, imprisonment, suicides, divorce, drug use and road accidents (i.e. negative values are shown in Figure 1.4), although the correlations are always below statistical significance. While not all of these outcomes can be unambiguously characterised as “bad” (e.g. divorce), there is generally consensus that they translate into significant social costs for affected individuals and societies at large.

Figure 1.4. Cross-country correlations between per capita income and different social indicators in OECD countries % 100 80 60 40 20 0 -20

Road accidents1

Crime1

Suicides1

Drug use1

Divorces1

Prisoners1

Income inequality1

Relative poverty rate1

Social isolation1

Share of happy persons

Life satisfaction

Student reading underachievement1

Adult literacy underachievement1

Avg. years of schooling

Infant mortality1

Healthy life expect. women 60

Healthy life expect. men 60

Life expectancy at birth

-60

Healthy life expect. at birth

-40

Note: Pearson coefficient of correlation. For variables denoted with an “1” (e.g. infant mortality, prisoners), higher values of the indicators denote worse social outcomes. For these variables, correlation coefficients with per capita income are shown with the opposite sign (e.g. countries with higher per capita income have lower infant mortality rate: shown as a positive correlation; and higher rates of imprisonment: shown as a negative correlation). Per capita income is measured as GDP in current prices and purchasing power parity exchange rates, divided by the total population. Data refer to 2000 or most recent year available. The number of countries considered varies among different pairs of variables. Bars in a darker colour indicate statistically significant correlations (at a 5% level). Source: Different issues of Society at a Glance – OECD Social Indicators, OECD, Paris.

Among the social variables mentioned above, income inequality is the one whose relationship to economic growth has been the more thoroughly investigated. On one side, economic growth may affect income inequality – depending on whether all individuals at different points of the income ladder benefit from it in equal proportion or otherwise. On the other side, income inequality may also affect economic growth in different directions: greater inequality may increase economic growth when it raises the incentives facing individuals to work, save and invest, while it may lower growth when it reduces access by the poor to financial markets (because of market failures), when it generates social and political unrest, or when it leads to the adoption of policies that are inimical to growth. In practice, much discussion about the relationship between social development and economic growth is dominated by the supposed existence of a trade-off between the goals of growth and equity. In practice, the results from the empirical research on this trade-off has moved back and forth: OECD analysis of this issue concluded that, while evidence leans towards suggesting that a wider income distribution is good for growth, the

24


I.1.


estimates as a whole explained so little of differences in growth rates across countries and over time that nothing, in fact, could be safely concluded (Arjona et al., 2001).

3. Conclusions This chapter has argued that, while elusive in several respects, the notion of social development can effectively be measured though a variety of social indicators. These indicators suggest that much social progress has been achieved in recent years, due to a combination of economic growth and effective social policy programmes. Nevertheless, social distress remains, and in some cases has even increased over time. While these indicators show significant associations between measures of social development and of economic growth, there are also areas where the positive association between the two is weaker or where higher material well-being is associated with lower social development. Policies and institutions play an essential role to assure that economic growth and social development move in tandem.

Notes 1. See Annex Table 58, in OECD (2004), OECD Economic Outlook, No. 75, Paris. 2. Measures of income based on national accounts, which do not consider the length of time over which income can be enjoyed, do not reflect well the benefits of lower mortality and morbidity for the well-being of individuals. 3. Layard (2004), in summarising results from studies on the relationship between happiness and work, concludes that, when a person becomes unemployed, “the pain caused by the loss of selfrespect is [...] at least as great as the pain which a person would feel if he lost half of his income”. 4. Data in Table 1.1 are based on the Pew Global Attitudes Project, a set of 44 national surveys based on a standard questionnaire that explore similarities and differences in public opinion on issues such as democracy, inequality, globalisation and terrorism. However, because of their small sample size, the survey data shown in Table 1.1 may differ from those reported in larger-scale national surveys. With respect to access to food, for example, a national survey conducted in Poland in 2002 (Social Diagnosis) shows that the percentage of households whose basic needs for food were unmet ranged between 6% for sugar and 36% for fish; in the United States, a supplement to the regular labour force survey estimated that 11% of households were “food insecure” at some time during 2002 and that 3½% of individuals were suffering outright hunger [M. Nord, M. Andrews and S. Carlson (2002), “Household Food Insecurity in the United States”, US Department of Agriculture, Economic Research Service, Research Brief]. 5. In 15 EU countries, on an average day in the early 1990s, 1.1 million people had to sleep rough or in squats or rely on emergency services providing accommodation; 1.8 million homeless people were dependent on public or voluntary services for temporary shelter over the course of the year; and close to 1 million people rotated between staying with friends and relatives, living in short-term private accommodation and using services for the homeless (Avramov, 1996). In Japan, national surveys of the number of homeless people reported by local government officials estimated the number of people living on the street or in public shelters at 24,000 in 2001, reflecting a 50% increase since 1997. In the United States, although the number of homeless people is widely perceived to have peaked in the late 1980s, 80% of major cities surveyed in 2000 reported an increase in requests for emergency shelters relative to the previous year [US Conference of Mayors (2001), A Status Report on Hunger and Homelessness in America's Cities, www.usmayors.org/uscm/ hungersurvey/2001/onlinereport/homelessness.pdf]. 6. As an illustrative example, Freeman (2001) calculates that – when income is assumed to be normally distributed – a 0.1 point increase of mean income (relative to its standard deviation) would reduce absolute poverty by 3.2 points when 30% of the population is in the bottom tail of the distribution, by 2.6 points when 20% of the population is in the bottom tail, and by 1.6 points when only 10% of the population is in that tail; he concludes that “the impact of an increase in income on [absolute] poverty falls roughly in half as poverty drops from 30% to 10%, due simply to the shape of the income distribution”.


25

I.1.


7. The only exceptions to this pattern are Japan and (to a lesser extent) the Czech Republic, where lower real income at the bottom of the distribution made for an increase in absolute poverty over the second half of the 1990s. 8. The notion of social exclusion is broader that that of poverty. Atkinson (1998) identifies three common elements shared by all definitions of social exclusion: relativity (i.e. exclusion can only be judged by looking at a person’s, or group’s, circumstances relative to those of others in a given place); agency (i.e. exclusion implies an act by some agents, either the excluded person herself or third parties); and dynamics (i.e. exclusion reflects not just current circumstances, but dim prospects for the future). 9. The poverty data used in this report are based on the concept of equivalised disposable income of individuals, i.e. the disposable income of households, adjusted using the square root of household size. 10. Current income is, however, an imperfect indicator of social distress. First, a greater variability of household income, at a given average level, can reduce the well-being of risk-averse individuals. Second, the overlap between income-based measures of poverty and direct measures of material deprivation is only partial: some persons may have income above the poverty line and still report having been unable to afford food or medical care, or vice versa – although income-poor individuals tend to have a much higher probability of experiencing material deprivation. Third, the vulnerability of individuals and their families is related to the assets that they can draw on in case of need. Information on asset holding by income level of households is rarely available. For the United States, however, data from the Survey of Income and Programme Participation show that median net worth of households in the 1st quintile of the income distribution in 2000 was equivalent to less than 6 months of household income, with most of this taking the form of home equity: when excluding this component, the net worth of households in the lowest quintile was equivalent to less than 1 month of income (Orzechowski and Sepielli, 2003). 11. In Australia, in 1998-2000, life expectancy at birth for Aboriginal and Torres Strait Islanders was 21 years less than for the total population in the case of men, and 20 years less in the case of women; in Canada, in 1998 life expectancy of registered Indians was 6.5 years less than for the entire population; in New Zealand, in 2000-2002 the differences for persons of Maori ethnicity were 7.3 years for men and 7.9 years for women.

26



PART I

Chapter 2

The Need for More Active Social Policies


27

I.2.

THE NEED FOR MORE ACTIVE SOCIAL POLICIES

1. Introduction The improvement in overall social conditions realised over the past few decades in all OECD countries has partly depended on an extensive set of social programmes. While there is much diversity in the institutional arrangements and in the scale of resources used to provide social support, in all OECD countries the dominant type of intervention has been publicly provided benefits. A large proportion of these programmes has been directed at older people, reflecting what was in the past a clear concentration of disadvantage in the later years of a person’s life. Today, as a range of structural factors that are common to all OECD countries increases the scale and complexity of the problems that social protection is called upon to address, further improvements in social conditions require that social policies become more ambitious in their goals – aiming to prevent disadvantage throughout the course of an individual’s life.

2. The traditional model of social protection Public provision has characterised social protection in OECD countries through much of the past century. Data on public spending allow us to track the pace of this expansion. In the early 20th century, very few countries spent more than 1% of their GDP on public social programmes, and by 1930 only Denmark exceeded 3% (Lindert, 2004). The great rise of public social spending came in the 1930s and 1940s and continued through the post-war period. Public social spending across 21 OECD countries almost doubled in the 20 years from 1960 to 1980 and continued to grow after that, although at a reduced rate, until reaching a maximum of 23% of GDP, on average, in 1993 (Figure 2.1).1 Since then, the level has declined by around 1½ points of GDP on average. This decline is accounted for entirely by non-health expenditures. The dominant form of public social spending is cash payments, which represent over 60% of the total throughout the period since 1980, but with significant differences across countries – ranging from between less than half in Canada, Iceland, Korea and Norway, to 70% or more in Austria and Poland. The largest share of these public spending flows has been directed to older people. The difference between the share of social spending targeted at the elderly and the share of elderly people in the total population is a good indicator of the extent to which social protection systems are today disproportionately “pensioner systems”. Figure 2.2 – which plots per capita benefits received by persons of a given age, relative to the per capita benefits of all persons, in 1998 – shows that persons aged less than 50 receive a share of benefits (net of the direct and indirect taxes levied on these benefits) that is always less than their population share. Conversely, persons aged 65 or more typically receive a share that is at least twice as large as their population share. The distribution is particularly skewed in the cases of the United States and Luxembourg, but less so in Sweden and the United Kingdom. While this pattern of higher social benefits among the elderly partly reflects the higher contributions that they have paid over their life-course, it also suggests that social protection systems mainly redistribute resources across different age groups

28


I.2.


Figure 2.1. Gross public social spending in OECD countries, as a share of GDP, 1960-2001 Australia

Canada

New Zealand

Japan United States

Mexico

Austria

Korea

Belgium

Germany

Luxembourg

Switzerland France

Netherlands

United Kingdom

% 40

% 40

35

35

30

30

25

25

20

20

15

15

10

10

5

5

0 1960 1980 1983 1986 1989 1992 1995 1998 2001

0 1960 1980 1983 1986 1989 1992 1995 1998 2001

Czech Republic Norway % 40

Denmark Sweden

Hungary

Finland

Ireland

Iceland

OECD-21

Spain % 40

Italy

Greece

Portugal

Turkey

Poland

Slovak Republic

35

35

30

30

25

25

20

20

15

15

10

10

5

5

0 1960 1980 1983 1986 1989 1992 1995 1998 2001

0 1960 1980 1983 1986 1989 1992 1995 1998 2001

Source: OECD Social Expenditure database.

rather than among persons of a given age confronting different contingencies.2 This pattern – which is based on evidence for individuals derived from surveys and microsimulation models – is consistent with that highlighted by aggregate spending data. Across 25 OECD countries, per capita public spending towards the elderly exceeds spending on the non-elderly by a factor of between 3 and 6; and, over the period 1990 to 2001, this ratio increased in around two-thirds of all OECD countries. The large amounts of public social spending are reflected in the large number of persons who rely on these benefits for their daily living. While most elderly people rely on public pensions, many working-age people also receive public benefits. Administrative data show that, across 16 OECD countries, close to 20% of the working-age population was receiving some type of income-replacing benefit at the end of the 1990s. This proportion ranged from around 10% in Spain and Japan to close to 25% in Denmark, France and Belgium, and to even higher levels in the Slovak Republic (Figure 2.3). The largest categories of these income-replacing benefits are incapacity, unemployment and old-age benefits paid to persons of working age, followed by social assistance and lone-parent benefits. Relative to 1980, recipiency of cash benefits among the population of working age increased, on average, by around 40%. In the 1990s, however, growth in the number of


29

I.2.


Figure 2.2. Net public social benefits per capita by age of recipients, 1998 Germany

Finland

France

Italy

United Kingdom

% 500

Luxembourg Sweden

% 500 400

300

300

200

200

100

100

0

0

18

18

-1 20 9 -2 25 4 -2 30 9 -3 35 4 -3 40 9 -4 45 4 -4 50 9 -5 55 4 -5 60 9 -6 65 4 -6 70 9 -7 75 4 -7 9 80 +

-1 20 9 -2 25 4 -2 30 9 -3 35 4 -3 40 9 -4 45 4 -4 50 9 -5 55 4 -5 60 9 -6 65 4 -6 70 9 -7 75 4 -7 9 80 +

400

Norway

United States

Note: The data refer to social benefits in cash, net of taxes paid on them, theoretically available to persons under the rules of various countries, as computed from income survey data and simulation models. Values above 100% indicate that per capita benefits paid to persons of a given age exceed those paid to the population as a whole (or, alternatively, that the share of benefits going to persons of a given age exceeds their population share). Source: Dang, T.-T., H. Immervoll, D. Mantovani and K. Orsini (2005), “Age and Economic Well-being in OECD Countries: Anatomy of incomes, taxes and benefits in the late 1990s”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

Figure 2.3. Recipients of cash public social benefits, persons of working age % 40

30 OECD average

20

10

ic bl pu

Re ak

ov Sl

m

ce an Fr

iu

ar

k

lg Be

y an

nm De

rm

ria

Ge

st Au

d

en ed Sw

lan Ire

m

Ki

ng

do

da na

d Un i

te

ds lan

Ca

lia

er

st

ra

th Ne

nd Ze

ala

at St

w Ne

d ite Un

Au

es

n pa Ja

Sp

ain

0

Note: Ratio of working-age recipients of income support benefits as a percentage of the working-age population, in 1999. Countries are ranked, from left (the lowest) to right (the highest), according to the rate of benefit dependency in 1999. The horizontal line shows the average value across the 16 countries shown. Source: OECD (2003b), “Benefits and Employment, Friends or Foe? Interactions Between Passive and Active Social Programmes”, Employment Outlook, OECD, Paris.

benefit recipients nearly stopped, with increases in the first part of the decade followed by declines in the second: when excluding the Slovak Republic, around 80% of the average (across countries) increase in benefit dependency rates recorded since 1980 occurred in the 1980-90 period.

30


I.2.


3. Changes in the context of social policies The concentration of social protection on the elderly has reflected a particular distribution of social risks among the population. In the past, full employment and growth in earnings ensured that the standard “male breadwinner” family was adequately provided for, while poverty was mainly concentrated in old age and in short periods of inactivity during working life (Esping-Andersen, 2002). Today, because of a range of factors, this distribution of social risks no longer holds. Social risks have become more varied, and they increasingly affect individuals over their entire life-course. The factors that are at the root of these changes manifest themselves in the most direct way in the form of shifts in the distribution of economic resources at the level of market income. Other factors that have contributed to changing the distribution of social risks include the altered demographic characteristics of OECD societies, changes in the role of families, and changes in the nature of employment. Each of these aspects is described below.

3.1. Trends in income distribution Figure 2.4 summarises evidence about trends in income inequality in several OECD countries, looking separately at market and disposable income. The most striking fact is that market-income inequality increased rapidly over the two decades from the mid-1970s to the mid-1990s. Indeed, it increased in every sub-period and every country during this period (Förster and Mira d’Ercole, 2005). This cannot be coincidence: it suggests that deeprooted causes – at work mainly in the labour market – have been affecting developed market economies. However, it is significant that the increase in market-income inequality has come to a halt on average in OECD countries since the mid-1990s. As discussed below, this was mainly because some countries made successful inroads into joblessness, so that a broader range of the population benefited from the successes of the market economy in improving living standards.

Figure 2.4. Trends in the distribution of equivalised household income, OECD average Index, mid-1970s = 100 Mid-1970s = 100 120 Market income 110

100

Disposable income

90

80

Mid-1970s

Mid-1980s

Mid-1990s

2000

Note: Gini coefficient of income inequality, average of 17 OECD countries in the mid-1980s, mid-1990s and 2000. Changes over the period from the mid-1970s to the mid-1980s are based on average values for only seven OECD countries. The Gini coefficient is defined as the area between the Lorenz curve (which plots cumulative shares of the population, from the poorest to the richest, against the cumulative share of income that they receive) and the 45o line, taken as a ratio of the whole triangle: its values range between 0 in the case of “perfect equality” (each share of the population gets the same share of income) and 100 in the case of “perfect inequality” (all income goes to the share of the population with the highest income). The value of the Gini coefficient in the mid-1970s is set equal to 100. Source: OECD questionnaire on income distribution and poverty.


31

I.2.


Changes in the distribution of disposable income (after taxes and benefits) were smaller over this period. This reflected two factors. First, there was an increase in the overall level of taxes and social transfers in most countries. Second, in many countries the distribution of taxation became more concentrated on richer groups, and that of benefits more concentrated on poorer groups. The offset provided by taxes and social transfers has however only been partial. More importantly, such policies cannot continue to offset an ever-continuing widening of underlying market incomes. Voters can reject further increases in taxes and transfers, particularly if they feel (probably correctly, as discussed in Chapter 6) that disadvantage (and hence social spending) is becoming more concentrated on a relatively small group of severely disadvantaged people and families. Furthermore, there are economic costs to higher taxes and transfer payments, which the economy may ill-support. Hence – because of financial constraints on the capacity of public policy to limit inequality through higher government transfers and of the risk that redistributive policies aimed at offsetting a widening in market income inequality will bring about lower economic growth – the widening of market income is a cause of concern to social policy. In addition to the widening of market-income inequality, there has also been a shift in the distribution of economic resources across population groups and stages of the individual life-course. In the mid-1970s, poverty was largely associated with old age (Figure 2.5). Since then, poverty in old age has declined significantly, while childhood poverty has tended to increase. Higher childhood poverty is not necessarily a problem in countries where it reflects earlier leave from the parental home, and the somewhat inevitable ups-and-downs before settling into the labour market; but it becomes one when it affects children’s formative years. Social disadvantage during childhood can affect economic and social outcomes throughout a person’s life and may well be transferred to the next generation (Chapter 4).3 Evidence on the social mobility of individuals – the probability that people move between different income classes, and the opportunities that go with this – suggests that this declined in the United Kingdom and the United States in the past decade, and that economic disadvantage has become more persistent over time.4 Child poverty rates therefore indicate that OECD countries are some way from achieving equality of opportunity.

Figure 2.5. Relative poverty rates by age of individuals, OECD average % 30

Mid-1970s

Mid-1980s

Mid-1990s

2000

25 20 15 10 5 0

0-17

18-25

26-40

41-50

51-65

66-75

76+

Note: Poverty rates refer to the average of seven OECD countries. Poverty rates since the mid-1980s, calculated on the basis of a larger number of OECD countries, display a similar age profile. Source: OECD questionnaire on income distribution and poverty.

32


I.2.


3.2. Demographic influences The demographic context of social protection has also changed drastically, because of the combined effect of the decline in fertility rates and the increase in life expectancy in old age. Changes in both variables have consistently outpaced the predictions of demographers and have led to a continued ageing of OECD populations. Population ageing, in turn, by increasing the number of elderly people relative to those of working age, has made the implicit contract among generations that underlies most pension systems less sustainable. These demographic drivers have been compounded by changes in individuals’ life courses, towards a reduced number of years during which individuals contribute to the financing of social protection (due in large part to increasing time spent in education as well as retiring at an earlier age) and an increased number of years during which they receive pension benefits. While the impacts of population ageing on the financial sustainability of social protection are well understood, social policies have only recently started to focus on the ways in which the collapse in fertility rates could be attenuated. Fertility rates have declined in most countries to levels that are well below those required to assure the reproduction of OECD populations (Figure 2.6). The most rapid decline occurred from 1970 to 1985. From the mid-1980s to the early 1990s, the average decline in total fertility rates experienced by OECD countries was smaller, and some countries (e.g. the United States and several Nordic countries) experienced a recovery,5 which since 1988 has extended to a broader range of European countries as well as Australia. Nevertheless – aside from Iceland, New Zealand, Mexico, Turkey and the United States – no country comes close to having a fertility rate capable of maintaining a stable population in the long-run. While delays in women’s childbearing decisions have contributed to the observed decline in fertility rates, evidence on the total number of births by women over their entire reproductive lifetimes suggests that below-replacement fertility levels are going to be a feature of OECD societies for some time to come.

Figure 2.6. Trends in fertility rates in selected OECD countries OECD Germany

Australia Japan

United States

OECD

Italy 3.5

3.0

3.0

2.5

2.5

2.0

2.0

1.5

1.5

1.0

1.0

0.5

0.5

0

0

France

Canada

Germany

Korea

19 70 19 75 19 80 19 85 19 90 19 95 19 98 19 99 20 00 20 01 20 02

3.5

19 70 19 75 19 80 19 85 19 90 19 95 19 98 19 99 20 00 20 01 20 02

Sweden

United Kingdom

Note: The total fertility rate in a specific year is the number of children that would be born to each woman if she were to live to the end of her childbearing years and give birth to children at each age in agreement with prevailing agespecific fertility rates. A total fertility rate of 2.1 children per women ensures broad stability of the population (on the assumptions of no migration flows and no declines in mortality). Source: The data are derived from annual publications of the Council of Europe, Eurostat and national sources, as presented in OECD (2005a), Society at a Glance – OECD Social Indicators, OECD, Paris.


33

I.2.


Demographic factors calling for changes in social protection systems go beyond fertility. Migration inflows towards OECD countries have increased in recent years, because of family reunification, humanitarian migration in the wake of civil and ethnic conflicts, and the pro-active policies adopted by some countries. Greater ease of international travel, high labour demand and continuing wage disparities between developed and developing countries, in the face of official restrictions on movement, have also led to increases in irregular migration. Recently available statistics suggest that conventional measures of the stock of immigrants in OECD countries significantly understate their numbers: on average, census data on the stock of the foreign-born population exceed by around 60% figures based on “citizenship” (Figure 2.7), which is the measure used in most European and Asian countries.

Figure 2.7. Percentage of foreign-born persons and of non-citizens in OECD countries, around 2000 %

Percentage of foreign-born

Percentage of non-citizens

40 35 30 25 20 15 10 5

M

ex ic Tu o rk Sl ov P ey ak ola Re nd pu b Fin lic lan Cz H ec un d h ga Re ry pu bl ic Sp Po ain rt De uga nm l Un a ite No rk d rw Ki ay ng do m Ne Fra th nce er lan d Gr s ee c Ire e la Be nd lg iu Un Sw m ite ede d St n Ge ates rm an Au y st ri Ne Can a w ad Z a Sw eala itz nd er l A an Lu ust d xe ral m ia bo ur g Ja pa n Ko re a OE CD

0

Note: Countries are ranked in ascending order of the percentage of the foreign-born population (except Japan and Korea, where this information is unavailable). Data are based largely on census records of place of birth of the resident population. Census data refer to a year around 2000. Source: OECD (2005), Trends in International Migration, OECD, Paris.

Higher migration has important consequences for social protection. While increased migration cannot be expected to offset the impact of population ageing fully, it nonetheless plays a role in alleviating some of its consequences, at least in the short run, through increased fiscal receipts and the contribution of immigrants to demographic and economic growth. On the other hand, immigrants in some countries are over-represented among the ranks of the unemployed, the poor and the beneficiaries of social assistance, as individual and structural characteristics hamper their integration into the host society; and their behaviour and eligibility for social protection will adapt over time to match that of natives. The integration of immigrants confronts OECD social protection systems with new demands – such as providing language skills, housing and specialised education – and concerns – such as the consequences on local workers and communities when immigrants are highly concentrated in specific regions.

3.3. Household structures The sharing of resources within families has a fundamental influence on the wellbeing of individuals and on the setting of social policies. Two main trends characterise the

34


I.2.


recent experience of most OECD countries: declining household size and increasing diversity in household structures. The average household size has shrunk by around onefifth over the past three decades. Smaller household size – while raising the income of each member, for a given household income – also reduces both the availability of care provided by other family members in case of need and the ability to share costs and goods (e.g. housing) across family members (i.e. it implies smaller economies of scale in household consumption, Ringen, 1998). Declining household size is particularly marked in countries where multigenerational households have traditionally been more important (e.g. Japan and Mexico). Smaller households are due to lower rates of partnership formation; the higher age at which this happens; lower fertility; greater geographical mobility of family members; preferences for independent living; higher income and better health, especially among the elderly. In some countries, other factors are partially offsetting this decline. For example, the difficulties faced by younger persons in getting established in labour markets and in finding suitable accommodation have led many to delay independent living, increasing the share of those remaining in the parents’ home. Also, a narrowing of the gap in life expectancy in old age between women and men in some countries is increasing the share of older persons living with a partner. These factors, however, appear unlikely by themselves to reverse the trend towards smaller households, leaving more individuals without the support provided by other family members. The second trend is the increasing diversity of families. This is associated with a decline in the importance of marriages relative to partnerships and free unions, and with an increase in the frequency of partnership separations and in the number of lone-parent families (Figure 2.8). One partial measure of these separations is provided by the annual number of divorces, which in most OECD countries were equivalent to around 45% of annual marriages celebrated in 2000, as compared to 14% in 1970. Whether this trend will continue in the future is more uncertain. Since 1995, the ratio of annual divorces to marriages has declined in Canada, France, Sweden and the United Kingdom, while it stabilised in Denmark; the frequency of divorce has continued to increase in Southern European countries, although from levels much lower than elsewhere.

Figure 2.8. Ratio of annual divorces to marriages in selected OECD countries United States

Denmark

Canada

Sweden

Germany

France

Japan

Spain

0.7

0.7

0.6

0.6

0.5

0.5

0.4

0.4

0.3

0.3

0.2

0.2

0.1

0.1

0

1980

1985

1990

1995

2000

0

1980

United Kingdom Ireland

1985

1990

Netherlands Italy

1995

2000

Note: Ratio of the number of divorces in a given year to the number of marriages in the same year. Source: Data from Eurostat, Council of Europe and national sources, as presented in OECD (2005a), Society at a Glance – OECD Social Indicators, OECD, Paris.


35

I.2.


The greater frequency of marital dissolutions in most OECD countries is translating into greater diversity in household arrangements. Couple families have declined in importance in all OECD countries, although to different extents. Figure 2.9 shows that the proportion of people living alone has increased significantly, in particular single parents and older persons and especially in the Nordic countries, the Netherlands and Germany. While couples with children remain the dominant type of household in all countries, their importance is declining steadily. Although Figure 2.9 does not distinguish between married and cohabitating couples, the share of the latter has increased rapidly in all countries, and is today the dominant type of partnership in Sweden and other Nordic countries (Martin and Kats, 2003). Because of this trend, and because cohabitating couples are more likely to separate than are married ones, out-of-wedlock births have increased to account for more than half of all live births in Sweden in 2000, and for one-third or more in Denmark, France, the United Kingdom and the United States.

Figure 2.9. Distribution of the population by household types, mid-1980s and 2000 Persons in households with a head of working age Persons living alone

% 100

Single parents

Couples without children

Couples with children

80 60 40 20

Au

st ra li Au a st ria Be lg iu m Cz C ec an ad h Re a pu b De lic nm ar Fin k lan d Fr an Ge ce rm an Gr y ee Hu ce ng ar Ire y lan d Ita ly Ja pa M n Ne exi th co e Ne rlan w d Ze s ala n No d rw a Po y lan Po d rtu ga l Sp a Sw in S ed Un witz en ite erl a d Ki nd Un ngd ite om d St at es

0

Persons in households with a head of retirement age Singles

% 100

Couples

80 60 40 20

C ec an a h Re da pu b De lic nm ar Fin k lan d Fr an Ge ce rm an Gr y ee Hu ce ng ar Ire y lan d Ita ly Ja pa M n Ne exi th co e Ne rlan w d Ze s ala n No d rw a Po y lan Po d rtu ga l Sp a Sw in S ed Un witz en ite erl a d Ki nd Un ngd ite om d St at es

a

m iu

lg

st ri

Cz

Be

Au

Au st

ra li

a

0

Source: OECD questionnaire on income distribution and poverty.

Patterns of household formation in a number of OECD countries also point to an increase in the probability of unions between persons with similar social-economic characteristics. Most of this evidence refers to individuals’ level of education, and shows that the gap between the educational levels of partners within a couple has decreased over

36


I.2.


time, even after controlling for the general increase in women’s education.6 Patterns of partnership formation are of special salience for social policies because of the role that education plays in the intergenerational transmission of disadvantage and in shaping individuals’ demographic and socio-economic outcomes: for demographic groups characterised by strong educational inheritance between mothers and daughters, the increase in the probability of unions between persons with similar education will lead to larger inequalities in education and earnings for future generations (Mare, 2000). However, while trends in patterns of partnership formation are relatively clear, research on their consequences remains scarce – because of the complexity of the problem, and because of the need to consider simultaneously the impact of intergenerational social mobility and partnership formation.

3.4. Labour-market trends Changes in family settings are interacting with trends in labour markets, and with the increasing feminisation and ageing of the paid labour force, to determine changes in the risks confronting workers. An increasing number of couples are opting for dual-earner households as the most effective hedge against the risks of life. Among couples with young children, the combination of a full-time employed man and a not-employed woman has declined in importance in all OECD countries (Martin and Kats, 2003). These changes in the demographic composition of employment (older and more feminised) have led to greater differentiation in working schedules and increased pressure on governments to adapt working-time regulations to individual circumstances. The demise of the male-breadwinner model has been accompanied by an increasing risk of joblessness, especially for adults living alone. There are, however, large differences across countries in the share of persons living in households where no adult works, with a range from less than 4% in Japan and Mexico to close to 16% in Poland and Germany (Figure 2.10). Moreover, in several OECD countries, the lower rates of labour-market inactivity among individuals recorded over the 1990s have not translated into lower levels

Figure 2.10. Persons in households with a head of working age and no adult working, 2000 Persons in jobless households (%) 20

Persons without jobs (%) 60 50

16

OECD average

12

40 30

8

20

4

10 0 M

Ja pa n Sw exi c itz o er la Po nd Un rt ite uga d St l at e Ca s na S da Lu we xe den m bo ur Tu g rk e Fin y lan Gr d e De ece nm ar Au k st ria Ne Irel w and Ze ala nd Ita No ly rw a F y Ne ran c t Cz he e ec rla n h Re ds pu Un Au blic ite st d ral Ki ia ng d Hu om ng Ge ary rm an Po y lan d

0

Note: Countries are ranked in increasing order of the share of the population in jobless households (shown on the left axis). Diamonds represent the share of the working-age population that is not employed, based on labour force survey data (shown on the right axis). Exact years are those mentioned in Figure 1.2. Source: OECD questionnaire on income distribution and poverty.


37

I.2.


of joblessness at the household level, leaving individuals in these conditions exposed to risks of social deprivation, poverty and benefit dependency (Gregg et al., 2004). These patterns in the distribution of paid work are increasing pressures on social policies to find effective ways of encouraging entry or re-entry into work and avoiding a concentration of joblessness into a few households. Social protection systems are also affected by labour-market trends towards the destandardisation and individualisation of work relationships. De-standardisation takes many forms. Looking first at contractual arrangements, part-time work, fixed-term contracts and self-employment have increased in most countries, even though permanent, full-time contracts remain the norm in most.7 While this diversification in arrangements allows increasing employment among groups with different preferences, it also implies greater variation in the protection that jobs provide to workers. Some jobs are poorly paid and offer poor protection, and workers holding them are sometimes trapped in them over prolonged periods of time, or may simply move between such jobs. This differentiation in contractual relations is also contributing to growing differentiation in individual trajectories in the labour market, leading to more frequent transitions between work and inactivity, between different types of jobs, and between different career paths. All of these changes undermine social programmes built on the assumption of continuous employment relations. Among the many factors that have shaped changes in the functioning of labour markets is the increasing international integration of product and capital markets. Much discussion has recently focused on the possibility that more open economies may reduce the scope for financing social protection through taxes and social security contributions that increase labour costs, and that this may lead to a race-to-the-bottom in levels of protection.8 Yet evidence also points to the opposite effect, i.e. that greater openness to international trade and investment makes a country rely more on social protection to deal with the greater vulnerability to external shocks brought about by globalisation (jobs displacement due to trade and outsourcing, Rodrik, 1997). The jury is still out on which of these two opposite effects will prove more powerful in shaping the future of social protection.

4. More ambitious goals for social protection Because of these structural forces, the traditional model of social protection is running up against the limits of its ability to deal with the scale of change sweeping through contemporary society.

38

●

With respect to the distribution of economic resources, trends towards a widening in the distribution of market income call into question the sustainability of polices that aim to counter this only through taxes and transfers. What is required is avoiding everwidening inequalities in market income through policies that pay greater attention to labour market integration. Shifts in economic disadvantage toward younger age groups – with greater risks that their consequences become permanent – also call for a redirection of social protection so as to reduce the current bias of social protection systems in favour of the elderly.

●

With respect to demographic changes, population ageing increases the costs of old-age pensions and other programmes towards the elderly, at the risk of crowding out other social programmes and of increasing the weight that their financing imposes on today’s


I.2.


working-age population. While population ageing reflects many factors outside the reach of social policies, the heavy burden it puts on the public purse is in part a consequence of the failure of social protection to adequately support families with children so as to create more favourable conditions for raising birth rates. Growing immigration towards OECD countries also confronts social protection with a range of new demands, while also offering benefits. ●

With respect to household structures, the growing diversity in family settings implies greater risks that some persons may grow up with special disadvantage, increasing the importance of programmes capable of moderating the effects of family background on individuals’ life-chances. Changes in household structure also imply that family support can no longer be taken for granted, increasing the importance of care for children, persons with disabilities and the frail elderly, and of adapting social programmes to the greater diversity in family needs.

●

With respect to labour markets, the rise of dual-earner families is leading to the demise of the traditional “male breadwinner” model underpinning social protection. Also, due to the increasing diversity of individuals’ experiences in the labour market, the traditional social insurance approach that awards benefits on the basis of previous employment risks leaving some of the groups most exposed to social risks without adequate social protection.

Because of these consequences, social policy needs to move in a new direction. It needs to shift its emphasis from an approach focused on insuring individuals against a few, well-defined risks, mainly concentrated in later life, towards a more ambitious set of goals to meet a continuum of needs throughout the life course. This is the reason why a lifecourse perspective – from childhood, to prime age and old age – is used in Parts II to IV of this report to describe the set of challenges confronting OECD countries and the range of policies most effective in dealing with them. The term active social policies is used to characterise this new model for social protection. Active social policies pursue three main goals. First, is to give children the best possible start in life, in particular by reducing childhood poverty and by helping parents (and potential ones) to reconcile conflicts between their family responsibilities and work obligations. Second, is greater emphasis on work as a way to ensure that individuals in their prime age are not excluded from mainstream society. Third, is adapting social protection to changes in conditions in old age, recognising new needs (e.g. dependency) and new roles that the elderly can play in today’s societies.

5. Conclusions The scale of ongoing labour market, family and demographic changes means that income redistribution alone will not suffice to counter the trend toward a more unequal distribution of market income. These changes are modifying the configuration of risks that people face during their life course. While traditional social policy continues to be needed to alleviate the distress of persons in particular contingencies, this is not enough – it is also essential to emphasise an active approach that focuses on investing in people’s productive potential. However, this will also require additional resources, making it necessary to reconsider not only the balance of current government spending but also the prospects for mobilising non-government resources to achieve a more ambitious set of goals for social protection.


39

I.2.


Notes 1. The OECD Social Expenditure Database covers the provision, by both public and private institutions, of benefits and financial contributions to households whose circumstances adversely affect their welfare. Social expenditures involve either redistribution of resources across households or compulsory participation; data are available for nine main categories – old-age, survivors, incapacity-related benefits, health, family benefits, active labour market programmes, unemployment, housing, and other social policy areas – and, for each of these categories, they distinguish between cash-benefits and services. 2. At any point in time, the age-profile of social benefits is the result of a combination of transfers between and within age-cohorts. A particular transfer may aim at smoothing incomes across the life-cycle or at redistributing in favour of those facing particular contingencies. Regardless of the purpose of transfer payments and their sources of financing, age-profiles show which age-groups are the focus of interventions by the social protection system (Dang et al., 2005). 3. Taking as reference the income differential between households with children in the top and bottom quintile of the distribution in the United States in 1999 (12 times higher in the former than in the latter), Corak (2004) estimates that values of the elasticity between parents’ and children’s incomes within the range typically found in empirical research for the United Kingdom and the United States (between 0.4 and 0.6) imply that children born to higher income parents can be expected to earn between 2.7 and 4.5 times as much as children born to lower income parents; even in countries where estimates of such elasticity are found to be lower, such as Canada, Sweden and Finland (around 0.2), the income advantage of children born to higher income parents is around 1.7 times. 4. In the United States, those at the top and bottom of the income distribution are less likely to move between income groups than those in the middle, and around half of the individuals in the bottom quintile of the income distribution in 2001 spent all or most years over the preceding decade in that quintile, with an increase relative to the 1970s (Bradbury and Kats, 2002). In the United Kingdom, estimates of the correlation between parents’ and children’s income and earnings, for two cohorts born in 1958 and 1970, point to a steep fall in the extent of inter-generational mobility, a decline which seems to be partly accounted for by the fact that the expansion of higher education benefited people from rich families more than those from poor ones (Blanden et al., 2002). 5. However, fertility rates declined steeply in Sweden in the first half of the 1990s, in the aftermath of a deep economic recession and of reforms in welfare programmes. 6. For the United States, measures of the probability of unions between persons of similar education that control for changes in the distribution of schooling suggest that the degree of resemblance – in terms of schooling years – between spouses increased by 25 to 30% from 1940 and 1990, and that the growth of educational barriers to marriage is larger and more sustained at the highest levels of schooling (Mare, 2000). 7. Evidence on different forms of non-standard jobs is reviewed in different issues of the OECD Employment Outlook (2002, for temporary employment; 2000 for self-employment; 1999 for parttime jobs). 8. According to Tanzi (2000), a number of pressures already evident today may increase tax competition between countries in the future. Because of these factors, Tanzi argues that “it would be prudent […] for the welfare states to begin preparing themselves for what could prove to be significant falls in tax levels in future years”.

40



PART I

Chapter 3

Broadening Roles for Social Protection


41

I.3.

BROADENING ROLES FOR SOCIAL PROTECTION

1. Introduction Social protection in OECD countries has always been delivered through a combination of public agencies, profit providers, not-for-profit groups and families. Because of this feature, a simple dichotomy between the public and the private sector cannot do justice to the complexity of arrangements that characterise social protection systems, nor to their evolution. Each of these actors can, in turn, play a variety of roles. Burchardt (1997) provides a useful typology for mapping different welfare systems that distinguishes among delivery, financing and decision-making. As this perspective recognises, social protection can be assured through a variety of different arrangements. For example, the public sector can finance and deliver a service, or contract with the private sector to secure provision, or limit its functions to regulating a private provider whose services are directly paid by individuals. The family plays a central role in social protection, sharing resources and providing care and nurturing. The other main pillar in every social protection system has traditionally been the government. While debate rages about which actors should deliver different components of social protection, there is a growing consensus in every country that the traditional model is running up against its limits – that it is neither financially sustainable nor optimally suited to meet the rapidly changing configuration of risks facing growing numbers of citizens. This has led to a search for ways to broaden the roles of actors, both within and outside government, and to achieve a better mesh between the different actors within this broadened social protection system. This chapter starts by reviewing the rationales for private and public responsibilities for social protection. It then describes differences in social protection arrangements across OECD countries and some recent reforms that have shifted the boundaries in the social protection system between the public and the private sector in terms of delivery, financing and decision-making. Last, the chapter describes those factors that may shape decisions about the roles of individuals, families, employers, trade unions, and not-for-profit organisations in achieving the goals of active social policies.

2. The theoretical case for and against private roles for social protection Much of the theoretical discussion about social protection has treated public and private responsibilities as alternative arrangements, as if they were in opposition to each other. These discussions have also often been confused by a lack of agreement on the fundamental criteria to assess the merit of different arrangements: for example, whether the underlying goals of reforms that expand the role of the private sector is to increase efficiency in the delivery of a given package of services, or to give individuals more choice over the services on offer, or to promote fiscal savings.

42


I.3.


2.1. The case for private roles The case for expanding private roles in social protection builds on arguments that focus on both the financing and the delivery of social protection. Private financing of social expenditure may improve transparency and be perceived by individuals as being more closely tied to the value of the service purchased, and less as a tax paid to provide services to others; this may reduce disincentives to work and save and promote economic growth. Advantages of greater private financing of social protection include: ●

Getting the appropriate level of provision. Private financing may help individuals face the true price of social protection, and thereby reduce the risk of excess provision. In publicly financed, social insurance systems, individuals are rarely aware of the actual costs of social protection. The value of the contributions that individuals pay rarely matches the value of the benefits they can expect to receive in the future: employers (through social security contributions) or the electorate (through taxes) fund the deficit. When workers perceive their personal contributions as being the only costs they incur, they may come to believe that increasing social provision through higher employer contributions or general taxation is an attractive alternative, even if, in practice, both employers’ social security contributions and most taxes ultimately fall on workers’ wages. At the microeconomic level, by hiding from individuals the true costs of social protection, publicly financed programmes may harbour an in-built bias towards expanding provision beyond its optimal size. At the macro-economic level, public provision implies a higher tax wedge that can damage employment and output.

●

More developed financial markets. Greater private financing of social protection often involves greater use of financial markets. Private health insurance, for example, increases the volume of funds paid to insurance companies, which then diversify their risk by purchasing other assets. Similarly, the development of private pension systems usually involves investing the pension funds in financial markets by, or on behalf of, the individual or the company concerned. When the returns from financial markets exceed those provided by pay-as-you-go public pension systems, the pool of savings available for productive investment is correspondingly greater, with beneficial effects on the economy as a whole.1 The ability of private pension funds to bring specialist knowledge to bear on individual investment decisions, and the shift in investments towards higherrisk vehicles that will spur innovation and growth, may also lead to economy-wide benefits (Leahy et al., 2002).2

●

Increasing pre-funding of social programmes. Greater private financing for social protection may increase savings and build up assets that could be used later to cope with future contingencies, such as population ageing. Pre-funding is particularly important for retirement income provision, but similar arguments can be used for individual health savings accounts. Pre-funding makes it possible to smooth lifetime consumption of contributing cohorts of the population, and to increase the productivity of later cohorts, who will have to fund future pensions. While a public pension system may also be managed so as to increase pre-funding in anticipation of unfavourable demographic conditions in the future (either by generating assets that could be sold when the scheme moves into deficit, or by cutting the public sector debt today so as to reduce the tax payments needed to service that debt tomorrow),3 there are good reasons why prefunding is better left to the private sector, to minimize the risk that government may pursue objectives other than achieving the highest rate of return on its investments.


43

I.3.


Other arguments in favour of greater private roles for social protection relate to the delivery of social services and insurance. Potential advantages here include: ●

More flexible products and practices. Regardless of who pays for the services provided, private delivery may provide for more choice in tailoring social protection to individual circumstances, and thus respond to concerns about the bureaucratic rigidity to which public programmes have often been subject. The public sector is not always very good at promoting innovation. Much of the case for private delivery in other areas of government is based on the argument that it frees managers from government interference. Governments are, intrinsically, risk-averse, as innovation comes with a risk of failure that few political systems value as an intrinsic part of the process of identifying “what works”. Some governments have addressed this problem by creating arms-length agencies responsible either for providing the goods and services in question or for regulating the providers, but the track record of many of these agencies is mixed.

●

Greater efficiency in administration. The profit motive is a powerful tool to cut costs and improve the value of the service delivered. Of course, some forms of cost-cutting are inappropriate: if people are entitled to a benefit, they should receive it, and very careful regulation may be necessary to assure private provision to eligible persons. Also, the case for using privatisation as a means of cutting costs has to take into account the fact that the public service ethos means that wages in the public sector may be lower than in the private sector and that more may be spent on marketing, leading to waste. Still, administrative gains are a possibility, not least because costs will become more transparent to users.

2.2. The limits of private roles The economic case in favour of greater private roles for the financing and delivery of social protection does not go unquestioned. With respect to private financing, many observers have disputed the frequent assumption that there are very high rates of return on financial investment: even when these high rates of return prevail in the short run, they do not imply that individual portfolios will offer such returns on a life-time basis. Moreover, assessment of the potential rate of return must also consider the risks associated with achieving this return. These risks can be reduced through more diversified investments, but only at the cost of lower expected returns, which makes the case for privatising pensions less compelling. Finally, with regard to pre-funding, unless assets are invested abroad (which is politically difficult in many countries) ultimately the consumption possibilities of the retired population will be limited by the supply of goods and services produced by the working population, and proceeds from asset sales will lead to higher product prices rather than to higher consumption levels. Experience has also shown that there can be significant additional administrative costs of privatising pensions, particularly if privatisation takes the form of the creation of individual accounts, and these additional costs can sharply reduce potential gains. Finally, any transition from existing pay-as-you-go pension systems to fully-funded ones implies a “double burden” for current workers – who have to finance pension benefits for the current generation of retirees and build up assets for their own retirement. Theoretical objections also relate to the very nature of the markets in question, which may make private provision of social protection unsuitable in some cases.

44


I.3.


●

Social protection (or at least, some elements of it) has the characteristics of a public good: the service purchased may deliver more benefits than those accruing to the person consuming it. As a consequence, when individuals (or firms) decide how much to spend on such goods, they are likely to take into account only their own private benefit and purchase “too little” of the services or social insurance in question. For example, childcare, in addition to the private benefits accruing to the parents and children directly affected, can benefit society as a whole if child-development outcomes improve. To take another example, people who take into account only their private interests might not get medical coverage against communicable diseases, leading to an increase in general morbidity.

●

Social protection may involve informational asymmetries: the consumer may lack information about the quality of the service provided, or about the conditions that will shape his needs in the future. Traditionally, the welfare state has offered a way around this problem: it promises to act in the public interest in ensuring that a quality service is provided. The private sector (with its concern about cutting costs and maximising profits) may not have the same incentive to offer quality services.

●

Social protection is often about achieving distributional objectives that cannot be guaranteed by the market. Most people object to seeing old people in poverty, or sick people going untreated, and are willing to provide them with help even though these elderly could and perhaps should have had the foresight to have saved for their retirement, or to have purchased health insurance. Also, informational asymmetries may raise concerns with regard to access to, and the level of, provision for individuals who (because of their health, or lower capacity to pay) may be excluded from private markets (Blank, 1999). This risk could increase with the diffusion of genetic screening of individuals that might allow private companies to exclude individuals with particular health risks from coverage.

These are important problems. However, they do not provide a blanket justification for exclusive reliance on the public sector, any more than the arguments listed in the previous section provide justification for exclusive reliance on the private sector. It is, therefore, difficult a priori to make a strong case for or against the public or the private sector in the field of social protection. Ultimately this is an empirical question, which hinges around whether it is possible to design a regulatory framework that both meets the public interest in social protection and encourages those aspects of private sector behaviour that lead to greater efficiency in service provision. In sum, while it is rather easy to outline theories under which greater private roles for social protection could lead to efficiency and welfare gains for individuals and societies, there is nothing inevitable about these outcomes: whether the public sector is second-best or not will depend on the specific circumstances of each country and on the area of social protection concerned.

3. Differences in social protection arrangements across OECD countries Over time and space, there have been substantial differences in views as to the appropriate role of government and non-governmental actors in social protection. The different approaches towards private roles in social protection followed by OECD countries can be highlighted using spending data – although this is in certain respects an inadequate yardstick to gauge the allocation of functions between the public and private sector.4 In this regard, while direct public spending represents the dominant model in all


45

I.3.


OECD countries, it is by no means the only one. First, gross public spending is often complemented by tax advantages provided for social purposes – while both direct and indirect taxation claw back from the client’s purse part of the cash benefits provided by governments. Second, social spending may be undertaken by the private rather than the public sector. Private social spending encompasses those expenditures for social purposes – spending to help families cope with sickness, joblessness, old age, etc. – that are undertaken by individuals, firms or non – governmental organisations rather than by governments and are funded through taxes or social security contributions.5 There are two main categories of private social spending: ●

Mandatory spending is undertaken by individuals or firms because the government legally requires this. Mandatory private spending plays a significant role in relation to incapacity-related benefits – such as sickness, disability and occupational injury – in Germany, the Netherlands, the United Kingdom and parts of the United States. In other countries, such as Denmark and Sweden, employer payments are used to top-up public benefits. Mandatory private social spending is also common in retirement income provision, in the form of mandatory contributions to employer-based and/or individual pension plans (in Australia, Finland and Japan).6

●

Voluntary spending is undertaken spontaneously by individuals and firms, though the government may encourage this through tax expenditures or other subsidies. Voluntary social spending is common in countries where public provision is more limited and in the area of old-age pensions – especially in the Netherlands, the United Kingdom and the United States (Adema, 2001). In this area, spending data do not fully reflect the importance of private pension programmes, as the data refer to the benefits paid under such programmes, rather than to the individuals’ contributions to these schemes, many of which are not mature yet. Other examples of voluntary social spending include private health insurance in the United States;7 employer-provided paid sick leave in Australia and Switzerland; employer-provided childcare support in the Netherlands; supplementary unemployment compensation in the United States; employers’ top-up to incapacity-related benefits in several European countries; and employer payments during parental leave in most countries.8

Some OECD countries have been quite effective in mobilising private spending to achieve social goals.9 Since the instruments used vary greatly between OECD countries, a more comprehensive assessment that takes into account all resources devoted to social protection invariably narrows cross-country differences in the overall level of support provided. 10 Consideration of tax expenditures and clawbacks, as well as of private expenditure,11 points to higher net social outlays, as a share of GDP, in the United Kingdom and the United States (both countries with below-average rates of gross public social spending to GDP) and lower levels in Austria, Netherlands, and Sweden (Figure 3.1). While voluntary private spending adds to social outlays in all countries, this effect is especially large in the United States and (to a lesser extent) the Netherlands. Despite differences across OECD countries, concerns about the importance of assuring universal access to social protection led to a preference for public provision in the 1960s and 1970s. In the 1980s and 1990s, the emphasis shifted to increasing competition among providers as a way to improve the quality and effectiveness of social protection and to make users more aware of its costs.

46


I.3.


Figure 3.1. Public and private social spending, 2001 Percentage of GDP at factor costs Gross public

%

Net public and mandatory private

Net total

35 30 25 20 15 10 5

8 -1 CD

De

OE

bl ic nm ar Ge k rm an y Ic ela nd Ire lan d Ko re a M ex ico Ne th er lan Ne ds w Ze ala nd N Sl o r ov wa ak y Re pu bl ic Sp ain Un Swe ite de d n Ki ng do Un m ite d St at es

da

pu

na Cz

ec

h

Re

Ca

lg iu m

Be

Au

st ria

0


4. The experience of reform in OECD countries Over recent years many OECD countries have experimented with reforms aimed at broadening responsibilities for social protection, at enhancing the role of civil society and mobilising agents and resources outside of government. As a result of these reforms, the boundaries between “public” and “private” in the domains of delivery, financing and decision-making have started to shift. Some countries have also instituted reforms to clarify the responsibilities of different levels of government and to improve the administration and quality of social programmes through improved co-ordination between government agencies, both vertically and horizontally, as well as through a clearer setting of programme goals and improved monitoring and evaluation of programme effectiveness. A further area of reform has been in relation to the rights and responsibilities of the clients of social protection programmes, particularly in relation to activation and personal effort. These reforms have been motivated by a range of objectives: the search for greater economic efficiency in social protection at the macro-level; improved administration and co-ordination of social programmes at the operational level; and more supportive behaviour by recipients at the individual level. The drive towards a greater role of the private sector within the social protection system of OECD countries has taken two main forms (Ascoli and Ranci, 2002). ●

The first has focused on the supply-side of welfare provision, whereby different private providers compete for public contracts to deliver different social benefits. This model retains public responsibility for the financing of social protection, while allowing for greater competition among private providers. The advantages of this model include the greater efficiency and innovation that competition could deliver, and the possibility of ensuring that individuals with less capacity-to-pay are not excluded from private provision. The drawbacks include that contracts are often difficult to specify when they involve multiple services, and that providers offering lower costs – but also offering lower quality – may come to dominate the market for social services. While a detailed specification of the services to be provided may help overcome these drawbacks, this would reduce the discretion of providers and, as a result, the scope for innovation. These


47

I.3.


policies also run the risk that providers will be less responsive to changes in client demand, as the agent financing the programme is neither the one delivering the benefit nor the one receiving it (these are often called agency problems). ●

The second form has focused on the demand-side of social protection. It aims to increase private responsibilities for the financing of provision. Vouchers given to individuals are one example of this approach; reforms that create personal entitlements to social protection are another. The advantage of this approach is that it avoids the agency problems of supply-oriented schemes, and that it is suited to the use of targeted subsidies so as to ensure that individuals with lower resources are not excluded from social protection. But problems of inadequate supply, limited competition and the possible sorting of clients remain.

In practice, most OECD countries have combined elements from both these approaches. This has taken the form of measures aimed at better distinguishing between funding and delivery of social protection; use of management techniques and procedures typical of the private sector within public programmes; tax allowances for clients and donations; and deregulation aimed at allowing new agencies to access sectors previously dominated by public monopolies. Reforms have also aimed to enhance the role of civil society, mobilise resources outside the public sector, clarify the responsibilities of different levels of government and improve co-ordination between agencies. Practical experience with greater reliance on the private sector in social protection – in the areas of health insurance, private pensions, and care provision for children and the frail elderly – suggests that gains in economic efficiency and cost savings have not always materialised to the extent expected (Pearson and Martin, 2004). Provision from a diversity of providers has indeed promoted choice and innovation, but efficiency gains have often been limited, because of the high costs of individualised provision, informational asymmetries between providers and consumers, and cross-subsidisation of particular groups. Moreover, private provision has raised problems related to the uneven coverage of the population. In other cases, however, these reforms have helped to increase the scale of resources devoted to social protection: for example, in the case of long-term care to the frail elderly in several OECD countries, reforms seem to have led to an increase in total provision in some sectors (Ascoli and Ranci, 2002). Despite differences in the balance of benefits and costs across social areas, these reforms have altered the function of governments, from tasks related to the day-to-day management of social programmes towards the more complex task of integrating the different components and ensuring oversight of the overall system.

5. Towards a better mesh of public and private roles in social protection Discussions about public and private roles in social protection often focus on the role played by individuals in programme financing and by commercial providers in service delivery. This focus is, however, too narrow: the range of actors involved is broader, and each of them can play different functions. Table 3.1 describes different arrangements for social protection based on the taxonomy proposed by Burchardt (1997), as applied to the areas of social security and employment and of community services and housing. In the first area, retirement pensions, social assistance and child benefits are some of the examples of the traditional model where the government plays the dominant role in delivery, financing and decision-making. The range of alternative combinations is wide.

48


I.3.


Table 3.1. Different arrangements of social protection Different functions Provision

Financing

Examples of social programmes Decision-making

Social security and employment

Community services and housing

Public

Public

Public

Retirement pensions, social assistance, Nursing home care, domiciliary services, child benefits, unemployment insurance, Housing benefits, services for homeless unemployment assistance

Public

Public

Private

Optional second pensions, public employment services, voluntary income-related unemployment insurance

Public

Private

Public

Compulsory child maintenance through Public housing rents government agency

Public

Private

Private

Private

Public

Public

Tax incentives and reliefs on pensions

Fees in municipal child care

Private

Public

Private

Vouchers and assistance with private employment services

Private

Private

Public

Mandatory private or occupational pensions

Private

Private

Private

Voluntary occupational and private pensions

Tax relief for owner-occupied housing, housing benefits for private renters, housing vouchers Private nursing homes and nursing services, imputed rents on owneroccupied housing, private rents

Source: Adapted from Burchardt, T. (1997), “Boundaries between Public and Private Welfare: A Typology and Map of Services”, CASE paper No. 2, Centre for Analysis of Social Exclusion, London School of Economics.

For example, in the case of optional second pensions the public sector assures financing and delivery, while decision on the use of funds is taken by the individual. At the other extreme, private nursing homes and services are examples of social care programmes where provision, financing and decision-making are all private responsibilities; even in this case, however, the government will generally establish a regulatory framework for the quality of services, which providers must meet. Getting the right mesh between the roles of all the actors concerned by, and involved with, social policy is crucial not only to mobilising actors not already involved, but also to optimising the levels and quality of protection provided.

5.1. Individuals Discussions about social policies have traditionally been divided on whether the clients of social policies are people who needed help through no fault of their own, or whether social support may make them worse off in the long run by trapping them in dependence (Lindert, 2004). While this debate has being ongoing, there is a growing consensus on one point: that the chances that social policy will be successful in addressing individual needs are enhanced when clients co-operate in collective efforts to ameliorate their own conditions. This responsibility increases as social policies move from a passive approach, limited to the payment of cash transfers, to an active one stressing integration and re-integration into work. If society is ready to invest more resources in the rehabilitation efforts and in preventing social problems, it is legitimate to expect individuals to co-operate in these efforts. Greater reliance on individual responsibilities within active social policies, however, raises new challenges for policies. Monitoring the behaviour of individuals is expensive, and the costs associated with tailoring interventions to individual conditions may be high. Both of these costs increase when clients are increasingly diverse in terms of their personal characteristics (e.g. their skills and labour market experience) and when the obligations


49

I.3.


imposed upon them try to differentiate according to their personal characteristics and the external conditions in which they live (e.g. the availability and affordability of alternative caring arrangements for those who are encouraged to move from benefits to paid work). Moreover, paid work may not be an option for some individuals who will always remain incapable of getting and retaining paid employment in the regular economy. In these circumstances – to avoid a cycle of marginalisation from parents to children – alternative ways to achieve social integration are required.12

5.2. Families Families provide social support to individuals in a variety of settings: mainly through the care provided to children and elderly persons, but also through informal exchanges of time and money between family members at all stages of the life course. Historically, much of the drive to develop formal systems of social protection derived from the need to relieve other family members (mainly women) of these obligations – so as to provide them with the same opportunities (in terms of financial independence and career progression) that are available to others – and from concerns that family networks are less effective in delivering social protection when external shocks hit a large number of families at the same time.13 Despite the development of social protection systems, however, the role of families in attending to the social needs of their members remains predominant in many OECD countries for some type of interventions (e.g. care of children and the frail elderly). An important issue relates to whether – and to what extent – public transfers should be used to support families for the care services they provide. Publicly-subsidised fostercare arrangements – under which families are paid for the care provided to children from other families, as an alternative to institutionalisation – exist in several OECD countries, and they have been growing in importance with the diffusion of kinship foster-care arrangements, which pay members of the extended family when children live full-time with them (Gilbert, 2003). Public programmes that remunerate family members for the care provided to frail elderly people are also becoming more important in several countries. Such arrangements are often a means to avoid more costly institutionalisation, and often better respond to the needs of clients. However, they also raise problems of agency, i.e. to what extent other family members can be trusted to make decisions that are in the best interest of the person being cared for. While traditional government regulations have been limited to extreme situations (i.e. removing children from abusive family settings), to the extent that the choices of other family members cannot necessarily be trusted to be in the recipient’s best interest, a role for government regulations is warranted.

5.3. Employers and trade unions Employers and trade unions are often very involved in decision-making about social protection; they sometimes deliver a range of social benefits to workers, and sometimes to the community where they operate; and, in addition, firms sometimes finance a substantial part of the social benefits of their workers. Because employers and trade unions can better monitor the services delivered; they are also often better placed to negotiate agreements for social protection at the workplace level and to reduce any risks of moral hazard among workers and employers (i.e. the possibility that workers and employers adopt more risky behaviour because of the coverage they are granted). The central role attributed by active social policy to integration into work gives employers broad responsibilities. Working practices that penalise specific groups – such as

50


I.3.


parents with young children, the disabled, or older workers – can generate significant costs to society at large: what may appear as a rational choice to an individual employer in a more competitive environment (e.g. preferring to lay-off a worker with a disability rather than modifying the work tasks and environment that would allow him or her to retain their job) will produce adverse outcomes for the individual and society at large when broader considerations are factored in. Also, working arrangements that result in excessive stress on family life will weigh heavily on individual workers and their families. More generally, if workers are asked to be more flexible in the workplace, firms will need to step up their efforts accordingly to adapt their working practices and work environment to allow employees’ integration. In some cases, employers may benefit from the social protection that they provide, in the form of more productive workers, lower job turnover, and greater effort and motivation. The business case for greater hiring and retention of less productive workers is, however, often weak. It is also different among categories of workers, being stronger for high-skilled workers than for low-skilled ones. However, population ageing may mean that the business case for integrating older workers, women with caring responsibilities and people with disabilities is set to grow stronger.

5.4. The not-for-profit sector Provision of social services represents around 20% of the employment in not-for-profit organisations, on average, and a higher share in Austria, France and Germany (Figure 3.2). Within the broad category of services, of special importance is care provided to pre-school children, the elderly and the disabled. The not-for-profit sector plays a large and rising role in providing emergency relief and shelter for the poor in the United States and other countries. It is the main provider of care for the frail elderly in Germany and France and in specific segments (home care) in Italy and Spain; it has a significant presence in countries

Figure 3.2. Employment in the non-profit sector in 1995 Recreation and advocacy

% 20

Education and health

Social services

16 12 8 4

-1 6 CD

OE

d er lan ds

lan

th

Ire

Ne

te

iu m

Be lg

s

m

St a

ite d

ng do

Un

Un i

te

d

Ki

ce

st ra lia

Au

Fr an

an y m

d

in

Ge r

Sp a

ria

an

Fin lan

Au st

Ja p

ic

y

pu bl

ar Cz ec

h

Re

ia

Hu ng

ov ak

Sl

M

ex ico

0

Note: Paid and volunteer work in non-profit organisations, excluding religious congregations, in 1995, as a percentage of non-agricultural employment. Social services include emergency relief, income support and maintenance. Recreation and advocacy include culture and recreation, environment, development and housing, advocacy, philanthropy, international, business and professional activities. Source: Salomon, L.M. and W. Sokolowski (2001), “Volunteering in Cross-National Perspective: Evidence from 24 Countries”, Working Papers of the Johns Hopkins Comparative Nonprofit Sector Project, Baltimore.


51

I.3.


characterised both by developed government programmes, such as Norway, and by widespread commercial provision, as in the United Kingdom (Ascoli and Ranci, 2002). The significance of the not-for-profit sector for social policies lies in both its advocacy and its service-delivery functions. Most not-for-profit organisations have their origins in civic commitments to defend the interests of the most vulnerable and excluded groups. Also, they often represent a direct expression of the “voice” of these groups in the media and political process. Anti-poverty groups have played a role in increasing the priority attached to poverty in the policy agendas of several OECD countries. In Ireland, for example, the adoption in 1997 of a 10-year National Anti-Poverty Strategy was partly the result of a ten-year campaign by anti-poverty groups and community and voluntary organisations that sought to create awareness of the nature and scale of the problem. Some of the areas where the not-for-profit sector has a large presence are characterised by services that are not very standardised, and where many beneficiaries have low income and little ability to make informed choices. This gives not-for-profit organisations a particular advantage, as their staff tend to care more about the level or quality of service being provided and perform tasks beyond their strict job description (Francois, 2002). The greater presence of the not-for-profit sector in the delivery of services to children, the disabled and the frail elderly also reflects the late and partial development of government programmes in this field. Outside the realm of service provision, private organisations have also been given a stake as intermediaries in the payment of cash benefits in some countries, for example by reimbursing them when they help eligible individuals to submit a relevant claim to a public programme (Currie, 2003). Many OECD countries have recently introduced measures aimed at increasing the role of not-for-profit groups. In the United States, for example, much discussion has centred on the role of faith-based organisations. The Australian experience in better involving community groups in the delivery of social services is described in Box 3.1. The case for greater involvement of the not-for-profit sector in social protection is set to become stronger as policies shift from the payment of cash benefits to delivery of rehabilitation and reintegration services. However, this shift raises new challenges for policy, for example when not-for-profit providers make support conditional on forms of behaviour in a way that the public sector would not. This raises the possibility that some citizen might receive worse treatment than others or even be denied help.

6. Conclusions The need to reach beyond the public sector in order to achieve a more ambitious agenda for social policy confronts all OECD countries. The trend towards more diversified institutional arrangements in social protection is likely to continue, although probably not to the point of overwriting current cross-country differences in the organisation of social protection or the diversity of approaches that underpins different types of programmes. While the recent experience with greater reliance on the private sector suggests that it is unlikely to relieve much pressure on the public purse, it also points to the scope for giving individuals a greater range of choice, leading to programmes that are better tailored to individual circumstances, and expanding total provision in some areas. At the same time, a greater role for private financing and delivery increases the importance of public oversight and decision-making to assure effective systems of social protection. A better mesh of private and public roles rests on recognising not only the advantages and

52


I.3.


Box 3.1. New partnerships across government, business and community: the Australian experience A range of reforms in service delivery has been implemented in Australia over the past seven years. These include contracting with NGOs to provide brokerage for early childhood services; partnerships across portfolios and levels of government, and with civil society, to deliver complex networks of services for very disadvantaged groups; the creation of a separate service delivery agency – Centrelink – to deliver the full range of social security benefits from a one-stop shop; contracting out entire services previously provided by government (such as the public employment service); and selective tendering in areas of high need such as reconnecting homeless young people with their families. NGOs have for many years delivered social services such as emergency interventions and accommodation and relationship counselling on behalf of government. This experience is now explicitly recognised in some government’s programmes, such as the Australian Government’s Communities for Children initiative, where responsibility for distributing funding to existing and new services has been transferred to selected nongovernment organisations. These partners engage the local community in identifying problems and solutions, and jointly develop a community action plan. New business-like relationships are emerging between the non-government and local government sectors. Similar models are in place at the state level. The Council of Australian Governments (COAG) has committed the federal, state and local governments to implement co-operative approaches on policy and service delivery to address Indigenous disadvantage. COAG has provided a common framework between levels of government that promotes tailored responses and helps to build stronger partnerships with Indigenous communities. At the federal level, the whole-of-government approach is given effect through a regional network of Indigenous Coordination Centres that work with local communities to “join up” and customise service delivery. These centres bring together, under the one roof, staff working in the main agencies administering government programs and services for Indigenous people. COAG is also sponsoring eight trials in Indigenous communities. In the Wadeye community in the Thamarrurr region (Northern Territory), for example, a pilot Shared Responsibility Agreement involves federal and state government agencies, the Indigenous Thamarrurr Regional Council and the people of Wadeye themselves. The agreement recognises the rights and responsibilities of each of the parties and provides a basis for co-operation. It uses the combined agencies to address the priority action areas of youth, housing and construction, women and families. Recent changes to the structure of government departments have also emphasised the importance of service delivery and policy implementation. The Department of Human Services was recently created to co-ordinate and improve the delivery of health, social security benefits, child support and rehabilitation services. At the federal level, the Cabinet Implementation Unit within the Cabinet Division ensures that implementation issues are considered when policy is developed.

drawbacks of different approaches to organising social protection but also their potential to work as complements rather than substitutes. Some of the specific areas where a better mesh between the public and private sectors could be achieved are discussed in subsequent chapters.


53

I.3.


Notes 1. Some economists have argued that the rate of return on equity investments can be high: for example, Feldstein (1998) takes a benchmark figure of 5.5% after inflation. This is significantly higher than current potential growth rates. The implication of such high rates of return is that if more savings were channelled to capital markets by, for example, reducing public provision of pensions and requiring individuals (or companies, on behalf of individuals) to save for their own retirement, investment would increase and growth rates would rise accordingly. For a discussion of the limits of this argument see Section 2.2. 2. For example, Davis and Hu (2004) find a positive relationship between the assets of private pension funds and economic growth in OECD countries, and a stronger one in emerging market economies. 3. In Ireland, for example, 1% of GDP must be “saved” annually by the public sector against future pension liabilities. The New Zealand Superannuation Fund, which commenced investing at the end of September 2003, is designed to partially provide for the future cost of New Zealand superannuation. 4. OECD social spending data classify flows as public or private according to the nature of the agents who retain control over them. 5. See Adema (2001) for a more detailed discussion of borderline cases. 6. Mandatory programmes here include those payments which, were they not to be made, would result in the employer incurring significant tax penalties. 7. In the United States, where the public health system does not provide universal coverage, employer-provided health benefits to workers, dependents and retirees were estimated at over USD 440 billion in 2004 (equivalent to over 4% of GDP) – and these expenditures do not include payments by individuals for health services. 8. Non-governmental organisations (NGOs) also provide important social support. However, comprehensive spending data on social services provided by NGOs that are comparable with the detailed information in the OECD Social Expenditure database are only available in a few countries. For example, in Sweden the value of social services provided by non-governmental organisations amounted to 0.4% of GDP in 1997. Estimates by the Australian Institute of Health and Welfare (1999) suggest that expenditures for welfare services by non-government community organisations was around 1.2% of GDP in 1997, with volunteer contribution to these organisations adding a further 0.2% of GDP, as compared to 0.9% of GDP in direct and indirect government assistance. 9. Mandatory private social expenditures, in the OECD classification, are significant in nine OECD countries, where they account for around 8% of GDP and one-third of gross public social outlays. Voluntary private social expenditures are smaller in scale but more widespread: in 21 OECD countries they represent a little more than 2% of GDP (with a small increase since the early 1990s, reflecting the maturing of private pension programmes) and around 5% of gross public outlays. 10. A measure of dispersion across countries narrows by around two-thirds when moving from a more narrow to a more comprehensive measure of social outlays. See also Adema and Ladaique (2004). 11. Private social expenditures are most significant in the areas of incapacity-related benefits, among mandatory outlays, and of old-age and health, in the case of voluntary ones. 12. These may include public support provided to persons caring for other family members, or to those who volunteer in not-for-profit groups or contribute to achieving collective goals in other desirable ways (e.g. after-school activities, tutoring students, and caring for children, the frail elderly or disabled persons). 13. The experiences of both Russia and East Asia in the 1990s provide ample evidence of how informal support provided by families, while crucial to protect individuals in the short-run, cannot easily cope with shocks that affect a great number of households simultaneously. These experiences also illustrate how – in the absence of effective social protection systems – the strategies used by families to adapt to adverse circumstances (e.g. taking children away from school, reducing the consumption of essential food and medicines, selling productive assets, increasing the use of barter) can translate a temporary setback in social conditions into a permanent one. See OECD (2001d) and OECD (2002b).

54


PART II

Giving Children and Parents a Brighter Future Investing in children has high pay-offs, both for the children themselves and for society in general. Children who grow up in disadvantaged households are more likely to have difficulty in school, to struggle to find jobs, to be unemployed, sick and disabled when adults. They are also more likely to be parents of poor children themselves, threatening an ongoing cycle of deprivation. Yet, for all the talk of investing in the next generation, social expenditure on children and families – leaving aside education – is dwarfed by other types of interventions, in particular towards the elderly. Low levels of social spending have remained possible because the family is still the best means of providing for children’s needs and care – but family structures are changing, and policy needs to change too. Furthermore, existing policies are already failing to prevent higher poverty among children in some countries, lower fertility rates in most of them, and a persistent gap in the division of paid and unpaid work between men and women. Active social policies towards families with children pay greater attention to ensuring adequate investment in the development of children, sharing part of the associated costs more broadly, and allowing parents to better reconcile work and family responsibilities. As current trends unfold, the need for active social policies towards families with children will increase, but so will the pay-off for adopting such policies. These policies hold the promise of helping adults overcome obstacles to have more children while giving mothers the support they need to remain in the labour market if they wish, thereby offsetting some of the anticipated effects of population ageing. They also help give children the best possible start in life, setting in motion a virtuous cycle of self-reliance and lifelong learning that prepares them for the greater demands of the future and averts the threat of a cycle of poverty and disadvantage. Achieving these results requires policies aimed at: ●

Investing in children.

●

Boosting maternal employment.

●

Reconciling work and family responsibilities.

●

Creating a framework that supports parents’ fertility decisions.



PART II

Chapter 4

Families with Children: Achievements and Challenges


57

II.4.

FAMILIES WITH CHILDREN: ACHIEVEMENTS AND CHALLENGES

1. Introduction This chapter describes the achievements and limits of current social policies targeted to families with children, and the importance of adapting these policies to a new configuration of risks. First, it describes some of the main social outcomes that characterise the experience of children and their families, and the range of social programmes that OECD countries have deployed in the past to support them. Second, it identifies the areas where performance has been less satisfactory, in particular the growing concerns about child poverty and other unfavourable outcomes affecting children, the sharp reduction in fertility rates to levels below those desired by parents, and the failure to close gender gaps in the distribution of paid and unpaid work.

2. Policies and achievements The attention paid to families with children in recent social policy discussions reflects a greater awareness that what happens during childhood is a critical determinant of outcomes in adult life. Overall, children in all OECD countries today enjoy better health and education than their peers did just a few decades ago. Infant mortality has declined in all OECD countries over the past four decades, from close to 30 deaths per 1 000 live births in the 1960s to less than 5 per 1 000 live births in 2002 (Figure 4.1, left-hand panel), although once this level has been reached it has proven difficult to obtain further reductions. Despite this progress, there are still large differences in infant mortality rates across OECD countries (in Turkey and Mexico levels are around five times higher than those prevailing in the three best-performing countries, and twice as high in Hungary, Poland, Korea, the United States

Figure 4.1. Mortality and immunisation rates among children, OECD average Infant mortality

Immunisation rates DTP

Per 1 000 live births 40 35

95

30 25

90

20

85

15 10

80

5 0

00 20 02

98

20

96

19

94

19

92

19

90

19

88

19

86

19

84

19

19

0

82

19

19 8

20 2000 02

19 90

19 80

75 19 70

19 60

Measles

% 100

Note: Infant mortality is the number of deaths of children under 1 year of age per 1 000 live births. Immunisation rates for measles and for diphtheria, tetanus and pertussis (DTP) are expressed as a percentage of children of a specified age having completed (or in some cases started) immunisation; in most countries rates refer to children reaching their first birthday, but in some countries a different age is used (2 years in Belgium, Canada, the Czech Republic, Finland and France; 3 years in Luxembourg, Switzerland and the United States; less than 5 years in New Zealand; less than 14 years in Mexico). Source: OECD Health Data 2004, OECD, Paris.


58

II.4.


and New Zealand). Other indicators of progress in protecting children’s health, such as immunisation rates for major infectious diseases (measles, diphtheria, tetanus and pertussis), are also high and rising (Figure 4.1, right-hand panel). Improvements have been equally impressive in education. Compulsory education has expanded in all OECD countries, and is almost universal in most. More children continue and complete higher education: on average, youths aged 15 in 2002 (who have been in school for 9 to 10 years) could expect to remain in education for 6 ½ additional years (OECD, 2004c). Maternal employment has also risen everywhere. In general, across OECD countries, higher numbers of employed mothers mean a reduction in the probability that children will experience poverty (Figure 4.2). There is, however, much variation in child poverty rates for a similar level of maternal employment (ranging from less than 4% in Nordic countries to over 20% in the United States), which suggests that employment is not the only factor at work: also important is how employment is distributed among households, the earnings jobs pay and public support provided to families with children.

Figure 4.2. Relative poverty among children and employment rates among mothers, 2000 Child poverty rates (%) 0.24 USA

0.20 0.16 0.12

ITA

NZL GBR PRT CAN DEU AUT

IRL GRC AUS

NLD

0.08

CZE

0.04 0 0.40

FRA

CHE NOR SWE FIN DNK

1.00 0.90 0.70 0.80 Employment rates of mothers (%) Note: Employment rates among women aged 25 to 54 with one or more children aged 15 or less (16 in the case of New Zealand and Sweden). Child poverty rates refer to the proportion of children (persons aged less than 18) in households with equivalised disposable income less than 50% of the median. 0.50

0.60

Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.

While the timing of the increase in female labour force participation has varied across countries, all countries have been affected. Nevertheless, differences in the levels of female participation remain large: among women aged 25 to 54 the level ranges from 60% or less in Turkey, Korea, Mexico and Southern European countries (with the exception of Portugal) to more than 80% in Nordic and some Eastern European countries. The differences in participation rates for this age group of women explain much of the variation in aggregate participation rates across OECD countries (Burniaux et al., 2003). Higher rates of employment among women have been driven by changes in both the structure of the economy (the shift from agriculture and manufacturing to services) and the characteristics of women (including rising education, changing preferences for work outside the home and higher female wages), as well as falling male earnings and lower certainty of continuous employment for men (Pettit and Hook, 2002). Entering the labour force has also represented the most effective way for mothers to ensure themselves and


59

II.4.


their children against the vicissitudes of relationships and of work. As divorce rates and job insecurity rise, the male breadwinner model has become less and less reliable as a guarantee for children’s future. Patterns of maternal employment vary widely across countries (Table 4.1).1 On average, nearly 71% of mothers with one child and 62% of mothers with two or more children are employed. When the child is under the age of 6, maternal employment rates are lower, but still exceed 56% in 13 OECD countries, and they have continued to rise over the 1990s. There are also cross-country differences with respect to the type of work-combination of parents that is most “typical”. In the United Kingdom, the Netherlands and Germany, full-time employment of fathers combined with part-time employment of mothers is most common, while in the United States and Finland both parents usually work full-time. Patterns of withdrawal from the labour market of mothers following birth also differ

Table 4.1. Employment and part-time work among women in 2003, by presence and age of children Employment rates (%)

Share of part-time employment (%)

Two or more Youngest child children under 6

Youngest child Youngest child under 6 6 and over

No children

One child

Australia

68.4

55.3

43.2

45.0

68.6

66.7

55.2

41.6

Austria

79.5

79.9

67.6

70.6

76.7

39.9

38.7

28.4 33.9

All women

All women

Belgium

70.4

71.2

65.3

63.5

67.7

39.0

41.6

Canada

76.5

74.9

68.2

..

75.7

..

..

..

Czech Republic

83.6

69.9

55.6

35.7

73.5

8.4

5.2

4.1

Denmark

78.5

88.1

77.2

74.3

78.9

5.1

8.3

15.5

Finland

79.2

78.5

73.5

..

78.9

..

..

9.4

France

76.1

77.0

61.6

62.2

71.6

26.3

26.6

21.6

Germany

77.6

70.3

56.5

56.2

72.0

53.3

55.1

37.6

Greece

56.1

54.4

47.4

50.2

56.6

11.5

8.9

9.9

Hungary

74.6

67.0

48.3

32.9

67.4

8.6

3.8

4.3

Iceland

89.1

89.3

80.8

..

85.7

..

..

..

Ireland

65.8

51.0

40.8

..

65.1

..

..

31.8

Italy

24.2

57.2

54.5

45.7

51.1

54.9

30.2

30.8

Japan

..

..

..

..

64.4

..

..

..

Korea

..

..

..

..

56.8

..

..

..

71.2

67.6

51.2

57.7

64.5

43.6

47.9

32.1

..

..

..

..

45.5

..

..

..

Netherlands

76.8

73.8

68.6

69.2

74.0

80.6

75.6

57.6

New Zealand

80.7

66.9

58.9

..

72.0

..

..

..

Norway

82.9

83.3

78.0

..

79.7

..

..

29.7

Luxembourg Mexico

Poland

..

..

..

..

62.1

..

..

..

Portugal

72.3

78.6

74.9

77.3

74.2

8.3

10.1

10.6

Slovak Republic

77.8

73.1

58.1

38.9

71.5

2.6

3.3

2.5

Spain

56.1

54.4

47.4

50.2

56.5

18.1

20.0

15.5

Sweden

81.9

80.6

81.8

..

81.7

..

..

16.4

Switzerland

84.3

75.5

65.5

..

77.2

..

..

47.7

..

..

..

..

27.4

..

..

..

United Kingdom

81.6

72.4

61.8

57.5

74.1

61.5

51.8

36.7

United States

78.6

75.6

64.7

61.2

72.0

..

..

18.2

OECD

75.1

71.3

61.7

56.1

63.7

31.5

30.2

24.1

Turkey

Note: Women aged 25 to 54. Data for all women in France, Iceland and Luxembourg refer to 2002; data by the presence of children refer to 2001 for Australia, Canada, Iceland, Ireland, New Zealand, Sweden and Switzerland. Apart from overall employment of women and part-time employment for all women, which refer to 2003, figures for Denmark refer to 1998, for Finland to 1997, and for the United States to 1999. Source: OECD (2002 and 2004), Employment Outlook, OECD, Paris; and European Labour Force Surveys.

60


II.4.


significantly: in some OECD countries (e.g. Portugal), employed mothers typically take a break of a few months from paid work while their children are infants and then return to work; in others (e.g. Sweden and Norway) mothers often withdraw from employment during the first one, two, or even three years of their children’s lives – through paid leave schemes, career breaks or both – and then return to paid work; in yet other countries (e.g. Japan) mothers may never come back after giving birth, or – if they do – are likely to be confined to jobs with reduced wages and career prospects. Part-time work is more common for women, and for mothers in particular, than for men.2 To some extent, these positive outcomes for children and their families have reflected the influence of social policy programmes in most OECD countries. Gross public spending on families tends, however, to be low compared to other types of social spending: on average, for 21 OECD countries, it accounts for less than 2% of GDP – and for around 3% or more in Australia, Austria, Denmark, Finland, France, Luxembourg, Norway and Sweden – a level only marginally higher than in the 1980s (Figure 4.3). Cash transfers represent the dominant component of gross public spending delivered to families (around 70%), although in-kind services represent more than half in Denmark, Iceland, Japan, Korea, Mexico and the United States (but very small absolute amounts in the latter four countries). Private spending on family benefits, both mandatory and privately undertaken, is also negligible in all OECD countries. Finally, spending by educational institutions (both private and public) on early childhood education and care represents less than 0.5% of GDP on average, but is significantly higher in Norway, Hungary, Denmark and France.

Figure 4.3. Gross public spending on families as a percentage of GDP OECD % 6

France

Finland

United States

Sweden

Italy

OECD % 6

5

5

4

4

3

3

2

2

1

1

0 1980

1985

1990

1995

2001

0 1980

Denmark

United Kingdom

1985

1990

Canada

Germany

1995

Japan

2001


3. The challenges today While families with children have shared the gains of greater prosperity in all OECD countries, performance has been disappointing in some areas, and new risk factors have emerged. The risk of poverty has shifted towards children in recent years, a trend that if left unchecked would threaten problems in the life courses of millions. Fertility levels have fallen not only below the replacement level but also below levels desired by many adults. And although millions of women have entered the world of work, gender gaps in earnings and in paid and unpaid work persist in all countries. Major changes are taking place in family structures and the world of work, and social policy needs to respond.


61

II.4.


3.1. Poverty and other risks for child development remain high in some countries Material deprivation and its impact on child development still pose significant risks. In most OECD countries, the risk of poverty (“relative poverty”, as measured with respect to 50% of median equivalised income) has shifted over the past 20 years towards children (aged less than 18) and their families. On average, more than 12% of all children in OECD countries fell below the 50% poverty threshold in 2000, with rates in excess of 20% in several countries. While poverty rates for children tend to be higher in countries with high overall poverty levels, there is substantial diversity across countries in the extent of child poverty for a given level of poverty of the entire population, suggesting that specific factors increase the risks of disadvantage for children in some countries (Figure 4.4).

Figure 4.4. Relative poverty rates among children Child poverty rate

% 30

Overall poverty rate

25 20 15

Mid-1990s

2000

10 5

Au

st ra l Au ia st r Be ia lg iu Cz C m ec an a h Re da pu De blic nm ar Fin k lan Fr d a Ge nce rm an Gr y ee Hu ce ng ar Ire y lan d Ita ly Ja pa M n Ne ex th ico Ne erla w nd Ze s ala n No d rw a Po y lan Po d rtu ga l Sp ain Sw Sw ed itz en er lan Un d T ite u d rke Ki y Un ngd ite om d St a OE tes CD -2 3

0

Note: Poverty is measured with respect to a threshold set at 50% of median equivalised household disposable income. Children are persons aged less than 18 years. The years to which the data refer are those listed in Figure 1.2. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.

Beyond income-poverty, a number of pathologies affect only small minorities of children and teenagers but often have extreme consequences. Their form differs among countries. Child abuse and neglect are a more visible concern in an increasing number of OECD countries. While difficult to monitor, deaths of children due to maltreatment (and other “unspecified” causes) remain higher than most citizens are willing to accept in several OECD countries (Figure 4.5, 1st panel). Every year a non-negligible proportion of youth comes into contact with the criminal justice system and increasing proportions are victims of drug abuse. While each of these pathologies affects only a small minority of children, many of them occur simultaneously, compounding their effect on the future lifechances of affected youths. Other unfavourable outcomes affect larger numbers of youths, but with less dramatic consequences. There are relatively high proportions of teenage mothers in the United States and, to a lesser extent, the United Kingdom and New Zealand, as compared to both Nordic and Southern European countries and Japan (Figure 4.5, 2nd panel). In the United States, this proportion was stable between 1948 and the early 1990s, followed by a steep decline since 1992. While teenage pregnancy and birth are not always a policy problem (early marriage still characterises more traditional OECD societies), teenage pregnancy is

62


II.4.


Figure 4.5. Different forms of child disadvantage in OECD countries, late 1990s Levels (left axis) Per 100 000 4.0

Difference from three best-performing countries (right axis)

Deaths from maltreatment

% difference 24 18

2.0

12

1.0

8

0

0

Po

rtu g Un Me al ite xic d St o at e Ne Fra s w nc Ze e a Cz H land ec un h ga Re ry pu Be blic lg iu Ca m na Au da st Sl ria ov ak Jap Re an pu b Un ite Po lic d Ki land n Sw gd itz om er lan d Ko Au r ea st Ge ralia rm De any nm ar k F Ne inl th and er lan Sw ds ed e Ire n lan No d rw Gr ay ee ce Ita ly Sp ain

3.0

Per 1 000 60 50

Teen pregnancy

% difference 12 10 8

30

6

20

4

10

2

0

0

U Un nite ite d S d ta K t Ne ing es Sl w dom ov Ze ak al Re and pu Hu blic ng a Ic ry el Po and rtu Ca gal na Ire da la Po nd Cz A land ec us h tra Re lia pu b Au lic s Ge tria rm a No ny rw Gr ay ee c B Lu elg e xe iu m m bo u Fr rg an Fin ce De land nm ar Sp k ain I Sw taly Ne e th de e n Sw rlan itz ds er lan Ja d pa Ko n re a

40

Note: Levels for each indicator are shown on the left-hand axis; point differences from the three best-performing OECD countries are shown on the right-hand axis. Deaths from maltreatment refer to the annual number of reported deaths of children under age 15 from maltreatment and unspecified causes, averaged over five years, per 100 000 children of the same age group (UNICEF). Teen pregnancy gives the number of births to women under 20, per 1 000 women aged 15-19 in 1998 (UNICEF). Source: Different issues of Society at a Glance – OECD Social Indicators, OECD, Paris.

often associated with an interruption of schooling, greater difficulties in integrating into the labour market, a higher risk of single parenthood, and dependence on social assistance for daily living. Youth also face risk factors related to education and health. Assessments of reading literacy of students aged 15, conducted through tests administered under the OECD Programme for International Student Assessment (PISA), measure the extent to which young people are able to construct, expand and reflect on the meaning of what they have read in a wide range of texts, both in and outside school. In the OECD area as a whole, 14% of students of this age seem capable of completing only the least complex (Level 1) reading tasks (such as locating a single piece of information, identifying the main theme of a text, or making a simple connection with everyday knowledge) and 6% perform below this level. While in Finland and Korea, fewer than 7% of students perform at or below this level, this proportion exceeds 20% in (in descending order) Mexico, Turkey, Greece, the Slovak Republic, Italy, Luxembourg, Germany, Portugal, Spain, Austria and Hungary (OECD, 2004e). Students performing below Level 1 may be at risk of difficulties not only in their initial


63

II.4.


transition from education to work but also in terms of failing to benefit from further education and learning opportunities in adult life. Beyond the age of 15, many youth fail to complete upper secondary education, and will be at a disadvantage when they enter the labour market.3 While it is difficult to identify the root causes of this very diverse set of problems, some of these risks are related to family experiences during childhood, which are now more diverse than in the past. In several OECD countries, more than half of all children are born outside marriage, either to a single parent or to an unmarried couple; more children grow up without siblings (in single-child families); and more experience family conflicts, possibly leading to dissolution, often at younger ages. One measure of family dissolution – the cumulative share of time that children under 14 have spent in different types of families (Figure 4.6) – suggests that in Sweden and the United States children typically spend one-third or more of their life with a lone mother, in step-families, or in consensual unions, while the comparable figure is around one-fifth in France and the western länder of Germany.

Figure 4.6. Cumulative years spent by children in different family types % 100

With both parents in marriage

With both parents in consensual unions

With lone mother

In step family

80 60 40 20

d lan

bl pu Re

Cz

ec

h

Po

ic

y ar ng Hu

ain Sp

ly Ita

y an rm Ge

ria st

St d ite Un

Au

es at

ce an Fr

d lan Fin

ay rw No

Sw

ed

en

0

Note: Percentage of cumulative time spent by children from age 0 to 15 in different family types. Data for Germany relate to the western länder. Data are based on special tabulations from the “Family and Fertility Surveys” project. Source: Andersson, G. (2001), “Children’s Experience of Family Disruption and Family Formation: Evidence from 16 FFS Countries”, Max Planck Institute for Demographic Research, Rostock.

Research suggests that children are especially vulnerable to the consequences of family distress, although it remains unclear whether these consequences reflect the experience of family breakdown per se or the difficulties that children may confront in retaining good relationships with their biological parents. Evidence for the United Sates suggests that family dissolution tends to lower the educational attainment of children and to lower the age at which they leave the family home, with the size of these effects depending on the characteristics of the families where children live following dissolution – effects being higher for single-parent households than for recomposed families. While results vary, separations (both their occurrence and form) also seem to increase the risks of poverty, learning and emotional difficulties, and reduce the probabilities that children will maintain contact and receive support from the non-custodial parent (Marcil-Gratton et al., 2000). In many cases, however, these negative effects may simply reflect the greater likelihood of conflicts between parents within couples that will end up separating, and the worse relationships of children with their parents in these families.4 More generally, not

64


II.4.


all – or even most – children experiencing family breakdown develop behavioural and emotional problems, and many of these problems may have their roots in factors other than family breakdown per se (Cherlin, 1999).5 Whatever the causal factors underlying these unfavourable child outcomes, their significance for social policy lies mainly in their capacity to affect conditions in adult life. Developmental psychologists think of child development as a sequence of distinct transitions, occurring at critical times in a child’s life; each of these transitions shapes emotional, behavioural and learning attitudes, and prepares the child for subsequent transitions (Chase-Lansdale, 1998). Because of the critical role played by the timing of events in children’s lives, some of the scars associated with unfavourable child outcomes may not heal easily, and their effects may be transmitted over time. In the case of poverty, research suggests that the proportion of children born to low-income parents who became low-income adults is around half in the United States, four in ten in the United Kingdom, one in three in Canada, and less in the Nordic countries (Corak, 2004). In the case of teenage births, it has been shown that the negative consequences are not limited to mothers but extend to the health and future socio-economic status of their children: the children of teenage mothers have less chance of educational, economic or family success, even when differences in the socio-economic characteristics of the mother are taken into account (Bélanger, 2002). When these effects persist over time and across generations, opportunities in later life are compromised.

3.2. Observed fertility rates are well below desired levels The childbearing decisions that have led to today’s low fertility rates are the outcome of a complex set of factors influencing the decisions of couples. The overall result may however pose difficulties for society at large, in terms of financing social protection and consumption for older generations, lower output growth, financial market disruptions (i.e. a reduction in asset prices as a larger cohort of elderly people sells the assets it has accumulated during working life, and a smaller cohort of younger people buys them), lower rates of return on infrastructure investment, and reduced availability of support from other family members in case of need. While changes in individuals’ preferences and life-styles are contributing to lower fertility, the effects of these changes are being exacerbated by the constraints that individuals and couples face in everyday life, by the emergence of new risk factors confronting them (such as labour market insecurity, difficulties in finding suitable housing, unaffordable childcare) and by the failure of social policies to provide adequate support. Evidence about the potential role of these constraints on women’s childbearing decisions can be derived from answers to questions about the “desired” or “ideal” numbers of children provided from opinion surveys. While interpreting survey evidence from these questions is not without problems,6 the evidence summarised in Figure 4.7 highlights a number of consistent patterns.7 ●

Women generally have fewer children than they actually want. Exceptions to this pattern – in Turkey (in all years) and Mexico and Korea (in the 1980s) – are limited to countries that are (or were) characterised by lower per capita income and lower diffusion of contraception.8

●

The gap between desired and observed fertility is higher in countries where fertility rates are lowest. Some of the OECD countries where fertility rates are lowest (Japan, Italy and


65

II.4.


Spain) in 2000 recorded the largest gaps between desired and actual fertility, while countries with higher fertility rates (United States and France) show smaller gaps. ●

The gaps between desired and actual fertility have increased over the past ten to twenty years. On average, across the countries for which observations are available in each of the three years shown in Figure 4.7, the gap between desired and actual fertility grew from 1980 to 1990 and from 1990 to 2000.

Figure 4.7. Desired and observed fertility in different years Desired fertility 1981

Desired fertility 1990


Total fertility rate

5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

Au

st ra l Au ia st Be ria lg iu Ca m Sw na Cz itz da ec erl h Ge Re and rm pu Ge an blic rm y ( an Eas y ( t) W De est) nm ar k Sp ain Fin lan Un ite Fra d d Ki nce ng d Hu om ng ar Ire y lan Ic d ela nd Ita ly Ja pa n Ko re M a Ne ex th ico er lan No ds rw a Po y l Sl P and ov or ak tug Re al pu b Sw lic ed e Un Tu n ite rke d y St at es

0

Note: Observed fertility is measured by the total fertility rate of each country in that year. The three bars shown for each country refer to data for 1981, 1990 and 2000, with the exceptions of Austria, the Czech Republic and the eastern länder of Germany (where data refer to 1990 and 2000), and of Switzerland, Poland and Turkey (where the data refer to 1990, 1995 and 2000). Source: : Data from the World Values Survey (1981, 1990, 2000) and Eurobarometer (2001) as presented in D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

●

66

The gaps between observed fertility and desired fertility have increased among different cohorts of women. Aggregate changes in the gap between desired and observed fertility are partly affected by changes in the demographic composition of women (i.e. a growing share of older women who are close to the end of their reproductive life). Information about changes in desired and observed fertility among different cohorts of women can be obtained by looking at women of the same age (29 to 39, and 39 to 49) at 10-year intervals. Figure 4.8 highlights a number of patterns. Among younger women (aged 29 to 39) the gap between desired and observed fertility increased strongly over time, as postponement of childbearing led to sharp falls in observed fertility. Among older women (aged 39 to 49) the gap between desired and observed fertility also rose, but by a smaller amount: most women in this group, who in several OECD countries in the 1980s had more children than they desired, declared in 2000 that they desired more children than they actually had. For women who are close to the end of their reproductive cycle, postponement of childbearing is a less plausible explanation of this widening gap between desired and actual fertility: despite the effects of medical advances in extending childbearing until higher ages, women in this age group are unlikely to realise fully their fertility intentions.


II.4.


Figure 4.8. Desired and observed fertility among women of different ages Desired fertility 1981 5



Observed fertility

Women aged 29-39 in 1981, 1990 and 2000

4 3 2 1

5

y

St at es

ke

en

Un

ite d

Tu r

Sw

ed

ay

s

No

rw

nd

ico

rla

Ne

th e

a

M ex

n

re Ko

pa Ja

Ita ly

d

d

lan Ic e

m

lan Ire

do

ce an

Ki ng

Un

ite d

Fr

k

ain Sp

y De

nm ar

da

rm an

na

Ge

Ca

Be

lg iu m

0

Women aged 39-49 in 1981, 1990 and 2000

4 3 2 1

es

y

at

ke Un

ite

d

St

Tu r

en ed

ay

Sw

rw

ds

No

lan

ico

er th Ne

M

ex

a re Ko

n pa Ja

ly Ita

nd

d lan

ela Ic

Ki Un

ite

d

Ire

m ng

do

ce an Fr

k

ain Sp

ar

y an

nm

De

Ge

rm

da na Ca

Be

lg

iu

m

0

Note: Observed fertility is measured by the number of children that women of different ages declared in the survey. Data for Germany refer to western länder only. Source: Data from the World Values Survey (1981, 1990 and 2000), as presented in D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

The views of governments about fertility levels largely match those of individuals: in 1976 most OECD governments regarded fertility levels as “satisfactory”, with only a few characterising them as either “too high” (Korea, Mexico and Turkey) or “too low” (Finland, France, Greece and Luxemburg); by 2003, a large majority of OECD governments regarded prevailing levels of fertility as “too low”, with only Mexico and Turkey characterising them as “too high” and a larger number (Australia, Belgium, Canada, Denmark, Ireland, Iceland, Netherlands, New Zealand, Sweden and the United States) as “satisfactory” (UN Population Division, 2003).

3.3. Countries have not closed the gender gap in work Substantial gender gaps continue to exist with regard to both paid and domestic work. The entrance of large numbers of women – and mothers in particular – into the labour market has eroded the traditional division of responsibility for care and paid work within households. However, changes in responsibilities for paid work within the family have not been associated with similar changes in responsibilities for care. Despite the higher labour force participation of women today, gender gaps in employment remain widespread, especially for women with children. Mothers are less likely than fathers to be employed and, when employed, they work fewer hours; they are more likely to take paid leave and/or career breaks to care for children and other family members; they are also less likely to work in upper-level occupations or jobs. Labour EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

67

II.4.


market losses associated with single parenting are experienced overwhelmingly by women. Women’s care-giving responsibilities also exert a major effect on their earnings, on an hourly, monthly, annual and lifetime basis: career breaks, periods of part-time work, parenting-related job choices, and firms’ discrimination based on parental status combine to lower the earnings of mothers, compared both to women without children and to fathers. As a result, following periods of childrearing, many mothers never catch up in terms of labour force participation, hours, job quality, earnings, and investments in human capital (education, training, work experience). In contrast, for men, having children generally leads either to no change or even to an increase in labour force participation, working hours, and pay. Gender gaps are most evident in the case of employment rates. In 2001, on average across 24 OECD countries, employment rates for all adult women were around 19 points lower than for men and 23 or 32 points lower, respectively, in the case of mothers of one or of two or more children. While these differences narrow at higher levels of educational attainment, there is still a 10 point difference between men and women with university and tertiary education, and an even larger gap in Japan and Korea (Figure 4.9). Gender gaps in employment narrowed significantly in the 1980s and the 1990s (due to a combination of entry or re-entry into the labour market of women of all ages and of those around childrearing age) in most countries, but not in Greece, Italy and Spain (where the gap is largest in 2000).9

Figure 4.9. Gender gaps in employment of persons aged 25 to 54 by educational attainment, 2000 % 50

Total

Less than upper secondary education

University/tertiary education

40 30 20 10

Au s

tra Au lia s Be tria lg Cz C ium ec an h Re ada pu De blic nm a Fin rk lan Fr d a Ge nce rm an Gr y ee Hu ce ng a Ic ry ela n Ire d lan d Ita ly Ja pa Lu Ko n xe re m a Ne bo th urg Ne erla w nd Ze s ala No nd rw a Po y Sl P land ov or ak tu Re gal pu bl ic Sp Sw ain S e Un wit den ite zer d lan K Un ing d ite do d m St OE ates CD -2 8

0

Note: Percentage-point difference between the employment rates of men and women. Source: OECD (2002), Employment Outlook, OECD, Paris.

Evidence on movements into and out of employment, and between full-time and parttime work, confirms that women spend less and more discontinuous time in employment than men, especially if they have children or less education. Career breaks or working-time reductions are particularly frequent immediately after the birth of a child, especially in Germany and the United Kingdom.10 The distribution of employment by occupation or sector also shows gender segmentation. Women are over-represented in clerical occupations, sales jobs, life-science/ health and teaching professions, whereas they remain under-represented in managerial and top administrative occupations, as well as in manual and production jobs. Women with low educational attainment or with children tend to be more occupationally

68


II.4.


segregated than highly educated and childless women. Furthermore, the degree of occupational segmentation tends to increase as women’s presence in the labour market rises, although there are some signs of falling occupational segmentation among younger cohorts. Finally, there are gender gaps with regard to earnings. While gender wage gaps have narrowed over the past two or three decades in virtually all OECD countries, full-time employed women still earn on average 16% less than men (Figure 4.10). As women also work fewer hours, their take-home pay is considerably less than that of men. Gender differences in observable characteristics that influence earnings – such as education, potential experience and job tenure – account for little of these gaps.11 In addition, except in a few countries, gender differences in hourly wages tend to apply to all women, whether they are mothers or not.

Figure 4.10. Gender gaps in earnings 1980

% 60

1990

2000

50 40 30 20 10

d

St

at

do Un

Ki d ite

Un

ite

ng

er itz

es

m

d lan

en ed

Sw

lan

nd

d

Sw

Po

ala

ds Ne

w

Ze

a

lan

re

er

Ko

th Ne

n

ly

pa Ja

Ita

y ar ng

an

y Hu

rm

an

d

ce

Ge

Fr

lan

ic bl pu

Re h ec

Cz

Fin

m iu lg

Be

Au

st

ra

lia

0

Note: Gender gaps in earnings are measured as the difference between male and female median full-time earnings, as a percentage of male median full-time earnings. Data refer to 1980 (1985 for Belgium, Germany, Korea and the Netherlands); 1990 (1995 for Korea, Poland and Switzerland; and 1996 for the Czech Republic); and 2000 or most recent available year (1999 for the Czech Republic, Finland, Japan, Korea, Netherlands and Poland; 1998 for France, Germany, Sweden and Switzerland). Source: OECD (2001), Labour Market Statistics, CD-ROM, OECD, Paris.

Although increasing numbers of women have entered paid work, this has not resulted in a significant change in the distribution of responsibilities for caring within the family. While mothers in paid work (especially those working full-time) devote less time to childcare and other unpaid family chores than those who are not, this time largely exceeds that devoted by fathers to the same tasks (Figure 4.11). Even mothers working full-time dedicate 3.5 times as many hours to children and other family tasks as do men in Italy, and between 2.7 and 1.3 times more in other countries. This higher caring-burden is only partly offset by the smaller number of hours spent in paid work by women working full-time. More recent trends for European countries suggest that gender patterns in the division of unpaid work have not changed much in more recent years (Istituto di Richerche sulla Popolazione, 2001). This “double burden” on working mothers puts pressures on family life, affects the amount of care children receive, and hampers the career prospects of working mothers.


69

II.4.


Figure 4.11. Time spent daily by women and men on childcare, paid and unpaid work Paid work

Child care

Other unpaid work

Hours 14 Women working full-time

12

Women working part-time Women not working

Men

10 8 6 4 2 0

Austria

Australia

Canada

Denmark

Finland

Germany

Italy

Netherlands Sweden

United Kingdom

United States

Note: Bars refer respectively to men, women working full-time, women working part-time, and women not working. Data refer to a couple family with a child aged less than 6. Time devoted to childcare includes that spent feeding children, dressing, bathing and changing them, and providing medication. Other unpaid work is defined broadly. Paid work includes working in a family enterprise and is averaged over the year, including weekends and paid leave. The data refer to 1985 for the Netherlands; 1987 for Denmark and Finland; 1989 for Italy; 1991 for Sweden; 1992 for Australia and Germany; 1995 for the United Kingdom and the United States; 1997 for Austria; 1998 for Canada. Data are drawn from time-budget surveys harmonised by researchers co-ordinated by Essex University, United Kingdom. Source: OECD (2001), Employment Outlook, Paris.

4. Conclusions Families with children have shared the gains of higher material prosperity in all OECD countries. As a result, children enjoy today much better health and education than their peers just a few generations ago, while maternal employment has increased everywhere, bringing with it greater financial independence for women and security for families. But, despite these achievements, more than 12% of all children live in relative-income poverty in 2000, and significant numbers are affected by several forms of distress such as drug abuse, involvement in criminal activities and teenage pregnancy. Also, a large gap has opened up between women’s desires for children and the number they will actually realise over their life course, while gender gaps in the distribution of paid and unpaid work persist. Ways in which active social policy can contribute to improve these outcomes are described in the following chapter.

Notes 1. Only in Spain are fewer than half of mothers with one child employed, and only in four countries (Spain, Australia, Ireland and Italy) do fewer than half of mothers with two or more children work for pay. 2. On average, around one-fourth of employed women and one-third of employed mothers work part-time; part-time work is particularly common in the Netherlands, the United Kingdom, Australia and Germany but is rare in Eastern Europe, Denmark and Portugal. 3. In the health area, there is a disturbing trend towards more overweight and obese children: more than 15% of children (of different ages) are overweight in many OECD countries, and many of them are obese. The most immediate consequences of being overweight include loss of self-esteem and depression among the children themselves; longer-term risks include higher rates of heart disease and Type II diabetes and becoming overweight or obese in adult life. 4. Evidence on the independent effect of separations on children outcomes is often contradictory. For example, Piketty (2003) reports that, within a large sample of French individuals, the school

70


II.4.


performance of children a couple of years before their parents separate is as bad as for children already living with a single parent, i.e. separation has no residual effect on school performance when controlling for pre-separation family characteristics (Piketty, 2003). Conversely, Cherlin (1999) reports that longitudinal studies in the United Kingdom show that, while children of divorced parents already experienced behavioural problems before the break-up, the emotional problems become aggravated after the event. 5. Among these factors, of special importance are genetic characteristics, which may affect the probability of behavioural problems leading to divorce. Cherlin (1999), which summarises evidence from a range of studies in the Unites States, concludes that heredity may account for about half of the variance in human behaviour (e.g. occurrence of major depressions among twins), with life and family experiences accounting for the other half. 6. Among these problems are the difficulty in distinguishing between personal desires for their own conditions and societal norms about what is considered to be the “ideal” family size; the dependence of responses on conditions that may change over the life-course of the individual; the adaptation of fertility intentions to actual experience; and the fact that survey questions may fail to specify the determinants of fertility intentions. 7. Survey evidence about desired fertility, as available for most OECD countries, is based on data from the various waves of the World Values Survey (1981, 1990, 1995/1997 and 2000) as well as from Eurofound (2004) for European countries (based on a Eurobarometer survey undertaken in 2001). Measures of “desired fertility” from both surveys refer to the “ideal number of children” that the respondent would like to have. The question in the World Values Survey is: “What do you think is the ideal size of the family – how many children, if any?”; the question in Eurobarometer is: “For you personally, what would be the ideal number of children you would like to have or would have liked to have had?” 8. Findings are similar when the observed number of children that survey respondents declare is used in place of the total fertility rates, whose movements are affected by changes in the timing of fertility. Data on completed fertility rates, as available for women who have reached the end of their reproductive cycle, have not been used as they fail to reflect the behaviour of younger generations. 9. Comparisons of headcount measures of employment by gender, moreover, overstate the degree of women’s presence in employment in all countries, as they do not take account of the higher incidence of part-time employment for women and of the fewer hours worked by women in fulltime work (because of less overtime, and higher absences from work). 10. In some OECD countries, women also have much higher unemployment rates than men. This is especially the case in Greece, Spain, Luxembourg, Italy, the Netherlands and the Czech Republic (where unemployment rates for women are 50% higher than for men). In other countries, women are less likely to be unemployed than men – notably in Canada, Australia, New Zealand, Turkey, Norway, Hungary, Ireland and the United Kingdom. 11. Cross-country differences in the wage structure and in women’s employment rates provide important proximate explanations of much of the variation in gender wage gaps: in a few countries, e.g. the United Kingdom, the gender wage gap would be considerably lower if the wage structure were as compressed as the OECD average; and larger wage gaps are found in countries where less educated and less skilled women are more integrated into the labour market. However, it is difficult to identify the most important factors underlying these associations.


71


PART II

Chapter 5

More Effective Policies for Families with Children


73

II.5.

MORE EFFECTIVE POLICIES FOR FAMILIES WITH CHILDREN

1. Introduction The greater attention being paid to children and their families has been accompanied by adjustments in the policy stance of most OECD countries, but not yet in major re-orientations. Public transfers to families with children remain important in all OECD countries for reducing child poverty. They are, however, expensive, and may reduce financial incentives to take up paid employment. Because of these considerations, public transfers are increasingly complemented by other measures. Active policies aimed at children and their families combine several ingredients. First, is the attention paid to investing in children, so as to improve their life-chances in adulthood and prevent a cycle of deprivation in future generations. Second, is greater emphasis on increasing the employment rates of mothers with children, which is one of the best ways of reducing childhood poverty. Third, is the emphasis given to reconciling work and family responsibility, so as to provide an environment that is less stressful for parents and more supportive to child development. Last are steps to share collectively some of the costs associated with having and rearing children, so as to allow a recovery in fertility rates towards the levels actually desired by couples. As in other areas of social policy, it is crucial to strike a proper balance between these elements – for instance, support for maternal employment needs to ensure that too much time working does not mean too little time spent with children. Also, achieving these goals requires finding ways to mobilise forces outside government, in particular to get firms to provide family-friendly environments and to give parents greater choice so that they can make the specific childcare arrangements best suited to their needs.

2. Public transfers and tax advantages can reduce child poverty Because of children’s vulnerability to the consequences of material deprivation, public transfers and taxes will continue to play a key role in reducing child poverty. Indeed, since the mid-1990s, public policies have increasingly focused on child poverty. In the United Kingdom, a quantified target has been set to halve the number of children living in relative low-income households between 1998-99 and 2010-11, and then to eradicatw child poverty by 2020. Among other EU countries, Greece has also set targets for the reduction of child poverty in its National Action Plan on Social Exclusion, while other countries (e.g. Germany) have set targets in areas that have important impacts on child poverty (such as halving by 2010 the number of youth who have not obtained vocational qualifications). Other OECD countries are less systematic in setting quantitative targets with respect to child poverty. Canada made a commitment to “seek to eliminate child poverty”, but no definition or indicator was settled upon. In New Zealand, the government made a commitment to eliminate child poverty in New Zealand’s Agenda for Children (June 2002). Varied means have been adopted to achieve these targets, and no single policy approach is likely to work on its own. Public transfers and tax advantages towards families with children play a significant role in reducing child poverty at a given point in time: on

74


II.5.


average, across 16 OECD countries, public transfers and taxes lift out of relative poverty around 40% of all households with children whose market income is below the poverty threshold, but this effect is much lower than in the case of households without children (where it is close to 70%). Also, the extent of this reduction ranges widely across countries, from over 70% in Nordic countries to negligible levels in Italy and Japan (Figure 5.1). While general assistance for families with children is often set at relatively low levels (Table 5.1), some OECD countries have recently introduced specific transfer programmes targeted to poor children, while others have increased tax advantages given to families with children. In the United Kingdom, assistance has taken the form of substantial increases in Child Benefit and the introduction of a Child Tax Credit. OECD countries have also introduced tax advantages to families with children that are conditional on one parent working: examples are, in the United Kingdom, the Working Family Tax Credit and more recently the Working Tax Credit. In other countries, such as Belgium, Greece, France, Ireland, and Luxembourg, the amount per child of universal child benefits increases with the number of children, which helps to target benefits towards poorer (and larger) households without relying on explicit means-testing. The Mexican experience with Oportunidades, an innovative programme targeted to families with children, suggests that cash-transfers can be effective in improving children’s material conditions when they are paid to mothers, and when payments are conditional on various activities that improve children’s education and health conditions (Box 5.1).

Figure 5.1. Effect of taxes and transfers in reducing relative poverty in families with children, 2000 Effect of taxes and transfers in reducing poverty

% 35 30

Poverty rate, disposable income

Households with children

Households without children

25 20 15 10 5

at

6 -1

OE

St

CD

es

m d

Un

ite

ng

do

nd Un

ite d

Ki

rla

en

tze

ed

Sw i

ga

l

Sw

rtu

ay

Po

nd

rw No

s Ne w

Ze

ala

nd

n

rla

pa

he

Ja

Ne t

ly Ita

y an

d

k

ce an

rm Ge

Fr

lan Fin

ar nm

na

da

De

Ca

Au

st

ra

lia

0

Note: Child poverty rates are measured with respect to a threshold of 50% of median equivalised disposable income. The years to which the data refer are those listed in Figure 1.2. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.

3. Investing in children has long-term pay-offs Because poverty is more than lack of income, it is also important to help children build up skills and motivation. Investing in children can avert the inter-generational cycle of deprivation that is only too prevalent today in the OECD countries. Historically, the main form of this kind of public investment has been in expanding universal education. In recent years, however, investment strategies towards children have extended beyond formal education. This is especially the case for early interventions, as evidence accumulates that well-designed programmes, particularly ones that closely involve parents, have positive effects that persist through adulthood. EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

75

II.5.


Table 5.1. Assistance for families with children, 2003 Tax support Country

Cash transfers

Total support

Net tax wedge

Type of support for children as % of the gross earnings of an average production worker

Australia

Cash transfers (refundable tax credit)

–

8.3

8.3

20.4

Austria

Cash transfers, tax credit

3.5

18.5

22

29.5

Belgium


2.6

9.4

12

39

Canada

Tax credit

–

6.4

6.4

23.3

4.3

12

16.3

30.6

–

6.6

6.6

30.1 37.8

Czech Republic

Cash transfers, tax deduction

Denmark

Cash transfers

Finland

Cash transfers

France

Cash transfers, tax quotient

Germany

Tax credit

Greece1

Wage supplement

Hungary


Iceland

–

8.3

8.3

2.9

5.6

8.5

40

11.4

–

11.4

33.5

9.3

34.3

8.2

1

12.5

(8.3)

20.7

30.5

Cash transfers

–

9.7

9.7

8.9

Ireland

Cash transfers

–

10.5

10.5

7.4

Italy


4.7

6.1

10.7

35.5

Japan

Tax deduction

3.6

–

3.6

23.2

Korea

Tax difference

0.2

–

0.2

13.6

Luxembourg


1.9

16.4

18.3

9.6

Mexico

–

–

–

–

17.3

Netherlands

Cash transfers

–

5.4

5.4

33.7

New Zealand

Low income tax credit

–

0.2

0.2

20.4

Norway


–

7.6

7.6

27.6

Poland

Cash transfers

–

–

–

41.3

Portugal


3.3

5.8

9.1

23.7

Slovak Republic

Cash transfers, tax allowance

4.5

9.3

13.8

32.3

Spain

Tax deduction

3.2

–

3.2

30.9

Sweden

Cash transfers

–

9.2

9.2

39.5

0.5

8

8.5

5.2

–

–

–

42.1

7.2

6.9

14.1

18.3

10.6

–

10.6

15.5

Switzerland

Cash transfers, tax deduction

Turkey

–

United Kingdom

Cash transfers

United States

Tax deduction, tax credit

Note: Support is estimated for a couple with two children in a single-income family earning the average production worker’s gross wage. The value of tax support is estimated as the difference between the average income tax rate paid by a couple with two children minus the average tax rate of a single person earning as an average production worker. The tax wedge includes income tax, employee and employer social security contributions, and state and local direct taxes. 1. Cash assistance in Greece is not provided as a transfer from government but as a supplement to wages paid by employers. Source: OECD (2004d), Benefits and Wages – OECD Indicators, OECD, Paris.

3.1. Giving children the best start in life Most OECD countries have recognised that making affordable and high-quality care available for young children is a sound investment in the future. Reviews of the experience of several OECD countries have stressed the value of early childhood education and care in easing the potential trade-off between parents’ employment and children’s development, and in improving the well-being of children (OECD, 2001c). While there is no consensus as to the appropriate age of attendance, a large body of research indicates that early childhood education and care brings benefits for children’s cognitive development (better literacy scores later in school), emotional growth (greater socialisation of children, less aggressive behaviour) and physical development (Kamerman et al., 2002). As the benefits are especially great for children from disadvantaged backgrounds, early childhood

76


II.5.


Box 5.1. Conditional benefits to poor families: the experience of Oportunidades in Mexico Oportunidades – introduced in 1997 as Progresa – provides cash grants to poor families to promote food consumption, encourage children to stay in school and visit health centres: cash-grants are conditional, as they depend on children staying in school and undertaking regular health checks. In 2004, the programme reached 5 million families, around 25% of the total population in 82 000 communities, with an annual budget of over 25 billion pesos (0.3% of GDP). Cash subsidies are paid periodically and can reach 22% of the beneficiaries’ total income; they are slightly higher for girls than for boys; and they increase with children’s school level. These benefits are paid directly to mothers, who are also responsible for reporting obligations. In addition to cash-transfers, the programme provides basic health services to all family members and special care services for pregnant and nursing women and to children under 2 years of age. Targeting is accomplished by first selecting communities with a high concentration of the population below the poverty line or inadequate coverage of public services and basic infrastructure (using an index based on census data) and then based on a proxy means-test of each household in these selected localities. These criteria allow for much better targeting than in earlier milk and food subsidy programmes, while keeping administrative costs low (at around 6%). Effects of the programme have been assessed through a randomised experiment whereby – within a sample of 506 communities covering 24 000 households – 320 communities were assigned to the programme, and 186 were left without treatment, regardless of their actual eligibility characteristics. The evaluation model is in its seventh year, with panel data in both rural and urban areas. Evidence from these evaluations, summarized by Skoufias (2001), indicates, among others that: ●

In terms of education, between 1998 and 2000, the programme increased attendance of children entering secondary school by almost 20% for girls and 10% for boys, with an increase in educational attainment relative to the control group of around 0.7 years by age 18.

●

In terms of health, the programme reduced the incidence of illness among children aged 1 to 5 by 12% between 1998 and 2000, increased the number of early medical visits of pregnant mothers by 8%, and lowered by 17% the number of sick-days of adult participants, relative to the control group.

●

In terms of nutrition, the programme increased median food expenditures and calorie intake by 13% and 11% respectively, and reduced the incidence of protein-malnutrition of children aged 1 to 3, relative to the control group.

These positive effects on child development were achieved with no significant impact on hours of paid work of fathers, and with an increase in paid work of mothers. In 2002, this programme received a 1 billion USD loan from the Inter-American Development Bank – the largest ever for a social program – to extend coverage to urban areas. More information is available on www.oportunidades.gob.mx.

education and care also helps to narrow the gap separating children, and to reduce the concentration of disadvantage later in life. The problem is that such care is not always available to respond to needs, and that its cost often puts it beyond the reach of many families. There are large cross-country differences in arrangements for care, especially for younger children. Coverage at younger ages remains limited in several OECD countries, while both the costs and quality of services supplied differ significantly across providers


77

II.5.


(institutionalised centres, family day-care, and individual care from relatives or babysitters). Investment strategies to reduce the risks of material deprivation among children have also focused on helping older youth make the transition from school to work. The relation between the structure of the education system and youth labour market outcomes is not a simple one. Cross-country differences in the structure of educational systems following compulsory education are not systematically related to different transition outcomes (OECD, 2000). However, greater diversification of educational pathways makes this transition easier for those who would otherwise have dropped out of school early. Because of this, successful reforms have aimed at making the school system more diverse, so as to meet the interests and aspirations of the widest possible range of youth. This has required working from both sides, by increasing the vocational content of the general education system and by strengthening general studies within the vocational stream, partially blurring the boundaries between the two. Reforms have also eased moving between these two paths in upper secondary education, often delaying specialisation until a later age and extending opportunities to access tertiary education for students in the vocational stream. Other types of investment programmes have targeted children at greater risk of dropping out of school (Box 5.2).

3.2. Involving families in programmes for children is critical for success Parents are the most important actors in the development of their children, but some may require special and tailored assistance to perform their task effectively. The issue of parents’ involvement and responsibility has special salience in the case of participation by children in early intervention programmes aimed at reducing the incidence of emotional, behavioural and social development problems among children. Evaluation studies of drug prevention programmes show that early childhood programmes are most successful over the longer-term if they closely involve parents at various stages of the programme’s activities (Normand et al., 2000). Yet the parents of disadvantaged children are also the most difficult to recruit and retain in these programmes. While these considerations suggest the existence of an unavoidable trade-off between programmes targeted to disadvantaged youths and parents’ participation, a range of practical steps have proved their worth in increasing parental involvement:

78

●

When planning the programme, parents’ willingness to participate can be strengthened by measures to increase awareness of the risk factors and potential consequences facing children before they arise, to provide opportunities for exchange with parents who have already participated in the programme, and to evaluate the programme’s outcomes.

●

In the recruitment phase, efforts to reach potential participants through advertising and personal invitations, as well as to standardise enrolment, are important. This implies avoiding stigmatising participants as “bad parents”, stressing the unavoidable nature of many critical transitions in the child’s development, and emphasising that the parents are not alone in this situation.

●

During the programme, it is important to take steps to make it flexible and accessible through a selection of evening, daytime and weekend meetings, or by repeating the programme several times over the course of the year, or by providing language facilitators in communities where language barriers may prevent parents from participating.


II.5.


Box 5.2. Investment policies targeting child poverty and disadvantage: Job Corps and Head Start in the United States The United States has developed a range of investment programmes targeted to reduce child poverty and disadvantage. While the programmes involve only a minority of children at risk, and evaluations are mainly limited to small-scale “model” programmes, evidence about their effects is suggestive of the outcomes that large-scale programmes could potentially achieve. Among programmes targeted at young children, the most important is Head Start, a pre-school programme for disadvantaged children that serves over 800 000 children in predominantly part-day programmes, at a budgetary cost of over USD 5 billion. A variety of programmes target older youths, most often with the goal of improving employment outcomes and educational achievement. One of the best known is Job Corps, a second-chance education programme that serves 60 000 disadvantaged school dropouts aged 16 to 24 and which is characterised by quite high costs (USD 20 000 per participant) and intensity (participants are sent for 8 months to residential campuses away from home). Analysis of the effects of such programmes on child development outcomes suggests a number of common patterns. In the case of early childhood education, experimental evaluations of model programmes that follow children at least until middle school suggest that well-designed and intensive interventions can make a positive difference in children’s lives. While these positive results cannot be automatically transferred to larger-scale programmes, Currie (2001) concluded that “… the short- and medium-term benefits (e.g. improving health and nutrition, preventing special education and grade repetition) could easily offset 40 to 60% of the costs of large-scale, publicly funded early intervention programmes such as Head Start”, and that “even relatively small long-term benefits (e.g. improving schooling and wages, reducing crime and teen pregnancy) of such programmes may be sufficient to offset the costs of public investment”. In the case of programmes targeted at older youths, an evaluation of 12 such programmes concluded that all “have demonstrated impacts on some of the outcomes they were designed to affect” (mainly in the fields of educational attainment and cognitive achievements, health and safety, social and emotional well-being, employment and welfare dependence); effects are however moderate in size, inconsistent among programmes, often limited in time and restricted to subgroups of older children (Hair et al., 2003). The same study also concludes that programmes that offer comprehensive services targeted at broader development outcomes (such as Job Corps) are more effective than limited interventions focused on specific outcomes.

Provisions concerning the role of non-custodial parents in relation to children also raise important issues for public policy. Receipt of child-support from the absent parent can provide an important addition to the income of lone-parent families and substitute for income-related benefits. In practice, often only a low proportion of non-custodial parents provides financial support to their children and maintains relations with them. This raises issues about the role of legislation in effectively leveraging the role of non-custodial parents and about their rights in relation to their children.

4. Increasing maternal employment reduces childhood poverty To ensure that public transfers do not reduce incentives for parents to seek gainful employment, many of the measures taken to reduce poverty among children have aimed to help parents find jobs, so as to increase their economic self-reliance. Boosting maternal EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

79

II.5.


employment is one effective way to do this, but will often require changes in existing tax and benefit systems as well as in childcare and family-friendly workplaces, which will be discussed in the next section.

4.1. Existing tax and benefit systems tend to discourage paid employment of mothers The tax system has a direct effect on the probability that second earners, usually mothers, enter paid work. In theory, joint taxation systems imply greater disincentives to work for second earners: in such systems, when taxes are progressive, the marginal rate for second earners will be higher than that of a single person at the same level of earnings (while, with separate taxation, couples would achieve higher disposable income when both spouses work). As shown in the second column of Table 5.2, in a majority (19) of OECD countries, spouses are currently taxed separately, while they are taxed jointly in five, and in the other six they can choose between separate or joint taxation (with joint taxation required over certain income levels). In practice, the labour force decisions of a second earner also depend on how the taxation system interacts with the benefit system. Table 5.2 compares the average effective tax rate applied to the additional income earned by the second earner for a couple with two children (3rd column) to that of a single person (i.e. the “tax penalty” for not being single, 4th column) and to that which would apply if the additional income were earned by the primary earner (i.e. the extent to which a couple has an incentive to become a dual-earner family, 5th column). Comparisons are made for different levels of earnings: primary earner paid at 67% of the earnings of an average production worker (APW), and second earner working part-time paid at 33%, in Panel A; primary earner paid at 100% of APW earnings and part-time second earner paid at 33%, in Panel B; primary earner paid at 100% of APW earnings and second earner paid at 67%, in Panel C. Effective tax rates include income tax, employee social security contributions and loss of income-related benefits (social assistance, housing and income-related family payments, excluding unemployment insurance benefits).1 Table 5.2 suggests a number of patterns: ●

For lower-income couples (Panel A), tax rates for second earners are in most countries much higher than for singles at the same earning levels (on average, second earners are taxed at a rate of 43%, 26 points higher than for singles). The tax penalty relative to singles is especially large in Portugal and the United Kingdom, but also in Australia, Canada, Denmark, Poland and the Slovak Republic; only in Greece and Hungary do second earners face the same effective tax rates as singles, while in Luxembourg and Switzerland their tax rates are lower, despite joint tax systems there. Also, in seven of 28 countries the effective tax rates for second earners are equal to or higher than those for primary earners, while they are substantially lower in six countries.

80

●

Where the primary earner’s income is higher, while that of the second earner remains unchanged (i.e. 100% and 33%, Panel B), differences in effective tax rates compared to singles are less marked: the effective tax rate on a second earner drops to 29%, on average, which is 12 percentage points higher than for singles and 10 points lower than for primary earners. While the average effective tax rate on a second earner is only marginally higher than for singles in some countries, the difference is 20 points or more in several others.

●

Where the second earner’s income is higher (i.e. 100% and 67%, Panel C), effective tax rates on second earners drop further on average (to 26%). In most countries, the tax rate


II.5.


Table 5.2. Effective tax rates on second earners at different income levels, and differences relative to singles and primary earners in couple families, 2002

Type of taxation system

A. Primary earner at 67% of the gross earnings of an average production worker, second earner at 33% Second earner, tax rate

Difference between second earner tax rate and Singles

First earner

B. Primary earner at 100% of the gross earnings of an average production worker, second earner at 33% Second earner, tax rate

C. Primary earner at 100% of the gross earnings of an average production worker, second earner at 67%


First earner

Second earner, tax rate


First earner

Australia

Separate

53

42

–3

29

18

–15

26

6

–26

Austria

Separate

21

5

–20

21

4

–21

25

2

–23

Belgium

Separate

44

26

–7

48

30

–7

48

15

–8

Canada

Separate

51

41

–7

37

27

–6

31

12

–9

Czech Rep.

Separate

33

16

–7

25

8

–10

37

15

0

Denmark

Separate

103

55

33

73

24

16

48

9

–12

Finland

Separate

50

30

–27

17

–4

–31

25

0

–25

France

Joint

51

30

–7

21

0

–3

24

4

0

Germany

Joint

55

33

0

49

28

0

51

15

5

Greece

Separate

16

0

–1

16

0

–13

16

0

–18

Hungary

Separate

18

0

–19

18

0

–41

18

0

–39

Iceland

Separate

45

15

1

47

17

0

44

25

–4

Ireland

Optional/Joint

34

34

–9

34

34

8

22

12

–10 –22

Italy

Separate

38

28

–13

38

28

–20

32

10

Japan

Separate

53

39

1

22

8

–8

20

3

–7

Korea

Separate

10

5

1

5

0

–16

6

1

–16

Luxembourg

Joint

14

–4

–49

14

–3

0

13

–4

–9

Netherlands

Separate

42

25

7

30

13

1

33

5

–4

New Zealand

Separate

56

33

–6

22

–1

–16

19

0

–19

Norway

Optional

27

9

–9

27

9

–15

30

5

–16

Poland

Optional

65

40

–3

31

6

–3

33

3

–1

Joint

87

76

26

56

45

–12

17

4

–29 –5

Portugal Slovak Rep. Spain Sweden Switzerland United Kingdom United States Simple average

Separate

69

56

–6

16

3

–6

25

7

Separate/Joint

16

10

–3

27

21

–2

20

8

–9

Separate

45

20

–10

28

3

–14

28

0

–19

Joint

4

3

–8

17

6

–6

24

6

1

Separate

68

62

–13

16

10

–29

27

7

–11

Optional joint

28

12

–25

20

4

–13

30

8

–1

43

26

–6

29

12

–10

26

6

–12

Note: The data compare the average tax rate applied to the additional income earned by the second earner for a couple with two children (3rd column) to that of a single person (4th column) and of a primary earner in a couple family with two children (5th column), at different earnings levels expressed relative to the wage of an Average Production Worker (APW). For example, in the case of Australia, the additional income of the second earner in a couple with two children is taxed at a rate of 53%; this is 42 points higher than that applied to a single, and 3 points lower than that applied to a first earner. Tax rates for second earners are calculated as 100 minus the change in the family’s disposable income given the change in individual income, where the change in disposable income includes changes in income-related family benefits, social assistance and housing benefits, but does not include unemployment insurance. The calculation is the same for primary earners, but assumes that the increase in income is associated solely with the employment of a single earner. The calculations for single people include only income tax and employee social security contributions, and do not take account of any income-related benefits or social assistance. Source: OECD (2004d), Benefits and Wages – OECD Indicators, OECD, Paris.

on second earners exceeds that of singles by only about 6 points (with the exception of Iceland, where the difference is 25 points, and of Belgium, Canada, Denmark, Germany, Ireland and Italy, where the difference is between 10 and 20 points). Differences in effective tax rates relative to primary earners are large and negative in most countries with individualised tax systems, while they are small in countries with joint or optional joint systems.


81

II.5.


Overall, this suggests that, at low levels of family income (e.g. when the second wage earner works part-time), most countries effectively operate “joint” tax-benefit systems that discourage the participation of a second earner (relative to the situation of a single person). This mainly reflects the impact of income-tested payments such as social assistance, housing benefits and income-related child payments that are withdrawn progressively as family earned income increases. An element of joint family taxation is also implicit in tax allowances that reduce the income tax paid by families with non-employed spouses, relative to that paid by singles. Other factors influencing these outcomes include employee social security contributions that, although individualised, operate in most countries from relatively low income levels, and the degree of progressivity of the income tax system. The net result overall is that current tax and benefit systems discourage mothers from entering or re-entering the labour force – a situation that needs to be remedied.

4.2. The impact of higher employment of mothers for poverty among children The scope for increasing labour force participation rates of women in general, and of those with children in particular, is potentially significant in most OECD countries. Figure 5.2 shows the simulated effect that three reforms would have on female labour force participation in 2025: i) equal tax treatment of second earners relative to single individuals; ii) an increase in public childcare spending per child; and iii) more favourable tax treatment of part-time work.2 The effects of these reforms on women’s labour force participation range from around 20 additional points in the case of Japan and Spain to negligible increases in Iceland and Hungary.

Figure 5.2. Potential increase in female labour force participation resulting from various policy reforms %

Favorable tax treatment of part-time work

Increase in childcare to Denmarks level

Equal tax treatment of second earner

Baseline

100 80 65

71

75

75

60

65

79

89

89

90

91

91

91

92

93

93

96

96

96

88

95

87

94

86

94

86

92

84

85

91

84

76

80

75

76

78

78

68

79

79

85

85

88

79

82

78

86

79

73

83

89

81

87

96

82

60 40 20

16

58

71

77

Tu rk M ey ex Hu ico ng ar Ko y re a Ita l y Gr ee Au ce st r Ne Swe ia w de Ze n a OE land CD -3 Po 0 Au land st ra l Lu Ja ia Un xe pa ite mb n Sl d K our ov in g ak gd Re om pu De blic nm a Fin rk l a Cz Po nd ec rt h ug Un Rep al ite ub d lic St at Ca es na d Fr a an ce Sp Be ain lg iu No m rw ay Sw Irela itz nd er lan d Ne Icel th and er lan ds

0

20

Note: Labour force participation rate of women aged 25 to 54. Countries are ranked in increasing order of the participation rate that could be achieved as a result of three sets of policies: i) equal tax treatment of second earners and single individuals (at 67% of APW), as is already the case in a few OECD countries; ii) an increase in public childcare spending per child to the highest level observed in the OECD (Denmark); and iii) more favourable tax treatment of part-time work, i.e. an increase of 11% in household disposable income (the maximum level observed in the OECD, Finland and Mexico) when moving from a situation where the husband earns all the market income (133% of an APW) to one where the husband and wife share market work (earning 100% and 33% of an APW). In the case of Turkey, for example, these three measures would increase female labour force participation from a projected baseline level of 16% in 2025 to around 20%. Source: Based on data shown in Burniaux, J.M., R. Duval and F. Jaumotte (2003), “Coping with Ageing: A Dynamic Approach to Quantify the Impact of Alternative Policy Options on Future Labour Supply in OECD Countries”, Economics Department Working Paper, No. 371, OECD, Paris.

82


II.5.


Increasing employment for mothers will reduce child poverty. Figure 5.3 shows the impact of higher female employment on child poverty under different assumptions regarding the allocation of the employment gains among different household types. While this (as well as other) “scenarios” presented in this report are based on very simplified assumptions, they highlight that policies can have an effect on outcomes that societies care about. For the purpose of these calculations, poverty rates in the various types of household are kept unchanged; what happens is that higher maternal employment leads to a greater share of children living in households with a higher number of wage earners. The effect on reducing child poverty is significant – an average increase in the employment rate of mothers of close to 12%, leading to a decline of around 10% in child poverty – although it is not that large in some countries. Beyond the size of the employment increase, the magnitude of the effect on child poverty also depends on how the higher employment is assumed to be allocated among different households. It is higher when this results in a decline in the share of children in jobless households – on average, the decline in child poverty increases from around 9%, when it mainly affects spouses of a first earner, to 13%, when it benefits jobless couples. However, this effect is sometimes muted in countries where the share of children in jobless households is already small (e.g. Mexico) and where poverty rates are higher in one-earner couple households than in those where no one works (e.g. Greece).3 Overall, this means that the average increase in female labour participation rates that could be achieved through the reforms simulated in Figure 5.2 would in many countries lead to tangible reductions in child poverty (as well as to a larger labour force).

Figure 5.3. Potential impact of higher employment among mothers on child poverty Poverty rate baseline

Spouses of working partners becoming more likely to work Spouses of working and non-working partners becoming more likely to work

% 30 25 20 15 10 5

Ita ly Ja pa n M Ne exic th o e Ne rlan d w Ze s ala n No d rw ay Po lan Po d rtu g Sw al e Sw de itz n er lan d Un T ite ur ke d y Ki n Un gd ite om d St at es

ria na Re da pu b De lic nm ar Fin k lan d Fr an Ge ce rm an y Gr ee c Hu e ng ar y Ire lan d Cz e

ch

Ca

st

Au

Au st

ra li

a

0

Note: The figure shows the relative poverty rates among children (when the poverty threshold is set at 50% of median equivalised disposable income) that are associated with the changes in female labour force participation shown in Figure 5.2. The first bar is the chid poverty rates in around 2000. The other two bars reflect the impact of higher female employment on child poverty produced by changing the distribution of children among different household types. Both alternatives assume the same aggregate increase in female employment (and of employment of single mothers), but they differ in the way this increase is allocated among different households. In the first alternative, the increase in employment for couples takes the form of an increase in second-earner employment (i.e. no changes in the share of children living in jobless couple households). In the second alternative, the increase in employment for two-parent families takes the form of an increase in both first- and second-earner employment (i.e. the share of children living in jobless couple households falls). Source: Data based on different waves of the OECD questionnaire on income distribution and poverty.


83

II.5.


5. Reconciling work and family life brings benefits to parents and children With the increase in women’s participation in the labour force, a growing proportion of families face the challenge of combining family and paid-work responsibilities. Balancing family and paid work is a challenge for both men and women and has important implications for child development as well. ●

For some women in their main childbearing years, having children and fulfilling labour market aspirations are mutually exclusive activities. In such conditions, they may adjust their family behaviour and decide to have children at a later age, or not to have as many as desired, or not to have any at all. Alternatively, they withdraw from the labour market, temporarily or permanently, at childbirth. In some cases, this can be because they wish to care for their children on a full-time basis, whatever their employment opportunities. In other cases, however, they do so despite their desire to resume paid work, or to work longer hours, because of constraints in terms of time, access to services or limited opportunities to resume career tracks after childbirth. As a result, the current labour supply is less than it could be, human capital is underused, and economic growth falls below potential.

●

Other parents may find themselves compelled to work to make ends meet, despite the lack of low-cost, high-quality childcare, putting their children’s development at risk. Some studies have indeed suggested negative effects from mothers’ employment on the cognitive development of children under the age of 9 months or one year – in particular when the mother’s employment involves long hours, irregular shifts, or work during weekends, and when high-quality and affordable early education and care is not available.4

To avoid these risks, a broad range of policies to reconcile work and family life can be used. These include ensuring the availability of quality and affordable childcare, leave provision that allows working mothers to withdraw temporarily from the labour market following childbirth, and flexibility in working-time and school-time arrangements. These provisions are particularly important in the case of children with special needs, such as chronic diseases and disabilities.

5.1. Childcare arrangements: the challenge of availability, quality and cost Many families lack viable childcare options during their employment hours. For many children informal substitute care options are lacking, and formal childcare is unavailable, unaffordable, and/or of inadequate quality. In particular: ●

Affordable, high-quality, full-day childcare is especially limited for children younger than age 3 and for school-age children after regular school hours.

●

Where childcare is unregulated, there are concerns about its quality.

●

Low-income families can be burdened by heavy out-of-pocket expenditures for childcare.

●

School hours are poorly synchronized with parents’ employment hours: some school systems, for example, still expect children to return home at lunch.

●

Children’s summer break exceeds parents’ annual vacation allowance by a large margin.

These kinds of problems undermine mothers’ labour market attachment. Some parents choose atypical work schedules, so that one parent is at work while the other is at home. Some young children – especially poor children – spend time in low-quality

84


II.5.


childcare (or may be altogether unattended), compromising their safety and placing them at developmental risk. Where public early education is not widely available, poor children may miss the early educational experience of their more affluent peers. Among the countries for which data are available, the proportion of children under 3 in formal government or market childcare is highest (40% or more) in Canada, three of the Nordic countries (Denmark, Sweden and Norway), the Slovak Republic, the United States and New Zealand. Very much lower proportions are found in the Southern and Central European countries. For children between 3 and the mandatory school age, enrolment is more uniform and much higher, reaching 90% or more in several countries. While facilities for pre-school children over 3 years old tend to be financed mainly out of public expenditure, there is more diversity in the financing of formal childcare for the under-3s. Childcare centres for this age group are mainly publicly financed (though not necessarily publicly operated) in all of the Nordic countries, as well as in a number of other European countries. The non-European countries, as well as Ireland, the United Kingdom, the Netherlands and Switzerland, rely mainly on commercial private sector provision of formal childcare services for children under 3. Greater access to formal childcare facilities, whether provided directly or subsidised by public authorities, can be expected to raise the employment participation rates of mothers. Indeed, some studies find significant positive effects (Gustafsson and Stafford, 1992; Kimmel, 1998; Powell, 1998). Conversely, high childcare costs can have a negative impact on the probability that the mother works. However, other studies find that the primary benefit of more generous childcare subsidies is to allow users of high-quality care to purchase services of even higher quality or to switch from informal to formal childcare (Dobbelsteen et al., 2000). The main issues in assessing the relative merits of public and private provision of childcare from the point of view of the household are availability, cost and quality. In countries relying on private provision, costs can be high. For example, in the United States, where parental fees constitute three-quarters of childcare financing, low-income families can devote about 25% of their family income to childcare. Canada, Ireland, New Zealand and the United Kingdom also have very high net childcare costs for those having to rely on market care. While some private day care is of very high quality, standards vary, and there is a case for requiring special scrutiny, for example of private networks of child-minders based in their own homes. Finally, a number of countries, including Denmark, Finland, France and Norway, have schemes to provide subsidies to parents looking after their own children at home. In these countries, benefit rates may decline with the hours of public childcare use (as in Norway) or may be conditional on parents not using public childcare facilities at all (as in Denmark and Finland). These schemes are closely linked, conceptually, with the paid parental leave schemes that have been developed in many countries. However, they do not necessarily carry any rights to return to a job.

5.2. Leave provisions: helpful for mothers’ employment if not too long While maternity leave, with employment protection, has been widespread in OECD countries for many years, paternity leave and parental leave are more recent developments. Parental leave is usually defined as leave in addition to maternity/paternity leave to allow parents to take care of an infant or young child. However, in countries where childcare leave systems combine individual and family entitlements, paternity and parental leave may be incorporated into childcare leave arrangements for the family as a EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

85

II.5.


whole. Entitlement to maternity and childcare leave is often conditional on previous work experience on a continuous and full-time basis as an employee over a certain period (usually for a year). Exceptions include the Scandinavian countries (where most women are covered), the Netherlands (where some temporary and part-time workers are covered) and Germany (where mothers in education or who are unemployed are also covered). In the Southern European countries, entitlement often depends on having a contract for permanent employment. There is statutory paid maternity leave, which is often remunerated at 100% of previous earnings, in almost all OECD countries (exceptions include the United States, Australia, and until recently New Zealand).5 Statutory paid maternity leave equivalent to 13 weeks of pay or more had been instituted before the end of the 1970s in Finland, Norway, Sweden, Italy, Austria, Germany and France. By the end of the 1990s, this level had been exceeded in 16 countries. The total duration of maternity/childcare leave (paid or unpaid) is currently a year or more in over 20 OECD countries (Table 5.3). Other recent changes include the extension of some forms of leave to part-time employees and the greater flexibility introduced into parental leave arrangements. Following the precedent set by Sweden, a number of countries (including Austria, Denmark, Finland and the Netherlands) now provide for some flexibility in the working hours of parents. As noted in OECD (1999), transitions from part-time to full-time work have been relatively common in Sweden, partly as a result. Germany facilitates the re-entry of mothers to work by means of employer subsidies for retraining programmes, childcare provision and wages. Specific paternity leave entitlements are still relatively uncommon, and often of short duration.6 In addition, fathers are increasingly eligible for paid leave under maternity/ childcare leave provisions, sometimes with a “father quota” available on a “use-it-or-loseit” basis. Until recently, fathers took up little of the paternity or childcare leave available to them, but this has been changing. In most Nordic countries, paternity leave take-up rates of at least some of their entitlement are close to 60% or more, and substantial proportions of fathers are taking up at least part of the childcare leave that is reserved for them – in particular younger and more educated men (Ellingsaeter, 1998; European Commission, 1998).7 However, the amount of leave taken is generally unknown – as is the proportion of time the fathers on leave spend looking after their children. Parental leave arrangements with specific “father quotas” also exist in Austria, Denmark and Portugal. In the Netherlands, fathers of young children are entitled to reduced hours and, according to a 1994 survey, 13% of fathers switched temporarily to a four-day week when their children were very young. Maternity leave entitlements that provide a right to return to the previous job can be expected to raise mother’s employment rates. Indeed, the main reason given by employers who offer extended periods of maternity leave is precisely to increase retention rates of mothers. Concern has been expressed, however, that long periods of maternity leave (or childcare leave, which is generally taken by the mother) may lead to detachment from the labour market and to lower employment rates and earnings for mothers in the longer term. In the Nordic countries, long parental leave entitlements, paid at almost a full rate, do not seem to have had a significant negative impact on women’s labour market opportunities compared with other OECD countries where leaves are shorter in duration and sometimes unpaid. Ruhm (1998), comparing data from 16 OECD countries, concludes that short spells of maternity leave are associated with higher female employment rates, but finds no consistent results regarding longer periods of leave. The special features of the

86


II.5.


Table 5.3. Maternity and parental leave, duration and benefits in 2003-2004 Statutory maternity and child care leave Maximum number of weeks Maternity leave

Total

Firm provision

Benefit levels

Maternity grants and allowances/ other payments

Paid

% of female employees with children reporting extra-statutory coverage

Australia

0

52

0

..

Lump sum of AUD 3 000 per child.

> 34

Austria

16

112

38

100%

–

85

Belgium

15

67

31

82% for 30 days, then 75% up to maximum.

–

65

Canada

15

27

14

55% of previous average insured earning plus a supplement for annual income less than CAD 25 921. Benefit paid for up to 15 weeks, plus up to 35 additional weeks for parental care on the birth or adoption of a child. Maximum weekly benefit is CAD 413.

–

..

Czech Republic

28

28

..

69% of the daily assessment base (as calculated for cash sickness benefit). Maximum benefit is CZK 419 a day.

–

..

Denmark

30

82

42

90% up to maximum.

–

40

Finland

52

164

55

100-60%, decreasing with earnings.

–

36

France

16

162

73

100% up to maximum.

–

58

Germany

14

162

38

100%

–

92 (Western Länder)

Greece

16

42

8

100%

EUR 698.70 for each child.

81

Hungary

24

180

58

70% of daily average net earnings in the previous year. In the absence of earnings, the current minimum wage is used.

Lump sum equal to 225% of the minimum old-age pension.

..

Iceland

..

26

..

80% of the insured’s wage or income for 1 year ending 2 months before the first day of parental leave for employees and self-employed. Minimum benefit depending on whether working full-time or part-time.

ISK 40 409 a month for parents inactive or in partial employment; ISK 91 200 a month for parents attending full-time education.

..

Ireland

14

42

10

70% with minimum and maximum.

–

68

21.5

64.5

30

80%

–

81

Japan

14

58

8

60% of average daily basic wage; the benefit is reduced if the mother is receiving a wage or cash sickness benefit.

Lump sum of JPY 300,000

10

Korea

8.5

61

..

100%

Luxembourg

16

68

32

100% with minimum and maximum.

Lump sum of EUR 2 954.77 payable for 16 weeks to persons with no entitlement to maternity benefit or whose benefit is less than the lump sum allowance.

82

Mexico

12

12

12

100% of average earnings; if unable to work 42 days after childbirth, the mother is eligible for sickness benefit.

Assistance in kind, payable for up to 6 months after childbirth.

..

Italy

..

Netherlands

16

68

16

100% up to maximum.

–

75

New Zealand

0

52

0

100% up to maximum.

Income-tested maternity benefit payable to single woman after 27th week of pregnancy up to 13 weeks after childbirth.

..

Norway

42

116

..

80-100% with maximum and depending on length of leave.

NOK 33 584 if not receiving maternity benefit; NOK 1 765 for giving birth at home. Additional benefit payable to widowed, divorced, separated, or unwed mothers.

..


87

II.5.


Table 5.3. Maternity and parental leave, duration and benefits in 2003-2004 (cont.) Statutory maternity and child care leave Maximum number of weeks

Poland Portugal

Benefit levels

Maternity leave

Total

Paid

18

122

35

100%

Firm provision Maternity grants and allowances/ other payments

% of female employees with children reporting extra-statutory coverage

Childbirth supplement (means-tested): lump sum of PLN 500 for each child.

..

24.3

128.3

17

100% with minimum.

Slovak Republic

28

184

59

90% of net wage, low maximum.

Grant for the purchase of necessities for a newborn child (if parents have permanent residence in the country).

49 ..

Spain

16

164

16

100%

Lump-sum of EUR 450.76 on birth or adoption of a third child (payable if family income does not exceed the level established for family allowances).

69

Sweden

64

85

40

80% with minimum and maximum, followed by flat rate.

–

7

Switzerland

16

16

..

Negotiable between insurer and insured.

CHF 600 to 1 500, according to canton.

..

Turkey

12

12

8

Two-thirds of earnings.

Lump sum nursing grant of TRL 500 000.

..

United Kingdom

18

44

8

90% for 6 weeks, then flat amount or 90% if lower

61

United States

0

12

0

Benefits in 6 states (75% of earnings in Rhode Island, 66.6% in Hawaii, 53% in New Jersey, a variable proportion of quarterly or annual earnings in California, New York, and Puerto Rico).

50

Note: Data refer to national legislated maternal and parental leave provisions. “..” means that data are not available. Source: See OECD (2005d), Reconciling Work and Family Life – Social Policies for Working Families, forthcoming, OECD, Paris, for detailed sources.

programmes, and the way they are funded, may matter more than their duration. The takeup of the schemes is likely to vary, and may be quite low among highly-skilled women.8

5.3. Fathers need to share responsibilities for children At the core of the debate about how best to organise parents’ employment and caregiving time lies a real conflict. Children, especially the very young, need ample time with their parents. If fathers continue to maintain high levels of labour market attachment – maximum employment rates, full-time hours, and little leave-taking – then, in practice, increasing maternal employment will imply a reduction in the total amount of time children spend with their parents. While the strong labour market attachment of fathers, as a group, has long been assumed to be immutable, in recent years this assumption has been subject to questioning. An alternative to the “male breadwinner/mom at home” model is the “dual-earner/ dual-carer” model. This envisions an arrangement in which men and women engage symmetrically in both paid work and unpaid care-giving in the home. Also, this model places primary responsibility for the care of very young children in the home, rather than in the hands of out-of-home carers; and, since there are now two “breadwinners”, it makes possible shorter average work hours (i.e. less than a 40-hour standard) throughout the life cycle. While the “dual-earner/dual-carer” model has not been fully realised anywhere, reforms in several Nordic countries have explicitly stressed paternal engagement in caregiving – via the enactment of “use-or-lose” portions of parental leave, the expansion of

88


II.5.


high-quality part-time work for both parents and a shortening of the standard full-time work week – as well as provision of a continuum of support provided to parents with young children (Box 5.3). In the Netherlands, the 2000 Work and Care Act aims to enable couples to hold “one and a half jobs” between them – with each holding a “three-quarter time job” – thus achieving both gender equality and time for care (European Foundation for the Improvement of Living and Working Conditions, 2000).

Box 5.3. Reconciling work and family life in Sweden: a continuum of public policy supports Public policy in Sweden, as in other Nordic countries, has long embraced the ideal of providing parents with a continuum of support, often until children are in their teens. The model involves flexible use of paid parental leave, affordable high-quality childcare, extensive out-of-school care, and entitlements to shorter working hours for both parents when children are young. ●

Employment-protected leave is an integral part of Sweden’s family-friendly policies. Parental leave can be taken for up to 18 months, at relatively high replacement rates (80%) for most of this period. In addition, following birth, parents are entitled to reduce working hours until children go to school: as a result, more than half of mothers in dualearner households work less than 35 hours per week.

●

Childcare. Sweden has a long-established system of municipal day care, whose emphasis has progressively shifted from merely “facilitating parents working” to include child development concerns. The Swedish authorities plan to further raise the already high quality standards, through staff training and reductions in group size, which will add to the already significant public childcare outlays (2% of GDP).

●

Out-of-school hours care services, provided at school facilities or elsewhere, cover about 80% of all 6 to 8 year-old children. Management of these services is closely integrated with that of primary schools.

●

Gender equity. Swedish policies try to induce more fathers to take parental leave for longer periods by reserving some weeks of parental leave exclusively for their use. These policies have contributed to some increase in the paternal share of days of parental leave, from 11% in 1994 to 17% in 2003. However, this increase does not reflect a fundamental behavioural change among Swedish men, as paternal leave is often tagged on to extend holidays.

A common feature of these measures is their strong employment focus for all adults, including sole parents. While it is difficult to disentangle the influence of each of these measures, there is consensus that – combined – they help parents a great deal in squaring their work and care commitments. Survey responses indicate that few women consider having a career and motherhood as mutually exclusive: the total fertility rate of 1.7 is above levels recorded in many other European countries, and only 15% of all women at age 40 (with or without university degree) are childless. Employment rates are very high (over 80%) among both mothers with children and lone parents, thereby contributing to low child poverty (less than 4%) and to small gender gaps in employment (3%). The downside to this policy model is its high cost (public outlays on childcare and parental leave payments was equivalent to 2.7% of GDP in 2001), which may negatively affect economic growth.


89

II.5.


5.4. Responsibilities of employers towards working parents Firms play a crucial role in work/family reconciliation, but there is still much uncertainty as to how much the “business case” for family-friendly workplaces can be relied on and how much other measures, such as compulsory legislation and tax incentives, are needed. Whatever government policies are put in place, the day-to-day reality of how easy or difficult life is for working parents depends largely on what happens in the workplace. Employers’ attitudes are often quoted by fathers as an important reason for their low take-up rates of parental leave; employers sometimes regard fathers who take parental leave as uncommitted to their jobs. Relatively low-skilled or easily-replaced employees may be the most likely to be working under conditions that put greater strains on family life, as the business case for using family-friendly employment practices to retain these workers is weaker than for more “highly-valued” employees. On the other hand, in some countries, many firms not only comply fully with national legislation but complement it through “family-friendly” arrangements – either for business reasons, or because of company values. Family-friendly work practices include leave provisions for extra-statutory maternity, paternity and parental leave, career breaks, leave to care for elderly relatives, and emergency leave to deal with a sick child or problems with childcare. Changes in work arrangements for family reasons include reductions in working hours (e.g. from full-time to part-time), working contracts for school-terms only, working at home for family reasons, and appropriate flexi-time arrangements. All these arrangements can ease work/family tensions, especially where alternative means to reconcile work and family life, such as quality and affordable childcare, are not available. Family leave is useful when children are ill or when childcare arrangements break down. Flexibility in working hours is important to deal with the emergencies of everyday family life. The most extensive information on family-friendly arrangements in firms comes from Australia, Japan, the United Kingdom and the United States. Because of low public provision of childcare and statutory leave (maternity, paternity and parental), a good deal of responsibility for work/family reconciliation in these countries depends on workplace practices. Employer surveys show that family-friendly arrangements are most common in the public sector, which usually employs a relatively high proportion of women. Familyfriendly arrangements are also more common in large firms, especially in Japan. However, the differences between firms of different size may not be as high as shown in surveys, as smaller firms may be more willing to allow informal arrangements. Family-friendly arrangements are also more common in firms with higher proportions of professional and technical workers, where a written “Equal Employment Opportunities” statement exists, where there is a structured hierarchical management system, or where there is a relatively high proportion of female managers (Evans, 2001). In the United States, there appears to be a link between family-friendly arrangements and a “highcommitment” management style, in which senior managers adopt a strategy of teamworking and job-rotation, and delegate more responsibility to lower-level staff. Regarding the type of family-friendly arrangements on offer and the benefits they bring, employers mention changes in working hours, such as part-time working and flexi-time, more often than extra family leave or help with childcare; only rarely do employers seem to provide benefits from all of the four categories mentioned above. Employers refer most commonly to better retention rates of valued staff with family responsibilities, and improvements in

90


II.5.


staff moral, as reasons for introducing these arrangements.9 However, while familyfriendly working practices can deliver benefits to firms, they also have costs, notably to cover for absences. While subjective evidence from employers suggests that net benefits of family-friendly arrangements are positive, there is little objective evidence that introducing them improves the financial situation of firms. One reason for this may be that such arrangements are often associated with other working practices (such as “highcommitment” practices) that have stronger beneficial effects. Flexible working hours, followed by various types of short-duration family leave schemes (such as sick-child leave) are reported most often in surveys of employees’ working conditions – work-place crèches and career breaks much more rarely. For the European Union, the proportion of women employees with a child under 15 who reported that extra-statutory family leave or childcare arrangements were available in the companies where they worked was between 66 and 75% in Austria, Italy and western Germany, and between 60 and 65% in Greece and Spain, while Nordic countries, Ireland and the United Kingdom are at the bottom (20-40%). Also, the Netherlands and Portugal stand out for having particularly high levels of firm-provided day-care (20-25%) relative to the amount of extra-statutory leave that their firms provide.10 Analysis of Australian and Canadian data shows that such flexible-hours arrangements are appreciated by employees: job satisfaction is increased, and stress reduced, when employees with family responsibilities do not have to work more hours than they desire and have some control over their starting and stopping times. However, data from the United Kingdom show that some forms of family-friendly arrangements are seldom used even when they are available. In particular, career breaks are a fairly common entitlement for “fast-track” women employees, but are hardly ever taken up, perhaps because women are not attracted by arrangements that might slow their career progression. In some countries, however, these changes have not stopped employees from becoming less content, overall, with the reconciliation between their work and family lives. Flexi-time shows much less national variation than does extra-statutory leave: it is more common in the United States and Australia (37 to 40% of all workers reporting flexitime) than in Europe (18 to 35%). Voluntary part-time work, as a proportion of total female employment, is most common in Japan, the Netherlands and the United Kingdom, while Finland and the Southern European countries have comparatively low figures. Another form of family-friendly working arrangement is working at home for family reasons. Despite considerable discussion of its potential, there is as yet little evidence that working at home is common or growing quickly. In most cases, however, the business case for “family-friendly” workplaces is likely to be weak. In these circumstances, the result of relying on the industrial relations system and voluntary initiatives is that only small numbers of employees (usually high-skilled workers, or public employees) will benefit. Legislation is one option to ensure that provision is available to workers. A less controversial alternative is to subsidise tailoredadvice to companies on how best to manage employees with parental responsibilities (Box 5.4).

6. Raising fertility: satisfying couples’ preferences, boosting the labour force Supporting the fertility and reproductive decisions of families has only recently re-entered the agenda of social policy. While some countries have used specific cash


91

II.5.


Box 5.4. Government policies to make business more family-friendly Governments can use a mix of policies to support the diffusion of family-friendly workplaces: encouragement, provision of financial assistance, or legislation. Several OECD governments have sponsored awareness campaigns on the merits of family-friendly workplaces: measures often include annual competitions to identify the firms that provide the best conditions for working parents and their families. Financial support for enterprises can also be linked to specific measures, for example, fiscal support to stimulate employer provision of childcare. A more innovative way of encouraging businesses to become more family friendly is to provide them with technical advice. In Austria, under the Work and Family Audit, a consultant identifies enterprise needs through discussions with both management and employees covering different areas (e.g. working time, available support services). An enterprise plan is then developed that identifies goals to be achieved in a given period. Enterprises are partly reimbursed for the costs associated with participation in the audit process through subsidies. Finally, at the end of this period, an external auditor assesses achievements, and, ideally the process is then repeated – with new goals identified to ensure that workplace practices continue to reflect the (changing) needs of both firms and working parents. Legislation in Japan identifies both individual workplace needs and long-term enterprise commitment to work and family reconciliation. The 2001 Childcare and Family Care Leave Act entitles parents of very young children to access at least one of several family-friendly measures, according to which is most suitable to employers: i) reduced working hours; ii) flexible working hours; iii) exemption from overtime work; or iv) operation of a childcare centre. The Act also encourages enterprises to appoint a “work and family life reconciliation promoter”, to oversee its implementation and advise, on a continuous basis, on family-friendly practices that best suit the workplace in question. Amendments to this act in 2004 expanded child-care leave to temporary workers, extended its length to 1½ years, and introduced a care leave to allow parents to look after their sick and pre-school children. Some countries have legislated the right of parents with children who are below school age to reduce their working hours (e.g. Sweden) or to change working hours for employees regardless of their family status (e.g. the Netherlands). Other governments are reluctant to introduce legislation, because it can result in large costs to employers and does not recognise differences in workplace conditions. Since 2003, legislation in the United Kingdom has granted parents with children not yet of school-age the right to request flexible working hours (including part-time work), thereby forcing both employees and employers to justify their case for or against a particular practice. While this Act does not establish a direct entitlement for parents, it will nevertheless increase costs to employers in certain cases.

benefits to encourage families to have more children, making a longer-term impact on reproduction decisions requires paying attention to providing families with more support overall, particularly with regard to financial affordability, work environment and childcare.

6.1. Several policies can help remove obstacles to childbearing In some OECD countries, the greater attention paid to the need to raise current recordlow birth rates has led to the introduction of specific cash benefits. In France, beginning in January 2004, mothers of each newborn child are awarded EUR 800; and in Italy a benefit is

92


II.5.


paid at the birth of the second child. In 2004, Germany introduced provisions to increase contributions to the nursing scheme paid by families without children (relative to those with children), thereby sharing the costs of raising children more broadly. However, because of the small amounts provided, it is unlikely that specific benefits will significantly change current reproductive decisions. Of greater importance is whether the full set of taxes and cash benefits provided to families actually reduces the cost of children. Most OECD countries provide tax allowances or benefits to families according to the number of children, on either a universal or targeted basis. The potential impact of these programmes on childbearing depends on the interaction of several features of the taxbenefit system, such as its progressivity and disincentives to second earners. Table 5.4 provides information on the size of these net transfers to families at different levels of gross household income (100% of earnings of an APW in Panel A, 133% of an APW in Panel B, 200% of an APW in Panel C) under different configurations of children (0 and 2), family type (singles and couples) and number of earners (1 and 2).11 The average effective tax rate (which includes both direct taxes and public benefits) that applies to a couple with two children where both parents work (2nd column) is compared to that of a two-earner couple without children (3rd column) and to that of a single parent with either no children (4th column) or two children (5th column). Average tax rates of families with children are generally lower than those for families without children, but this varies according to the level of household income and other characteristics: ●

In the case of middle-income households (gross earnings at 100% of those of an APW, Panel A), the average tax rate of a two-earner couple with two children (at 11% on average) is 8 points lower than that of a couple without children (while there are no differences between the two in Korea and Poland). Single parents with two children are taxed at rates that are 3 points higher than couples with children, with the exception of the Nordic countries, where tax rates on single-parent families are lower than those for couples. Singles with no children, at the same gross income level, are taxed at a rate that is, on average, 15 points higher than for couples with two children, and more than 25 points higher in several European countries.

●

At income levels that broadly correspond to those of a household with one spouse working full-time and one working part-time (gross earnings at 133% of those of an APW, Panel B), couples with two children are taxed, on average, at a rate of 16%, 3 points lower than single parents with two children, 6 points lower than couples without children, and 13 points lower than persons living alone at the same income level.

●

At higher levels of gross income (gross earnings of 200% of those of an APW, Panel C), the average tax rate of a two-earner couple with two children rises to 22% on average, and the advantage relative to a childless couple declines to 4 points on average (while the difference between the two rates is zero in Australia, Korea, New Zealand, Poland and Turkey). In all countries except Canada, Iceland, the Netherlands, Poland and Spain, single parents with two children are taxed at higher rates than couples with the same number of children, while the disadvantage of singles without children rises to 12 points on average.

While the preferential tax advantages granted to families with children may seem significant, they are smaller than estimates of the higher household costs of larger families that are implicit in the elasticity generally used in comparative research to “equivalise” household income, i.e. to adjust for the additional costs of larger families. Figure 5.4


93

II.5.


Table 5.4. Effective tax rates on couples with two children at different income levels, and differences relative to singles and childless couples, 2002 A. Primary earner at 67% of the gross earnings B. Primary earner at 100% of the gross earnings C. Primary earner at 100% of the gross earnings of an average production worker, of an average production worker, of an average production worker, second earner at 33% second earner at 33% second earner at 100% Difference in tax rates relative to: Couple with 2 children, 2 earners

Australia

Couple without children, 2 earners

Single without children

Difference in tax rates relative to:

Couple with 2 children, Single 2 children 2 earners



Difference in tax rates relative to:

Couple with 2 children, Single 2 children 2 earners



Single 2 children

13

–4

–11

–1

17

–3

–12

–4

24

0

–11

Austria

3

–16

–25

–7

13

–12

–19

–5

20

–8

–17

–3 –5

Belgium

20

–11

–21

–8

29

–9

–17

–7

37

–6

–14

–2

Canada

8

–10

–16

–3

18

–4

–8

–2

24

–1

–5

2

Czech Rep.

9

–11

–14

–4

13

–9

–12

–6

18

–6

–10

–4

Denmark

32

–7

–12

6

35

–6

–13

0

40

–4

–14

–1

Finland

17

–9

–16

6

23

–7

–13

3

28

–4

–13

–2 –2

France

15

–7

–12

0

17

–7

–12

–1

21

–6

–12

Germany

17

–12

–24

–9

26

–9

–20

–7

35

–6

–13

–1

Greece

13

–2

–5

0

14

–3

–6

–3

17

–1

–10

–4

Hungary

4

–18

–26

–5

12

–15

–25

–9

20

–10

–22

–10

Iceland

6

–9

–21

–7

16

–5

–14

–6

26

–1

–9

0

Ireland

–4

–8

–20

–6

5

–6

–18

–5

12

–4

–18

–5 –1

7

–12

–20

–10

19

–5

–12

–6

25

–3

–11

Japan

Italy

13

–5

–7

–2

15

–4

–6

–4

18

–1

–6

–1

Korea

5

0

–2

–2

6

0

–5

–5

7

0

–8

–4

Luxemburg

–3

–16

–25

0

2

–14

–26

–4

9

–11

–25

–9

Netherlands

19

–5

–15

–1

25

–4

–7

3

31

–2

–7

2

Norway

15

–8

–14

1

20

–6

–13

–1

25

–4

–14

–3

New Zealand

18

–1

–3

–2

20

0

–4

–4

21

0

–8

–2

Poland

28

0

–3

–3

30

0

–2

–2

31

0

–2

2

5

–6

–11

0

7

–7

–14

–4

12

–5

–13

–5

10

–2

–10

–3

6

–12

–17

–11

13

–6

–13

–8

0

–16

–18

–3

15

–1

–7

4

17

–2

–7

4

18

–10

–13

9

22

–7

–11

5

26

–5

–14

–3 –2

Portugal Slovak Rep. Spain Sweden Switzerland

6

–8

–14

–3

10

–7

–12

–2

16

–5

–9

United Kingdom

4

–11

–19

–10

11

–8

–15

–7

17

–6

–12

–5

United States

12

–7

–13

–4

15

–7

–13

–4

17

–6

–15

–5

Unweighted average

11

–8

–15

–3

16

–6

–13

–3

22

–4

–12

–3

Note: The data compare the average tax rate of a two-earner couple with two children (2nd column) to that of a two-earner couple without children (3rd column), of a single person without children (4th column) and of a lone parent with two children (5th column), at different earnings levels. For example, in the case of Australia, the income of a two-earner couple with children, with a household earned income of 100% of the earnings of an APW (average production worker), is taxed at a rate of 13%, which is 4 points lower than the rate paid by a two-earner couple without children, 11 points lower than that paid by a single person without children, and 1 point lower than the one paid by a lone parent with two children. Source: Computation from the OECD Tax-Benefits model database, as shown in D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

illustrates this with respect to the average OECD values shown in Table 5.4, for three levels of gross earnings (100%, 133% and 200% of APW earnings) and four family types. Despite their lower tax rates, couple families with two children have a lower equivalised disposable income relative to couples without children and, to a larger extent, to singles without children; their equivalised net income is also lower than that of single parents with two children, although the differences narrow at higher levels of gross income.12

94


II.5.


Figure 5.4. Non-equivalised and equivalised household income for different family types and gross earnings, OECD average 2002 Gross non-equivalised income 200

Net non-equivalised income

Net equivalised income

Gross income at 100% of APW

160 120 80 40 0

200

Couple with Couple without 2 children, children, 2 earners 2 earners



200

160

160

120

120

80

80

40

40

0



Single 2 children

0

Single 2 children




Single 2 children

Note: APW = Average Production Worker. Source: Calculations based on OECD Tax-and-Benefit models database. See D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris, for details.

Of specific importance for policies aimed at making children more affordable is their relationship to policies that encourage greater female employment. As work and childbearing represent alternative uses of mothers’ time, policies in these two fields could conflict. This risk is especially important in the case of cash-transfers granted to families at the birth of each child that, if large enough to alter reproductive decisions, may lead mothers to withdraw (or not enter) the paid-labour market in order to rear their children instead. There is evidence, however, that the nature of the relationship between paid work and fertility has changed over time. At the level of individuals, time-series studies suggest that the negative coefficient linking paid work and childrearing has declined in several countries (OECD, 2003e). Across OECD countries, the relationship between female employment and fertility rates has also reversed its sign (Ahn and Mira, 2002): while in 1980, total fertility rates tended to be higher in countries with lower female employment rates (and a more traditional gender division of work and care responsibilities), and lower in those with a higher proportion of women in paid employment, the opposite pattern occurs in 2000 (Figure 5.5). The fact that countries where female employment rates are higher also display higher fertility rates is consistent with the view that financial security and autonomy are required by women before they contemplate having children, and that having a foothold in the labour market is the best assurance of securing both.


95

II.5.


Figure 5.5. Cross-country relations between total fertility and female employment rates 2000

1980 Total fertility rate 3.4

Total fertility rate 2.6

IRL

MEX

3.0 KOR

2.2

2.6 ESP

2.2 1.8

ITA BEL NLD

1.4 1.0

20

PRT

GRC NZL

AUS GBR FRA

AUTDEU CHE 30

FRA

1.8 USA

JPN

40 50 60 Employment rates of women

NLD FIN

NZL AUS

USA

ISL

NOR

DNK GBR CAN KOR SWE CHE HUN DEU AUT POL SVK JPN ITA ESP CZE PRT

SWE FIN

IRL

1.4

1.0

20

30

40

60 50 70 Employment rates of women

Source: Data on employment and fertility rates, as published in various issues, Society at a Glance – OECD Social Indicators, OECD, Paris.

The existence of a positive relationship between work and fertility suggests that the same policies that have proved their worth in increasing female employment and facilitating the reconciliation of work and family life may also be effective in helping parents to overcome obstacles to childbearing. OECD (2001b) shows evidence of a positive relationship between a synthetic index of the extent to which labour markets are “familyfriendly” and fertility rates across countries. A simple cross-section analysis13 identifies some policy variables as having significant impact on fertility rates: these include the share of children below mandatory school-age that are attending formal childcare, the length of parental leave, direct costs of children, and availability of part-time employment for women.14 Similar results are reported in other studies of the determinants of women’s reproductive decisions.15

6.2. The impact of higher fertility rates on the size of the population and labour force Both the scope and pay-off of policies aimed at supporting fertility rates are potentially significant. Figure 5.6 provides one indication about the level that fertility rates could reach in several OECD countries if various policies were to be set at the level prevailing in the three countries where these are currently highest. These results are not predictions of “most likely” outcomes, but simulations of the possible effects on fertility rates of various policies based on a very simplified set of assumptions.16 The policies considered are taxes and transfers that lower the direct costs of children, greater availability of part-time employment for women, longer periods of parental leave and greater availability of formal childcare for pre-school children. While these policies are considered as independent levers to affect fertility decisions, longer periods of parental leave to care for children at home and greater availability of formal childcare for pre-school children are considered as alternatives – e.g. in countries where parental leave is already the longest (e.g. Germany) the scope for supporting fertility through greater childcare is assumed to be zero.17 Despite these limitations, Figure 5.6 suggests that different policies can help parents overcome the obstacles that prevent them from attaining their desired fertility, and that the effects of these policies are potentially significant.

96


II.5.


Figure 5.6. Potential impact of various policy reforms on total fertility rates Increase in formal childcare for pre-school children Increase in availability of part-time employment Current level

Lower direct costs of children Increase in length of parental leave

3.5

1.5

1.4

1.4

1.2

1.4

1.7

1.9

1.9

1.9

1.9

1.9

2.0

2.0

2.0

1.7

1.6

1.3

1.2

1.1

1.5

1.3

1.7

1.7

2.1

1.8

2.5

2.3

2.3

2.3

2.4

2.4

1.3

1.9

1.7

2.0

1.5

1.4

re a

2.0

2.1

ga l

2.5

es

3.0

Ko

rtu

Po

Un

ite

d

St

at

ds

d

lan

Ne

th

er

ce

lan

Ire

m

ce

ee

Gr

an

do ng

Fr

d lan

Ki

Fin

d ite Un

da

pa n Ja

ic bl

na Ca

ly

pu

Ita Cz

ec

h

Re

m

ria st Au

k

iu

ar

lg Be

De

nm

en

1.5

ed

an rm

Ge

Sp

ain

0

Sw

0.5

y

1.0

Note: Countries are ranked in increasing order of the total fertility rates that could be achieved as a result of four sets of policies: i) a reduction in the direct costs of children (measured as the difference between the equivalised disposable income of a two-earner couple without children and that of a two-earner couple with two children, where the principal earner earns 67% of the earnings of an APW, and the spouse 33%.); ii) an increase in the availability of part-time employment to the level achieved in the three OECD countries where it is highest (Japan, the Netherlands and the United Kingdom); iii) an increase in the availability of formal childcare (the share of children below 3 years of age attending formal childcare) to the levels of the three countries where it is highest (Denmark, Sweden and the United States); and iv) an increase in the length of leave (both maternity and parental) to the levels of the four countries where it is highest (Germany, France, Spain and Finland). The simulations allow for the possibility of substitution between longer parental leave and greater childcare availability. The combined effect of these policies, e.g. in the case of Japan, is an increase of the total fertility rate from a level of 1.3 in 1999 to around 2.0. Source: D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

The most obvious benefit of higher fertility rates is on the size of the population and labour force. Population projections to 2050, based on these higher levels of fertility, are shown in Figure 5.7 for Canada, Denmark, Germany, Italy, Japan, Sweden and the United States.18 These projections suggest that the set of policies described above could lead to an increase in the population size by 2050 that is quite large for low-fertility countries. For example, in the case of Japan, where national projections anticipate a decline in the size of the population by 2050 to 79% of the level prevailing in the 2000 base-year, a recovery in fertility rates to the level of 2.0 (shown in Figure 5.6) would leave the population in 2050 at 94% of the 2000 level. Higher fertility also has significant effects on the old-age dependency ratio: in the case of Japan, the ratio between the population 65 and over and that of working age could decline by 2050 from around 50% in the national “central” projections to 33% in the alternative. A larger population would in turn affect employment. While the impacts are relatively small when assessed in terms of the employment levels that could be reached by 2050, they are significantly larger when considering the cumulative effects of higher fertility in terms of employment-years.19

7. Conclusions Providing more effective support to families with children is a widely shared priority of active social policies. This chapter has argued that a range of policies will be effective in lowering poverty and securing better development for children, allowing parents to


97

II.5.


Figure 5.7. Potential impact of a recovery in fertility rates on population size and structure Population size, index 2000 = 1.00

Elderly dependency ratio Projection 2050

Projection 2050

Alternative 2050

1.7

1.60 1.18

1.20

1.01

0.94

0.80

50 0.98

1.48 1.16

1.01

0.79

0.40

54

50

60

1.37 0.94

40

1.20

0.93

39

34

33

30

12

22

16

14

20

20 26

26

10 0

47

42

33

30

0.88

Baseline 2000

Alternative 2050

% 70

22

19

27

25

18

y an rm

ly

da

Ge

na Ca

Ita

k ar nm

es at St d

De

en ed

ite

n

Un

Sw

pa Ja

y an rm

ly

da

Ge

na Ca

Ita

k ar nm

es at St d

De

en ed Un

ite

Sw

Ja

pa

n

0

Note: Population size in 2050 is expressed relative to the level prevailing in the base-year (2000 for most countries). Elderly dependency rates are the ratios between the population 65 and over and the population of working age. These scenarios are produced using the “Rural Urban Projection” model of the International Programme Center of the US Bureau of the Census. The baseline projection uses the same assumptions – on fertility, mortality and migration – that underlie the most recent national demographic projections of member countries (“central”, or more likely, scenario); the alternative projections use the total fertility rates shown in Figure 5.6. In the case of Denmark, where the national “central” projection is based on a total fertility rate in 2025 that is very close to that shown in Figure 5.6, the population projection under the alternative scenario is the same as in the baseline. Source: D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, forthcoming, Social, Employment and Migration Working Paper, OECD, Paris.

reconcile their work and family responsibilities, and removing obstacles to realise their fertility intentions. These policies include: ●

Investing in children – including early childhood intervention programmes, particularly well-designed ones that closely involve the family.

●

Boosting maternal employment – including by adjusting tax and benefit systems so as not to discourage second earners.

●

Reconciling family and work responsibilities – through co-ordination across a range of areas such as childcare provision, parental leave and family-friendly workplaces.

●

Creating a framework favourable for raising fertility rates – through policies to share the costs of childrearing more broadly, and to allow young couples to obtain access to affordable housing and a secure footing in labour markets.

Notes 1. For a single person, all income-related benefits are ignored – i.e. the tax rate is only the income tax and employee social security contributions that a person would pay if earning either 33% or 67% of APW – even if in some countries they may be eligible for housing and social assistance benefits. 2. The results compare “baseline” labour force participation rates in 2025 – which incorporate demographic and cohort effects, i.e. an autonomous increase of female participation as older cohorts of women are replaced by younger ones – with an “alternative” scenario incorporating the effects of policy reforms. See Burniaux et al. (2003) for details. 3. Results are based on both labour force survey and household income data. They reflect a number of simplifying assumptions. First, employment rates of mothers, as available in labour force

98


II.5.


surveys, generally refer to women with a child under 15, while household income data define households with children as those including at least one person below the age of 18. Second, information on the distribution of children among different household types was based on 1995 data, adjusted to reflect the actual number of children around 2000. Third, all single-parent household are assumed to be headed by women. Fourth, in two-parent households, all women are assumed to be second-earners (with the exception of the second “alternative scenario” of Figure 5.3). 4. Research in the United Kingdom and the United States suggests that maternal employment in the first (or pre-school) year of a child’s life may reduce the child’s cognitive development and later educational outcomes (Kamerman et al., 2002). For example, Ruhm (2002), using data from the National Longitudinal Survey of Youth for the United States, separates the effects of maternal employment from other characteristics: the results suggest large negative impacts of early maternal employment for the reading and mathematics achievements of children aged 5 and 6, and smaller reductions in the verbal ability of those aged 3 to 4 (partially offset by improvements in the next two years of the child’s life). 5. In the United States, unpaid leave is available under the Family and Medical Leave Act covering industrial and commercial firms with more than 50 workers, as well as private schools and public agencies irrespectively of their number of workers. In Australia, permanent employees who have worked with their employer for at least 12 continuous months have a minimum entitlement of 52 weeks of unpaid leave following the birth or adoption of a child; also, as a result of negotiations, around 38% of female employees report being entitled to some form of paid maternity leave. 6. Statutory paternity leave does not exist in Australia, Austria, Germany, Ireland and the United States, and it is very limited in Italy, Poland and the Slovak Republic. It varies from three days or less in Greece, Luxembourg, Portugal, Spain, and the Netherlands, to five days in Hungary, one week in New Zealand, around two weeks in Sweden, France, Belgium, Denmark and Norway and the United Kingdom, three weeks or more in Finland and Norway, and three months in Iceland. 7. In Sweden, more educated and younger men with well-educated wives and one or two children are most likely to take childcare leave. In the Netherlands, fathers taking up part-time parental leave are generally well-educated and often work in the public sector. 8. Schemes to pay parents to look after their own children at home, without any guarantee of employment or re-employment, may also encourage labour market detachment if they continue over a long period of time. For example, in Finland, a high rate of take-up of the “home-care allowance” available to parents who do not use public childcare services has apparently led to a reduction in female employment rates (Ilmakunnas, 1997). Similar results have been found for France (Fagnani, 1998). 9. There is indeed evidence that mothers who are entitled to extra-statutory leave or pay are more likely to return to work after child-birth, even after controlling for a range of other factors. The same is true of mothers entitled to part-time work. 10. The high figures for the Netherlands reflect its system of partnership between parents, firms and the government, in which firms are encouraged to buy places in privately-run childcare centres, which they then provide to employees at reduced rates. 11. The structure of this table is similar to that of Table 5.2, but uses a higher earnings level for highincome households (100% of the earnings of an average production workers for both the primary and second earner in Table 5.4, as compared to 100% and 67% in Table 5.2). 12. Estimates of the additional costs of children are based on the square root of household size. Based on this scale, a couple with two children bears additional child-related costs of 42% of household income. Estimates of the actual costs of children, as available in a few OECD countries based on expenditure data, suggest that these costs increase with the age of the child and decline in percentage terms with family income. In Australia, for example, the estimated average cost of two children is around 23% of gross household income, ranging between 17%, in the case of two children aged less than 10 in a high-income family, and 75%, in the case of two children aged more than 15 in a low-income family (Percival and Harding, 2002). 13. The model estimated does not allow for country-specific and dynamic effects, for interaction of the various policies, and for the possible endogeneity of fertility and employment decisions. Results should then be interpreted with care. 14. Details on the specification of the variables that have significant coefficients are provided in the notes to Figure 5.6. Other variables that appear with the predicted sign but with a non-significant coefficient include the female employment rate, opportunity costs of children – i.e. the difference


99

II.5.


between the equivalised disposable income of a one-earner couple with two children, with earned income of 100% of an APW, and that of a two-earner couple without children, with earned income at 167% of an APW – and a variable expressing the difference in values with respect to the family and women's role in society between young (15 to 34 years old) men and women. The level of pay during parental leave (the share of previous earnings paid during the parental leave) was also included, but the coefficient was not significant and had the wrong sign. Estimates are based on a simple cross-section model applied to 1999 data for 19 OECD countries. D'Addio and Mira d'Ercole (forthcoming 2005) provide additional details. 15. Evidence on the influence of various policies on fertility that is consistent with the results reported in the text is provided by cross-country studies that use both cross-section (Castes, 2003) and panel data (Gauthier and Hatzius, 1997; Adserà, 2004). Similar results are reported in studies that rely on individual data for specific countries (e.g. Del Boca, 2002; Rønsen, 2004). For a survey see Sleebos (2003). 16. The assumptions underlying Figure 5.6 are those from the simple model described in note 13. 17. For countries with intermediate levels of parental leave and formal childcare, the simulations refer to the policy that has the largest impact on fertility. 18. In these projections, total fertility rates are assumed to converge to the levels shown in Figure 5.6 by 2025, and to remain constant thereafter. Assumptions for mortality and migration remain unchanged at the level used in national projections. 19. For example, in the case of Japan, higher fertility would lead to a level of employment by 2050 that exceeds the baseline level by 15%. However, when the total number of employment-years is considered (i.e. the cumulative sum of the additional employment due to higher fertility rates up to 2050) the gain is of around 115% of the initial stock, implying GPD gains of the same size.

100


PART III

Combating Poverty and Exclusion Among Prime-aged Persons Preventing poverty and exclusion among prime-aged persons is a key goal of social protection. While poverty and exclusion manifest themselves most directly in the form of inadequate income, their principal roots lie in a lack of skills, in a range of personal characteristics and structural factors that create obstacles to the use of skills, and in the reinforcement of these obstacles during periods of benefit receipt. Active social policies to avoid poverty and exclusion among persons of working age aim at helping benefit recipients overcome the obstacles to entering into paid work. They combine early and tailored interventions, greater focus on integration services, mutual obligations on both clients and providers to co-operate in the rehabilitation process, and reforms of benefit systems to remove disincentives to work. Active social policies, however, are not a silver bullet. While substantial progress has been achieved in many OECD countries by putting employment integration at the heart of social policy, not everyone can be expected to participate in the labour market, and getting people into jobs will be insufficient to avoid exclusion if people do not keep the jobs, if the wages that the jobs pay are not high enough to escape poverty, or if they offer little prospects of skills development and career progression. Hence, policies aimed at integration into employment must be complemented by measures to “make-work-pay” and to assure adequate income for those for whom integration or re-integration into the labour market is more difficult to achieve, in particular single parents and people with disabilities. A consensus is also emerging in a number of countries on applying an approach based on the principle of “mutual obligations” – that the commitment and effort society makes to assist beneficiaries requires that they in turn do their best to take steps to find work or engage in other productive activities. The pay-off of these policies in the future is higher income and selfsufficiency among traditional client groups, as well as higher employment rates for the economy as a whole. Reaping these benefits will require: ● ● ● ●

Completing the welfare-to-work agenda. Making progress with welfare-in-work. Moving beyond work as the only focus of policies. Strengthening the effectiveness of programmes targeted to persons for whom work is less feasible.



PART III

Chapter 6

Poverty and Exclusion in Prime-age: Achievements and Challenges


103

III.6.

POVERTY AND EXCLUSION IN PRIME-AGE: ACHIEVEMENTS AND CHALLENGES

1. Introduction Preventing poverty and exclusion among prime-aged persons has always been among the goals of social protection. However, the attention paid to this goal, the way poverty and exclusion are understood and conceptualised, and the range of policies used to reduce them have undergone radical changes over the past few years. This chapter reviews the progress made and the set of interventions – mainly in the form of social assistance and disability programmes – that OECD countries have put in place to address risks of poverty and exclusion among the adult population. It then identifies, as areas of continued policy concern, the large number of persons of working age who depend entirely on public benefits for their daily living, the shifts in poverty risks towards groups with easily identified characteristics, and the growing awareness about the geographical concentration of poverty.

2. Policies and achievements The progress achieved by OECD countries in the fight against poverty and exclusion among prime-aged persons largely mirrors that described in Chapter 1 for the entire population: extreme forms of destitution among prime-aged persons have declined in all countries, as education, health and labour-market conditions have improved. Social policies have contributed to these outcomes by supporting those exposed to a variety of risks, including mental and physical impairment, behavioural problems, lack of skills, and an unfavourable family background. Risks of relative income poverty among persons of working age are generally lower than among children and the elderly, and the extent of this advantage has not changed much over time (Figure 6.1).

Figure 6.1. Relative poverty rates among the population of working age Working-age population

% 25

Total population

20 15

Mid-1990s

2000

10 5

Au

st ra l Au ia st r Be ia lg i Cz C um ec an h Re ada pu De blic nm a Fin rk lan Fr d a Ge nce rm an Gr y ee Hu ce ng a Ire ry lan d Ita ly Lu Ja xe pa m n bo u M rg Ne ex th ico Ne erla w nd Ze s ala No nd rw a Po y la Po nd rtu ga Sp l S ain Sw wed itz en er lan Un ite Tu d d rk K e Un ing y ite do d m St OE ates CD -2 3

0

Note: Data for Canada and Sweden are adjusted for a break in series. Data for Germany refer to old länder. Exact years are those noted in Figure 1.2. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris

104


III.6.


Because earnings represent the largest component of the disposable income of the working-age population, having a job plays a critical role in shaping their risk of poverty and exclusion. Over the second half of the 1990s, the improved labour market performance has contributed to reversing trends towards a widening in the distribution of market income in most OECD countries, increasing the share of earnings that goes to the lower quintile of the working-age population (Förster and Mira d’Ercole, 2005). The cross-country relationship between employment levels and poverty rates suggests that countries with lower non-employment rates (in particular for women) tend to experience lower poverty rates at the level of market income among the population of working age (Figure 6.2, Panel A). The relationship is stronger when account is taken of the way in which employment is distributed among households, as countries with similar levels of employment (e.g. Japan and Australia, where employment-to-population ratios are slightly above 70%) may show large differences in the share of the working-age population living in households where no adult works (ranging from less than 3% in Japan to more than 13% in Australia).

Figure 6.2. Relative poverty rates among the working-age population, non-employment rates and joblessness A. Poverty rates at the level of market income Inactivity among individuals

Jobless households

Poverty rate (%) 40

Poverty rate (%) 40 POL

POL 30

30 NZL GBR AUS DEU

USA CZE SWE DNK IRL MEX CAN NOR JPN NLD FIN PRT CHE

20

10

0

FRA GRC

15

25

35

GRC

ITA

HUN

45 55 Non-employment rates

20 JPN 10

0

ITA

FRA

HUN GBR DEU DNK USA NZL CAN CZE AUS IRL NOR PRTSWE FIN NLD CHE

MEX

0

5 10 15 20 Share of the population in jobless households

B. Poverty rates at the level of disposable income Inactivity among individuals Poverty rate (%) 20 MEX

16 12 8 4 0

USA

JPN

25

IRL

35

MEX

16

GRC ITA CAN PRT NZL DEU POL GBR AUS AUT HUN FRA NOR CHE FIN NLD DNK SWE LUX CZE 15

Jobless households Poverty rate (%) 20

45 55 Non-employment rates

JPN USA

12

CAN GRC PRT

8 CHE

4 0

SWE

IRL

AUT FIN

ITA NZL

AUS DEU GBR POL NOR FRA NLD HUN

LUX DNK 0

CZE

5 10 15 20 Share of the population in jobless households

Note: Poverty rates are measured with respect to a threshold set at 50% of median equivalised disposable income and refer to persons aged 18 to 64. Non-employment rates refer to persons aged 15 to 64, from labour force surveys. Joblessness is the share of the total population living in households with a working-age head and where no one works. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.


105

III.6.


The relationship between employment and poverty rates is however weaker when the latter are measured in terms of disposable income (Figure 6.2, Panel B). This is because, beyond the influence of labour markets, government policy in all OECD countries helps to achieve low poverty among prime-age persons through a variety of interventions. The relationship between government policies and poverty outcomes is striking: across countries, poverty rates among the working-age population are lowest where (non-health) social spending on the working-age population is highest (Figure 6.3). Within each country, the combined effect of the tax and benefit systems is to lift out of relative income poverty more than half of the population at risk, on average (Figure 6.4). This effect, which ranges between around onefourth of those below the poverty threshold before taxes and transfers in the United States and more than two-thirds in Denmark, declined over the second half of the 1990s in most OECD countries, most often reflecting an increase in real benefits lagging that of median disposable income rather than cuts in the value of benefits in real terms.

Figure 6.3. Relative poverty among the working-age population and social spending in 2000 Poverty rate (%) 20 MEX 15

JPN

USA

TUR

ITA

GRC

10

PRT

OECD-25 GBR CAN DEU NZL

IRL

POL AUS HUN AUT

CHE

5

CZE 0

0

5

LUX

FRA NOR NLD

FIN

SWE

DNK

10 15 Non-health public social spending towards working-age population (% GDP)

Note: The poverty rate is the percentage of persons of working age with equivalised disposable income less than 50% of the median. Social spending refers to public non-health social spending on the population of working age (i.e. total spending less spending for health, old-age and survivors’ benefits), as a share of GDP. Source: OECD social expenditure database; calculations from OECD questionnaire on income distribution and poverty.

Figure 6.4. Effects of taxes and transfers in reducing relative poverty among the population of working age Effect of taxes and transfers in reducing poverty

% 40

30

Poverty rate, disposable income

Mid-1990s 2000

20

10

m Cz Ca na ec h d Re a pu bl De ic nm ar k Fin lan d Fr an Ge ce rm an y Gr ee ce Hu ng ar y Ire lan d Ita ly Ja pa n M e Ne xic o th e Ne rlan d w Ze s ala n No d rw ay Po lan Po d rtu ga Sw l e d Sw e Un itz n ite erla d Ki nd Un ngd ite om d St a OE tes CD -1 7

iu

tra

Be

Au s

lg

lia

0

Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.

106


III.6.


Social assistance and disability programmes are particularly important means for reducing the poverty and exclusion of prime-aged persons. All combined, gross spending on these programmes (as well as housing programmes) accounts for around 4% of GDP. Services (such as reintegration and labour market training) represent slightly less than half of this public spending, with the remainder representing cash-transfers. In terms of individual spending categories: ●

Gross public spending on disability benefits represented 2.5% of GDP on average in 2001, and has remained broadly stable over the past two decades (Figure 6.5). There are, however, large differences across OECD countries in both spending levels – ranging from more than 4% in the Netherlands, Norway and Sweden to less than 1% in Canada, Japan, Korea, Mexico and Turkey – and trends – with large declines since the mid-1990s in the Netherlands, Belgium, Italy and Ireland, and further increases in Switzerland and Australia.

Figure 6.5. Gross public spending on disability benefits, as a percentage of GDP OECD Canada

Germany

United States

Japan

OECD

Turkey

3.0

Sweden

Netherlands Italy

Poland

France

United Kingdom

8.0 7.0

2.5

6.0 2.0

5.0 4.0

1.5

3.0

1.0

2.0 0.5 0

1.0 1980

1985

1990

1995

2001

0

1980

1985

1990

1995

2001

Note: The “OECD average” is computed with respect to the 21 OECD countries with more extended time-series. Source: OECD Social Expenditure database.

●

Gross public spending on social assistance benefits represents, on average, half a point of GDP in 2003 – but higher levels in the Netherlands, Switzerland and Nordic countries – with a reduction of around one-fourth point since the peak reached in 1993 (Figure 6.6). Stronger declines have been recorded in some countries (Canada, the Netherlands, the United Kingdom and the United States) while little sign of reduction is evident in others (Australia, Austria and Ireland).

In addition to these programmes, gross public spending on housing programmes also plays an important role in several OECD countries (e.g. France and the United Kingdom) to support persons at greater risk of poverty and exclusion, and accounts on average for 0.4% of GDP. Conversely, public spending on unemployment (1.1% of GDP on average) and active labour-market programmes (0.7% of GDP on average) have been used to respond to the needs of those persons who have maintained a stronger degree of attachment to the labour market. Over time, the boundaries between programmes targeted towards persons with greater attachment to the labour market and those targeted to persons at greater risk of poverty and exclusion have become more blurred as a result of two sets of factors. First, as remaining at the margin of the labour market is hardly a satisfactory outcome for both clients and their families, many social programmes targeted to persons traditionally


107

III.6.


Figure 6.6. Gross public spending on social assistance, as a percentage of GDP OECD

Germany

Finland

OECD

France

Japan

Turkey

Sweden

0.8

1.6

0.7

1.4

0.6

1.2

0.5

1.0

0.4

0.8

0.3

0.6

0.2

0.4

0.1

0.2

0

1980

1985

1990

1995

2001

0

1980

1985

Netherlands

Denmark

United States

1990

United Kingdom

1995

2001

Note: The “OECD average” is computed with respect to the 21 OECD countries with more extended time-series. Source: OECD Social Expenditure database.

deemed as being “less employable” have aimed to increase their degree of attachment to the labour market so as to increase their well-being (welfare-to-work). Second, as having a job does not always provide shelter from the risk of poverty, labour-market and social programmes specifically targeted at the working poor (welfare-in-work) have been developed.

3. The challenges today Labour market integration has so far benefited only small categories of benefit recipients; tackling the larger numbers of single parents and people with disabilities will inevitably demand greater and more creative efforts. A better understanding is also being gained of how risk factors combine to compound disadvantage and interact in geographic clusters of poverty, where entire neighbourhoods suffer social exclusion – a situation that cries out for greater steps to eliminate poverty by co-ordinating policy and joining-up efforts by different institutions.

3.1. A significant minority of the population is at risk of social exclusion The notion of social exclusion conveys the idea that some groups of the population are ill-equipped to take advantage of changes in the structure of labour markets and economic conditions. In some cases, social exclusion may result from a single severe disability. More commonly, however, it is the result of multiple barriers, such as low education, poor skills, and ethnic and household characteristics. Because of these multiple barriers, some primeage individuals risk being left on the margins of mainstream society, where they could experience a decline in their social standing and conditions. While social exclusion is often the result of multiple factors, each of which compounds disadvantage, exclusion from the labour market is of special salience. For example, while the relative equivalised household income of disabled persons is not much lower than that of other people (Figure 6.7, Panel A), this is no longer the case when the disabled persons are also out-of-work: the equivalised disposable income of disabled persons without a job is, on average, only half of that of non-disabled people (Panel B). The same applies among single parents without work, a group that accounts for a small share of the total population in all OECD countries but for more than 3% in Australia and the

108


III.6.


Figure 6.7. Relative income of non-employed disabled persons and single parents, late 1990s A. Income of disabled relative to non-disabled persons

1.0

25

7 -1

R

A

OE

CD

KO

US

ES

GB

IT A

A FR

R

PO

NO

N

BE

T

CA

AU

DE

CH

DN

NL

SW

P

0

PR T

0

R

5

L

0.2

L

10

U

0.4

E

15

K

0.6

D

20

E

0.8

B. Income of non-employed disabled relative to non-disabled persons

3

0

0

4 -1 OE

CD

M

US

ES

A FR

NO

T

GB

AU

L

DE

BE

K DN

NL

E SW

EX

0.2

A

6

PR T

0.4

P

9

R

0.6

IT A

12

R

0.8

U

15

D

1.0

C. Income of single parents relative to all two-parent households

2

0

0

5 -2 CD OE

NZ

IR

NL D AU S GB R CA N PO L US A JP N

CZ

FI

NO

DN

LU

T

AU

GR

L

0.2

L

4

E

0.4

N

6

R DE U SW E FR A

0.6

K

8

X

0.8

IT A TU R

10

C CH E PR T M EX HU N

1.0

D. Income of non-employed single parents relative to all two-parent households

1

0

0

CD -2 OE

NZ

LU X

FR

AU

FI

NL D SW E NO R M EX JP N

K

DN

C GR

4

0.2

IR L PR T HU N US A

2

IT A PO L CA N

0.4

L

3

A CZ E GB R AU S

0.6

T

4

N

0.8

DE U

5

TU R

1.0

Note: Countries are ranked in decreasing order of relative household disposable income (left-hand axis). In panels a and b, the equivalised disposable income of disabled persons (all, and those without jobs) is expressed relative to that of non-disabled persons. In panels c and d, the equivalised disposable income of single parents (all, and those without jobs) is expressed relative to that of two-parent households. Diamonds in each panel refer to the population shares of people in each category (i.e. disabled persons, non-employed disabled persons, persons in single-parent households, and in households headed by a non-employed single parent, right-hand axis). Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris; and Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.


109

III.6.


United Kingdom: the equivalised disposable income of single-parent households is close to 70% of that of two-parent families, on average (Panel C), but this falls to around 40% in the case of single parents without work (although with much variation across countries, Panel D). Prolonged lack of a job is a primary driver of social exclusion for prime-aged individuals. Its consequences are especially dire because many of these prime-aged persons have important family responsibilities. Research on 11 European countries in 1996 shows that unemployed persons are, on average, close to three times more likely to fall into (relative) poverty than the working-age population at large, and more than two times more likely to experience different forms of material deprivation (Gallie and Paugam, 2004). Also, while joblessness does not systematically reduce the extent of social interactions (in terms of likelihood of living alone and of frequency of contacts with friends and relatives), it does change the nature of people’s social networks, increasing the extent to which unemployed persons socialise with other unemployed and raising the degree of network segregation. The likelihood of inactivity increases exponentially among groups exposed not just to joblessness but to multiple disadvantages as well, such as being disabled, being a migrant and living in problem areas.1 Poverty is of special concern for policy when it extends over prolonged periods of time. Data on the experience of individuals and families over three consecutive years (around the mid-1990s) suggest that, on average, about 5% of the population not previously poor fall below the poverty line each year, and almost 50% of those under the poverty line will rise above it in the same year. However, and of greater significance for the range of policies discussed here, a large proportion of the poor in a given year have incomes just above the poverty line even in the best of times and experience repeated spells of poverty over a multi-year period (OECD, 2001a).

3.2. Too many people are trapped in social assistance and other benefit programmes The experience of poverty and exclusion is reflected in the high number of persons of working-age who are dependent for their daily living on different social programmes (Figure 6.8). This proportion is significant in most OECD countries, and has remained stubbornly high despite falling some in recent years. In the English-speaking countries, at least, lone parenthood and benefit receipt remain closely entwined, while in several countries the number of recipients of disability benefits has overtaken that of unemployment benefits. Trends in the number of working-age recipients of various income replacing benefits differ among the different benefit categories.

110

●

Recipiency of disability benefits across the 16 countries covered increased strongly until the mid-1990s, stabilising thereafter except in Australia, Belgium and France. In addition, dependency rates for sickness benefits were broadly unchanged over this period. The declines recorded in some OECD countries were in most cases accompanied by improved incentives for employers to monitor sickness absences (e.g. by obliging employers to pay benefits for the first weeks of leave).

●

Recipiency of unemployment benefits increased sharply until the mid-1990s, followed in many countries by strong declines. When comparing years corresponding to cyclical peaks, dependency on unemployment benefits stabilised on average; and in four of the 16 countries shown dependency rates for unemployment benefits were lower in 1999 than in 1980.


III.6.


Figure 6.8. Full-time recipients of various income-replacing benefits among the population of working age, 1980-1999 Disability benefits Denmark Germany % 8

Australia

Canada Austria

Ireland

France Belgium % 8

Netherlands United States Sweden United Kingdom Spain Japan Slovak Republic New Zealand

6

6

4

4

2

2

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

Unemployment benefits Denmark Germany %

Australia

Canada Austria

Ireland

France Belgium %


12

12

10

10

8

8

6

6

4

4

2

2

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998

Lone-parent and social assistance benefits Denmark Germany % 10

Australia

Canada Austria

Ireland

France Belgium % 10

8

8

6

6

4

4

2

2


0 0 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 Note: Number of benefit recipients are expressed in full-time equivalents (benefit years) to account for part-time receipt of benefits, benefits received from multiple schemes, and benefits received by families rather than individuals. Percentage of the working-age population. Source: NEI-SZW database, partially revised and augmented by OECD. See OECD (2003b), “Benefits and Employment, Friends or Foe? Interactions Between Passive and Active Social Programmes”, Employment Outlook, OECD, Paris.


111

III.6.


●

Growth in recipiency rates for lone-parent and social assistance benefits came to a halt in most countries after 1997, with strong declines in Canada, the Netherlands, the United Kingdom and the United States, but little sign of abatement in Australia, Austria and Ireland.

Many working-age persons depend on benefit not just for a period of time, but more or less permanently, as the probability of exit from the benefit roll declines rapidly at higher durations of receipt. For example, in the case of disability benefits, exit rates are typically less than 1% of the stock of beneficiaries (see OECD 2003a), and this holds both in countries with a strong focus on avoiding inflows into these programmes through vocational rehabilitation and training and in countries that have given priority to providing better economic incentives to get benefit recipients off the rolls (Figure 6.9). The “quasi-permanent” nature of benefit recipiency weakens the rationale of insurance programmes that aim to protect individuals against random and temporary contingencies. First, the most successful groups will find that the hedging strategies that they have adopted (in the labour market, family choices, savings decisions) provide adequate cover against the risks that they face, making them less willing to contribute to social programmes whose benefits accrue to others. Second, coverage of these insurance programmes is limited to a subset of the population, often excluding those in greatest need. These elements suggest that persons of working age who are continuously dependent on benefits are badly served by existing systems; at best, these systems are not generous enough to provide adequate income; at worst, they make it more difficult for recipients to escape poverty and exclusion.

Figure 6.9. Outflow rates from disability benefits 1995

1999

Percentages 10 8 6 4 2

re a Ko

wa y No r

Au

st ra

lia

n ed e Sw

ad a Ca n

ga l rtu Po

a st ri

St d ite Un

Au

s at e

an y rm Ge

ds lan he r Ne t

Un ite

d

Ki

ng

do m

0

Note: Annual rates of outflows from disability benefits in percentage of the outstanding stock. Persons exiting from disability rolls may move to employment, shift to other benefit programmes, or be in a condition where they neither work nor receive benefits. Countries are ranked in decreasing order of the 1999 rate. Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris.

112


III.6.


3.3. Poverty and exclusion are geographically concentrated The causes and consequences of poverty and exclusion cluster in particular areas, compounding each other. Evidence about the geographical concentration of disadvantage is suggestive of common patterns across the OECD area. In the United Kingdom in 2000, the 25% most deprived districts accounted for 57% of all poor individuals claiming income support and Job Seeker’s Allowance (Tunstall and Lupton, 2003). In France in 1999, persons living in “sensitive urban areas” – neighbourhoods characterised by degraded housing and social conditions, which account for around 8% of the French metropolitan population – were, in comparison with residents of the metropolitan area surrounding these sensitive neighbourhoods, three times as likely to be living in social housing, twice as likely to be headed by immigrants, 1.6 times as likely to be unemployed and 1.9 times as likely to rely on welfare (Fitoussi et al., 2004). In the United States in 2000, persons living in extreme poverty neighbourhoods (census tracts with poverty rates of 40% or more) were, in comparison with all metropolitan residents, 3.4 times as likely to be receiving public assistance, 2.3 times as likely to lack a high-school degree, and 2.6 times as likely to be unemployed (Kingsley and Pettit, 2003). Research on selected US cities also shows that deprived neighbourhoods have worse health (e.g. higher frequency of child obesity, tuberculosis, use of emergency health services) and other negative social outcomes (violent crime and violence-related school suspension, Pettit et al., 2003). Despite a lower geographical concentration of poverty in some countries during the 1990s (e.g. Australia, the United States), these differences in social disadvantage have remained disturbingly high in most countries. The significance of these patterns of geographical concentration of disadvantage goes beyond the lower well-being of the affected communities at a point in time. The intrinsic characteristics of poor neighbourhoods (in terms of location, housing, and job opportunities) worsen over time, as better-off families move out of them and those worseoff move in, lowering the quality of the environment, services and facilities, weakening community ties and increasing crime. Geographical concentration of disadvantage also implies a loss of productive potential – and a diversion of productive resources to treat these social problems – for the country as a whole. The strength of the forces contributing to urban segregation may lead to “spatial hysteresis” that increases the persistence of mass unemployment even after its root causes have disappeared (Fitoussi et al., 2004). At the root of this spatial persistence is the physical distance between the areas of residence and those where jobs are available, the higher commuting costs faced by their residents, the separation of these individuals from the information channels that are required to gain employment, the discrimination by employers in their hiring and personnel practices, the development of underground activities in these neighbourhoods, and low job creation in these areas.

4. Conclusions While prime-aged persons are less exposed to risks of income-poverty than children and the elderly, risks of exclusion from mainstream society remain too common for a significant minority. These risks are most often a function of a failure to successfully integrate into the labour markets, because of the combined effect of personal and structural characteristics. While a range of social programmes protect individuals from the most severe consequences of poverty and exclusion, many people are trapped in benefit programmes more or less permanently, as the probability of exit from caseloads declines EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

113

III.6.


rapidly with the duration of the benefit spell. Poverty also clusters in particular geographical areas, compounding risks of persistence of social exclusion even after its root causes have disappeared. This combination of factors calls for innovative policies tailored to the specific conditions of individuals and of the communities where they live, which are described in Chapter 7.

Note 1. In the United Kingdom, for example, more than 50% of those with three or more disadvantages (being a lone parent or a single person; having low qualifications and skills; having a physical impairment; being over 50; being from an ethnic minority group; living in a region of high unemployment) are non-employed, as compared to 3% of those with only one of these disadvantages. The likelihood of non-employment is above 90% when persons cumulate all six of these disadvantages (Social Exclusion Unit, 2004).

114



PART III

Chapter 7

More Effective Policies to Reduce Poverty and Exclusion


115

III.7.

MORE EFFECTIVE POLICIES TO REDUCE POVERTY AND EXCLUSION

1. Introduction In all OECD countries, policies to reduce the extent of poverty and exclusion rest on three different pillars: ●

First, delivering cash assistance to families in need.

●

Second, providing services to address those special problems that are responsible for the difficulties faced by some individuals and their families in entering mainstream society.

●

Third, assisting labour-market re-integration, so as to allow families relying on benefits to regain financial independence.

Each of these pillars is essential to avoid marginalisation, yet their underlying goals may sometimes conflict with each other. Poorly designed programmes may leave recipients with little incentive to move into employment, generating poverty and poverty traps, while – conversely – inadequate benefit levels may worsen the material conditions of poor households. An imbalance between these three policy pillars may end up reducing exits and increasing benefit rolls. Reforms pursued in several OECD countries over the 1990s have addressed this imbalance by giving greater emphasis to reducing benefit dependency and welfare traps and increasing the share of persons holding jobs. These policies give priority to creating incentives to take up jobs and to developing the skills and earnings potential of persons relying on benefits. While all three pillars of poverty reduction policy are vital, labourmarket integration is key because it is the dynamic factor, the one capable of changing individuals’ conditions and enabling them to become self-sufficient over the long term. But the foundation of a country’s poverty reduction strategy will not be solid if policies to take up jobs are not combined with policies aimed at “making-work-pay” and at strengthening the effectiveness of social programmes targeted to persons for whom work is less feasible.

2. Labour market integration is the key Integration into the labour market for persons at higher risk of poverty and exclusion seemed an impossibly optimistic goal just a few years ago. Today, nearly all OECD countries have employment integration at the very heart of their policies for tackling poverty and exclusion. This is a big change.

2.1. Policies towards social-assistance clients Active policies were introduced first for unemployed job-seekers, where they have involved a combination of more frequent and more personalised interventions of the public employment service during the individual’s unemployment spell, stricter eligibility criteria, benefit sanctions, and stronger obligations to participate in labour market programmes (OECD, 2003b). The same criteria are today finding their way into general

116


III.7. MORE EFFECTIVE POLICIES TO REDUCE POVERTY AND EXCLUSION

social assistance programmes targeted to persons without easily-identifiable obstacles to work. Active policies towards social assistance clients combine several features: ●

Tailored interventions of public agencies during the individual’s benefit spell, aimed at increasing the frequency of contact between benefit recipients and the agencies responsible for benefit payment and for provision of integration services. While practices have traditionally differed, in several countries, after initial contacts to establish the person’s right to benefits, the responsible agency intervened very little. To avoid the risk of insufficient and ineffective efforts by individuals to find suitable jobs, reforms in several OECD countries have introduced obligations for the individual to maintain regular contact with the responsible agencies and to engage in other jobsearch activities. Scheduled interventions by the responsible agencies can take a variety of forms and occur at different times of the benefit spell. More frequent interventions by the agencies have also been achieved through “individual action plans”, which usually involve an intensive interview at a defined month of the spell and then follow-up activities. These “individual action plans” may also involve profiling the participants according to characteristics such as age, skills and length of service, so as to better anticipate their service needs and employability.

●

Stricter eligibility criteria and benefit sanctions have also featured prominently in active policies. Changes in eligibility criteria have differed across countries. Some countries have shortened the maximum duration for receiving benefits or introduced benefits that decrease with the length of the benefit spell. Several countries have also introduced an obligation to accept any offers of “suitable work”, or broadened its definition. With respect to lone parents, reforms in several countries have made assistance conditional on availability for work: such work tests exist in 17 OECD countries and are generally based on the age of the youngest child1 and, more rarely, on the availability of child care (Denmark). Failure to meet these criteria is often accompanied by sanctions, such as reductions in the amount of benefits or their outright termination. In the Netherlands and the United Kingdom, an increase in the incidence of benefit sanctions occurs at some stages of the activation strategy, even if – in general – sanctions are rarely applied to the letter (Gilbert, 2002).

●

Obligations to participate in labour market programmes have been established in some countries. In Sweden and Denmark, where rates of participation in such programmes are high, reforms have aimed at improving programme quality by strengthening their training and educational content.2

●

Changes in the nature of interventions and greater co-ordination in institutional arrangements. In many countries, emphasis on activation of benefit recipients has led to greater provision of integration services to accompany benefit payments. The range of services provided has progressively extended, from advice and counselling services to more tangible services, such as day-care and job-training.

While there is still only limited rigorous evidence about the results of these policies, some of the countries that have implemented them most fully have experienced large drops in caseloads. The number of people receiving certain key welfare benefits in the United Kingdom and the United States fell by more than half from their peak levels in the mid-1990s; in Canada and the Netherlands, levels fell by one-third or more, and in Finland by one-fourth (Figure 7.1). In the United States, a majority of leavers are working, often fulltime, and at wage rates that are close to those of similar groups in the labour force (Loprest,


117

III.7.


Figure 7.1. Changes in welfare caseloads following welfare reforms % 8 7

Unemployment

% 8

Finlanda

7

6

6

5

5

4

4

3

3

2

2

1

1

0 1970 % 8 7

74

78

82

Unemployment (RWW)

86

90

94

98

02

Unemployment and social assistance (RWW + ABW)

Netherlandsc

7

6 5

d)

e)

f)

90

94

98

02

Unemployment and lone parent assistance

United Kingdomd Jobseekers allowance

Stricter benefit regime

2 1

0 1970

c)

86

3

1

a) b)

82

5

RWW abolished

2

6

78

Unemployment

4

3

7

74

6

4

% 8

Unemployment and social assistance (ASS + RMI)

Franceb

0 1970 % 8

Activation measures: sanctions and suitable work requirements

Unemployment

74

78

82

86

90

94

98

Domestic purposes benefit New Zealande Lone-parent part-time work requirements

02

0 1970 % 8 7

74

78

82

86

90

94

98

02

AFDC/TANF United Statesf

6

5

5

4

4

3

3

2

2

1

1

EITC and activation measures

TANF reform

0 0 1970 74 78 82 86 90 94 98 02 1970 74 78 82 86 90 94 98 02 State unemployment assistance, created in 1971, and labour market support. Allocation de Solidarité Spécifique (ASS), created in 1984, and Revenu Minimum d’Insertion (RMI) created in 1988. For RMI, beneficiaries in metropolitan France (excluding the “DOM” overseas territories) on 31st December. Rijksgroepsregeling werkloze werknemers (RWW, unemployment assistance) was created in 1963 and abolished in 1996, with transfer of beneficiaries to the general social assistance benefit (Algemene bijstandswet, ABW, National Assistance Act). Unemployment assistance benefit named National Assistance (from 1949), Supplementary Benefit (from 1966), Income Support (from 1988) and Job Seekers Allowance (income-based) from 1996. Data relate to a specific week of the year (a week in May as from 1987) and do not include people also receiving an unemployment insurance benefit. Domestic Purposes Benefit, introduced in 1973, June figures from 1990 onwards. A relatively small number of beneficiaries of Widows’ Benefit who are lone parents are not included. Unemployment assistance (not shown) also grew sharply over these years. Data refer to adult beneficiaries (i.e. not including caretaker recipients) of Aid to Families with Dependent Children (AFDC) and Temporary Assistance for Needy Families (TANF). EITC stands for the Earned Income Tax Credit.

Source: See OECD (2003), Employment Outlook, OECD, Paris for detailed sources.

118



2002). In all these countries, the declines followed reforms that increased the emphasis on the labour-market integration of benefit recipients. These reductions are all the more significant when compared to previous claims about the existence of ratchet effects in the dynamics of welfare caseloads, i.e. the notion that these caseloads would naturally increase during recessions while failing to decline to previous levels during recoveries. In a number of countries, the ratchet is no more. Not all those who exit welfare caseloads, however, move to employment. Evaluations in the United States suggest that lower participation among eligible families accounted for about one-third of the decline in the welfare caseload (and for more than half of the decline in the food stamp caseload) recorded in the 1995-2000 period (Zedlewski et al., 2002). Similarly, random-assignment evaluations (at the 11 sites of the US National Evaluation of Welfare-to-Work Strategies) suggest that participation in programmes aimed at humancapital development and job-search assistance had an impact in terms of higher employment that was between 60 and 70% of the impact on reducing welfare recipiency. Evidence about the effects of these policies on the employment rates of lone mothers shows that these largely mirror those for other categories of social assistance recipients. In the United States, the (administrative-based) number of single mothers receiving ADFC/ TANF fell by 2.4 million (around two-thirds of the 1993 stock) from 1993 to 2001, while the (survey-based) number of single mothers employed was up by 1.4 million in the same period (Figure 7.2).

Figure 7.2. Single mothers receiving welfare benefits and in employment, United States Neither earnings nor welfare income (right-hand axis) Millions 10

ADFC/TANF beneficiaries (left-hand axis)

Employed (left-hand axis)

Percentage of single mothers 16-45 10

8

8

6

6

4

4

2

2

0 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2001 2002 Note: Data on the number of single mothers receiving AFDC/TANF are based on administrative records. Data on single mothers employed are based on the Current Population Survey. Data on the percentage of single mothers aged 16 to 45, excluding disabled persons and students, with non-zero annual income, who report neither work nor labour income, are based on the March supplement to the Current Population Survey. 0

Source: OECD (2003b), “Benefits and Employment, Friends or Foe? Interactions Between Passive and Active Social Programmes”, Employment Outlook, OECD, Paris.

Figure 7.3 illustrates the potential impact of higher employment among lone mothers on reducing their poverty rates. The simulation is based on very simplified assumptions: it simply assumes an increase in the employment rate of single parents (as measured in household income surveys of member countries) to the level prevailing in the five OECD countries where this rate is currently highest (i.e. a level of 85%). This higher employment rate of lone parents increases the share of the population living in households headed by a lone parent with a job, and reduces the share living in one-adult jobless households. The effect is to lower the poverty rates of all lone parents, by close to 20% among the countries shown in Figure 7.3.3 EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

119

III.7.


Figure 7.3. Potential impact of higher employment among lone parents on relative poverty rates of single-parent households Poverty rate, baseline

% 80

Poverty rate, when employment of lone mothers increases

60

40

20

Ita ly

Ja

pa n M e Ne xic th o e Ne rlan d w Ze s ala nd No rw ay Po lan Po d rtu g Sw al ed en Un ite Tur ke d y Ki Un ngd ite om d St a OE tes CD -2 3

Cz

ec

h

Au

Au

st ra

lia st ria

Ca na Re da pu b De lic nm ar k Fin lan d Fr an Ge ce rm an y Gr ee Hu ce ng ar y Ire lan d

0

Note: The figure shows, for each country, the relative (50% of median household disposable income) poverty rate for persons living in households with a single parent, and the level that would prevail if the employment rate of single parents were as high as in the five OECD countries where it is currently highest (Austria, Denmark, Japan, Sweden and the United States). The calculations assume that the number of children in each single-parent household remains unchanged from its 1995 level. Employment rates of single parents are kept unchanged in the five countries where they are currently highest. Source: OECD questionnaire on income distribution and poverty.

While labour market integration is the key to lower poverty for persons with limited obstacles to work, changes in the state of the economy may result in new challenges. During a recession, new welfare leavers will face weaker prospects of finding jobs, while former welfare recipients with weaker skills and work experience may lose their jobs first and fail to qualify for unemployment insurance because of shorter contribution periods. While these difficulties increase with the length of the cyclical slowdown, even a shortlived recession might translate into limited employment options for welfare recipients for years to come, as employment typically lags economic output, and welfare rolls lag employment (Holzman, 2002). In the longer-term, those who remain on welfare programmes may face greater personal barriers to employment, making job integration more difficult to achieve. Because of these features, even reforms that succeed in improving the conditions of a majority of clients who move to jobs may worsen conditions for people for whom employment integration is more difficult to achieve. This accentuates the importance of a co-ordinated set of policies to maximise the inroads made in nonrecessionary times among groups facing the greatest obstacles (Box 7.1). At the same time, reducing risks that prolonged dependence on public benefits might increase the probability of poor educational and employment outcomes of the next generation – the children of jobless parents – requires encouraging those who have repeatedly failed to integrate into the labour market to participate in other useful activities – such as volunteering in the not-for-profit sector, or caring for other family members – that may help to break the cycle of welfare dependency.

2.2. Policies towards people with disabilities The same principles that have been guiding reforms of social assistance programmes are finding their ways into disability programmes. Policies with respect to disabled persons have typically given priority to compensating rather than activating clients. In recent years,

120



Box 7.1. Increasing employment among lone parents: the United Kingdom experience Despite a sharp rise in employment of married or cohabiting mothers in the United Kingdom over the past thirty years, the employment rate of lone mothers, at just under 40% in the early 1990s, was much lower than in the late 1970s and 25 points below that of married mothers. Partly as a result, child poverty was also higher in the United Kingdom than in most other OECD countries, and the proportion of children living in low-income households had more than doubled in 20 years. In 1997 the government initiated a series of reforms aimed at increasing employment among lone parents as part of its strategy to reduce child poverty. These reforms included a target of getting 70% of lone parents into employment by 2010. A first element of this strategy is the New Deal for Lone Parents, launched nationally in April 1998, aimed at encouraging job search and easing the transition into work among lone parents who volunteer to participate. From April 2002, Job Centre Plus has further increased the extent of contact with lone parents, requiring them to attend an interview. The government also introduced outreach services to reach those living in isolated communities: these include advisers working with local organisations and community groups to inform lone parents about the help available to move into work. A further element of this strategy was the Working Families Tax Credit (WFTC), which replaced the Family Credit in October 1999. The WFTC raised both the maximum benefit payable to working lone parents, and reduced the rate at which the tax credit was withdrawn. The generosity of support for childcare was also increased: funding for the National Childcare Strategy was trebled and the childcare tax credit element of WFTC was increased (to cover 70% of childcare costs, up to GBP 135 per week for a family with one child; and up to GBP 200, for a family with two or more children). Overall, the Government increased financial support for children through tax credits, child- and other benefits by GBP 10.4 billion since 1997, a real term rise of 72%. Gregg and Harkness (2003) estimated that these policy reforms have contributed to close to half of the 11 point increase in the employment rate of lone parents recorded between 1992 and 2002. This increase in employment occurred in spite of significant rises in the level of support for non-working lone parents claiming Income Support. Benefit increases, combined with success in helping sole parents into work, meant that the trend toward higher child poverty has been stopped and reversed. Between 1998-99 and 2002-03, the number of children in relative low-income households fell by around 500,000, and the government is on track to achieve its target to reduce by a quarter the number of children in relative low-income households by 2004-05. Because of further policy changes introduced since 2002, the increase in the employment rate of lone parents is likely to have continued.

however, reforms in all OECD countries have strengthened the integration dimension of policy while, in some cases, also tightening access to, or reducing the level of, benefit payments (Table 7.1). While overall the scale of this re-orientation of interventions remains limited, these steps signal a change in the way OECD governments think about disability and reflect an effort to better distinguish between disability conditions and benefit receipt. The emphasis on the labour market integration of persons with disabilities has taken several forms: ●

Strengthened vocational rehabilitation and training. Most OECD countries offer disabled persons some types of vocational rehabilitation and training, ranging from initial-needs


121

III.7.


Table 7.1. Changes in disability policies in the period 1985-2000 Compensation policy Strong contraction

Some contraction

No change

Australia Strong expansion

Netherlands

Denmark

Poland

United Kingdom Austria Integration policy

Germany Intermediary expansion

Italy

Norway Spain

Canada France Korea United States

Sweden Belgium Mexico Small expansion

–

Portugal

Switzerland

Turkey Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris.

assessment to training for several years. There are, however, major differences in the form of these programmes, particularly with respect to the following features: the mandatory or voluntary nature of participation;4 the timing of vocational intervention (which starts early in only a few countries, e.g. Germany and Sweden); the number of persons who undergo vocational rehabilitation; 5 and the per capita costs of rehabilitation services relative to disability benefits (which are very low in the Netherlands, Poland, the United Kingdom and Austria). ●

Special employment programmes for disabled people. People with disabilities might be helped into work through a range of employment programmes, including: subsidies to firms for employing disabled people (e.g. Sweden, France, Austria, Denmark and Norway); special schemes offering extensive on-the-job support through individual job coaches (especially in Australia and the United States); and programmes providing employment in special businesses or protected segments in ordinary companies (in Switzerland, Sweden and Norway). The number of participants in all such special programmes, as a share of disability benefit recipients, lies between 15 and 20% in France, Sweden and Austria, but is less than 5% in the United Kingdom, the United States and the Southern European countries.

Significant changes have also been made to compensation policies for disabled persons so as to provide greater incentives to take up paid employment, through a combination of “carrots” and “sticks”: ●

Periodic review of benefit entitlements. While in more than one-third of OECD countries disability benefits are granted permanently, at least de facto, in other countries several examinations are conducted before benefits become permanent, and revision tests are possible at any time (Portugal) or every few years. 6 Periodic reviews of benefit entitlement are particularly important in the case of disabilities whose effects may lessen over time, such as lower back pain and mental problems. Consistent re-testing of medical conditions can increase the rate of outflow from disability benefits (e.g. the Netherlands), although this effect is often only temporary (e.g. the United States).

●

122

Specific in-work benefits. Poor design of the benefit system may also reduce outflows from benefits and inflows into work. A number of countries have introduced specific in-work supplements for people who forgo their disability benefit (Denmark and the United EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005


Kingdom) or tax abatements for workers with disabilities (Turkey). In other countries, benefit entitlement (but not payment) is safeguarded during a trial period so that benefit recipients who accept a job do not have to go through a cumbersome award procedure again should their work experience prove unsuccessful (the Netherlands, the United Kingdom and the United States). Take-up of all these schemes is, however, generally low. ●

Reforming benefit access. When determining benefit eligibility, the range of jobs that could be accepted has been broadened in many countries. Other reforms have changed the medical process leading to long-term benefit awards, through a combination of stricter reviews of sickness status (Norway, Denmark, Spain and Poland) and tighter disability assessment procedures (Portugal, the United Kingdom, Poland, Sweden, Australia and the Netherlands).7 These measures have reduced the risk of granting benefits to people who do not need them – albeit at the cost of refusing benefits to some who do – or have moved to granting eligibility on only a temporary basis. While few countries have chosen to reduce disability payments themselves, these changes in benefit access have frequently shifted recipients from (more generous) contributory schemes to (less generous) means-tested programmes.

Notwithstanding these reforms, no single OECD country can be said to have a distinct integration focus in its policies for disabled people (OECD, 2003a). Existing integration programmes are also characterised by a striking age bias against older people: vocational rehabilitation and training is offered predominantly to people below the age of 45, and sheltered- and supported employment tend to benefit primarily young (and more severely) disabled people. While it is too early to assess whether this policy re-orientation is having significant effects on employment outcomes for disabled people, the recent reforms in Luxembourg and the United Kingdom provide good examples of the greater attention being paid to the labour market integration of persons with disability (Boxes 7.2 and 7.3). When well designed, such targeted schemes have proved useful in improving the employment prospects of disadvantaged groups, although these gains generally occur at the expense of other groups, and with limited net employment creation for the economy as a whole. As in the case of lone parents, it is possible to illustrate the potential impact of higher employment among disabled persons on their relative income. Figure 7.4 shows the increases in the relative personal income of disabled persons that could be associated with an increase in the probability that they held a job. Results are based on the assumption that employment rates of disabled persons increase to the level now prevailing in the bestperforming countries (62% in Switzerland and Norway). Changes in the employment rates of disabled persons can be used to project the relative income of disabled persons (relative to non-disabled): in this example, an 18 point (exogenous) increase in the employment rate of disabled persons among the 14 countries included in Figure 7.4 raises their relative income by around 12% on average, and by more in some countries.

2.3. The involvement of individuals and firms in efforts to secure labour market integration A consensus is emerging in several countries that, if society agrees to invest more in helping beneficiaries in more effective ways, it is reasonable that it requires them to respond accordingly. This principle of “mutual obligations” has proved effective in reforming unemployment benefits and welfare systems, and it is finding its way into other policy areas. This principle is applied in different areas in different ways, with varying degrees of effectiveness in assisting beneficiaries to find work, and varying degrees of EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

123

III.7.


Box 7.2. Policies to promote the rehabilitation of disabled persons: a new reform strategy in Luxembourg While several OECD countries have started to reform their disability programmes so as to move them away from a passive approach towards a stronger emphasis on activation, Luxembourg has introduced one of the most comprehensive approaches. The main goal of this reform, introduced in 2004, is to improve the labour market (re)integration of disabled persons. The law envisages four different stages, each accompanied by different benefits: ●

In the first stage, workers on sick leave undergo an in-depth medical exam. If they are found able to return to work, payment of sickness benefit is stopped and administrative procedures in case of contest are accelerated; if the medical exam finds that sickness is likely to continue, the sickness benefit continues to be paid and another exam is scheduled at a later date; if the person is found to be disabled (in the sense of being unable to continue in their current job) application for a disability pension is launched.

●

In the second stage, for workers who have been found to be unable to continue in their current job, either the worker is eligible for a disability pension (because found to be “disabled in a general sense”) and his/her previous work contract is dissolved, or the worker is found to be ineligible, and further exams are conducted to see whether the worker can return to the previous job or not.

●

In the third stage, for workers who have been found to be unable to continue working in the previous job, two outcomes are possible. In a first case, the worker is re-classified to a new job within the same enterprise (which is compulsory for all companies with more than 25 employees): in this case the worker enjoys special protection from dismissal during one year and receives compensatory benefits (subject to a ceiling) if the new job pays less than the previous one, while the employer is entitled to special support and tax deductions (and subject to sanctions in the case of non-compliance). In the second case, the worker is registered as unemployed and entitled to unemployment benefits while the search for suitable employment continues; when a suitable job is found, disabled workers and their employers are entitled to the same benefits as in the case of internally re-classified workers.

●

In the fourth stage, for disabled persons who could not be placed in suitable employment during the legal duration of unemployment benefit, the worker is paid a waiting allowance (which corresponds to the level of the pre-reform disability pension) by the pension insurance scheme until a suitable job is found, during which period he or she has to remain available for placement attempts.

In the short to medium term, the reform is expected to lead to higher public spending to improve medical assessments of disability, speed up administrative processing of benefit applications, and better assist disabled workers in the reclassification process. In the longer term, while higher public spending will be required by the new compensatory benefit, waiting allowances and tax deductions for employers, these are expected to be offset by lower payments on sickness benefits and disability pensions (in the case of successful re-integration into the labour market). Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris.

compulsion on those out of work. However, the degree of pressure that should be put on benefit recipients to co-operate with efforts to re-integrate them into the labour market remains a sensitive and controversial issue in many countries. Difficult political questions are raised when considering not whether benefit recipients should be asked to co-operate

124



Box 7.3. Reforms of invalidity benefits: Pathways to Work pilots in the United Kingdom The Pathways to Work pilots aim at improving opportunities for people on incapacity benefits. The process involves a series of mandatory Work Focused Interviews together with the Choices package, a range of provisions aimed at improving labour market readiness and opportunities. Pathways to Work is primarily aimed at new incapacity benefit customers: for existing customers, it is currently available on a voluntary basis, but will become mandatory in 2005. The programme includes: ●

Six mandatory Work Focused Interviews. Through the use of a screening tool, those who are unlikely to return to the labour market without additional help are required to undergo five further interviews.

●

A Choices package. This include the NHS Condition Management Programme (to help customers manage their pain, improve their physical fitness and mood, and cope with the uncertainty of their illness); the New Deal for Disabled People, a job-broking service; the standard Jobcentre Plus Provision, including Work Based Learning for Adults and the Advisor Discretion Fund; and a Return To Work Credit, of GBP 40 per week for the first twelve months after returning to work, to assist in the transition from benefit to work. Early evidence on the performance of the Pathways to Work Pilots is quite encouraging:

●

In the phase 1 pilot sites, the outflow rate from the Incapacity Benefit appears to have increased by between 8 to 10 percentage points in the first four months (compared to expectations of a 4 percentage point increase in outflows in the 12 months following reforms).

●

Over 5 500 job entries have been recorded among Pathways to Work customers. Furthermore, the number of recorded Jobcentre Plus job entries for sick and disabled customers in the pilot districts has doubled over the same period last year.

●

Following attending the initial work-focused interview, take-up of elements of the Choices package is between 15 and 20%. Around 3 800 people have taken up elements of the Choices package, with over 1 000 customers having been referred to management programmes.

●

The number of people entering the pilots has now reached a steady state level of around 8 000 new customers per month, including around 800 existing customers entering on a voluntary basis.

On December 2004, the Chancellor of the Exchequer announced that Pathways to Work pilots will be extended, starting from October 2005, to cover one-third of the country. This expansion will focus on support so that Pathways provision is available to the Jobcentre Plus Districts within the 30 most disadvantaged Local Authority Districts. As a result of this extension, Pathways will cover around 900 000 people receiving incapacity benefits.

with efforts to reintegrate work, but the nature of their obligations, and how these should be differentiated according to recipients’ personal and labour market conditions. For example, expectations that lone parents should be available and ready to work require the existence of a range of policies and services; training opportunities should be available to allow low-skilled workers to earn enough to support their families; childcare must be available and affordable; and school- and after-school hours must guarantee lone parents that their children are adequately cared for and supervised while they are at work. If these


125

III.7.


Figure 7.4. Potential impact of higher employment rates among disabled persons on their relative income Relative personal income, baseline Relative personal income, with higher employment rates of disabled persons

% 120 100 80 60 40 20

4 CD -1

OE

St

at es

m ite Un

Ki d

d

en

ng do

ed Sw Un

ite

ain Sp

l ga Po rtu

er th

No rw ay

ds la n

ico ex Ne

M

Ita ly

y

ce

an rm Ge

an

k

Fr

ar nm De

lg iu m Be

Au st ria

0

Note: The figure shows, for each country, the personal disposable income of a jobless person aged 20 to 64 with a selfreported disability, relative to that of non-disabled persons of the same age. The first bar refers to the observed level of relative income of disabled persons in the late 1990s. The second bar is the relative income of disabled persons that would prevail if their employment rate were as high as in Norway and Switzerland (62%), the OECD countries where it is currently highest. Source: Calculations based on OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris.

services are not available, or are not affordable and of adequate quality, work obligations falling on parents may have negative consequences for children. An approach based on mutual obligations relies on establishing closer links between rights and responsibilities. This is especially important for people with disabilities. Several OECD countries have coupled greater emphasis on the individual responsibilities of disabled people with greater promotion of civil and social rights through antidiscrimination legislation. In Australia, the United States and the United Kingdom, public policy is shaped by anti-discrimination legislation that was introduced during the early 1990s – first in the United States, and then in the other two countries – while Canada pursues the same goals through its federal and provincial Human Rights Acts. All these laws have special chapters prohibiting discrimination based on race, gender, age and disability status in all aspects of employment and work relations, unless this would cause “undue” hardship or costs.8 During the past few years, several European countries have also adopted general anti-discrimination laws (e.g. Sweden, Norway and Germany), while others have introduced anti-discrimination clauses into their constitutions (e.g. Austria and Switzerland). However, even when effective in enforcing individual rights in the workplace, attention has to be given to the possibility that employers may respond to these provisions by discriminating in the hiring process. The principle of mutual obligations in labour market integration is not limited to individual clients and government agencies but extends to employers. The most extensive initiatives in this respect concern people with disabilities and low-paid workers and training provisions for disadvantaged groups. With regard to people with disabilities, some OECD governments (e.g. Austria, France, Germany, Italy, Poland, Spain and Turkey) have formalised employers’ obligations through mandatory quotas for the employment of disabled persons, while others rely on general employment legislation (Belgium, the

126



Netherlands, Norway, Spain and Sweden), anti-discrimination acts (the English-speaking and some European countries) or voluntary approaches (Denmark, Table 7.2).9 In practice, however, many of these regulations are difficult to enforce, despite financial sanctions for non-compliance. Even in countries that have mandatory quotas for the employment of disabled persons, fines for non-compliance are often so low that employers may find it easier to pay than to comply. Another reason for the ineffectiveness of these measures is that employers may need help to fulfil their obligations. While workplace and job adjustment generally involve small financial investments, companies often require technical assistance, including in assessing the problems and developing an intervention strategy, help that is often in short supply.

Table 7.2. Legislative framework and extent of employer responsibility towards disabled persons Legislation shaping the framework Strong antidiscrimination legislation Major employer responsibilities

Extent of employer responsibility

Some employer responsibilities, less important for new applicants

Few employer obligations

Australia Canada United Kingdom United States

–

Mandatory employment quotas

Legislation on employer obligations

Voluntary approaches

(Germany)a Italy

Sweden

–

(Austria)a France (Germany)a Poland Spain

Netherlands Norway

Denmark

(Austria)a Korea Turkey

Belgium

Mexico Portugal Switzerland

a) The classification of Austria and Germany regarding employer obligations straddles two groups. Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris.

Financial incentives can strengthen employer responsibilities towards persons at greater risk of poverty and social exclusion. Some countries have recently introduced measures that make employers partly responsible for continued wage payment in the case of sickness absences, in the hope that they will invest in prevention and retention measures that avoid these people becoming permanently disabled. Other countries have made the continued availability of “make-work-pay” subsidies conditional on an employer’s provision of in-work training or on conversion of subsidised contracts into regular ones. Evidence about the effects of these financial incentives in changing employers’ behaviour remains sparse. In the case of employer obligations for wage payments during sickness periods, the number of full-time equivalent sickness cases fell dramatically in the short-term in the Netherlands and the United Kingdom, followed by renewed rises in the longer term (OECD, 2003). Results are more encouraging in other areas: in the United States, firms achieved sharp reductions in the frequency and severity of work accidents partly because of the existence of high insurance premiums on these risks, and also due to premiums that are differentiated according to the accident records of individual firms (Askenazy, 2004).


127

III.7.


3. Increasing the rewards of work makes labour market integration more sustainable Poverty and social exclusion reflect more than lack of jobs. Many of the jobs available may not pay enough to lift households out of poverty, or may not provide career prospects to the workers who hold them. While the risk of falling into poverty is much higher for households with no adult in employment than for those where someone works, households with one or more workers represent a very substantial proportion of the income-poor in all OECD countries. Since 1995, the share of income-poor households with one worker has increased in around half of the countries shown in Figure 7.5. Even households with two or more workers are not immune from the risk of inadequate income, especially in Austria, Greece, New Zealand, Portugal, Switzerland and the United States. Also, because of the importance of intermittent employment, the overlap between work and poverty is even higher when poverty is assessed over a multi-year period rather than a single year (OECD, 2001a). The fact that many of the “poor” hold jobs, at least for some part of the year, goes a long way towards explaining the lack of a significant cross-country association between relative poverty, at the level of disposable income, and employment rates shown in Figure 6.2.

Figure 7.5. Relative poverty rates among households with a working-age head around 2000, by number of adults working %

Two or more workers

One worker

No workers

20 16 12

Mid-1990s

2000

8 4

Au

st ra li Au a st r Cz Ca ia ec h nad Re a pu De blic nm ar Fin k lan d Fr an c Ge e rm an Gr y ee Hu ce ng ar Ire y lan d Ita ly Ja pa M n Ne exi th co e Ne rlan w d Ze s ala n No d rw a Po y lan Po d rtu ga l Sp a Sw in S ed Un witz en ite erl a d K nd Un ingd ite om d St a OE tes CD -2 1

0

Note: The height of each bar represents the poverty rate (using a 50% median income threshold) of persons living in households with a head of working age in each country. Data for Germany refer to old länder. Exact years are those specified in Figure 1.2. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.

Because many poor households are characterised by low-paid or precarious employment, rather than persistent lack of jobs, all OECD countries confront the dual challenge of promoting both “welfare-to-work” and “welfare-in-work”. Welfare-in-work policies may be grouped in two main categories, according to whether what they aim to encourage is labour supply or labour demand for low-paid workers.

3.1. Employment-conditional benefits Labour supply of low-skilled workers can be supported through employmentconditional tax credits or benefits. In-work benefits are long-standing pillars of the social protection system in the United Kingdom and the United States. The Working Family Tax

128



Credit in the United Kingdom (recently replaced by the Working Tax Credit) and the Earned Income Tax Credit in the United States reach, respectively, 5% and nearly 20% of all households (Hotz et al., 1999; US Census Bureau, 2000), representing total spending of 0.6% and 0.3% of GDP. Both programmes are directed at families with low earned income, particularly those with children, and provide relatively generous benefits that are administered through the tax system. Similar programmes have been introduced in other OECD countries in more recent years (Table 7.3). Several differences distinguish these programmes: ●

Benefits are based on the individual rather than the household in Australia, Canada, Belgium, France and the Netherlands.

●

Benefit eligibility often depends on some combination of household income and the recipient’s prior status (e.g. welfare leavers in Canada, long-term unemployed in Ireland) or household composition (e.g. families with children in the United Kingdom, the United States, New Zealand and several other countries).

●

In-work benefits have been introduced in the context of general reforms of the tax system – rather than as specific programmes – in some countries (in the form of the lower employee social security contributions for low-paid workers in Belgium, and of phased reduction of the burden imposed by taxation on low-wage earners in Germany).

3.2. Wage subsidies to employers hiring or retaining low-paid workers Because of wage floors and fixed-components in labour costs, high labour costs may limit firms’ willingness to hire low-paid workers. Measures to increase firms’ demand for low-skilled workers fall into two main categories: ●

Broad measures, covering all those in low-paid work, have been introduced in several countries (e.g. Belgium, France and the Netherlands). Most commonly, these measures take the form of reductions in employers’ social security contributions on all workers with earnings below a given level, with no limits on duration.

●

Specific subsidy programmes, targeted to groups that are less easily employable, exist in most OECD countries. Public spending on such targeted employment schemes represents – on average – close to 0.2% of GDP and around one-fourth of public outlays on active labour market programmes.

3.3. Effects of welfare-in-work policies While employment-conditional benefits and wage subsidies share the common objective to make work pay, the two approaches suit different circumstances (Pearson and Scarpetta, 2000): ●

First, they differ in their impact on poverty: subsidies to low-paid workers, awarded on an individual basis, are less effective in reducing poverty than employment-conditional benefits based on household income, as low-paid workers may not always live in lowincome households.

●

Second, they differ in their effects on the employment of low-skilled workers. Tax credits have been an effective means of encouraging entry or return to employment for loneparent families and households where no-one works. Several evaluations of countries that have implemented broad-based wage subsidies to low-paid workers also report significant employment effects, in particular for low-skilled workers, but also high deadweight effects (i.e. part of the additional employment would have been created even in the absence of specific financial support). 10 With respect to specific subsidy


129

III.7.

130 EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

Type of recipient

Maximum benefit

Minimum hours worked per week

Description of benefit

Unit

Australia

Employment Entry Payment

Lone parents or long-term welfare recipients (12 months or more)

AUD 104/year

–

Lump sum paid when entering employment (eligible every 12 months)

Individual

Belgium

Refundable tax credit (being introduced)

Wage earners or self-employed with an earned income of between EUR 3 260 and EUR 14 140 per year

EUR 440/year (in 2004)

None

Lump sum

Individual

Canadab

Provincial programmes (e.g. Ontario’s Start Up Benefit)

Social assistance recipients

CAD 253/year

None

Lump sum paid to people who begin/change employment or join training programme (once every 12 months only)

Individual

Earned Income Tax Credit

Working families with children and individuals with low salaries

USD 2 428/year

None

Entry phase: increases with income Families with children, individual Exit phase: decreases with income

United Statesc

USD 4 008/year USD 364/year France

Employment allowance (Prime pour l’emploi)

Individual with earnings between EUR 3 265 and EUR 15 235 (or EUR 23 207 for a married person with an unemployed spouse)

EUR 720/year (in 2003)

None

Entry phase: increases with income Exit phase: decreases with income. Supplement for dependants.

Individual

Irelandd

Family Income Supplement

Families with children and low wages (threshold from EUR 62 to 539 per week depending on number of children)

75% of previous allowance

19 hours

Amounts to 60% of the difference between effective income and eligible income

Families with children

Back-to-Work Allowance

Long-term unemployed (min.12 months) and welfare recipients

20 hours

Limited to three years, decreasing over time (50% then 25% of previous allowance in 2nd and 3rd years).

Individual

New Zealand

Family Tax Credit

Families with children and low earned income (under NZD 18 368 before tax, p.a.)

30 hours, couples; 20 hours, singles

Provides a minimum income of NZD 18 368 p.a., or NZD 286 after tax per week

Families with children

Netherlandse

Employment-conditional tax credit

Wage-earners or self-employed

None

Entry phase: increases with income

Individual

EUR 920/year


Table 7.3. Examples of employment-conditional benefits Namea

United Kingdomf

Type of recipient

Maximum benefit

Working Family Tax Credit

Working parents with low income (threshold depends on household composition)

GBP 88.95 per week (one-child family)

Minimum hours worked per week

Description of benefit

16 hours (supplement of 30 hours Exit phase: decreases with income or more) (depends on hourly wage). Supplements for extra dependent children

Unit Families with children

a) The benefits presented here fall into two categories: tax credits for earned income; and cash supplements to offset benefits foregone upon return to employment. Other countries (e.g. Spain, Finland and Japan) grant wage-earners tax relief, which amounts to exempting a share of their earned income (the relief is usually higher for low incomes). Such exemptions are not included here. b) The old Working Income Supplement has been replaced by the National Child Benefit supplement, a new nationwide measure targeted at all low-income families; it does not really constitute an employment-conditional benefit. However, each Canadian province has special benefits which it pays to welfare recipients when they enter or re-enter employment. c) The first figure applies to one-child families, the second to families with two children or more, and the third to families with no children. d) Ireland has other employment-conditional subsidies, including an allowance for returning to part-time work (less than 24 hours per week) after at least 15 months of unemployment. e) In 2001 the Netherlands tax system underwent radical change, including the introduction of a variety of tax credits. Some families, for instance, can deduct some of their childcare costs from their taxes. Welfare recipients are also eligible, under certain conditions, for a back-to-work allowance. f) From April 2003, this tax credit for working families is being replaced by two separate benefits, one being employment-conditional and the other covering childcare costs. Source: See OECD (2003c), “Making Work Pay – Making Work Possible”, Employment Outlook, OECD, Paris for detailed sources.

131



Table 7.3. Examples of employment-conditional benefits (cont.) Namea

III.7.


programmes, evaluations suggest that direct job creation in the public sector is less effective than subsidies of regular employment in the private sector, as the public-sector jobs typically provide limited opportunities to gain experience that can be transferred to regular jobs. Also, targeted schemes have proved useful in improving the employment prospects of disadvantaged groups, although these gains generally occur at the expense of other groups, and with limited net employment creation. ●

Third, they differ in their (unintended) effects on other people not directly targeted by these programmes. Employment-conditional benefits tend to lower the employment of second earners: when attempts to limit the costs of these programmes lead to benefits being clawed-back at higher levels of household income, this may induce other family members to withdraw from the labour market or reduce their incentive to work longer hours. Employment subsidies awarded on an individual basis are not prone to such negative effects.

Whatever their effects in getting people into work, the big challenge for welfare-in-work policies is getting low-skilled people into careers. While evidence about what else might be cost-effective in helping low-paid workers keep and progress in their jobs remains limited, there has been a shift among in-work programmes in recent years towards more comprehensive and individually-tailored schemes that pay specific attention to the career development of disadvantaged workers: in the Netherlands, complementary schemes reduce the higher tax burden for employers who raise the pay of workers receiving in-work benefits; in the United States, the Joint Training Partnership Act offers employers temporary subsidies for the training of new recruits, and monitors employer behaviour so as to strengthen the link between the person hired and employer’s investment in training; in still other countries, some schemes contain clauses that exclude from in-work benefits firms that refuse to extend the contract of workers hired in this way (OECD, 2003c). While many of these programmes are still in their infancy, early evaluations suggest that a balanced approach, which combines makingwork-pay policies with other measures, is effective in relieving poverty and increasing employment among the less skilled, especially when combined with a broader set of interventions to build their human capital and earning potential.

4. More effective social assistance makes a difference to the lives of the poor While labour market integration policies have resulted in greater employment of some disadvantaged groups, traditional income support policies remain indispensable for reducing poverty and exclusion. Still of great importance is the extent to which programmes reach those they are supposed to insure against a given contingency (coverage rates); the extent to which those entitled to a given benefit actually receive it (take-up rates); and the adequacy of the support provided to escape poverty and exclusion (benefit levels). These critical dimensions of the benefit system are interrelated (e.g. a low benefit level may lead some eligible persons to not claim it). Moreover, variations in benefit systems with respect to these characteristics partly relate to differences in underlying goals and philosophy. While many “welfare-to-work” and “welfare-in-work” measures are driven by the belief that robust behavioural incentives will allow more significant inroads into fighting poverty and exclusion, a range of social programmes are grounded in the belief that no-one should suffer the deprivation and disadvantage that result from dire poverty, irrespective of their behaviour and life-style. It is not possible to measure directly the coverage of existing social protection systems, mainly because of the difficulties in identifying the conditions that establish programme

132



eligibility. In most OECD countries, the majority of the income-poor receive some types of benefit income. Partial coverage is, however, common when looking at some of the contingencies against which the social protection system is supposed to protect. Evidence in the previous section suggests that a significant proportion of lone mothers not-at-work (at a given point in time) in the United States report that they did not receive any benefit income in the 12 months preceding the interview. This extends to other types of programmes. For example, more than one-third of all non-employed persons reporting a disability in OECD countries also report that they did not receive benefit income. Similarly, in European countries, only one-third of the persons classified as unemployed in labour force surveys – and even fewer of those unemployed for two years or more – report having received unemployment benefits.11 Conversely, both types of benefit are often received by persons that do not seem to fit the conditions targeted by the programme. In the case of disability, close to one-third of those receiving disability benefits did not classify themselves as “disabled”; in the case of unemployment, close to half of those receiving unemployment benefits are not classified as unemployed according to ILO criteria (Table 7.4). While these differences partly reflect under-reporting of benefits in household surveys, misclassification

Table 7.4. Proportions of disabled and unemployed persons receiving benefits, and proportion of benefit recipients who are not disabled or unemployed Disability benefits Proportion of disabled people receiving a disability-related benefit

Unemployment benefits

Proportion of disability benefit recipients who do not report themselves as disabled

Proportion of ILO unemployed receiving an unemployment benefit

Proportion of unemployment benefit recipients not classified as ILO-unemployed

Australia

28.6

15.2

–

–

Austria

20.9

27.7

72.7

32.2

Belgium

39.0

43.4

69.8

74.0

–

–

31.5

8.4

34.3

26.2

53.8

59.2

Czech Rep. Denmark Finland

–

–

53.9

43.2

France

22.8

33.3

46.7

41.8

Germany

–

–

74.6

20.0

Greece

–

–

12.3

21.3

Hungary

–

–

37.3

57.8 55.2

Iceland

–

–

26.5

Ireland

–

–

–

–

20.7

43.9

1.4

80.2 –

Italy Japan

–

–

–

Korea

10.3

0.0

–

–

Mexico

0.3

91.4

–

–

Netherlands

29.0

30.6

33.8

–

Norway

28.4

28.4

30.5

77.7

Poland

49.1

–

13.3

21.8

Portugal

21.2

28.6

29.6

58.5

Spain

39.6

18.3

20.9

45.6

Sweden

35.6

48.9

42.2

55.7

Switzerland

18.5

29.8

34.8

51.4

United Kingdom

37.0

43.3

38.9

26.5

United States

39.2

46.7

–

–

Note: Data on disability benefits refer to the late 1990s. Data on unemployment benefit receipt refer to 2001. Source: OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris; and Immervoll H., P. Marianna and M. Mira d’Ercole (2004), “Benefit Coverage Rates and Household Typologies: Scope and Limitations of Tax-Benefit Indicators”, Social, Employment and Migration Working Paper, No. 20, OECD, Paris.


133

III.7.


of benefits, and differences in the period over which benefit recipiency and risk-occurrence (e.g. unemployment) are assessed, they are also likely to reflect genuine gaps in coverage rates, for example, because of a lack of contribution records to qualify for benefits. Information on take-up rates is sparse, but what does exist suggests that low take-up is a common feature of both means-tested and social-insurance programmes. Estimates of take-up rates typically range between 40 and 80% in the case of social assistance, and 60 to 70% in the case of unemployment benefits (Hernanz et al., 2004). While non take-up is not per se evidence of a policy problem (eligible persons may not take up benefits because their amount and expected duration fall short of application costs), it may become one when these costs (e.g. stigma or transaction costs) fall disproportionately on those with the greatest needs: for example, immigrants, although more likely to be eligible for welfare, are less likely to take it up, all else being equal (Currie, 2004). When this happens, poverty is much higher than what would prevail under full take-up (Figure 7.6). Unfortunately, existing research does not provide much guidance on how best to increase take-up of existing programmes: the mix of interventions will depend on the specific programme and client group considered. Making information about existing programmes more easily available, reducing the stigma associated with benefit receipt, lowering transaction costs and better co-ordinating programmes are all likely to be important.

Figure 7.6. Impact of full take-up of benefits on poverty in the United States All families Millions 25.0 20.0

Poverty

Single-parent families Millions 3.5 3.0 2.5

15.0

2.0

10.0

1.5

5.0

Extreme poverty

1.0 0.5

0 1996 1998 1996 1998 Note: The height of each bar indicates the number of people living in families with children counted as “poor” or “extremely poor” in the two years shown. The dark part of the bar indicates the reduction in the number of poor people that could be achieved if all those entitled to benefits (Food Stamps, Supplementary Security Income, and Temporary Assistance for Needy Families) took full advantage of them. Poverty is here measured with reference to a broad notion of household disposable income, which includes all government assistance and deducts federal payroll and income taxes, as well as out-of-pocket childcare expenses. Poverty is measured with reference to the federal poverty threshold. Extreme poverty is defined as income below 50% of the federal poverty level. 0

Source: Zedlewski, S.R. et al. (2002), “Extreme Poverty Rising, Existing Government Programmes Could Do More”, Urban Institute, New Federalism, Series B, Washington DC.

The success of policies in reducing poverty and exclusion also depends critically on the adequacy of the benefits provided. Only rarely, however, do last-resort benefits provide an income level adequate to escape poverty, at least when the latter is measured with respect to the (relative) poverty thresholds – 40 to 60% of median equivalised household income – conventionally used in comparative research. Figure 7.7 shows net income levels for persons relying only on social assistance benefits for their daily living over a full year, calculated under two assumptions on receipt of housing benefits (no housing-related benefits and costs, and housing benefits available to households whose rental costs are

134



Figure 7.7. Net income of social assistance recipients, 2001 No housing benefits

With housing benefits

% equivalent median household incomea 70 Single person 60

Un

Hu

ite d

Gr

ee

ce Ita ly

St at es ng ar Po y rtu ga Un Sw l ite ed d Ki en ng do Ge m rm an y Ca na da Sp ain Fin lan d Fr an De ce n Sw mar itz k er lan d Ire lan d Cz A ec ust r h Re ia Ne pub li w Ze c ala n Au d st ra li No a rw ay Po Ne lan d th er lan d Be s lg iu m

50 40 30 20 10 0

% equivalent median household incomea 70 Lone parent, two children 60

d lan Ge d rm an No y Ne rwa y th er lan ds Au st ria Ca na De da nm ar k P Ne ola nd w Cz Zea ec lan h Re d pu bl Be ic lg iu Au m st ra lia

m

itz

Fin

Sw

d Un

ite

Un

er

lan

d

do

lan

Ki

ng

Ire

l

ce

an

Fr

en

ga

ed

rtu

Sw

Po

y

ain

ar

Sp

Hu

ng

es

ce

at

St

d

ite

Gr

ee

Ita

ly

50 40 30 20 10 0

k ar

d

nm

De

ds

lan

Po

ic

lan

th

er

nd

m

bl

pu

h

Ne

Re

ala

Ze

w

Cz

ec

d

iu lg

Be

Ne

ay

lan

Ire

d

rw

No

ria

lan

st

er

Au

itz

Sw

l

d

ga

lan

Fin

da

Po

rtu

ce

na

Ca

Fr

an

y an

Ge

rm

en

ed

Sw

Au

st

ra

lia

m do

Un

ite

d

Ki

ng

Sp

ain

y ar

Hu

ite Un

ng

es

ly

at St

Ita

d

Gr

ee

ce

% equivalent median household incomea 70 Married couple, no children 60 50 40 30 20 10 0

% equivalent median household incomea 70 Married couple, two children 60

Ita

ly

Gr ee c Hu e Un ng a ite ry d St at es Sp ain Fr an c Sw e Sw ede itz n er lan Un Ger d ite ma d Ki ny ng do Po m rtu ga l Ire lan d Fin Ne lan d th er lan ds Ca na d No a rw ay Au st ri Be a Ne lgiu m w Ze ala De nd nm a Au rk Cz ec stra l h Re ia pu bl ic Po lan d

50 40 30 20 10 0

a) Calculation based on OECD tax-benefits models. Housing benefits that do not depend on housing expenses are included in both cases. Percentages are calculated by comparing actual benefit entitlement for each household type to the equivalised median household income in each country. Household are here considered as poor when their equivalent income is in the range 40-60%. Source: OECD (2004d), Benefits and Wages – OECD Indicators, OECD, Paris.


135

III.7.


equivalent to 20% of their gross earnings): it suggests that the net household income of persons receiving social assistance (with no other income sources) is generally below the income range that qualifies them as poor, although they are higher when children are present and when housing benefits are also available. In practice, as last-resort benefits are rarely the only income source over the entire course of the year, tax-and-benefit systems play a much larger role in reducing poverty at any given time (cf. Figure 6.4).

5. Improving policy coherence can reduce unintended adverse effects Increasing the effectiveness of policies aimed at reducing poverty and exclusion also requires improving the coherence of policies at the local, national and regional level. These governance issues are important because single types of interventions are unlikely to suffice to eradicate poverty and exclusion, a task which calls for more comprehensive strategies (Box 7.4).12 The experience of those countries that have been most successful in reducing poverty and social exclusion points to the critical importance of a long-term commitment and of greater policy coherence. The drive to strengthen the coherence of policies affecting poverty and exclusion has several dimensions. One dimension of policy coherence relates to horizontal co-ordination within governments. Different government agencies typically share responsibility for different interventions (paying cash benefit, addressing some of the social problems that affect social assistance clients such as disability and poor health, and helping labour market re-integration). A proliferation of different social programmes, each focused on a specific problem, may lead to insufficient attention being paid to the needs of the individual client as a whole person. When institutional coherence is low, eligible clients will distrust the system and choose not to participate in programmes, leading to a further worsening of poverty as existing programmes fail to reach their potential beneficiaries. Administrative barriers between different programmes, poorly integrated information systems, and different (and contradictory) eligibility requirements may end up preventing the full participation of all potential clients in the full range of services available. To improve horizontal co-ordination, the functions of job placement, benefit payment and service provision have been brought closer in some countries through the establishment of a single office. In other countries, measures have been taken to bring closer the financing and delivery of different services. Other measures have aimed at aligning eligibility requirements among programmes and granting automatic access for persons participating in specific programmes to other types of benefits (i.e. “passporting”). Centralising the administration of the same type of benefits has also reduced unintended interactions and adverse incentive effects, and has improved benefit take-up. Australia, Canada and the United Kingdom have introduced measures to better integrate tax concessions and social benefits. Measures to improve horizontal co-ordination within governments also include ensuring that labour market services are provided to social assistance clients. A second dimension of policy coherence concerns vertical co-ordination between central and local government. Issues of inter-governmental management and decentralization are high on the policy agenda of social policy makers in many OECD countries. In the United Kingdom, for example, the geographical concentration of deprivation has led the authorities to introduce specific social inclusion partnership programmes (Social Exclusion

136



Box 7.4. Ireland’s National Action Plan against poverty and social exclusion Building on the commitments made at the UN World Summit for Social Development in Copenhagen in 1995, the Irish Government approved the development of a National AntiPoverty Strategy (NAPS). The ten-year strategy, developed by an inter-departmental committee, was launched in April 1997 following wide-ranging consultation with the social partners. A NAPS Unit was established to co-ordinate implementation of the Strategy. The global target of the National Anti-Poverty Strategy is to reduce the numbers of those who are “consistently poor” from 9-15% of the population to less than 5-10% over the period 1997-2007.* Beyond this general target, the strategy also embodies subsidiary targets and timeframes in the five areas of unemployment, income adequacy, educational disadvantage, urban disadvantage and rural poverty. Following the launch of the NAPS, much of the early work has aimed at setting up the structures necessary to support and oversee implementation. A set of structures is today in place to ensure that social inclusion is central to policy formulation and implementation in all areas of policy-making – from the high-level strategic view to the co-ordination and executive functions necessary to ensure its ongoing implementation. Following agreement with the social partners in 1998, the Government adopted a “poverty proofing” system for official Cabinet procedures. According to this system, all major policy proposals must indicate their impact on groups in poverty or at risk of falling into poverty. “Poverty proofing” ensures that Departments are kept aware of the possible implications of policy on those most in need. All Memoranda for Government and key policy initiatives (e.g. the National Development Plan), upon which significant policy decisions have to be made, are required to be poverty proofed, and departments have put in place the arrangements necessary to make all staff working on the preparation of government initiatives fully aware of these requirements. To assist in the implementation of “poverty proofing”, the NAPS-Unit has produced guidelines that were distributed to all Government Departments. The Local Government Anti-Poverty Learning Network is a key support for the local implementation of NAPS. In addition, the Practice to Policy Programme was established in 2004 to strengthen the voice of people living in poverty at local, regional and national levels: this programme will provide funding and support to a number of initiatives aimed at engaging people living in poverty in the development and implementation of the NAPS. Following achievement of the original poverty target (down from 9-15% in 1994 to 7 to 10% in 1997), in June 1999 the Government set a new target of reducing consistent poverty to below 5% by 2004. Data from the 2001 Living in Ireland Survey shows that consistent poverty is now down to 5.2% of the population; in 2001, according to the same survey, one in fifteen children were living in “consistently poor” households, as compared to one in four in 1987. * “Consistent poverty”, as measured by the Economic and Social Research Institute, is defined as being below 50-60% of average household income and experiencing enforced basic deprivation (i.e. the presence of at least one of eight indicators, e.g. not having two pairs of strong shoes; having recourse to debt to meet ordinary living expenses, etc.).

Unit, 2004). More generally, interest in decentralisation as a tool of effective implementation of social policies reflects a number of arguments: ●

First, there are often strong reasons for giving local governments a prominent role in the delivery of social assistance: they can better tailor interventions to the needs of their clients, assess genuine needs requiring exceptional support, verify that clients behave


137

III.7.


in ways that justify continued assistance, and promote policy innovations. Decentralisation may also offer the opportunity to reduce unnecessary duplication and overlap, so resulting in better value for taxpayers’ money; it may improve service delivery across levels of government and co-ordination between public and private service providers; it may increase policy flexibility and improve client and service orientation; and, finally, it may provide opportunities for policy experimentation, by providing a basis for evaluation of alternative policy reforms. ●

Second, decentralisation may improve the quality of democratic participation and the involvement of local communities and of welfare clients. When services are more responsive to local needs, it may be possible to empower citizens to participate more in decision making. Hentschel et al. (1998) note that underlying some of the arguments in favour of decentralising poverty alleviation efforts is the notion that local communities are best placed to identify the interventions that are most beneficial to the poor in their area, and that, when locally managed, public resources are less likely to be captured by non-poor households. A related argument is that centrally managed social programs may have sometimes promoted a culture of “welfare dependency” among clients. This calls for initiatives to promote the active participation of low-income individuals and families in valued social and economic activities, initiatives that are almost invariably focused at the local level.

●

Third, interest in decentralisation also arises because of a renewed interest in the geographical concentration of social problems – what Kleinman (1998) has called “the new politics of place and poverty”. For example, Jalan and Ravaillon (1997) discuss “spatial poverty traps”, as situations where otherwise identical households experience different gains in living standards as a result of living in different areas, with different endowments of “geographic capital”: if spatial poverty traps can be shown to exist, then there may be efficiency and equity arguments for investing in these areas by developing local infrastructure. Location may provide a good proxy for greater need, and thus provide a basis for targeting scarce resources.13

Because of the geographical concentration of individuals facing the greatest risk of exclusion, targeted local policies add a specific territorial dimension to social policies and increase the importance of ensuring the coherence across a range of fields (urban development, social housing, health and education). Nevertheless, it is difficult to design mechanisms that secure adequate financing to regions where needs are higher, while retaining incentives for local governments to engage in activities to remove people from poverty (such as rewarding them for reducing poverty). Split financing between central and local governments may give the latter incentives to shift social assistance clients (often financed by local government) to programmes for which they have no direct financial responsibility.

6. Conclusions To reduce poverty and exclusion in the future, a broad range of policies is required. ●

138

Completing the welfare-to-work agenda. Much has been done to tackle long-term benefit receipt for those receiving unemployment benefits, and a fair amount has been undertaken to help those in receipt of social assistance. Progress is less evident in the case of lone parents and, in most countries, people with a disability: policies in these



areas – by and large – remain passive income support programmes that are not coherent with attempts to tackle social exclusion. ●

Making progress with welfare-in-work. A challenge for social and labour market policy is to soften the often unwarranted distinction between those without work, who are provided with various forms of financial support and services, and those in work, who generally receive far less help. This requires policies to make work pay and to increase job retention and career prospects of low-paid workers.

●

Strengthening the effectiveness of social programmes targeted to persons for whom work is less feasible. This includes extending coverage and take-up of existing programmes to all categories of persons that are in conditions of need, and assuring adequacy of the benefit provided. In a dynamic setting, it also requires moving beyond “work” as the only focus for social policies. Most countries have no “plan B” for those who have been through welfare-to-work programmes once, but who still do not find work: programmes that recycle these persons through the system will increase the costs of such programmes, with only limited results; while programmes that simply put these persons back on passive income support, with no real attempt to promote inclusion, are hardly satisfactory either. Policies to promote social integration, independently of labour market integration, remain a priority in all OECD countries.

●

Promoting the coherence of different policies affecting poverty and exclusion across government departments and levels of governments is important to reduce unintended effects of policies in specific areas and increase clients’ trust in the system. Making longterm commitments to poverty goals may also increase the effectiveness of policies in this area as – despite better knowledge about what works in this area – all policies will take time to deliver and no “quick fix” exists for complex problems.

Notes 1. In the 13 OECD countries where work tests depend on children's age, this varies from 3 months in several states in the United States to 16 years in Australia; in a majority of countries this age is 3 years (OECD, 2005d). 2. The effects of these labour market courses in improving the chances of moving into jobs, however, are not always clear-cut: when participating in a programme, job-search intensity and rates of entry into non-subsidised jobs tend to be lower than in periods of non-participation; after participation, evaluations often suggest that prospects for unsubsidised employment are not necessarily higher, especially in the case of public-sector job-creation schemes. Even in these cases, however, obligations to participate in labour market programmes can ease public willingness to finance these programmes, force government agencies to pay attention to disadvantaged groups and reduce fraud. 3. Figure 7.3 relies on information about the share of persons living in households headed by a lone parent, with and without work; the poverty rates of these household types, derived from the income distribution questionnaires for the year 2000; and the average number of children in lone parent households in 1995. 4. Some OECD countries, where a benefit claim is automatically treated as a request for vocational rehabilitation, effectively apply the principle that “rehabilitation goes before pension”. 5. In Denmark, Norway and Korea, the stock of people receiving vocational rehabilitation services exceeds the annual inflow of benefit recipients, while in around half of the countries this ratio is less than 5%. 6. Between two to three years, in Austria, Germany and Italy, or up to five years, in Australia and the Netherlands.


139

III.7.


7. In a large majority of countries, the medicalised assessment is nowadays performed by specialised (insurance) doctors, with a small role accorded to medical certificates prepared by treating doctors; other countries have introduced more formal central medical commissions. Despite these changes, the higher frequency of more difficult-to-diagnose diseases – such as mental illness and stress-related physical conditions like lower-back pain – is making the assessment of disability more difficult. 8. There are, however, important differences across these countries as to the category of person to which legislation applies, the consequences of breaking the law, and the definition of “undue” hardship. 9. For example, in Italy employers are required to assign the disabled person to “equivalent” tasks or, if this is not possible, to lower-graded tasks with the same remuneration; in Sweden, employers must equip the workplace appropriately for any worker who becomes disabled and, if possible, provide them a different job in the company; in Germany, employers have a general obligation to promote the permanent employment of severely disabled people (via provision of adequate workplaces, preferential selection for in-house training and support for external measures). 10. Evidence, summarised in OECD (2003c), for selected programmes shows net increases of market employment of 3% (7% for low-skilled workers), at an annual cost of 1% of GDP, in Belgium; of between 0.7 and 3%, at an annual cost of 0.5% of GDP, in France; and of close to 1% (5% for lowskilled workers), at an overall cost of 0.5% of GDP, in the Netherlands. 11. In most European countries, respondents to labour force surveys are asked whether they are receiving unemployment benefits in the same period used to assed labour force status (Immervol et al., 2004). 12. In the European Union, this task is being accomplished through the use of the so-called “open method of co-ordination”. This requires each member country to prepare a National Action Plan on Social Exclusion – containing quantified targets and a description of the measures to attain them – that are coherent with Community guidelines; based on these national action plans and on their evaluations by other countries, the European Council and the Commission prepare a Joint Report on social inclusion (two such reports have been released so far, in 2001 and 2004). 13. It should be noted that this is in the context of poor alternative mechanisms for targeting to the poor.

140


PART IV

Responding to Shifts in the Risks Confronting Older Persons As recently as the 1970s, the words “pensioner” and “poverty” were inextricably linked. This is no longer true: during the past few decades, older persons have experienced significant improvements in their economic well-being. They are also living longer and healthier lives in all OECD countries. But, in a context of rapid population ageing, this improvement in performance is weighing heavily on government budgets, threatening the financial sustainability of the programmes that underpinned this progress. Moreover, social policy has been less successful in addressing other risks facing the elderly: the barriers that older workers face in retaining work, the risk of disability and isolation in old age, and the disturbingly high rates of poverty among some specific groups of elderly, in particular those with no pension rights or with interrupted careers. The changing configuration of risks confronting older persons, and the threat to the financial sustainability of the relevant programmes, calls for a shift in policy intervention. Active policies strive to address the obstacles to the continued participation of older persons in economic and social life, with a view to improving their well-being and making better use of their capabilities in activities that benefit both the whole community and themselves. Raising the labour-force participation rates for older persons is important not only for labour-market reasons and to put pension systems on a more solid footing, but also to avoid the risk that, in countries where the value of public pensions has been reduced recently, lower replacement rates will lead to higher poverty in old age. One specific concern is to ensure that the burden of assisting older persons with disability does not weigh disproportionately on women. Because many of them will have entered the labour market, they may no longer be available to provide the care needed by an increasing number of frail elderly. This requires adapting the system of formal care provision and its relationship to informal care, redistributing costs, and taking measures to lower the burden falling on family carers. Priorities for reform include: ●

Lowering the costs of old-age pensions while diversifying retirement income.

●

Promoting a longer working life.

●

Providing more quality services for an increasing number of frail elderly.



PART IV

Chapter 8

The Social Conditions of Older People: Achievements and Challenges


143

IV.8.

THE SOCIAL CONDITIONS OF OLDER PEOPLE: ACHIEVEMENTS AND CHALLENGES

1. Introduction This chapter reviews changes in the configuration of risks confronting elderly persons. It summarises the achievements of OECD countries with respect to the income and health situations of elderly persons. It then examines areas of continued policy concern, such as the rising costs of retirement provision, the barriers to continued employment for older workers, the persistent risk of poverty in old age for some vulnerable groups, and a potential gap between the need for and availability of quality long-term care.

2. Policies and achievements While only a few decades ago getting old meant facing the spectre of “scraping by” for many – if not most – elderly, risks of low income in old age are no longer the most important concern for most elderly persons, though they remain critical for some groups. Yet the pension systems and social policies that largely underpin this achievement are under threat. Moreover, health and dependency present major risks for increasing numbers of older people. Demand for long-term care services is growing at the same time as the supply of informal care within extended families is shrinking. The high costs of long-term care represent a significant risk not only for poorer households but also for middle-income pensioners with modest assets, who may find their newly-won financial security lost as they “spend down” to the level where long-term care costs are met by public programmes. Policymakers in all OECD countries confront a conflict between improving access to, and the quality of, long-term care while also maintaining the financial security of users.1 Over the past few decades, elderly citizens have enjoyed steady gains in their incomes relative to the population at large; poverty rates in old age have dropped, both in absolute terms and relative to other age groups; and income inequality among the elderly population has tended to be lower than among people of working age (Förster and Pearson, 2002). In most OECD countries, older people today have disposable incomes that correspond on average to about 80% of that of persons aged 55 to 64 (Figure 8.1). The income of older people, relative to that of persons in later working life, is even higher among persons in the lowest quintile of the income distribution (OECD, 2002a). Lower work-related expenditures (such as transport) and a higher rate of home ownership among retirees mean that, in most countries, people experience no or only minor reductions in their standards of living when moving from working life to retirement. There are indications, however, that these long-term gains in retirement income have been stalling in the second half of the 1990s. During this period, the income of persons aged 65 to 74 (relative to that of persons aged 55 to 64) declined marginally, both on average and in most countries, while the improvement in the position of the very old (75 and over) also came to a virtual halt (Förster and Mira d’Ercole, 2005). Moreover, pension arrangements in all OECD countries are coming under pressure as the consequences of population ageing are felt more strongly and policy makers strive to make pension systems financially sustainable without jeopardising past successes in terms of income adequacy.

144


IV.8.


Figure 8.1. Mean disposable income of the elderly relative to persons aged 55 to 64, 2000 Ratio 1.0 Average

0.8 0.6 0.4 0.2

No

rw a Sw y ed De en Ne nm w ar k Ze ala n Fin d lan d Un ite Irel an d Cz Kin d ec gd o h Re m pu bl Po ic rtu g Au al st ra lia Un I t ite aly d St at Hu es Sw nga itz ry er lan d Fr a Ne nc th er e lan d Gr s ee c M e ex i Ge co rm an y Ja pa n Ca na d Au a st ria Sp ain Po lan d Tu rk ey

0

Note: Countries are ranked in increasing order of relative income of the elderly. Relative income of the elderly refers to the equivalised household disposable income of persons aged 65 and over, relative to that of persons aged 55 to 64. Differences in this ratio across countries reflect a range of factors, such as the living arrangements of elderly people and the share of persons aged 55 to 64 receiving pension income. Exact years to which the data refer are given in Figure 1.2. Source: OECD questionnaire on income distribution and poverty.

To a large extent, relatively high income during retirement has reflected the operation of mature pension systems. Gross public pension expenditure represents close to 8% of GDP in most OECD countries – and more than 10% in Austria, Denmark, France, Germany, Italy and Switzerland – around 2 points higher than in 1980 (Figure 8.2). In addition, there are significant resources that take the form of private spending: in countries with mandatory private old-age pensions, pay-outs represent close to 0.4% of GDP, while spending on pension arrangements voluntarily undertaken by individuals represents an additional 1.7% of GDP (close to three times higher than the level in 1980). While the high level of incomes in retirement highlighted in Figure 8.1 have been achieved in countries with very different public pension systems, public pensions account on average (across 25 OECD countries) for almost all of the disposable income of the bottom quintile of the elderly population, and close to 80% of that of the middle 60% of the elderly population. In each country, only elderly persons in the top quintile of the distribution have managed to achieve a significant degree of differentiation of income sources.

Figure 8.2. Gross public spending on old-age pensions, as a percentage of GDP OECD United States

United Kingdom Japan

Canada

OECD

Australia

France

Per cent of GDP 9

Germany

Sweden

Italy

14

8

12

7

10

6 5

8

4

6

3

4

2

2

1 0

Netherlands

1980

1985

1990

1995

2001

0

1980

1985

1990

1995

2001



145

IV.8.


Older people today are living longer and healthier lives than previous generations. In most OECD countries they enjoy universal coverage by public programmes against the risk of illness and injury. Medical progress over the past twenty years has greatly improved the feasibility and safety of acute interventions even for people in their 80s. Improvements in health conditions in old age have affected spending programmes in obvious ways: when pension systems were first established, most elderly people could look forward to only a few years of life in retirement; today, life expectancy at age 65 is close to 16 years for men and to 19 for women (Figure 8.3). Most of these additional years are spent in good health, with disability and limited capacity to perform everyday functions often concentrated towards the end of a person’s life. While the gap in life expectancy in old age between men and women has been widening on average during much of the past four decades, it has started to narrow among persons aged 65 and above since the late 1990s, allowing a larger number of older persons to live with their spouse for a longer time, and increasing the potential number of male care-givers in the family. To a significant extent, improvements in the health conditions of the elderly reflect the functioning of developed health systems.2

Figure 8.3. Trends in life expectancy in old age, OECD average Women Life expectancy at 65

Years 20

Men

18

8

16

7

14

6

12 1960

1970

1980

1990

Life expectancy at 80

Years 9

2002

5 1960

1970

1980

1990

2002

Note: Average of 22 OECD countries, excluding Denmark, Iceland, Ireland, Korea, Portugal, the Slovak Republic, Switzerland and the United Kingdom. Source: OECD Health Data 2004, OECD, Paris.

Despite these achievements, marked differences in health and disability status between different socio-economic groups persist in old age. Health and disability outcomes are worst for elderly persons with lower income. Women in their 80s living alone, in particular, are most vulnerable to the risk of poverty and dependency, and account for the vast majority of residents in nursing homes. In general, the prevalence of disability increases exponentially with age, and is highest for persons aged 80 and older. Although the prevalence of severe functional limitations among the elderly is declining in most countries, this is often accompanied by an increase in the prevalence of some chronic diseases, adding to the demand for health-care services.3

3. The challenges today The capacity to sustain this overall improvement in the income and health conditions of older people is at risk. Most OECD countries face the prospects of a rapid escalation of old-age expenditures. If left unchecked, such spending growth would crowd out other

146


IV.8.


components of public and social spending, weigh heavily on the income and employment prospects of future generations of contributors, and compromise macro-economic stability. Other areas where performance has been less satisfactory relate to the barriers to continued employment faced by older workers and to the persistence of risks of inadequate retirement income for some groups of the elderly population. Beyond income and employment, many countries are concerned about the low quality of services available to the frail elderly, the costs of improving access, and the risk of a widening gap between care needs and availability.

3.1. Rising financial costs of programmes targeted to older people The pay-as-you-go pension systems set up in the early and mid-20th century were not merely sets of contribution scales, vesting rules and accrual rates: they embodied a social contract between generations. This contract was based on two fundamental principles. The first was that retirees were a particularly vulnerable group within society that needed special protection. The second was that, in a context of strong economic growth, younger generations would ensure the provision of retirement income to older persons, allowing pensioners to share the benefits of economic growth through appropriately-designed public pension systems, in the expectation that they would then be provided for in turn. The favourable demographic and economic conditions that permitted this contract to work in the past have gone. As a result of declining fertility and increased longevity, nearly all OECD countries are confronting a rapid ageing of their population. For the OECD as a whole, the share of the population aged 65 and over, as a proportion of that aged 20-64, is projected to rise from 22% in 2000 to around 46% in 2050. OECD countries differ substantially in both the projected pace of ageing as well as in their point of departure (Figure 8.4). Because of population ageing, the burden of financing pension systems is today falling on a shrinking working-age population. High unemployment and early withdrawal from the labour market have compounded demographic pressures, threatening the

Figure 8.4. Considerable country differences in the projected pace of population ageing Ratio of population aged 65 and over to the population 20-64 2000

2025

2050

Population aged 65 and over as a percentage of the population aged 20 to 64 80 70 60 50 40 30 20 10

CD OE

Ja

pa n Ita Gr ly ee c Ko e Po rea rtu Cz g ec S al h pa Re in pu b Po lic lan Sl ov F d ak ra Re nce pu b Au lic Ge stri rm a a Fin ny l a Au nd st r Be alia Un H lgiu ite un m d ga K Ne ing ry w do Ze m ala No nd rw Ca ay n De ad nm a ar Sw Irel k itz and er la Lu Sw nd xe ed m en bo ur Ne Icel g th an Un er d ite lan d ds St a M tes ex i Tu co rk ey

0

Note: Countries are ranked in decreasing order of the ratio in 2050. Source: National population projections and projections by Eurostat and the United Nations.


147

IV.8.


sustainability of pension policies. OECD projections – which take into account the effects of reforms still being phased in – suggest that public spending on old-age pensions, various transfers to retired people below pension age, and health and long-term care could increase on average by some 6 to 7% of GDP over the period to 2050 (Figure 8.5). Moreover, simple projections that merely extrapolate spending rates by age in line with demographic changes – like the ones shown below – understate the spending pressure that could result in countries with less developed welfare systems as they converge to the higher standards of service quality and access that prevail in other countries. Drastic tax increases to pay for such expenditure hikes, while reducing fiscal imbalances, would come at the price of further reducing incentives for market work and saving. The effects of population ageing will be most evident for spending on old-age pensions. OECD projections – which already assume an increase in labour-force participation rates for women and in the effective age of retirement, and a decline in unemployment rates – suggest that old-age pension spending could rise, on average, by around 3 to 4 percentage points of GDP in the period to 2050. Spending on health and longterm care will also increase. For most countries, projections of health spending – based on projected per capita health care expenditures by age group, multiplied by the number of people in each group – point to an average increase over the next 50 years of between 3 and 3.5% of GDP. Similarly, public expenditure on long-term care services in OECD countries is projected to double as a share of GDP over the next 50 years, adding around 1 point of GDP to public expenditure (Bains and Oxley, 2004). As its importance within total public expenditure grows, long-term care will assume a greater importance among the priorities of health and finance ministers in all OECD countries.

Figure 8.5. Projected effects of ageing on public spending, 2000-2050 Old-age pension and early retirement programmes

Health care and long-term care

Increase in per cent of GDP 11 9 7 5 3 1 -1

pa n ed e Hu n ng ar Au y Un stra lia ite d St at es Au st r De ia nm ar Be k lg iu m Fr an c Po e rtu ga l Fin lan Ge d rm a Ne th ny er l Ne and w s Ze ala nd Cz Can ec ad h Re a pu bl ic Ko re a Sp ain No rw ay

Ja

Sw

Ita ly

gd

d Un

ite d

Ki n

Po lan

om

-3

Note: These projections of the public spending implications of population ageing are based on demographic projections undertaken in 2000-2001 (a new set of projections is currently in preparation) as presented in Dang et al. (2001); projections of health and long-term care for Austria, France, Germany, Italy and Spain, which were not part of the original projections, have been elaborated based on the same macroeconomic assumptions. Countries are ranked from left to right based on the projected expenditure increases in old-age pension and early retirement programmes. Source: Dang, T-T., P. Antolin and H. Oxley (2001), “Fiscal Implications of Ageing: Projections of Age-related Spending”, Economics Department Working Paper No. 305, OECD, Paris; and Bains, M. and H. Oxley (2004), “Ageingrelated Spending Projections on Health and Long-term Care”, Towards High-Performing Health Systems: Policy Studies, OECD, Paris.

148


IV.8.


On the other hand, other public programmes that permit early exit from the labour market through early retirement schemes, disability and special provisions for the longterm unemployed, and which currently account for about 1.5% of GDP on average, are projected to decline marginally in most countries because of lower unemployment and of reforms already undertaken. Spending on education and family benefits, which currently account for around 6% of GDP on average, are projected to fall by about 1 point between 2000 and 2050: even these modest declines, however, could fail to materialise, because of competing demands arising from longer periods of education, training needs for older workers, and higher demand for publicly-subsidised childcare as paid employment among mothers with young children becomes more common.

3.2. Barriers to continued work in old age The diminishing amount of time that people spend in work during their lifetime has aggravated the fiscal impact of population ageing. In most OECD countries, people today spend a greater portion of their life in education and retirement than in work. Under assumptions of increased longevity and unchanged retirement ages, the proportion of life spent in paid work is set to continue to decline in the future among men, while increasing moderately among women (OECD, 2000b). Removing barriers to continued work in old age plays a crucial role in adapting to population ageing. If participation rates by age and gender were to remain at current levels, population ageing would imply that the OECD labour force, which had grown strongly in the past four decades, would increase only marginally up to 2020, and start declining thereafter (Figure 8.6). This would reduce economic growth, give rise to growing labour shortages and put considerable pressure on the fiscal sustainability of social protection systems. Both the speed and the size of this slowdown will depend largely on how the labour-force participation rates of elderly people change in the future.

Figure 8.6. Population ageing could lead to a sharp slowdown in the growth of the OECD labour force Millions of persons

Historical

Constant participation rate

Maximum participation rate

Millions of persons 800 700 600 500 400 300 200 100 0 1960

1970

1980

1990

2000

2010

2020

2030

2040

2050

Note: The “constant” scenario assumes that participation rates by 5-year age groups and gender in each country remain constant at their 2000 levels. The “maximum” scenario assumes that participation rates by age and gender in each country converge by 2030 to the corresponding maximum rate observed across OECD countries in 2000 and remain constant thereafter. Source: OECD estimates.


149

IV.8.


There are currently large cross-country differences in the extent to which the potential labour supply of older people is mobilised: the employment-to-population ratio for men aged 50 to 64 is currently under 50% in Poland and Hungary, but 80% or more in Japan, Switzerland, Mexico and Iceland (Figure 8.7). Differences are even larger among men above 65 years of age, the official age to be eligible for a full public pension in many countries. While 50% or more of men aged 65-69 are employed in Japan, Korea, Mexico and Iceland, this proportion is less than 20% in most OECD countries. Differences are also large for older women, partly reflecting the wide variation in employment rates for prime-age women across countries. Removing obstacles to a longer working life holds the promise of delivering multiple benefits. Although, according to their statements, governments, unions and employers’ associations generally accept this reality, this has not yet translated into changes in individuals’ and firms’ behaviour. Both public and private employers continue to lay off older workers first; and, even though many older people claim they want to continue to work, few actually do so. Some of the reasons for these discrepancies are easy access to early-retirement benefits, the availability of alternative pathways out of the labour market, a lack of suitable employment opportunities for older workers, low skills, poor health conditions and, in some countries, a general culture of early retirement. The way that pensions are calculated also discourages work at older ages: in many countries, working an additional year, even well before the official retirement age, leads to little or no increase in the value of pension benefits, while lowering the number of years over which pensions are paid; the result is a large drop in the present value of pension wealth (the sum of future pension payments, appropriately discounted). It thus pays for workers to retire at the earliest possible age. Partly as a result, in nearly all OECD countries the average effective age of retirement is well below the official age for entitlement to a full pension (Figure 8.8).4 Beyond pension systems, special early retirement programmes, disability, long-term sickness and unemployment benefits are also used as pathways to inactivity. During the 1970s and 1980s, early retirement programmes were introduced with the dual aim of alleviating the labour market problems of older workers and – it was hoped – of increasing the labour market chances of young people. Most OECD countries have at least one scheme that allows early withdrawal from the labour market, typically five years before the statutory retirement age if certain eligibility conditions are met. Countries differ, however, in the importance of each of these pathways. Early retirement schemes are particularly extensive in Belgium, where close to 30% of the population aged 50 to 64 states that they are inactive because of “retirement”. In contrast, in Sweden and the United Kingdom – where the absence of extensive early retirement schemes has led to disability and longterm sickness benefits being used as an alternative – a majority of the older population states that they are inactive because of either “illness” or “disability”. The prevalence of early retirement is also higher in countries relying on age-profiling of their reintegration programmes for disabled workers (OECD, 2003a). This creates a culture of early exit among disabled people, which makes it easier for employers to encourage their workers at risk of job losses to request a disability benefit (for those at risk of severe disability) or an early retirement benefit (for others). The willingness of many older people to carry on working is thwarted not only by the features of benefit systems, but also by a large range of working practices and arrangements. First, in some countries, earnings rise strongly with the age of the worker: these seniority wages raise the costs of older workers, and discourage employers from

150


IV.8.


Figure 8.7. Employment rates by age and gender, 2003 Percentages Men aged 25-49

Men aged 50-64

ISL MEX LUX CHE JPN NLD AUT CZE GRC KOR PRT GBR NZL DNK IRL FRA ITA USA ESP NOR CAN AUS BEL SWE FIN DEU HUN TUR SVK POL 0

20

40

60

80

100

%

0

20

Women aged 25-49

TUR 0

20

40

80

%

Women aged 50-64

ISL SWE NOR DNK FIN AUT CHE CAN PRT NLD GBR CZE DEU FRA SVK USA NZL BEL AUS HUN IRL LUX POL JPN ESP GRC KOR ITA MEX 60

Men aged 65-69

ISL MEX JPN CHE NZL NOR KOR SWE DNK USA GBR IRL PRT CAN AUS NLD CZE ESP GRC FIN DEU FRA ITA LUX SVK AUT BEL TUR HUN POL 40 60 80 100

100

SWE NOR NZL DNK CHE USA FIN GBR CAN JPN AUS PRT KOR FRA CZE DEU NLD IRL HUN AUT POL SVK MEX GRC BEL LUX ESP ITA TUR %

0

20

40

60

80

KOR JPN TUR PRT USA NZL NOR CHE IRL CAN AUS DNK GBR GRC SWE POL CZE NLD ITA FIN DEU AUT ESP HUN LUX BEL FRA SVK 0 20 40 60

ISL MEX

80

100

%

Women aged 65-69 ISL

100

%

ISL KOR JPN USA PRT MEX NOR TUR NZL CHE DNK GBR SWE CAN AUS POL IRL GRC CZE NLD DEU FIN AUT ITA HUN ESP FRA SVK LUX BEL 0 20 40 60

80

100

%

Note: 2002 data for Iceland and Luxembourg. Source: European Labour Force Survey and national labour force surveys.


151

IV.8.


Figure 8.8. Large country differences in the effective age of retirement Age 75

Effective

Official Men

70 65 60 55

Ic ela M nd ex ic Ko o re Ire a lan Ja d Po pan Sw rtu g Un itze al ite rla d nd St a No tes r De way nm Sw ar k ed e Tu n rk e OE y C Ca D na d Ne S a Un w pa ite Ze in d ala Ki nd ng do Gr m ee Au ce st ra lia I Ge taly Lu r m xe an m y bo u Fin rg lan Ne Fra d th nc er e lan Au ds st Cz r ec Po ia h la Re nd pu Sl ov Be blic ak lgi Re um pu Hu blic ng ar y

50

Age 75

Women

70 65 60 55

Ic

ela M nd ex ic Ko o re Ire a lan Ja d Po pan Sw rtu g Un itze al ite rla d nd St a No tes r De way nm Sw ark ed e Tu n rk e OE y C Ca D na d Ne S a Un w pa ite Ze in d ala Ki nd ng do Gr m ee Au ce st ra lia Ge Italy Lu rm xe an m y bo u Fin rg lan Ne Fra d th nc er e lan Au ds st Cz r ec Po ia h la Re nd pu Sl ov Be blic ak lgi Re um pu Hu blic ng ar y

50

Note: The effective age of retirement refers to the average age at which persons aged 40 and over left the labour force during the period 1997-2002. The official age of retirement refers to the earliest age at which workers are entitled to a full old-age pension. Source: OECD estimates as presented in OECD (2005a), Society at a Glance – OECD Social Indicators, OECD, Paris.

hiring them. Second, in some OECD countries, strict employment protection has also been linked with either a relatively low mandatory age of retirement (e.g. Japan and Korea) or extensive early retirement schemes (e.g. Belgium). Third, a lack of opportunities for vocational training and skills development during prime age also implies greater barriers to continued work in later life. While the share of workers aged 50 and above who receive job-related training is in all countries lower than among workers aged 25 to 49, this share is especially low in countries where the incidence of job-related training for young workers is also comparatively low (Figure 8.9). This suggests that the negative effects of a failure to maintain skills will become more important with age. Last, OECD reviews of older workers have also found evidence of age discrimination by employers in several countries, although it is difficult to determine how pervasive this is.5

152


IV.8.


Figure 8.9. Percentage of employees by age who received job-related training over the previous year 50-64

% 60

25-49

50 40 30 20 10

d lan Po

l.) (F iu m

Be lg

s

d lan Ire

nd rla

th e

Ita ly

Ne

rla

nd

bl ic

itz e

pu Re h

Sw

lia st ra

ec Cz

da

d ite Un

Au

na

m do Ki ng

ala

Ca

nd

d

Ze w

Fin

lan Ne

St at es

ay rw

Un ite d

No

De

nm ar

k

0

Source: International Adult Literacy Survey, 1994-1998; and, for Korea, the 2000 Social Statistics Survey.

3.3. Persistent poverty for some groups of elderly In 2000, across 25 OECD countries, around 14% of the elderly population had a net household income of less than half of the median, with considerably higher percentages in some countries (Figure 8.10). Most OECD countries recorded significant falls in pensioner poverty until the mid-1990s, followed by small increases (on average and in most countries) since then. As public pensions constitute the dominant income source for the lowerincome groups, minimum pensions and safety-net arrangements for older people can be expected to have a strong impact on their poverty rates. The design of public pension systems affects old-age poverty in different ways. Ireland, for example, has a flat-rate pension set at a low level of 30% of gross average earnings (36% of net earnings) and a slightly lower means-tested pension. In a context of

Figure 8.10. Relative poverty rates among persons aged 65 and above, 2000 % 40

30

20 OECD average 10

Ne w

Ze Ne ala n t h Cz e d ec rlan h Re ds pu bl Ca ic na Hu da ng De ary nm ar Po k lan Sw d ed Ge en rm an Au y st ria Fr an c Fi e Sw nlan itz d er lan No d rw ay Un ite Sp d a Ki in ng do m Ita ly Tu rk ey Ja p Au an Un str ite alia d St at e Gr s ee ce M ex i Po co rtu ga Ire l lan d

0

Note: Relative poverty rates of persons aged 65 and over are measured with respect to a relative income threshold set at 50% of median household disposable income. Countries are ranked in increasing order of relative poverty rates in 2000. The exact years to which data refer are those indicated in Figure 1.2. Source: Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris.


153

IV.8.


rapid GDP growth, this pension level, even though indexed to earnings, has not prevented a rapid increase in the gap between the disposable income of the elderly and that of the population at large.6 In other cases, a low pension coverage – in terms of both access to the system and years of contributions – can still result in high poverty rates among the elderly.7 Safety-net arrangements for older persons who are exposed to greater risks of incomepoverty include targeted, minimum and basic pensions. In targeted schemes, benefits are either flat-rate or depend on years of service (but not on individual earnings); eligibility is limited to persons satisfying a pension test (i.e. eligibility depends on the value of other pension income), an income test (which considers all other income sources) or a broader means test (which considers the value of assets). In minimum pension schemes, benefits depend only on pension income, but are usually limited to persons with a minimum period of past contributions. In basic pension schemes, a low benefit is paid to all elderly persons irrespective of their resources and contribution history. All these redistributive schemes are publicly provided and financed on either a pay-as-you-go basis (targeted and minimum pensions) or from general taxation (basic pensions). Table 8.1 shows the combined entitlements from general social assistance (available to all persons irrespective of age) and targeted, basic and minimum contributory pensions, as a percentage of the gross earnings of an average production worker living alone, in various OECD countries. On average, this overall entitlement is just under 29% of the gross earnings of an average production worker. There is however considerable variation among countries in the form and levels of these benefits, with benefits ranging from 12% of average earnings in the Czech Republic to over 40% in Luxembourg and Canada. Most countries rely on one instrument to prevent old-age poverty, but some rely on a combination of minimum and targeted pensions. The

Table 8.1. Benefit entitlement from safety-net arrangements for older people Social Targeted assistance

Basic

Minimum

Overall entitlement

Social Targeted assistance

Basic

Minimum

Overall entitlement

Australia

–

23

–

–

23

Korea

Austria

–

37

–

–

37

Luxembourg

Belgium

23

–

–

*31/38

38

Mexico

–

19

–

–

19

Canada

–

16

14

–

30

Netherlands

–

34

34

–

34 38

Czech Rep.

–

–

30

–

30

36

–

12

46

46

10

–

8

12

12

New Zealand

–

–

38

–

Denmark

–

17

17

–

34

Norway

–

33

18

–

33

Finland

–

21

–

–

21

Poland

–

–

–

24

24 44

–

31

–

29

31

Portugal

–

20

–

44

Germany

France

24

–

–

–

24

Slovak Rep.

–

–

–

*22

22

Hungary

–

–

–

22

22

Sweden

–

34

–

–

34

Iceland Ireland Italy Japan

9/16/7

–

–

25

Switzerland

–

26

–

19

26

–

28

31

–

31

Turkey

–

6

–

28

28

22

–

–

–

22

United Kingdom

–

26

20

13

33

–

–

19

–

19

United States

–

20

–

–

20

Note: Benefit entitlement for a full career worker, as a percentage of the average earnings of an average production worker. “–” means not relevant. * Belgium: values for minimum pension/minimum credits. Mexico: 5.5% of minimum wage contribution to individual accounts. The Slovak Republic and the United Kingdom: minimum pension calculated from minimum credits. Source: OECD (2005b), Pensions at a Glance – Public Policies across OECD Countries, forthcoming, OECD, Paris.

154


IV.8.


United Kingdom, with its very complex structure of public pensions, is the only country that combines all types of instruments. Despite this range of anti-poverty schemes, some groups of elderly people are still exposed to a high risk of poverty in some countries. The reasons are many. First, not all countries have safety-net benefits that extend to all elderly (in some cases, these are limited to persons with contribution records). Second, the level of safety-net benefits may not be sufficient to avoid poverty. Third, not all eligible pensioners take up means-tested old-age benefits, especially if the means test extends to other family members. Last, retirees may become increasingly poor in relation to persons of working age due to the price-indexation of benefits. Because of this range of factors, relative poverty rates among the elderly can vary significantly across OECD countries that have a similar “overall entitlement” as measured in Table 8.1 (Förster and Mira d’Ercole, 2005). To minimise the risk of poverty, several OECD countries offer additional benefits and services to pensioners, such as housing benefits, free public transport or medical services (e.g. the United Kingdom, Ireland, Denmark), or rely on the use of special means tests for pensioners (Germany) and on discretionary increases to price-indexed pension benefits (e.g. Belgium and the United Kingdom). People who rely only on safety nets run a high risk of poverty. This group is mainly composed of persons with interrupted work histories or with jobs that were not covered by formal pension arrangements. Single women above the age of 75 and living alone are most exposed to risks of poverty (OECD, 2002a): most of them had no or only a weak labourmarket attachment in their younger years and are often not entitled to survivors’ benefits (as opposed to older widows living alone). Most OECD countries, however, regard this problem as a legacy of the past, which can be solved through ad hoc interventions, rather than as a structural problem requiring fundamental reforms: women today spend more time in the labour market earning their own pension entitlements. Also, periods spent out of paid work caring for children or elderly relatives, which disproportionately affect women’s careers, are covered – at least for some years – by pension credits in most countries. Other groups of elderly are, however, exposed to high risks of poverty, and their importance could increase in the future, depending on the impact of three types of changes: ●

First, there is the impact of changes in living arrangements, particularly the extent to which divorces and family breakdowns will result in a new class of older people living alone. As the situation with respect to widows improves over time, this new group may fill their place. However, the effects are complex, as both partners are now increasingly earning their own pension entitlements, and household re-formation is increasingly frequent.

●

Second, there are the potential consequences of recent pension reforms. Most OECD countries have adjusted their pension arrangements to establish a closer link between contributions and benefits. Reforms have also promoted the role of private pension schemes and correspondingly reduced that of public benefits. The downside of these reforms is that they make it more difficult for workers with less than full careers to earn an adequate pension: workers frequently moving in and out of dependent employment, those with non-standard contracts and the long-term unemployed could end up with


155

IV.8.


very low entitlements to old-age pensions, and have to rely on safety-net arrangements for their daily living upon reaching pension age. ●

Third, there are the effects of measures that have closed down avenues into early retirement. While these steps are welcome, for some particularly vulnerable groups later retirement is hardly an option. These groups tend to include people with work-limiting health problems, older unemployed workers who lose their jobs before sufficient pension entitlements have accrued, minority groups, migrants not entitled to full public pensions, persons without access to supplementary employer pensions and those with low educational attainment.

3.4. A potential gap between demand and supply for long-term care Population ageing and the higher number of elderly living alone risk increasing the need for care at home at the very time that the growing participation of women in paid employment is reducing the availability of such care within the family. This poses the challenge of adapting informal and formal care to this new situation, and of developing systems to monitor and enforce the quality of care.

Higher labour force participation of women may reduce the availability of informal care services In all OECD countries, informal care provided by family and friends accounts for the dominant share (80% or more) of the long-term care provided to the frail elderly (OECD, 2005c). Most informal care-giving takes the form of help with instrumental activities of daily living provided to elderly people with lower caring needs, but it may also involve intense assistance provided to persons with more severe disabilities, such as dementia patients. Most informal care is provided by women (Table 8.2). Women are predominant among those with the heaviest caring commitments, and their share increases in line with the number of hours spent on care. Care-giving is also highest among women aged 45 to 65; women in this age group often have multiple care responsibilities, such as caring for elderly parents and for a spouse with age-related health problems (OECD, 2005c). The availability of long-term care at home is linked to several factors. A first factor is the living arrangements of older people. Older persons living with their partners or other family members are more likely to receive informal help in the form of personal services (e.g. washing, toileting, help to get in and out of bed) and routine help with home-making, shopping, meals and transportation. Unfortunately, over the past decade, living alone has become more frequent among the elderly in most OECD countries, with the exceptions of New Zealand, the United Kingdom and the United States (Figure 8.11). Living alone is of course more frequent as people age and among women.8 There is more uncertainty on whether living alone will become more or less common in the future, as the impact of the greater numbers of elderly without descendents, higher frequency of divorce and partnership separation, and higher geographical mobility (all of which will tend to raise the share of elderly people living alone) may be offset by a projected decline in the mortality differential between older men and women: a modelling exercise for the United Kingdom suggests that, between 1996 and 2031, the number of dependent elderly people living with others will increase faster than the number of persons living alone, with the proportion of dependent elderly people living alone declining from 43% in 1996 to 38% in 2031 (Pickard et al., 2000).

156


Gender of care-giver Country (source)

Australia

Year

1998

(ABS Survey of Disability, Ageing and Carers, 1998)

Austria

2002

(Microcensus 2002)

Canada

1995

Germany

1998

(Schneekloth and Müller, 2000)

Ireland

1993

(Survey of older persons, 1993)

Total

Gender of care-giver Country (source)

Male

Female

Year

Relationship

Total Male

Partner

43

19

24

Parent

22

3

19

Child

24

6

19

Other

13

Other

11

2

9

Total

100

Total

100

30

71

Partner

18

7

11

Child

38

14

24

Korea

2001

(Survey on long-term care needs of the elderly, 2001)

Spain

1999

(Survey on impairment, disabilities, and handicaps)

Partner

32

Child

55

Partner

23

Child

38

Other

43

12

32

Other

39

Total

100

34

66

Total

100

Partner

20

7

13

Child

46

Parent

9

Other

53

Child

35

9

26

Total

100

Other

45

11

34

Total

73

Sweden

2000

(Survey of aged care, 2000)

100

27

Partner

32

12

20

Parent

13

2

11

Child

28

5

23

United Kingdom

2000

(General household survey, 2000)

Partner

7

Child

43

Other

27

1

26

Other

35

100

20

80

Totala

100

22

5

17

Parent Child

United States (National long-term care survey, 1994)

48

13

35

Other

1994

7

49

6

32

13

33

15

Parent

Total Partner

Partner

23

10

13

Child

41

15

27

Otherb

35

11

24

100

36

64

Total

Total Japan (Comprehensive survey of living conditions, 2001)

2001

Partner

36

12

25

Parent

1

0

1

Child

60

11

48

Other

3

1

3

100

24

76

Total

Note: Definition of carers and care recipients may differ between countries. a) National data on care-recipients in the different categories, which include persons receiving care from more than one care-giver, have been recalculated to add up to 100. b) Missing values are included in the category “Other”

157

Source: OECD (2005c), Long-term Care for Older People, forthcoming, OECD, Paris.

Female


(Survey on informal caregivers to adults in British Columbia)

Relationship

IV.8.


Table 8.2. Relationship between care recipient and informal care-giver

IV.8.


Figure 8.11. Older persons living alone, 1990 to 2000 Persons aged 65 and over United Kingdom United States

New Zealand Australia

Japan

Hungary

Germany

Spain

Austria

Mexico 40

30

30

20

20

10

10

1980

1985

1990

1995

Sweden Canada

Ireland

40

0

Norway Netherlands

2000

0

1980

1985

1990

1995

2000


A second factor that will affect availability of informal care at home is the labour-force participation of women. In countries where social pressures on family-carers are weak, higher labour force participation among women will lead to the substitution of informal with formal care, resulting in additional costs for families and public budgets. In other countries, where social pressures lead to women being carers by default, caring would take priority over – and in some case preclude – paid work. While, in these countries, there may be no shortage in the supply of home-care, the consequences for public policies remain important, as the projected increases in the demand for long-term care will reduce labour supply and gender equity. Survey evidence for European countries suggests that prime-aged women who report having started or increased care-giving reduce their hours of paid work as a consequence, and that many of these prime-aged women do not return to employment, nor resume their hours of work, when they stop or reduce their provision of care (Spiess and Schneider, 2003). Information about the potential impact of a reduction in the availability of informal care (as well as of a range of other factors) on private and public spending on long-term care in four European countries is provided in Figure 8.12.9 Within each panel, the first bar shows the actual level of spending on long-term care in 2000 and the level that could prevail in 2050 in the “central baseline” scenario; the remaining bars show how changes in various assumptions – about the availability of informal care, onset of dependency, and growth of unit costs of longterm care – might alter spending levels relative to a “comparative baseline” based on zero growth in unit costs and GDP (shown as a line). When the share of elderly persons exclusively relying on informal care declines by 0.5% per year (i.e. a cumulative fall of 25 points by 2050), spending for long-term care increases by small amounts in all countries when care is transferred to formal home care services, but by larger amounts – from between 0.2% of GDP in the United Kingdom and 0.8% in Spain – when care is transferred to institutions. The effects of lower availability of informal care on total spending are large in Italy and Spain, where informal care arrangements are unusually predominant.10 Even this increase in total spending understates the economic costs of long-term care. Most frail elderly receive care from several sources at the same time; for older people receiving both formal and informal services at home, a lower supply of care provided by family members will increase the amount of formal care services demanded and its associated spending. More

158


IV.8.


Figure 8.12. Sensitivity of long-term care spending to different factors in four European countries Spending as a percentage of GDP in 2050 5 4 3

5

United Kingdom

2.89

2.99

Comparative baseline projections (2.75)

3.69

4 3 2.36

2.36 2

2

1

1

0

1.36

2.82

1.98

2.26

Central baseline scenario

Lower availability of informal care

Delay in onset of dependency

Higher growth in unit costs

5

2.18 1.62

0.99

2.07

1.26

1.96




1.23

1

4.42

Germany

3 2.06

3.02


3.32

Comparative baseline projections (1.39)

Comparative baseline projections (1.94) 1.53

4

3

0

0

2.55

5

Spain

4

2

Italy

3.07

Comparative baseline projections (2.72) 2.11

2 1

0.65

1.52

1.06

1.26





0

1.24

2.81

1.58

2.59





Note: Public and private spending on long-term care. For each country, the first bar shows the level of spending in 2000 (in dark) and the projected level in 2050 under assumptions on GDP growth and unit costs of care that are consistent with those used in projections by the EU Economic Policy Committee (“central baseline scenario”). The three remaining bars show sensitivity results relative to a “comparative baseline projection” based on zero growth in unit costs and GDP (shown as a line). The second bar shows the impact of a decline of 0.5% per year in the share of dependent elderly people receiving only informal care, under the assumption that this decline is offset either by an increase in formal home care (shown in dark) or by an increase in institutionalisation (shown in light colour). The third bar shows the impact of delayed onset of dependency associated with a one-year increase in life expectancy, a delay of either ½ a year in the first case (shown in dark) or of one year in the second (shown in light colour). The fourth bar shows the impact of alternative growth rates of unit costs for long-term care, either (0.5%) slower growth relative to the “central baseline scenario” (shown in dark) or (0.5%) higher growth (shown in light colour). Source: Based on results reported in Comas-Herrera, A.A. and R. Wittenberg (2004), European Study of Long-term Care Expenditure. Investigating the sensitivity of projections of future long-term care expenditure in Germany, Spain, Italy and the United Kingdom to changes in assumptions about demography, dependency, informal care, formal care and unit costs, report of the European Commission, Personal Social Services Research Units Discussion Paper 1840, London School of Economics, London.

generally, looking at monetary spending of long-term care is a narrow way of appraising the true economic costs of long-term care, as it neglects the opportunity costs of informal care provided by family members who would otherwise be in paid work.

Quality of formal care services is becoming a more important concern To reduce the burden falling on families, supply of formal care has expanded in recent years. OECD countries are at various stages of development of formal systems of provision of long-term care. Universal systems exist in many countries,11 while other countries (Hungary and Korea) are discussing ways to introduce public programmes of long-term EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

159

IV.8.


care insurance. Still other countries rely on social assistance programmes to assure provision of long-term care services (such as nursing home care) to poor elderly people, leaving those with incomes and assets above eligibility levels to meet most of the cost.12 These patterns of long-term care policies are partly related to labour-force participation rates of women at higher ages: care allowances now exist in Austria and Germany, where labour-force participation rates among women traditionally have been low, while longterm care services play a more important role in Nordic countries, where the labour-force participation of women is higher. As a result of these differences in the supply structure of long-term care, total expenditure on long-term care across OECD countries spans a range from 0.2 to around 3% of GDP. While less comprehensive information is available for private expenditures, public funding appears to be the most important component of formal long-term care expenditures in all OECD countries.13 Public spending on long-term care services varies across OECD countries by much more than in the case of public spending on health or old-age pensions, and the largest share of this spending takes the form of outlays linked to care in institutions (Figure 8.13).

Figure 8.13. Expenditure on long-term care, 2000 Public

Private

Expenditure on long-term care as a percentage of GDP 3.0

Total expenditure

2.5 2.0 2.0 1.5 1.0

en ed Sw

ay rw No

lan er th Ne

d Un

ite

Un

ds

m Ki

Ge

ng

rm

do

an

y

es ite

d

Ca

St

na

at

da

lia Au

st

Ja

ra

pa

n

nd ala

Ne

w

Ze

Ire

Sp

lan

d

ain

0

Home care (including services in support of informal care) Care in institutions (nursing homes and the like) Expenditure on long-term care as a percentage of GDP 3.0

Public expenditure

2.5 2.0 2.0 1.5 1.0

n ed e Sw

ay rw No

ds er lan

Ne

th

da Ca na

y an rm

gd o Ki n

Un

ite d

Ge

m

a ra li st Au

an Ja p

es te

d

St

at

d lan

Un i

Ire

ur g bo xe m

Lu

Ze ala nd

Ne

w

Sp

ain

0


160


IV.8.


In addition to differences in the amount of expenditure associated with long-term care, there are large variations in the quality of care provided across OECD countries. Problems of access to good-quality services for older people with functional limitations manifest themselves in different ways. Some countries have broad coverage of long-term care by public programmes, but provide too limited services to support family carers, leading to problems for acute health system facilities (such as hospitals) that act as lastresort suppliers of long-term care.14 Other countries have a very limited supply of formal long-term care services (Mexico, Turkey) or are undergoing structural changes in their health systems that expose older people to the risk that their health needs might remain unmet (e.g. several countries in Central Europe).15 The quality of care provided in institutions is of special concern.16 Table 8.3 lists the quality aspects that were listed among the top three concerns of policy makers in replies to an OECD questionnaire on long-term care in nursing homes (2nd column). Human and physical resources rank high. In many OECD countries, the qualifications of care workers need upgrading, and those with adequate skills must be retained in the long-term care sector. In other countries, a basic concern is quality of buildings, which is often related to the need to move nursing-home residents to single- and double-bed rooms so as to provide them more amenities and privacy. Also important are the financial constraints facing patients: because of these, inadequate service fees often lead to insufficient staff to care for individual residents (e.g. Korea, New Zealand and the United Kingdom). Insufficient co-ordination of care services and a poorly diversified supply of services tailored to patients’ needs are also identified in several countries as factors leading to low quality of care for individual patients. Finally, a poorly developed quality assessment and monitoring systems is mentioned as an important concern by countries at very different stages in developing their systems (e.g. Austria, Korea and the United States).

Table 8.3. Policy concerns about the quality of long-term care Nursing home care

Home care

Countries

Countries

All countries that replied to this question

Most countries that replied to this question

No or poorly developed quality assessment and monitoring systems

Austria, Korea, United States

Australia, Austria, Korea

Insufficient co-ordination of care services

Canada, Hungary, Germany

Australia, New Zealand

Norway, Austria

Canada, Norway, United Kingdom

Issues mentioned

Difficulties in recruiting and retaining an adequately educated and skilled workforce

Insufficient differentiation of services provided Poor building quality and amenities Other supply constraints (indequate fees, lack of time for staff)

Hungary, Japan

–

New Zealand, United Kingdom, Korea

Korea, United States

Low skills of care managers

–

Canada, Japan

Inadequate care supply for dementia cases

–

Germany, Japan

Lack of information about services

Japan, United Kingdom

Note: Data are based on replies from national administrations to the following question(s): “What are the top three concerns in your country in terms of quality of institutional (home) care? If possible, please answer this question both from a policy perspective and from the consumers’ perspective”. Source: Questionnaire of the OECD long-term care study.

Although frail elderly cared for at home report higher satisfaction levels than those in institutions, quality concerns are also important with respect to home care (Table 8.3, 3rd column). Surveys of dependent people at home show that only a small share of them EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

161

IV.8.


receives care that is grossly insufficient, or that puts the person cared for at risk. However, these surveys have also highlighted widespread health concerns and a risk that the heavy burden on informal carers might have negative consequences on care quality. The quality concerns most noted in the case of home care are remarkably similar to those reported for nursing home care and also include large regional differences in service availability (OECD, 2005c). Supply constraints are also mentioned, both in general (Korea, United States), and for dementia patients in particular (Germany, Japan). However, many countries still have only limited evidence on the quality of home care, and most of the research in this area measures satisfaction and unmet need rather than the quality of care per se (Institute of Medicine, 2001). This underscores the importance of moving from ad hoc surveys to ongoing quality monitoring, and of relying on survey instruments that have undergone a thorough testing and validation process. Austria and Germany have recently taken steps in this direction.

4. Conclusions Thanks to a combination of strong economic growth and mature pension systems, most elderly people experience only small falls in living standard when moving from work to retirement and can expect to enjoy relatively high income over an increasing span of their life. But, in a context of population ageing, the financial costs of pension programmes have escalated rapidly, at the same time as some groups of elderly people face difficulties in retaining a foothold in the labour market, confront persistent risks of relative income poverty, and risk finding their financial security eroded by disability in old age. Policies aimed at addressing this new configuration of risks confronting older persons are described in the following chapter.

Notes 1. The objectives of “access”, “quality” and “sustainability” figure prominently in the Joint Report by the EU Commission and Council on Supporting National Strategies for the Future of Health-care and Care for the Elderly (March 2003). A detailed analysis of the policies related to long-term care is provided in OECD (2005c), Long-term Care for Older People, OECD, Paris. 2. Expert judgements on per capita expenditure by age, for 14 OECD countries at some point during the 1990s, as presented in OECD Health Data, suggest that per-capita health spending on the elderly tends to be more than twice as large as that for non-elderly persons in most OECD countries. There is however uncertainty as to the balance between the health costs associated with dying (which will be only postponed by higher life expectancy in old age) and those associated with higher use of health care services by “healthy” elderly people (which would increase as life expectancy rises). 3. In the United States, for example, the prevalence of diagnosed diabetes among persons aged 65 and over increased from around 13% in 1997 to 16% in 2002 [US Department of Health and Human Service (2003), Health, United States, 2003, National Center for Health Statistics, Washington DC.] 4. However, older men and women both retire much later than the official retirement age in Iceland, Japan, Korea and Mexico (in the latter two countries, also because of less-than-fully mature pension systems). 5. For example, a survey conducted in 2002 in the United Kingdom reveals that “age” is reported by respondents as the most common form of discrimination – ahead of gender, disability and health conditions, ethnicity and religion; older workers also report that such discrimination is most common during the recruitment process (OECD, 2004f). 6. The income-related data, however, may overestimate poverty among the elderly, as Irish pensioners (as in many other OECD countries) receive several benefits in kind. 7. In Greece, for example, pension entitlements for workers with full careers are very generous (reaching replacement rates of almost 100% for workers with a 40-year contribution career), but

162


IV.8.


very few workers have contributed to the pension system for a full career: many retirees are therefore entitled only to the (much lower) minimum contributory benefit or to means-tested benefits. In Portugal, net replacement rates of workers earning an average wage throughout their career are close to 80% (and above 100% when earning only half average earnings); but the meanstested benefit for the elderly is only 20% of average earnings, which is around half of a poverty threshold set at 50% of median equivalised income. In Australia, the means-tested public pension is set at 27.5% of male average weekly earnings (which is equivalent to 23% of the earnings of an average production worker shown in Table 8.1); private pensions, however, are mandatory and their importance in preventing old-age poverty will increase as the private system reaches maturity in the future. 8. Around 50% of all persons aged 75 and over live alone in Canada, and around 40% in New Zealand, Sweden and the United Kingdom. Women 75 and over have the highest risk of living alone, especially in Canada, Sweden and the United Kingdom. 9. Results are based on the 2003 European Study of Long-term Care Expenditure (Comas-Herrera et al., 2003). 10. Figure 8.12 highlights other results as well. First, the proportion of GDP spent on long-term care is projected to more than double between 2000 and 2050 in each country under the central projection. Second, a delay in the onset of dependency of one year for each year of additional life reduces projected costs by over one-third in all the countries; recent evidence from a number of OECD countries suggests that disability rates among the elderly are declining, although this is not universal (e.g. Australia). Third, growth in the unit costs of long-term care of 0.5% faster than in the baseline in each year leads to spending levels in 2050 that exceed baseline levels by one-fourth or more in all countries. 11. These include Austria, Denmark, Finland, Germany, Japan, Luxembourg, Netherlands, Norway and Sweden. 12. This applies to Canada, Ireland, Spain, the United Kingdom and the United States. In the case of acute health care, co-payments are common in these countries, but limited to a fraction of total costs. 13. In Germany, Switzerland, and Canada, public spending on institutional care has also grown faster than spending on acute health care, though it has remained constant or declined as a share of public health spending (including long-term care) in countries for which time-series data are available (e.g. Germany and the United States). Also, public spending on long-term care provided in the home setting has increased faster than public health spending in general. 14. This was the case until quite recently in Japan, before specific long-term care services became more developed. Even in countries with a longer history of developing long-term care services, there have still been difficulties with “bed blockers”, i.e. hospitals unable to discharge elderly patients because the relevant services are not available in the community to care for them. 15. Evidence on quality deficits comes from hospital records and the scrutiny of mortality among long-term care patients. Studies using the tools of forensic medicine have turned up evidence on severe care deficits among long-term care patients at the end of their life. 16. Cases of lapses from quality standards in nursing homes have been reported in many OECD countries. The most frequently reported concerns include pressure sores; malnutrition, in particular in the case of dementia care; inadequate prophylaxis and treatment of incontinency; inappropriate use of physical and pharmaceutical restraint; pain management; health risks from poor food sanitation; neglect and abuse; accident hazards; and problems related to a lack of privacy and the violation of basic patient rights. Many of these quality problems are interlinked: for example, pressure sores – which have been found to increase the mortality rate in elderly patients and the cost of medical and long-term care – can be revealing about a number of underlying problems, such as malnutrition, dehydration, inadequate time devoted to individual residents, incontinence, and the use of physical restraints.


163


PART IV

Chapter 9

More Effective Policies for Elderly People


165

IV.9.

MORE EFFECTIVE POLICIES FOR ELDERLY PEOPLE

1. Introduction Active social policies have a specific meaning when they refer to persons at the end of their working age. For people approaching retirement, active policies encompass interventions aimed at removing obstacles to continued work and at delaying labour force withdrawal. Following withdrawal from the labour force, the focus of active policies shifts to sustaining the autonomy of elderly persons in the event of disability, so as to allow them to continue living independently. A third potential role for active social policies – that of encouraging participation in unpaid activities, such as volunteering and caring for grandchildren – appears to have resulted in fewer initiatives and less success so far. This section discusses the range of reforms that OECD countries have launched to address shifts in the risks facing older workers and the elderly: it examines the measures taken to extend working life and ensure retirement income while lowering the cost of pension promises, and to improve access to and the quality of long-term care. In reviewing these measures, however, a comprehensive view of the full range of measures affecting older persons is important. This is because policies excessively focused on any single issue risk shifting costs onto other parts of the welfare system (e.g. from old-age to welfare programmes in the case of pensions) or to other agents (e.g. from the public sector to individual families in the case of long-term care). Lack of a comprehensive approach to reforms in this area may also compound political resistance, as those most affected by reforms in a specific area (e.g. pensions) may fear that reform will make it more difficult to satisfy their needs in other domains (e.g. long-term care).

2. Reducing the burden of public pensions systems while diversifying retirement income Starting in the mid-1990s, most OECD governments have taken steps to lower the costs of their pension promises. These reforms have taken the form of reducing both the level of benefits and the number of beneficiaries.1 However, while essential to preserve the financial sustainability of public pension systems, these measures also increase the risk of inadequate income in retirement. As a result, reforms have often been accompanied by steps to increase the role of the private sector in old-age income provision, and to reduce the risks that are associated with greater reliance on private sources of retirement income.

2.1. Lowering public costs by changing the parameters of the pension system Most OECD countries rely on earnings-related, defined-benefit schemes in the provision of their public pensions, in which benefits are calculated by applying an “accrual rate” to the past earnings of each worker. However, reductions in the benefit levels provided by public pension systems have rarely been made through explicit cuts of this

166


IV.9.


accrual rate: rather, they have been mainly implemented through adjustments in the parameters of pension systems. Some of the main changes include: ●

Changes in the number of years used in benefit calculation. Individual earnings are measured in different ways: measures include earnings in the last (few) year(s) of employment, earnings over a number of best years, or life-time earnings. When earnings increase with age, as is the case in most OECD countries, pension systems that use earnings in the final – or last few – years of employment yield higher benefits than those that include the worker’s early years of work. In the past, the last few years of earnings were commonly used to calculate public pension benefits. Today, most OECD countries have moved towards the use of life-time earnings.2

●

Changing the valorisation of past earnings. Pension benefits have also been made less generous through changes in the way in which wages earned in the past enter into the benefit calculation. In all public pension systems of OECD countries, past earnings are re-valued to take account of changes in living standards between the time pension rights accrued and the time they are claimed. While there is obviously no need for such valorisation in final-salary schemes, such provisions are common in defined-benefit schemes where the earnings measure in the benefit formula is calculated over a longer period. The majority of OECD countries with earnings-related schemes valorise past earnings in line with economy-wide wage growth. However, several OECD countries have recently moved from earnings to price-valorisation (e.g. France in 1993) or to a mix of wages and prices (e.g. Finland).3 Changes in valorisation can strongly influence benefit levels, since prices tend to rise more slowly than wages.

●

Changing the indexation of pensions in payment. Indexation refers to the up-rating of pensions in payment from the point the pension benefit is claimed onwards. In recent years, most OECD countries have moved or are moving away from indexation of pension benefits based on earnings towards full or partial indexation to prices. This means that the purchasing power of pensions is preserved, but that pensioners are not participating in the increasing standards of living enjoyed by workers. When poverty thresholds are set in relation to household income, price indexation leads to higher relative poverty rates among pensioners as the economy grows.4, 5

●

Linking pensions to higher life expectancy. Several OECD countries have changed their benefit formulae to include a factor reflecting increases in life expectancy at retirement. Incorporating such a factor is particularly simple in the notional account systems that have been introduced in recent years in some OECD countries (Box 9.1). Explicit links between life expectancy and pension benefits have also been introduced in definedbenefit systems. In Germany, for example, the pension value, which determines the uprating of pensions in payment, will include in the future a “sustainability factor” linked to the ratio between the number of contributors to the pension system and the number of pensioners: this system effectively combines the effects of demographics and the financial balance of the pension system. Other countries are discussing the introduction of similar elements in their benefit calculation.

While the impact of these measures has varied depending on the parameter chosen and the design of pension systems, they have generally lowered the amount of public pension benefits being paid to current and, more often, future generations of retirees. Table 9.1 describes some of the main changes implemented in selected OECD countries.


167

IV.9.


Box 9.1. Linking pensions to life expectancy: notionally defined pension systems in Italy, Sweden and Poland In notionally-defined pension systems, each worker is assigned an individual account in which contributions are recorded but not actually paid in. The system thus remains payas-you-go financed. At retirement, assumptions about life expectancy are used to convert the notional capital in each account into a stream of pension payments. As life expectancy rises, for a given notional capital in each personal account the annual pension payment falls, thus preserving the financial sustainability of the system. OECD countries that have introduced such systems differ, however, in the frequency with which the parameters of the notional systems are revised: ●

Italy uses a “transformation coefficient”, which is akin to the annuity rate in a funded defined-contribution scheme. This coefficient – which varies with the age at which the pension is claimed, with values determined according to a formula based on actuarial equivalence – is reviewed every ten years in line with changes in mortality rates at different ages.

●

Poland and Sweden use a factor called “g-value”, which is revised annually: in Sweden, this “g-value” is linked to individual retirement age and contemporaneous life expectancy (based on unisex mortality rates in the previous five years); in Poland, it is based on average life expectancy at retirement age.

Depending on the notional rate of return used to credit individual accounts, notional defined-contribution systems will also have different implications in respect to valorisation of past earnings. In Italy, contributions are up-rated in line with the five-year moving average of nominal GDP growth, and in Sweden with earnings growth; in Poland, a new rule adopted in 2004 stipulates valorisation of notional accounts in line with real growth of the wage bill (a rule that could imply, in a context of lower growth in the labour force, significant falls in pension entitlements).

To what extent do the reforms described above risk compromising the insurance goal to provide an adequate pension relative to the pre-retirement earnings of each individual? Figure 9.1 shows the replacement rates of OECD public and mandatory private pension systems that male, full career workers entering the labour market today can expect in the future. The rates are shown net of taxes and social security contributions. On average, a worker on average earnings can expect a net replacement rate of just under 70% – in other words, his or her pension income after tax will be around 30% less than individual net annual earnings over their lifetime. High income workers (earning twice the average) will receive less than 60%. Even the least generous of the countries with defined-benefit, earnings-related systems (Korea, the United Kingdom and the United States) pay a net replacement rate of around 50% at average earnings. If, as seems likely, this level of payment represents the lower bound of acceptable replacement rates through mandatory systems, the scope for further cuts in pension entitlements seems limited, particularly for low-income workers. Unless countries become willing to abandon earnings-related pensions altogether and move to a flat-rate benefit for the elderly – like New Zealand and Ireland – these levels probably represent the lower limits of benefit reductions in countries with earnings-related pension systems. Even though the average OECD low-income worker (on half average earnings) will receive a net replacement rate of about 85%, pensions for low-income workers are very low

168


Country

Change of years in benefit formula

Change in valorisation

Change in indexation

Link to life expectancy

Adjusted retirement incentives

Austria

Best 15 years to 40 years

From price to earnings valorisation under discussion

From price indexation to discretionary adjustment

No

4.2% reduction p.a. for early retirement

Finland

10 last years to lifetime average

From 50% prices and 50% earnings valorisation to 80% earnings and 20% price valorisation

From 50% prices and 50% earnings indexation to 80% price and 20% earnings valorisation

Life expectancy multiplier will be introduced in 2010

4.8% reduction at 60-64 in basic pension scheme, high accrual rate of 4.5% at 63-67 for earnings-related pension

No

Change from earnings to earnings minus contribution rate increases and contribution rate to private plans (2001); further reduction as system dependency ratio worsens (from 2005)

Change from earnings to earnings minus contribution rate increases and contribution rate to private plans (2001); further reduction as system dependency ratio worsens (from 2005)

Pension valorisation and indexation reduced as pension system dependency ratio worsens

3.6% p.a. reduction for retirement at 63-64; 6% p.a. increase for late retirement after age 65

Lifetime contributions to notional account

From earnings to 5-year moving average of GDP growth

From general price indexation to 100% prices for low pensions, reduced price indexation for higher pensions

Yes, through transformation coefficient used to convert balance into stream of pensions

Actuarially equivalent reductions through corresponding transformation coefficients from age 57

Portugal

Best 10 out of last 15 years to lifetime average earnings

From earnings to 75% prices and 25% earnings valorisation

Unchanged: prices

No

4.5% p.a. for retirement at 55-64, lower for workers with 30 year contribution record; increase of 10% p.a. for late retirement at 66-70 years

Sweden

Best 15 years to lifetime contributions to notional account

From prices to three-year moving average of earnings; brake according to reserve balance mechanism

From prices to average earnings less 1.6% growth norm; brake according to reserve balance situation

Yes, through g-value used to convert balance into stream of pensions

Actuarially equivalent reductions through corresponding g-values from age 61

Germany

Italy

Source: Information collected by the OECD Secretariat.

169


Note: The so-called “g” value used in the reformed Swedish pension system is linked to individual retirement age and contemporaneous life expectancy.

IV.9.


Table 9.1. Examples of pension reform measures, 2002

bo u Gr rg e Hu ece ng ar y I Po taly rtu ga Ne Tur l th ke er y lan d Sp s ain Au st Fin ria la Sw nd ed P en Sl G olan ov er d ak m Re any p OE ubli CD c -3 Fr 0 an Ic ce ela No nd rw a Ja y pa M n ex Be ico Sw lgi u Un itze m ite rla d nd St Au ates st Cz De rali ec nm a h Re ark pu bl i Ko c Un r ite Ca ea d na K d Ne ing a w do Ze m ala Ire nd lan d

em

Lu x

170 bo u Tu rg rk Gr ey ee Au ce s Hu tria ng ar y Ita ly Ne Sp th ai er n la Po nds rtu g Fin al Ge land rm a Po ny lan Fr d OE anc CD e Sw -30 Sw e itz den er la Ic nd ela No nd rw Sl ov Be ay ak lgi Re um pu b Cz ec J lic h ap Re an pu b Ca lic na De da nm a Un Aus rk Un ite tral ite d S ia d ta Ki tes ng do M m ex ic Ne K o w ore Ze a ala Ire nd lan d

m

xe

Lu m bo Po urg rtu g Tu al rk Gr ey ee Fr ce an Ic ce e De land nm a Au rk st Fin ria la Sw nd ed Ca en na da Ita Cz ec S ly h pa Re in pu Hu blic ng No ary OE rwa CD y B -3 Ne elg 0 th ium er lan Un ite J ds d ap K a Ne ing n w do Ze m a Au land Sw str itz alia er la Po nd lan Ko d r Ire ea Ge land Un rm Sl ited an ov S y ak ta Re tes pu b M lic ex ico

xe

Lu

IV.9. MORE EFFECTIVE POLICIES FOR ELDERLY PEOPLE

Figure 9.1. Net replacement rates of OECD pension systems for workers at different earnings levels Per cent of individual pre-retirement earnings

125

125

125 Low earner (half average)

100

75

50

25

0

Average earner

100

75

50

25

0

High earner (twice average)

100

75

50

25

0

Note: Countries are ranked in decreasing order of net replacement rates.

Source: OECD (2005b), Pensions at a Glance – Public Policies across OECD Countries, OECD, Paris.


IV.9.


in some countries when expressed relative to economy-wide earnings. In Mexico, the Slovak Republic and the United States, for example, pension benefits can be worth 20% or less of average economy wide earnings. In Germany and Poland, pensions for low-income, fullcareer workers are worth less than a quarter of average economy-wide earnings. Such low levels of old-age pensions relative to the general living standard of each country warn about risks of higher old-age poverty in the future. While some countries still need to cut back on pension entitlements, they also need to maintain safety nets for poor pensioners; in other countries, where pensions are now tightly linked to previous earnings, safety-net elements appear inadequate.

2.2. Diversifying retirement income, changing risks Cuts in public pension entitlements require individuals to step up their own efforts to provide for their retirement income. This increases the role of the private sector in financing old-age pensions, in the form of both employer-based occupational pensions and personal pension savings. However, it also shifts risks from governments and firms towards individuals, who have less capacity to hedge against these risks, and whose degree of aversion to risk increases with age. The shift towards greater individual responsibility in the financing of retirement income has occurred through either explicit provisions embodied in reforms of public pension systems, or as individuals adapt to these reforms with the aim of making up for cuts in public pensions.6 Such a shift can be expected to be especially strong among those in middle and higher-income groups. OECD calculations of prospective pension entitlements (presented above) show that, under existing rules, middle and higher-income workers should anticipate substantial falls in their standard of living unless they make additional savings to supplement their retirement income. These reforms have changed the risk confronting individuals in various ways. Prereform systems generally provided a high degree of old-age insurance: full benefits were provided under generous conditions, and were not tightly linked to earnings over an individual’s career. As a consequence, in most OECD countries even individuals with short or interrupted contribution records and low earnings could expect pension benefits that largely exceeded poverty thresholds. This is no longer the case. Pension benefits are more closely aligned to lifetime earnings, shifting the risks of interrupted contributions towards individual workers. This transfer has been complete in countries that have moved towards notional defined-contribution systems, where the individual bears all the risks of higher longevity and of shorter contribution periods.7 Risks have also been transferred towards individual contributors by parametric changes in public pension systems that blur the distinction between defined-benefit and defined-contribution schemes, and that shift the insurance goals of such schemes towards safety-net programmes. Changes in employer-based pension arrangements also result in shifting risk towards individuals. One manifestation of these changes is the move from defined-benefit to defined-contribution plans in the United States since the mid-1980s (Figure 9.2). Many companies found defined-benefits plans convenient during times of high equity returns, when these plans allowed employers to take “contribution holidays”. These contribution holidays, however, have led to the emergence of huge funding gaps for occupational pensions: on a worldwide basis, this shortfall was estimated at close to USD 1 000 billion at the end of 2003.8 Although some analysts believe that most companies will still be able to meet their pension obligations, this could occur at the cost of a serious drain on the EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

171

IV.9.


Figure 9.2. Defined-contribution and defined-benefit plans in the United States, 1985-2002 Shares of the total stock of assets % 80

Defined benefit plans

Defined contribution plans

60

40

20

0

1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002

Source: US Flow of Funds Account and OECD Global Pension Statistics.

resources available for productive investment. The largest funding shortfalls are expected to be among firms operating in mature industries such as manufacturing and transport. In today’s context of low financial market returns and more stringent accounting rules, employers in other OECD countries have followed the example of their US counterparts. In defined-contribution plans, risks related to returns on investment fall entirely on individuals, even when investment management is delegated to the company or to external financial experts. While delegation reduces the individual’s risk of making uninformed choices, it implies greater costs for administration and fund management, which ultimately are borne by the worker. Under such systems, individuals – when they reach retirement – also bear the risk involved in converting the balance of their definedcontribution account into an annuity, as the value of the annuity depends on the interest rate specified in the annuity contract, on life expectancy at retirement, and on the price charged by the insurance company for the annuity contract. The only risk for individual workers that is reduced through a shift towards defined-contribution schemes is that of benefit losses due to job changes. Private defined-benefit plans based on final salary often imply the loss of a substantial part of entitlements when workers change job and pension plan: no vesting losses occur in defined-contribution plans.

2.3. Government regulations are required to reduce the risks confronting individuals Many of these risks can be reduced through appropriate regulation. Rules may stipulate, for example, that employer-based defined-contribution plans must guarantee a minimum rate of return, or that all contributions made on behalf of workers must be preserved in the event the firm experiences financial problems. All OECD countries already have rules regarding the investment of pension assets, varying from flexible frameworks (such as “prudent person” rules) to detailed limitations on different asset classes and mandatory standards on the portability of pension rights. The importance of such rules is set to increase sharply as private provision becomes a more important component of the retirement income package. This is altering the nature of governments’ tasks from directly providing comprehensive pension coverage to a number of looser governance functions. These include: ●

172

Informing workers about their prospective pension entitlements from public systems, so as to raise awareness about the need for additional private savings. To recognise the need for


IV.9.


greater personal provision, workers need a clear understanding of what they can expect from their public pensions. Even though prospective pension entitlements are subject to uncertainty, model calculations can help workers plan their retirement savings. Sweden, Germany, France and the United Kingdom regularly send to contributors an individualised statement showing future benefits under different assumptions about future economic growth and individual wage progression. In Sweden, the so-called “orange envelope”, which is sent out by the pension administration every year, contains a personalised statement of each worker’s prospective benefit entitlements under alternative economic and financial assumptions. ●

Educating people to make informed decisions on investment and saving strategies. Trends towards defined-contribution pension plans and towards greater choice for workers on how to invest their retirement capital imply that the financial education of workers is becoming more important. Workers need to be alerted not only about investment and return risks but also about the uncertainty regarding their own life expectancy, their current and future earnings potential, and their personal circumstances and attitudes toward risk. A recent survey to assess the financial knowledge of consumers in the United States found that only 6% of persons aged 65 and older, and 12% of those under that age, could answer correctly four simple questions about investments.9 A British survey of 1200 adults conducted in 2002 found that only 13% of respondents felt that they had a good knowledge about pension issues, and only 9% felt they had a good idea of their future retirement income (even among respondents less than 15 years away from retirement).10 These examples, from countries where private pensions are already widespread, underscore the need for governments to step up efforts in financial education. While many OECD countries have launched such programmes (e.g. Australia), few evaluations exist.

●

Enforcing and monitoring private pension schemes. Governments have several options to encourage increased coverage of private pensions. Some countries (e.g. Switzerland, Iceland, Finland, Sweden and the Netherlands) have made the provision of occupational pensions, by employers or groups of employers, statutory or quasi-mandatory. Other countries (e.g. Poland, Hungary and Mexico) have made individual retirement-savings mandatory. Still other countries have attempted to stimulate private pension savings through tax breaks. However, the net cost – in terms of forgone tax revenues to the government – of a dollar invested in a tax-favoured pension scheme may be as high as 40 cents (Figure 9.3).11 This implies that, when private savings are reallocated from taxed vehicles to tax-favoured schemes, public and national savings deteriorate.12 This raises a number of questions concerning the policy framework for private pensions. For example, higher private savings are more likely where low-income persons participate in such schemes; however, the returns to low-paid workers from participating in such schemes are typically low, as income from private pensions interacts with incometesting of public transfers, and because tax allowances for private pension savings tend to benefit high-income individuals; in addition, the high administration costs of personal pension plans may include fixed components that are particularly unfavourable to low-income people. Thus, if private savings are to be raised, some form of collective organisation of such schemes may be required. Employer-based schemes may also be more effective than individualised schemes in generating participation by low-income individuals, but they need to be regulated carefully to avoid vesting losses.


173

IV.9.


Figure 9.3. Net tax cost per dollar of contributions to tax-favoured pension schemes Cents 45 40 35 30 25

OECD average

20 15 10 5 N Me Sl ew xic ov Ze o ak ala R n Lu epu d x e bl m i Ne bo c th urg er lan Sw ds e Be de n lg iu Gr m ee c Ko e re a Ita Ic ly ela nd Sw Pola itz nd er la Au nd st ri Sp a ain Fin lan Ja d De pan nm Po ark rtu g Un Tu al ite rk d ey St a Hu tes ng Au a r y st ra Ire lia l an Un ite No d d rw Ki a y ng do Ca m na d Fr a a Cz Ge nce ec rm h Re any pu bl ic

0

Note: Discounted effects on tax revenue, over the lifetime of an individual, of switching from saving in a taxed vehicle to a tax-favoured pension scheme. Based on employer-sponsored schemes (except for Italy and Korea) and annuity pension income. Source: Antolin, P., A. de Serres and C. de la Maisonneuve (2004), “Long-term Budgetary Implications of Tax-favoured Retirement Saving Plans”, Economics Department Working Paper No. 293, OECD, Paris.

●

Monitoring replacement rates and poverty among pensioners. Finally, governments will need to monitor carefully whether the reformed pension systems are effective in delivering adequate income protection in old age. As public and private pensions are supervised by different agencies in most countries, more co-ordination is needed. Data on the coverage and benefit entitlements of private pension arrangements remain scarce in many OECD countries, and these are crucial for projections of total pension benefits. Better information on poverty at the household level would also help in designing more effective measures targeted to individuals – in particular workers with interrupted contributions who are exposed to a greater risk of poverty – and it would allow policy makers to prepare for the greater pressure on safety nets that could emerge in the future.

3. A longer working life reduces both old-age poverty and pension costs Working longer is a highly effective step to deal with the consequences of population ageing. OECD simulations have suggested that, under a set of simplified assumptions, a rise in the effective age of retirement of about one year – sustained through time – would assure that population ageing does not lead to a higher ratio of public debt to GDP over the period to 2050 (Dang et al., 2001).13 In the past, many governments deliberately encouraged early retirement, thinking this would free up jobs for youth, but this strategy has turned out to be ineffective in attaining that goal, and very expensive in financial terms. Reforms introduced in several countries to make “working longer” pay have followed a number of parallel tracks.

3.1. Pathways to early exit from the labour market must be closed Special early retirement provisions have traditionally existed in most OECD countries, and are usually available to workers with special characteristics (e.g. those in hazardous jobs) or in specific sectors (e.g. those affected by structural adjustment). In addition, disability and unemployment benefit schemes have traditionally been important pathways

174


IV.9.


to labour-market inactivity in many OECD countries. In the past, many countries explicitly used disability programmes to enable the withdrawal of older unemployed or sick workers who had little chance of reintegrating the labour market. Closing pathways to early retirement across the full range of benefit programmes is important for the success of reform in this area: if early retirement programmes are closed down but other schemes are left unreformed, the problems are likely to resurface in other parts of the social protection system. Most OECD countries have recently taken measures to close these pathways to early retirement. These measures, however, involve difficult trade-offs. Both population ageing and higher standard retirement ages can be expected to increase the number of both disabled and unemployed older workers. Working longer will not be possible for all workers, and rising numbers will need disability or unemployment benefits until they become eligible for a standard old-age pension. In France, for example, where the standard retirement age in the public pension system is 60, more than half of all workers are already inactive when they start receiving their old-age pension, with most of them relying on unemployment and special early retirement benefits. Further, most private pension schemes also offer early retirement options. In many countries, employers use these schemes to shed older workers, often supplementing these early pensions with bridge financing until the workers become eligible for the public pension benefit.

3.2. Increasing the standard retirement age A higher standard age of retirement, beyond reducing the number of beneficiaries of old-age pension in any given year, also indirectly reduces benefit entitlements: workers need to contribute for more years to receive the same benefit as before or, if they prefer to retire at the same age, must accept a lower pension. The cost of this measure for workers depends on how exactly benefits are adjusted in the event of retirement below the standard age. Increases in the standard age of retirement were implemented in recent years, for example in Italy (with respect to “seniority” pensions), the United States, Hungary, the Czech Republic and Japan. Several other countries are increasing the retirement age for women, which in most countries is lower than that for men. Only Denmark has recently reduced the standard retirement age from 67 to 65 years. Not all OECD countries have followed the path of increasing the standard retirement age. In fact, the continued trend towards higher life expectancy raises the issue of whether the notion of a fixed standard age of retirement is consistent with new demographic realities. In this respect, some OECD countries have taken steps that effectively move away from the notion of a standard retirement age towards flexible retirement (within a range of years) so as to provide workers with greater choice as to the timing of their labour market withdrawal; such steps are usually part of broader reforms linking pension benefits to life expectancy at retirement.

3.3. Incentives to retire early should be corrected Incentives to retire have been at the centre of the pension reform debate for most of the past decade. In many countries, the way that pensions were calculated in pre-reformed systems discouraged older workers from staying in the labour market: benefits were not reduced much (or at all) when retirement was taken early, and working beyond the standard retirement age did not add to the level of the pension. Given such incentives, it is hardly surprising that many workers chose to retire as soon as possible. To avoid such EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

175

IV.9.


distortions, many OECD countries have introduced reforms that adjust the parameters of the pension system, reducing early pension benefits by an amount that corresponds both to the lower amount of contributions paid by the worker and to the increase in the period over which the worker will receive pension payments. Under such conditions, workers would be financially indifferent as to the moment of retirement, as a shorter contribution period (and longer retirement period) would be fully reflected in a lower annual benefit level, while retiring later would result in a higher benefit. Several OECD countries have taken steps in recent years to correct the incentives to retire by adjusting the increments and decrements for late and early retirement, respectively. The debate about these adjustments, however, has often been confused by a failure to distinguish between the notions of actuarial fairness and actuarial neutrality (Box 9.2). Moreover, even when the steps taken by governments have implied a move towards actuarial neutrality as a solution to early retirement – under which workers’ decisions about when to retire would not depend on the parameters of the pension system 14 – in practice, the translation of these principles into policy has been less straightforward than theory suggests: ●

First, actuarially neutral benefits are calculated on the basis of average life expectancy for the entire population. In reality, as shown in Chapter 1, life expectancy varies substantially among different socio-economic groups. Low-income workers, who generally have a shorter life expectancy, are disadvantaged in such a system: they often cannot afford to retire earlier, and even if they work longer they will receive a pension that is lower than an actuarially neutral benefit that takes into account their shorter remaining lifespan.

●

Second, strict actuarial neutrality would require lower pension benefits for women than for men, because of women’s higher life expectancy. In practice, most countries willingly subsidise women’s pensions by applying single-sex life-tables, often to compensate them for the time spent in unpaid work caring for children and elderly relatives, including their retired husbands. Also, many OECD countries explicitly exclude gender discrimination in their pension systems.

●

Third, it is especially difficult to introduce actuarial neutrality into those pension systems that offer a minimum pension guarantee or any other safety-net provision directed to low-income retirees. Actuarially neutral benefit reductions for such workers would simply push them under the threshold of income poverty, and transfer public expenditure to means-tested programmes. In other words, actuarial neutrality will always reach its limits once pension benefits come close to the poverty line where safety nets kick in.

Despite these difficulties, several OECD countries have introduced reforms aimed at removing the most obvious labour market distortions from their pension systems. But, while these measures can improve the financial sustainability of the pension system, they may not be enough to reverse trends towards early withdrawal from the labour market. Measures to correct labour market distortions bring to light the crucial question of whether the aim of such reforms should be to correct distortions or actually to skew incentives against retirement, at least up to some age, in recognition that working longer provides positive externalities (the most obvious being on public budgets). While measures that skew incentives against retirement have been introduced in some OECD countries recently,15 they run the risk of representing gains for relatively well-off individuals who

176


IV.9.


Box 9.2. The notions of actuarial fairness and actuarial neutrality applied to pension systems Governments, policy advisors and pension experts have loosely used the terms “actuarially neutral” and “actuarially fair” to describe the characteristics that reformed systems should have. But in most cases these characterisations of reformed systems encompass a variety of situations. To help better understand the characteristics of a reformed system, it may be useful to define “actuarially fair” and “actuarially neutral” in the following way: ●

An actuarially fair system is one that matches individual entitlements to lifetime contributions, with no redistribution in favour of any subgroup within any generation. Fairness here refers to the relation between what a person pays into the system and what he or she will get out of it at retirement: when contributions are fully matched by future benefits, workers will perceive them as a form of savings rather than as a tax, and this will reduce labour supply distortions. Under realistic assumptions, the condition of actuarial fairness is only fulfilled by a mandatory retirement savings programme with individual accounts (Disney, 2004).

●

An actuarially neutral system is one in which an individual’s pension wealth (the discounted stream of future pension payments) does not change when a worker retires earlier or later than a defined standard age. This means that the implicit tax on working longer (i.e. the change in pension wealth) is zero. In theory, such a system allows for full flexibility in the choice of retirement age. In practice, however, a standard retirement age is often retained under such systems due to macroeconomic considerations (to avoid large exits from the labour market that could lower economic growth), administrative and other reasons (to ensure that very early retirement and very small old-age pensions do not lead to payment of social assistance benefits).

The two notions of actuarial fairness and actuarial neutrality have different implications in terms of the characteristics of pension systems and of the behaviour of individuals. In particular, the notion of actuarial fairness is relevant to an analysis of inter- and intragenerational redistribution but not to discussion of incentives to retire early.

would have continued working anyway, with (at most) small positive effects on labour force participation and a regressive impact on income distribution.

3.4. Employers have responsibilities for lengthening working lives Governments can encourage longer working lives through a number of measures aimed at raising the supply of older workers: they can adjust work incentives in public benefit programmes; launch active labour market programmes for older people; and foster lifelong learning policies. But these actions alone will not result in the higher employment of older workers if labour demand does not rise as well. The role of employers is crucial for success: if employers do not retain or hire older workers, changes in government’s policies and in workers’ attitudes, essential as they are, will not be enough. There are several reasons why employers are often reluctant to hire or retain older workers. First, in the presence of seniority wage profiles, wages may fail to reflect the productivity of older workers. Second, strict employment legislation may encourage employers to use early retirement schemes as a convenient way of laying-off older workers when they need to reduce their workforce. Third, employers may have negative attitudes


177

IV.9.


about the abilities of older workers and discriminate against them in favour of younger workers, even though such attitudes may well have less of a basis today than in an age when manual labour was dominant. In recent years, some firms in many OECD countries have voluntarily adopted a range of initiatives to remove some of the labour market obstacles faced by older workers. In some cases, these initiatives are undertaken in co-operation by governments, employers and unions. They may involve the setting of standards for non-ageist approaches to the different aspects of human resource management (e.g. the Code of Practice on Age Diversity and Employment in the United Kingdom), training programmes of firms’ managers and human resource staff (e.g. in the context of the National Programme for Ageing Workers in Finland) and the establishment of goals, obligations and specific institutions to promote a better use of older workers’ competences in the context of broader agreements (e.g. the four-year tripartite agreement on More Inclusive Workplaces in Norway). Despite these initiatives, however, employers have often remained reluctant to hire or retain older workers, and a range of policies is needed to strengthen firms’ incentives:

178

●

First is to offer wage subsidies (e.g. cuts in employers’ social security contributions) for older workers, a route followed by several countries. 16 Other countries have experimented with subsidised part-time schemes combining work and pension receipt. The experiences with such measures, however, have not been very encouraging. Wage subsidies have a rather mixed record. They are subject to large deadweight and displacement effects, so that the net employment gains may be quite low. Similarly, subsidised part-time schemes run the risk of reducing effective labour supply rather than increasing it.17

●

Second, are measures aimed to better match the earnings of older workers to their productivity. In Sweden, seniority wages have become less important over time as a result of explicit changes in wage-settings in the public sector. In Korea, the government has encouraged the introduction of a “wage-peak” system, whereby firms hold onto older workers in exchange for a cut in wages.

●

Third, are measures aimed at protecting older workers against lay-offs. Concerns that older workers often bear the brunt of layoffs as a result of company restructuring and bankruptcies, while also facing the biggest hurdles in finding new jobs, have prompted some governments to introduce penalties for firms that lay off older workers, either in the form of having to pay the costs of outplacement services (e.g. Belgium) or in terms of a tax or higher social security contribution (e.g. Austria, Finland, France and Spain). However, there is a risk that these measures may have a negative impact on the employer incentives to hire older workers.

●

Fourth, are measures to tackle negative age stereotypes. These have included specific legislation banning age discrimination (e.g. the United States), general information campaigns (e.g. the Age Positive campaign in the United Kingdom) and employer guidelines (e.g. Japan, Korea, Norway and the United Kingdom). In some countries, these measured have included provisions that restrict or ban mandatory retirement either in general (as in the United States since 1986, with the exception of certain occupations) or before a specific age (67 in Sweden; 60 in Japan, where the 2000 revision of the law called for a rise in the age of mandatory retirement and for specific measures by firms to guarantee employment until age 65).


IV.9.

●


Last, employers may be induced to adjust working conditions to the capacities and constraints of older workers through technical assistance from public authorities (e.g. the National Programme for Ageing Workers in Finland). A number of measures have also been taken to strengthen employer-provided training of older workers (e.g. through tax incentives in the Netherlands, the national collective agreement on Employee Lifelong Access to Training in France).

3.5. Higher employment among the elderly will reduce the risk of falling into poverty Working until later in life will help to increase both labour supply and living standards in old age. Burniaux et al. (2003) provide estimates of the impact on the labour-force participation of older workers (separately for persons 55 to 64, and 65 and over) of three of the reforms discussed earlier: i) removal of early retirement schemes; ii) moves towards actuarial neutrality in public pension systems; and iii) an increase of the standard retirement age to 67 (currently the highest level observed in OECD countries).18 Figure 9.4 shows the increase in participation rates for persons aged 65 and over associated with these reforms. The increase in participation rates, relative to the levels that would have prevailed in 2025 in the absence of reforms, ranges from less than one-third in some of the OECD countries where they are already high (e.g. Mexico and the United States) to more than fivefold in some of the countries where they are relatively low (e.g. Austria, France and Spain).

Figure 9.4. Potential impact of various reforms on labour-force participation of the elderly % 40

Standard age raised to 67

Early schemes removed

Actuarial equivalence

Baseline 35 35

35 30 25

23

20 15 10 5

2 1

3 1

3 2

8

8

8

8

4

4

2

6

3

9

11 10 11

7

3

9

2

7

8

5

6

13

6

8

3

11 10

9

19 19 19

3

14 17 20 18 20 33

Sl

ov Den ak m Re ark pu Cz H blic ec un h ga Re ry pu Ge blic rm a Sw ny ed e Lu Po n xe lan m d bo u Ca rg n Be ada lg iu Gr m ee c Fr e Un A anc ite us e d tra Ki lia ng do Fin m lan Ne Nor d w wa Z y Ne eal th an er d lan ds Ita ly Sp ain Sw Irela itz nd er OE land CD Au 30 st r Un Tu ia ite rk d ey St at Ic es ela n Ja d pa n Ko r M ea ex ico

0

7

17 17 17 15 16 16 14 15 12 13 13

25

Note: Labour force participation rates of persons aged 65 and over. Countries are ranked in increasing order of the participation rate that could be achieved as a result of three sets of policies: i) removal of early retirement schemes; ii) moves towards actuarial neutrality in public pension systems; and iii) an increase of the standard retirement age to 67. In the case of Mexico, for example, these three measures would increase the labour force participation rate of elderly persons from a projected baseline level of 33% in 2025 to around 35%. Source: Based on data shown in Burniaux, J.M., R. Duval and F. Jaumotte (2003), “Coping with Ageing: A Dynamic Approach to Quantify the Impact of Alternative Policy Options on Future Labour Supply in OECD Countries”, Economics Department Working Paper, No. 371, OECD, Paris.

These higher labour force participation rates will also translate into a lower risk of falling into poverty for elderly people. Figure 9.5 shows how changes in employment rates among people aged 65 and above could impact on their poverty rates.19 The first bar shows


179

IV.9.


the poverty rates among the elderly around the year 2000; the second shows the poverty rates of older persons when employment rates increase. The decline in poverty rates is small in most countries, i.e. almost doubling the employment rates of the elderly reduces their poverty rates by only 10% on average. This relatively small impact reflects two main factors. First, while work status makes a large difference on poverty rates of the elderly living alone, this group only represents a small proportion of the elderly population in all countries. Second, within two-adult households, poverty rates of households with one worker are not always much lower than those of jobless households. The reduction in the poverty rates of elderly people is however larger in some European countries where their employment rates are currently low, and where the impact of reforms on labour force participation would be greater.

Figure 9.5. Potential impact of higher employment rates for elderly people on their relative poverty rates Baseline

% 30

Alternatve

25 20 15 10 5

St d

ite Un

Ki d ite

Un

at

do

es

m

d

ng

er

lan

en Sw

itz

ed

ay

Sw

rw

ds

No

lan er

Ne

th

ly

ico ex M

Ita

d

ce

lan Ire

y an

ee Gr

d

k

ce an

rm Ge

Fr

lan Fin

ar nm

da

De

ria

na Ca

st Au

Au

st

ra

lia

0

Note: In this scenario, the employment rates of elderly people increase in line with the rise in participation rates shown in Figure 9.4. The additional employment generated is distributed among categories of households with a head of retirement age according to the relative share of persons in single-adult and two-adult households. The additional employment among the elderly implies a shift from no-worker to one-worker households. As employment rates of older people change, the distribution of older people among household types is modified, while the poverty rates of each household type are kept unchanged. Source: Data based on different waves of the OECD questionnaire on income distribution and poverty.

4. Better care for increasing numbers of the frail elderly In most OECD countries, long-term care has traditionally been regarded as a responsibility that falls primarily on families. As a result, informal care continues to be the main form of care provided to the frail elderly, with governments intervening only when families cannot meet their obligation. OECD governments, however, give mixed messages about the duty of families to look after their elderly relatives. Because several recent reforms to social protection strongly promote paid work as the central goal of policy, encouraging people to drop out of the paid labour market to care for their relatives is difficult, and unlikely to succeed. At the same time, transferring these care responsibilities to the formal sector can generate excessive costs for the individuals affected, as well as costs for the public sector. Policies in this field have hence tried to expand supply, inter alia by providing support to informal carers, to increase consumer choice, to improve quality and to diversify the financing of care.

180


IV.9.


4.1. Specific policies are needed to support informal carers of older people Because of the greater emphasis on encouraging labour market participation among women, and their crucial role in providing informal care, it is important to ensure that caring responsibilities can be combined with paid work. Formal home-care services initially were mainly directed at older people living without a co-resident family carer, with the goal of delaying or preventing a move into a nursing or social home. This approach still dominates policies in countries where formal care services are less available or nonexistent (e.g. Korea, Mexico and Spain) and remain important also in other countries because of the increase in the share of older people living alone in recent years. This approach is however changing. Older people living alone are less likely than other older people to avoid institutionalisation in the event of a severe disability. Indeed, success in maintaining severely disabled people at home depends heavily on the existence of informal carers, living at a close distance or sometimes in the same house (Jacobzone, 1999). This has prompted a reconsideration of policies in this area, with greater emphasis now being given to supporting the main informal care-giver through services. Evidence of the heavy burden falling on family carers is an additional reason for developing services targeted to family carers. Policies in this area were often inhibited in the past by the concern that a greater supply of formal services to carers would lead to the lower availability of family care. However, there is little evidence that families reduce their caring when formal services are available: if anything, there seem to be an increase in the hours of care provided by family carers when formal services are available. Policies to support informal carers take several forms: ●

Some countries (e.g. Australia, the United Kingdom and the United States) have published national strategies setting out the needs of carers and the role of various types of formal services in reducing their burden and making their care more effective.

●

Other countries (e.g. the United Kingdom) have given carers a statutory right to receive an assessment of their need for formal services, in addition to those needed by older people.

●

Many countries have introduced respite-care services to provide carers with a break from caring responsibilities. While, in most countries, the availability of respite care depends on an assessment and on local resources, Germany has introduced a statutory right to four weeks of respite as an element of its public long-term care insurance programme.

●

Some countries (e.g. Germany and the United Kingdom) give pension credits to those out of the labour market because of long-term care responsibilities.20

●

Last, some OECD countries have introduced various forms of payments to informal caregivers, so as to compensate them in part for their income loss and enable them to reduce other work activities. Cash benefits paid under these schemes are higher in Sweden and lower in Australia and Japan. These schemes are not meant to compensate fully caregivers for the value of their work but rather to allow a minimum income level to persons who, because of their caring obligations, cannot hold a full-time job. In the cases of Australia (Carer Payment) and Japan (Allowance for Families Caring for an Elderly Person), payments are limited to carers with low income, while most other schemes are not.

Respite care is one of the most important services for carers. It can take the form of daily respite or of short-term institutionalisation for the person in care. More intensive home-respite services and group-living homes have been found to be particularly


181

IV.9.


important for informal carers of people with dementia (Moise, 2004). Provision of respite care has increased significantly in recent years; for example, expenditure on respite care increased by a factor of four in Australia between 1997 and 2003. However, demand for respite care remains higher than provision in most countries (e.g. around four times higher than current availability in Canada). Most services to carers are also being expanded from a low base: for example, around three-quarters of all carers in Austria reported feeling over-burdened at times, and only 14% of them were receiving any formal service help (OECD, 2005c). While recent reforms typically consider the needs and views of carers, services remain in short supply. The availability of services for carers can also vary significantly between jurisdictions, as has been reported for the United States.

4.2. Policies should provide more formal services to disabled older people at home Developing a continuum of care services, tailored to different situations and degrees of disability, would allow a greater proportion of elderly people to remain in their homes rather than in care institutions. The potential pay-off would be improved care, greater personal satisfaction for the person in care, and a lower burden on the public purse. OECD Social Policy Ministers agreed to the objective of “ageing in place” over a decade ago (OECD, 1992) and reiterated this goal in the context of their Reforms for an Ageing Society (OECD, 2000a). Over the past decade, the OECD governments have devoted significant resources to developing home care services and to improving knowledge about the most effective ways to deliver them. Despite these measures, however, there are still large differences across countries in the extent of receipt of home care services – which includes receipt of cash benefits to pay for these services (Table 9.2).

Table 9.2. Persons aged 65 and over receiving institutional and home care services Year

Share of older persons receiving long-term care in an institution

1995

5.7

2000 Austria

1996/97

Canada Germany

1997

3.1

1997

6.9

2003

3.6

2003

6.5

Australia

Year

Share of persons aged 65 and over receiving home care

5.5

2000

14.7

3.8

1997

14.4

2000

3.6

2000

14.8

1995

5.9

1995

16.0

1998

3.7

na

2000

6.0

2000

18.0

1991

6.4

1992

13.4

1995

8.8

1995

8.9

2000

7.9

2000

9.1

7.0

2000

5.4

5.1

2002

20.3

Hungary

Ireland

Luxembourg Norway

Sweden

Year

Share of older persons receiving long-term care in an institution

Year

Share of persons aged 65 and over receiving home care

2001

3.4

2001

3.9

2004

4.3

2004

7.0

1991

6.2

1990

17.6

1995

5.1

Switzerland

2000

2000

n.a.

2000

4.5

United Kingdom

2000

2000

4.6

2000

Around 5

United States 1973-74

4.5

Japan

2000

3.2

2000

5.5

1985

4.6

1992

3.0

Korea

2000

0.2

2000

0.2

1995

4.2

1996

5.3

Note: Home care services include persons receiving cash benefits to pay for these services. Data relating to receipt of home care services may be affected by differences in the definitions of home care across countries. Source: OECD (2005c), Long-term Care for Older People, forthcoming, OECD, Paris.

While it is difficult to separate the impact of specific policies from general economic and social changes, use of nursing homes by persons aged 65 and over has declined in most of the OECD countries where information is available. The decline has been

182


IV.9.


continuous since the early 1970s in the United States, and has extended to Australia, Austria and Sweden since the mid-1990s. Such declines are significant because of the simultaneous increase in the share in the population of persons aged 80 and more, who have a greater probability of being institutionalised. While this development is partly attributable to the better health and the greater income and independence of older people, as well as to better housing conditions where care may be delivered in situ, it also likely to reflect the impact of greater availability of home care services and benefits and the growth of “supported” living arrangements. Home care provision is more extensive in Norway and Sweden, while its expansion has been more recent in Australia, Austria and the Netherlands. Both Sweden and the United Kingdom have recently taken steps to target their public home services to those who are the most disabled. In Austria, the “cash allowance” scheme represents a significant investment in maintaining older people at home by supporting informal carers. Other countries have taken steps to provide more intensive home care as an alternative to institutionalisation. Australia has for some years provided Community Aged Care Packages, designed for people who may need low-level institutional care, and it recently introduced Extended Aged Care at Home to assist those in need of nursing home care. In the United States, the Medicaid programme has allowed an increasing number of “waivers” that permit states to use this programme (primarily intended to support the costs of nursing homes) to provide alternatives to institutionalisation. Austria, Germany, Japan and Luxembourg have also recently introduced schemes to direct resources to disabled older people at home (in various forms, including cash payments in Austria, service provision in Japan, and a combination of the two in Germany and Luxembourg).

4.3. Consumer choice needs to be strengthened A number of measures have been taken by OECD countries to allow more consumer direction and choice for persons receiving long-term care at home. Although differing in many respects, these measures can be subdivided into two categories: ●

Personal budgets and consumer-directed employment of care assistants. To strengthen the position of frail elderly persons as active consumers, some OECD countries have allocated each of them a personal budget to purchase care from competing agencies, or to employ directly a personal care assistant. Experience from the most long-lived of these schemes shows that relatives and friends typically provide more hours of care than they are paid for. Personal budgets may also allow the elderly person to combine the purchase of care services and of physical aids (e.g. a special bed or chair).

●

Cash payments to frail elderly who can spend it to acquire care services. In Germany, persons receiving support from public long-term care insurance can choose between care services in-kind, the Cash Allowance for Care, or a combination of the two. In Austria, all public long-term care support given to persons living at home is given in cash; a substantial part of these transfers is used to compensate informal care-givers, or simply contributes to the household budget when care is provided by a cohabiting relative. While there are no restrictions on how the German Cash Allowance for Care is spent, quality controls are periodically conducted to verify that a sufficient amount of care is provided. Compared to personal-budget and consumer-directed employment, both the Austrian and German programmes are extensive, covering respectively around 20% and 6% of the population aged 65 and over, and payments vary depending on need. When used to pay informal carers, such cash transfers are exempted from income tax.21


183

IV.9.


Early findings from demonstration programmes in the United States indicate that greater consumer satisfaction can be achieved at no additional cost to mainstream service providers; in other countries, however, greater consumer direction and satisfaction has involved substantial additional costs.

4.4. Quality of long-term care should be better monitored – and improved Although the quality of long-term care has only recently emerged as a major policy concern, several countries have already implemented measures to monitor and improve it. Approaches that have been adopted to improve the quality of long-term care include: ●

Standard setting. Many OECD countries are trying to improve the quality of long-term care by setting minimum requirements on providers (e.g. staffing ratios, minimum space per resident in institutions) as a precondition of licensing or awarding contracts. Many countries have introduced or strengthened regulations on staffing levels and qualifications, and imposed procedural requirements to improve routine care (in areas such as the use of restraint and pain control, where strict adherence to standards is key).

●

Linking monitoring of performance to quality improvement. Independent agencies to monitor the performance of service providers have been set up in both Australia and the United Kingdom, while in the United States this function is fulfilled by surveys and by a certification process administered under contracts by state agencies (Institute of Medicine, 2001). To improve transparency and reliability, several countries have recently moved away from the use of initial inspections towards a combination of inspections, self-assessment and documentation prepared by providers. Other countries have strengthened and diversified sanctions.

●

Self-regulatory approaches. Many professional societies, trade associations and other organisations have set voluntary standards that operate alongside government regulations. In the Netherlands, all firms that are members of the major care service association have to comply with quality standards developed by the industry, and the compliance of members is certified by independent agencies. Other associations of private providers rely on self-regulation to raise the quality of care (e.g. Spain and the United States). Government agencies providing long-term care may also rely on selfregulation (e.g. Norway and Sweden). In Canada, professional bodies share responsibility for enforcing care standards, assessing the quality of long-term care and health-care institutions, and making recommendations on areas for improvement (Romanow, 2002).

184

●

Regulating home care. Specific measures to assess and improve the quality of home care have recently been introduced in Australia, Canada, Germany, Japan and the United Kingdom, and are being considered in other countries (e.g. Hungary). Standards are mandatory in several countries (e.g. the United Kingdom). In the United States, states are required to certify to the Federal Government that they have methods for assuring the quality of home and community-based services. In Japan, prefectures have the authority to reduce care fees and cancel the accreditation for service providers that fail to meet national standards.

●

Information systems for monitoring quality. While most OECD countries use indirect indicators of care quality (e.g. staffing levels, physical facility standards and/or accreditation processes), a few have taken steps to develop more reliable instruments to evaluate care outcomes. One example is the Minimum Data Set for long-term care developed in the United States; this system, which is now required for all nursing home


IV.9.


residents in the United States, helps care providers to monitor the quality of care and has led to the development of a national reporting system. Similar “resident assessment instruments” have been tested in over 20 countries; they have been made mandatory in Iceland and introduced on a pilot basis in two Canadian provinces. Information systems for home care are also being developed and tested in several other countries. Information on the impact of these various regulatory and monitoring instruments on improving the quality of long-term care is still limited. In several countries, the accreditation process for nursing homes under these new regulations has revealed numerous shortcomings. It is not uncommon to see failure rates on the initial assessment of 40% or more, and there are few institutions that rank high in all areas. However, reforms are producing positive results in several countries. In Australia, the number of private homes that meet regulatory standards is rising, and the number of residents per room is falling (Box 9.3). In the United States, care in nursing homes improved substantially during the 1990s, with evidence of lower average number of deficiencies per facility and a lower frequency of inappropriate use of physical and chemical restraints. Despite these improvements, however, recent surveys found that one in five nursing homes nationally had shortcomings that could put residents in jeopardy (US General Accounting Office, 2003).

Box 9.3. Improving quality through better co-ordination of long-term care policies: Aged Care Assessment Teams in Australia In the early 1990s, Aged Care Assessment Teams (ACATs) were introduced in Australia to improve co-ordination among different types of care services. This co-ordinating mechanism has today become an established feature of the Australian care system, bringing a multi-disciplinary perspective to bear on those who may need to enter institutional care. The functions of these teams have evolved over time. Originally, they simply managed the transition among different care arrangements. Today, they play a central role in many more areas: they manage the stock of institutional care places that are subsidised by government; they have pioneered a new system of integrated assessment and funding for institutional care; and they have shifted the balance of care towards more intensive home care. Two main lessons stand out from the ACAT experience. ●

First is the importance of involving the patient’s own general practitioner in the assessment process, so as to improve trust by the frail elderly in the decision recommended by such teams.

●

Second is that the presence of a single co-ordinator provides users and their families with greater assurance and more security as to the quality of the services provided.

4.5. Cross-country differences in financing of long-term care Meeting higher demand for quality long-term care confronts policy makers with the difficult issue of how to pay for it. Answers to this question have continued to differ across countries. Higher taxes and social security contributions remains the preferred options in several OECD countries (Box 9.4). It is often argued that the use of social security contributions to pay for long-term care is the most efficient way of insuring against this risk, and that it relieves clients of the high personal costs of care or of the need to apply for social assistance when savings have been depleted. The OECD countries that have relied on social security contributions, however, have generally been able to use good-quality information to predict with some accuracy the expected costs of long-term care. While this EXTENDING OPPORTUNITIES: HOW ACTIVE SOCIAL POLICY CAN BENEFIT US ALL – ISBN 92-64-00794-6 – © OECD 2005

185

IV.9.


Box 9.4. Changes in the financing of long-term care: the experience of France The need to secure adequate financing has critically shaped the quest for a viable system of long-term care for the frail elderly in France. The establishment of such a system in France has proceeded in stages, with several changes in financing provisions: ●

In 1997 – following a series of pilot programmes at the local level launched since 1994 – a specific programme (prestation spécifique dépendance, PSD) was created. The critical choice made at this stage was that its financing should be assured by local government rather than through compulsory social security contributions. The programme relied on tight means testing for determining access, and on partial recuperation of the programme’s costs from beneficiaries’ bequests. At end-2000, the PSD covered only 15% of persons who needed external help, providing an annual benefit of FRF 41 000 (around 40% of the net earnings of an average production worker).

●

In 2002, a new programme (allocation personnalisée à l’autonomie, APA) was established. It created a universal benefit, extended to a broader range of persons with disabilities, and covering both persons cared for at home and in institutions. Most financing was still assured by local governments but also complemented by existing social security contributions and integrated by the national government (through a fond national de financement de la prestation autonomie) to allow a similar level of service across the country. Benefits increased with the severity of disability (up to a maximum of 78% of the net earnings of an average production worker for persons with the greatest disability).

●

In 2003 and 2004 – as take up of the programme was higher than expected –, reforms were introduced to control the programme’s costs. These reforms have restricted the conditions of access for persons receiving formal services at home, while striving to maintain the universality of benefits: to that end, the beneficiary’s contribution to the costs was increased through a combination of a higher income threshold for exoneration and of higher user-fees for persons with higher income. In addition, a new payroll tax – 0.3% of wages, counterbalanced by the cancellation of one bank holiday – was introduced to finance the programme. At the end of 2004, the number of beneficiaries was around 850 000 (17% of the population aged 75 and over), a level that is close to that expected by the authorities at the launch of the programme.

has solved the short-term problems by generating additional finance, these countries now face the commitment to maintain service levels in both good and bad times. The size of these commitments will partly depend on how the health of older people will evolve in the future: unless their health status improves significantly, living within a fixed financing envelope imposes difficult political choices. Sweden, for example, had to introduce meanstesting of long-term care and reduce the scope of services in order to restrain the demand for care. Other countries have taken steps to obtain more financial contributions from the frail elderly and their families. Private households in most countries share the burden of care not only through provision of informal care at home but also through co-payments and out-of-pocket expenditure for care provided in institutions. Even in universal social insurance systems, long-term care services provided in institutions are only partially covered by public programmes. In most countries, users are given a means test to determine how much they should be charged for nursing and personal care. Moreover, households that can afford to pay may decide to buy additional services from private

186


IV.9.


providers, e.g. personal care assistance, or respite care for a limited time as a relief to informal care-givers. Because of these co-payments and additional outlays, private households are an important source of financing for long-term care services, in particular in institutions. As these private expenditures are concentrated on a small number of households, they can represent a heavy financial burden on the households concerned. A high user payment is never an easy political option, especially when those who need care have few resources. Other countries have focused on taxing the asset holdings of old people. Even when old people have low current incomes, they often have significant assets, mainly in the form of their own homes. In so far as policies try to encourage “living in place”, forcing people to sell their homes is hardly an effective measure. However, such asset holdings are typically bequests to their adult children, and taxing them after the death of the frail elderly (e.g. through inheritance taxes) may re-coup part of the costs to the public purse of the long-term services provided during their lifetimes (Casey, 2003). While some countries have this type of scheme (e.g. the United Kingdom, at the local level), their provisions are typically inadequately applied, most often because of resistance from potential heirs. One alternative to public financing is private insurance. Making insurance for longterm care voluntary has, however, typically resulted in very limited coverage in most OECD countries (Colombo and Tapay, 2004). This suggests that, where private and public financing are combined, either private insurance should be made mandatory for middle and higher-income groups, or the insurance premia have to be reduced through public subsidies. The German reform has promoted private long-term care insurance by making it compulsory for persons not covered by the general health system and affordable, as a supplementary insurance, for people under the public health and long-term care system. Some OECD countries have also relied on mandatory private insurance for long-term care: in countries that already mandate private insurance for acute health care or retirement income, such systems can easily be extended to cover long-term care costs. Whether countries choose to have financing that is largely public (with some complementary private insurance) or largely private, the case for mandatory coverage is strong if high and unpredictable costs on users are to be avoided, and if social assistance programmes are to be freed from a role that they are ill-equipped to meet.

5. Conclusions Most OECD countries have already undertaken major reforms to respond to the new configuration of risks faced by the elderly. Reforms need to be pursued in order to: ●

Limit the costs of old-age pensions on public budgets. To ensure that further benefit cuts do not lead more and more retirees to depend on old-age safety nets, however, reforms need to promote greater diversification of retirement income among middle and higherincome retirees and to better target public pensions on lower-income retirees.

●

Favour a longer working life, through a combination of steps to close pathways to early retirement, increase the standard age for entitlement to a public pension, correct disincentives to retire early that are embedded in pension systems and encourage employers to hire and retain older workers.

●

Improve quality and access to long-term care, through policies to support informal carers, increase the availability of formal services to frail elderly cared for at home, increase the choice of users among alternative types of provision and better monitor and improve care quality.


187

IV.9.


Notes 1. A different approach followed by some OECD countries (e.g. Belgium and Ireland) has been to increase the degree of pre-funding of future pension liabilities: under this option, primary budget surpluses are used to build up assets that are then used to finance higher spending on old-age pensions. 2. Several OECD countries, however, still rely on earnings gained in a smaller number of years for their benefit calculation: France (which is now moving to 25 years), Norway (best 20 years), Spain (last 15 years) and Greece (final 5 years). All other OECD countries use periods of 30 to 40 years, or life-time average earnings. In Austria, for example, the 2003 pension reform lengthened the period used for calculating benefits from the best 15 years to 40 years’ earnings. Finland is moving from the last 10 years to life-time earnings; the same holds for Hungary and Portugal. The effect of these changes is a marked reduction in benefit generosity. 3. In Finland, 80% of past earnings are valorised in line with economy-wide wage growth and the remaining 20% in line with price inflation. In other OECD countries (e.g. Belgium, Iceland, and Spain), the price valorisation of earnings has been a feature of the pension system for a longer period of time. 4. In the United Kingdom, for example, the price-indexed basic state pension is expected to fall from 17% of average earnings today to just 7.5% by 2050. For this reason, the basic state pension was increased by amounts closer to earnings than to prices between 1997 and 2003 (OECD, 2004b). 5. Sweden has chosen a different mechanism for indexation, linking benefits not just to wages but also to the financial health of the pension system. After retirement, pensions are up-rated with average earnings minus a “growth norm” of 1.6% (which is “paid” in advance when calculating the annual pension at retirement): real wage growth short of this norm implies that pensions will fall in real terms. The Swedish system also relies on a “balance mechanism” whereby if the assets of the pension system (the “buffer fund”, plus the estimated value of future contribution revenues) fall below pension liabilities, then both the indexation of pensions in payment and the returns credited to notional accounts are reduced in order to restore the balance between liabilities and assets. 6. In Sweden, the shift to a notionally-defined contribution system in 1999 has been accompanied both by a clear statement that provision will partly be a personal responsibility and by the mandatory diversion of 2.5 points of a worker's contribution rate from the public pension to private individual accounts managed by the social security system. The German pension reform of 2001 introduced generous tax privileges for individual pension plans, which – in the political debate – served to justify gradual cuts in the public pension level; pension replacement rates are projected to remain high based on the assumption of a high take-up of private pensions. France also introduced new financial instruments for individual and collective retirement savings in the 2003 pension reform. In general, however, these tax incentives are very costly and not very efficient in reaching the target population (see Antolin et al., 2004) 7. When notional accounts are not valorised with wage growth, individuals also bear the risk of labour-market changes (Poland) and economic changes (Italy). 8. OECD estimate, based on national data on funding deficits in major OECD countries with established private pension systems. 9. Questions included whether risk increases or decreases with diversification, and whether most brokers are paid on the quality of advice they offer or on the amount and type of investments they sell to clients (Esther Canja, President, AARP, testimony on “The State of Financial Literacy and Education in America”, US Senate Committee on Banking, Housing, and Urban Affairs, Washington, DC, February 6, 2002.) 10. Also, most respondents expected their retirement to last 20 years or more, yet a third of them did not have a private pension [Victoria Mayhew (2003), Pensions 2002: Public Attitudes to Pensions and Saving for Retirement, Department for Work and Pensions, Research Report No. 193, London]. 11. The results for New Zealand and Mexico shown in Figure 9.3 are driven by specific factors. In New Zealand, all types of accrued income from pension funds are taxed at a rate of 33%, while dividends and capital gains accruing from regular investment receive favourable tax treatment; in addition, employers' contributions are taxed at a rate of 21%, which is lower than the marginal income tax rate. In Mexico, income accruing from regular investment is tax exempted. 12. Current budgetary costs of private pension savings currently range between 0 and 2% of GDP across OECD countries.

188


IV.9.


13. These simulations refer to a “stylised” country characterised by median values of pension spending, primary surplus, public debt and other economic and demographic parameters. 14. Under such a system, not all workers would necessarily work longer, but at least their choice to retire at younger ages would not affect the system’s financial equilibrium. 15. Recent examples of countries that are offering higher accrual rates for older workers include Finland (where accrual rates increase gradually from 1.5% up to age 52, to 1.9% up to age 62, to 4.5% for workers between 63 and 67) and Luxembourg (where the accrual rate varies between 1.85% and 2.05%, for workers over 55 and with long contribution careers). 16. In Sweden, for example, employers can get a subsidy of 75% of the wage if they hire older longterm unemployed persons; Norway has wage subsidies for those aged 60 and at risk of becoming unemployed; Japan subsidises workers between the ages of 60 and 64; and Norway, Spain, and Belgium have reduced social security contributions for older workers. 17. In Germany, for example, a special scheme allowed part-time work to be organised in “blocks” of time rather than requiring older workers to work part-time each week; this scheme effectively amounted to subsidising early retirement, and it has since been modified. 18. Estimates of the impacts of various policies on the labour force participation rates of elderly people in Burniaux et al. (2003) are derived from both panel regressions (“low case” scenarios) and simple cross-country correlations (“high case” scenarios). Figure 9.4 refers to the “high case” scenarios. 19. As in the case of women's employment (Figure 5.3), results are based on both labour force survey and household income data for different years. Results refer to elderly people living in households with a retirement age head (i.e. they exclude elderly people living with their offspring in households with a working-age head). Information on the distribution of elderly people among different types of households in 1995 is used to calculate the number of elderly in households with a retirement-age head in around 2000. 20. The impact of disrupted employment on pension rights will be considered in OECD (2005), The Effects of Partial Careers on Pension Entitlements, forthcoming, OECD, Paris. 21. Because the Austrian Cash Allowance for Care is also the channel for support to persons with care needs so intensive as to require institutional care, the highest level reaches 132% of average private consumption per person; but only 1% of older persons receive support at this highest level. The lowest payment level is available for persons needing care for 12 hours or less per week.


189

REFERENCES

References Adema, W. (2001), “Net Social Expenditure 2nd Edition”, Labour Market and Social Policy Occasional Papers, No. 52, OECD, Paris. Adema, W. and M. Ladaique (2005), “Net Social Expenditure, 3rd Edition”, Social, Employment and Migration Working Paper, forthcoming, OECD, Paris (www.oecd.org/els/workingpapers). Adserà, A. (2004), “Changing fertility rates in developed countries. The impact of labour market institutions”, Journal of Population Economics, Vol. 17. Ahn, N. and P. Mira (2002), “A Note on the Changing Relationship Between Fertility and Female Employment Rates in Developed Countries”, Journal of Population Economics, Vol. 15. Andersson, G. (2001), “Children’s experience of family disruption and family formation: Evidence from 16 FFS Countries”, Max Planck Institute for Demographic Research, Rostock. Antolin, P., A. de Serres and C. de la Maisonneuve (2004), “Long-term Budgetary Implications of Taxfavoured Retirement Saving Plans”, Economics Department Working Paper No. 293, OECD, Paris. Arjona, R., M. Ladaique and M. Pearson (2001), “Growth, Inequality and Social Protection”, Labour Market and Social Policy Occasional Papers No. 51, OECD, Paris. Ascoli, U. and C. Ranci, eds. (2002), Dilemmas of the Welfare Mix. The New Structure of Welfare in an Era of Privatization, New York. Askenazy, P. (2004), Les désordres du travail. Enquête sur le nouveau productivisme, Seuil, Paris. Atkinson, A. B. (1998), “Social Exclusion, Poverty and Unemployment”, in Atkinson A. B. and J. Hills (eds), Exclusion, Employment and Opportunity, Centre for Analysis of Social Exclusion, London School of Economics, Case Paper No. 4, London. Australian Institute of Health and Welfare (1999), Australia’s Welfare, AIHW, Canberra. Avramov, D. (1996), The Invisible Hand of the Housing Market: A Study of Effects of Changes in the Housing Market on Homelessness in the European Union, European Federation of National Organisations Working with Homeless (FEANTSA), Brussels. Bains, M. and H. Oxley (2004), “Ageing-related Spending Projections on Health and Long-term Care”, Towards High-Performance Health Systems: Policy Studies, Paris. Bassanini, A. and S. Scarpetta (2001), “The Driving Forces of Economic Growth: panel data evidence for the OECD countries”, OECD Economic Studies, No. 33, 2001/2, OECD, Paris. Blanden, J., A. Goodman, P. Gregg and S. Machin (2002), “Changes in Intergenerational Mobility in Britain”, Centre for Economic Performance Discussion Paper No. 0517, London School of Economics, London. Blank, R. (1999), “When Can Public Policy makers rely on Private markets? The Effective Provision of Social Security”, National Bureau of Economic Research Working Paper No. 7099, Boston. Bauman, K.J. (2003), “Extended Measures of Well-Being: Living Conditions in the United Sates, 1998”, Current Population Reports, US Census Bureau, Washington DC. Bradbury, K. and J. Katz (2002), “Are Lifetime Incomes Growing More Unequal?”, Regional Review, September. Burchardt, T. (1997), “Boundaries between Public and Private Welfare: A Typology and Map of Services”, CASE paper No. 2, Centre for Analysis of Social Exclusion, London School of Economics, London. Burniaux, J.M., R. Duval and F. Jaumotte (2003), “Coping with Ageing: A Dynamic Approach to Quantify the Impact of Alternative Policy Options on Future Labour Supply in OECD Countries”, Economics Department Working Paper, No. 371, OECD, Paris.


191

REFERENCES

Castles, F. G. (2004), The Future of the Welfare State. Crisis Myths and Crisis Realities, Oxford. Casey, B.H. (2003), “How Will We Provide and Pay for Long-term Care?”, European Journal of Social Security, Vol. 5, No. 1, March. Cherlin, A. J. (1999), “Going to Extremes: Family Structure, Children’s Well-being, and Social Science”, Demography, Vol. 36, No. 4, November. Colombo, F. and N. Tapay (2004), “Private Health Insurance Study: The Benefits and Costs for Individuals and Health Systems”, Heath Working Paper No. 15, OECD, Paris. Comas-Herrera, A.A. and R. Wittenberg (eds, 2004), European Study of Long-Term care Expenditure. Investigating the sensitivity of projections of future long-term care expenditure in Germany, Spain, Italy and the United Kingdom to changes in assumptions about demography, dependency, informal care, formal care and unit costs, report of the European Commission, Personal Social Services Research Units Discussion Paper 1840, London School of Economics, London. Corak, M. (2004), “Do Poor Children Become Poor Adults? Lessons for public policy from a cross country comparison of generational earnings mobility”, in M. Corak (ed.), Generational Income Mobility in North America and Europe, Cambridge. Currie, J. (2001), “Early Childhood Education Programs”, Journal of Economic Perspectives, Vol. 15, Spring. Currie, J. (2003), “The Take Up of Social Benefits”, National Bureau of Economic Research, Working Paper No. 10488. D’Addio, A. and M. Mira d’Ercole (2005), “Fertility Trends and Determinants in OECD Countries: The Role of Policies”, Social, Employment and Migration Working Paper, forthcoming, OECD, Paris. Dang, T.-T., P. Antolin and H. Oxley (2001), “Fiscal Implications of Ageing: projections of age-related spending”, Economic Department Working Paper, No. 305, OECD, Paris. Dang, T.-T., H. Immervoll, D. Mantovani and K. Orsini (2005), “Age and Economic Well-being in OECD Countries: Anatomy of incomes, taxes and benefits in the late 1990s”, Social, Employment and Migration Working Paper, forthcoming, OECD, Paris. Davis, P. and Y. Hu (2004), “Is There a Link Between Pension-fund Assets and Economic Growth? A cross-country study”, Public Policy Discussion Papers, Department of Economics and Finance, Brunel University. Del Boca, D. (2002), “The Effect of Child Care and Part Time Opportunities on Participation and Fertility Decisions in Italy”, IZA Discussion Paper No. 427, Bonn. Diener, E. and E. Suh (1997), “Measuring Quality of Life: Economic, Social and Subjective Indicators”, Social Indicators Research, No. 40, Kluwer. Disney, R. (2004), “Pensions and Employment”, Economic Policy, July. Dobbelsteen, S.H.A.M., S.S. Gustafsson, and C.M.M.P. Wetzels, (2000), “Child Care in the Netherlands: Between Government, Firms and Parents. Is the Deadweight Loss Smaller than in the Public Daycare System of Sweden?”, Paper presented at the Conference on Families, Labour Markets, and the Well-Being of Children, University of British Columbia, Vancouver, Canada. Ellingsaeter, A.L. (1998), “The Future of the Worker-Carer Model in Norway”, Paper prepared for the Conference “Changing labour market and gender equality: the role of policy”, Oslo, October. Esping-Andersen, G. (2002), Why We Need a New Welfare State, Oxford University Press, New York. Eurofond (2004), “Fertility and family issues in an enlarged Europe”, European Foundation for the Improvement of Living and Working Conditions, Dublin. European Commission (1998), Men within Family and Work, European Network on Family and Work, No. 2/98, Brussels. European Foundation for the Improvement of Living and Working Conditions (2000), Part-time Employment Act seeks to promote combining work and care, European Industrial Relations Observatory Online. Evans, J.M. (2001), “Firms’ Contribution to the Reconciliation between Work and Family Life”, Labour Market and Social Policy Occasional Papers No. 48, OECD, Paris. Fagnani, J. (1998), “Recent Changes in Family Policy in France: Political Trade-offs and Economic Constraints”, Chapter 5 in E. Drew, R. Emerek and E. Mahon (eds), Women, Work, and Family in Europe, Routledge, London.

192


REFERENCES

Feldstein, M, and A. Samwick (1998), “Two Percent Personal Retirement Accounts: Their Potential Effects on Social Security Tax Rates and National Saving”, National Bureau of Economic Research, Working Paper No. 6540, Cambridge, Mass. Fitoussi, J-P., E. Laurent and J. Maurice (2004), Ségrégation urbaine et intégration sociale, Conseil d’Analyse Économique, Paris. Förster, M. and M. Pearson (2002), “Income Distribution and Poverty in the OECD Area: Trends and Driving Forces”, OECD Economic Studies, No.34, OECD, Paris. Förster, M. and M. Mira d’Ercole (2005), “Income Distribution and Poverty in OECD Countries in the Second Half of the 1990s”, Social, Employment and Migration Working Paper, No. 22, OECD, Paris. Francois, P. (2002), “Not-for-profit Provision of Public Services”, CMPO Working Paper No. 03/060, University of Bristol. Freeman R. B. (2001), “The Rising Tide Lifts….?”, National Bureau of Economic Research Working Paper 8155, March, Cambridge. Gallie, D. and S. Paugam (2002), Social Precarity and Social Integration, report written for the European Commission, Directorate-General Employment, Eurobarometer 56.1, Employment and Social Affairs, Brussels. Gallie, D. and S. Paugam (2004), “Unemployment, Poverty and Social Isolation: An Assessment of the Current State of Social Exclusion Theory”, in D. Gallie (ed.), Resisting Marginalisation. Unemployment Experience and Social Policy in the European Union, Oxford. Gauthier, A.H. and J. Hatzius (1997), “Family Benefits and Fertility: An econometric analysis”, Population Studies, Vol. 51. General Accounting Office (2003), Nursing Homes: Prevalence of Serious Quality Problems Remains Unacceptably High. Despite Some Decline, United States General Accounting Office, GAO 03-1016T, Washington DC. Gilbert, N. (2002), Transformation of the Welfare State. The Silent Surrender of Public Responsibility, Oxford. Gregg, P. and S. Harkness (2003), “Welfare Reform and Lone Parents Employment in the UK”, The Centre For Market And Public Organisation, University of Bristol, Working Paper No. 03/072. Grout, P. and M. Stevens (2003), “Financing and Managing Public Services: An Assessment”, CMPO Working Paper No. 03/076, University of Bristol. Gustafsson, S. and F.P. Stafford, (1992),“Child Care Subsidies and Labour Supply in Sweden”, Journal of Human Resources, Vol. 27, No. 1. Hair, E., T. Ling and S.W. Cochran (2003), “Youth Development Programs and Educationally Disadvantaged Older Youths: A Synthesis”, Child Trends, Edna McConnell Clark Foundation, July, Washington DC. Hentschel, J. Lanjouw, J.O., Lanjouw, P. and Poggi, J. (1998), “Combining Census and Survey Data to Study Spatial Dimensions of Poverty: A Case Study of Ecuador”, World Bank Working Paper, Washington DC. Hernanz, V., F. Malherbet and M. Pellizzari (2004), “Take-up of Welfare Benefits in OECD countries: A Review of the Evidence”, Social, Employment and Migration Working Paper, No. 17, OECD, Paris. Holzman, H. (2002), “Do We Need a Stronger Welfare Policy for a Weaker Economy?”, Short Takes on Welfare Policy, Urban Institute, Washington DC. Hotz, V.J., C.H. Mullin and J.K. Scholz (1999), “The Earned Income Tax Credit and Labour Market Participation of Families on Welfare”, Working Paper No. 102, Joint Center for Poverty Research, Northwestern University and University of Chicago. Ilmakunnas, S. (1997), “Public Policy and Childcare Choice”, Chapter 9 in Persson, I. and Jonung, C. (eds), Economics of the Family and Family Policies, Routledge, London. Immervoll H., P. Marianna M. Mira d’Ercole (2004), “Benefit Coverage Rates and Household Typologies: Scope and Limitations of Tax-Benefit Indicators”, Social, Employment and Migration Working Paper No. 20, OECD, Paris. Institute of Medicine (2001), Improving the Quality of Long-Term care, National Academy Press, Washington DC. Institute of Public Health (2003), Health Inequalities in Europe and the Situation of Disadvantaged Groups, Report on Socio-Economic Differences in health Indicators in Europe, Bielefeld.


193

REFERENCES

Istituto di Ricerche sulla Popolazione (2001), “Main Results on the Paid and Unpaid Studies in Finland, Germany, Greece, Italy, the Netherlands, Portugal, Spain, the United Kingdom”, Demotrends, Consiglio Nazionale delle Ricerche, Rome. Jacobzone, S. (1999), “Ageing and Care for Frail Elderly Persons: an overview of international perspectives”, Labour Market and Social Policy Occasional Paper No. 38, OECD, Paris. Jalan, J. and M. Ravaillon (1997), “Spatial Poverty Traps?”, World Bank Working Paper, Washington DC. Kamerman, S. B., M. Neuman, J. Waldfogel and J. Brooks-Gunn (2002), “Social Policy Aspects of Child Development: A Literature Survey”, Social, Employment and Migration Working Paper No. 6, OECD, Paris. Keating, M. (2004), Who Rules? How government retains control of a privatized economy, Federation Press, Sydney. Kimmel, J. (1998), “Child Care Costs as a Barrier to Employment for Single and Married Mothers”, The Review of Economics and Statistics, Vol. 2. Kingley, G.T. and K.L.S. Pettit (2003), “Concentrated Poverty: A Change of Course”, Neighborhood Change in Urban America, No. 2, The Urban Institute, Washington DC., May. Kleinman M, (1998), “Include Me Out? The New Politics of Place and Poverty”, CASE paper No. 11, Centre for Analysis of Social Exclusion, London School of Economics, London. Layard, R. (2004), “Good Jobs and Bad Jobs”, Centre for Economic Performance Occasional Paper No. 19, London. Leahy, M., S. Schich, G. Wehinger, F. Pelgrin and T. Thorgeirsson, (2001), “Contributions of Financial Systems to Growth in OECD Countries”, Economics Department Working Paper No. 280, OECD, Paris. Lindert, P.H. (2004), Growing Public – Social Spending and Economic Growth since the Eighteenth Century, Cambridge University Press, New York. Lindsay Chase-Lansdale, P. (1998), “How Developmental Psycologists Think About Family Process and Child Development in Low Income Families„, mimeo, paper presented at the Conference “Family Process and Child Development in Low Income Families, May 7, Chicago. Maddison, A. (2001), The World Economy. A Millennial Perspective, OECD, Paris. Mare, R.D. (2000), “Assortative Mating, Intergenerational Mobility and Educational Inequality”, California Center for Population Research, CCPR-004-00, University of California. Martin, G. and V. Kats (2003), “Families and Work in Transition in 12 Countries, 1980-2001”, Monthly Labour Review, September. Moise, P. (2004), “Dementia Case Study”, Health Working Paper, No. 13, OECD, Paris. Mor, V. (2003), “Henderson State-of-the-Art Lecture: health Care Quality in Nursing Homes”, Annals of Long-Term Care, Vol. 11(8), Ministry of Health, Welfare and Sport, The Hague. Nordhaus, W. D. (2002), “The Health of Nations: The Contribution of Improved Health to Living Standards”, Cowles Foundation Discussion paper No. 1355, Yale. Normand, C. L., F. Vitaro and P. Charlebois (2000), “Involving Parents in Early Prevention”, ISUMA, Autumn. OECD (1999), Preparing Youths for the 21st Century. The Transition from Education to the Labour Market, OECD, Paris. OECD (2000a), From Initial Education to Working Life. Making Transition Work, OECD, Paris. OECD (2000b), Reforms for an Ageing Society, OECD, Paris. OECD (2001a), “When Money is Tight: Poverty Dynamics in OECD Countries”, Employment Outlook, OECD, Paris. OECD (2001b), “Balancing Work and Family Life: helping parents into paid employment”, Employment Outlook, OECD, Paris. OECD (2001c), Starting Strong – Early Childhood Education and Care, OECD, Paris. OECD (2001d), The Social Crisis in the Russian Federation, OECD, Paris. OECD (2002a), Ageing and Income. Financial Resources and Retirement in 9 OECD Countries, OECD, Paris.

194


REFERENCES

OECD (2002b), Towards Asia’s Sustainable Development – The Role of Social Protection, OECD, Paris. OECD (2003a), Transforming Disability into Ability. Policies to Promote Work and Income Security for Disabled People, OECD, Paris. OECD (2003b), “Benefits and Employment, Friends or Foe? Interactions Between Passive and Active Social Programmes”, Employment Outlook, OECD, Paris. OECD (2003c), “Making Work Pay – Making Work Possible”, Employment Outlook, OECD, Paris. OECD (2003d), “More and Better Jobs? Aggregate Performance During the Past Decade”, Employment Outlook, OECD, Paris. OECD (2003e), Babies and Bosses. Reconciling Work and family Life, Vol. 2, OECD, Paris. OECD (2004a), “Recent Labour Market Development and Prospects”, Employment Outlook, OECD, Paris. OECD (2004b), OECD Economic Survey. United Kingdom, OECD, Paris. OECD (2004c), Education at a Glance – OECD Indicators, OECD, Paris. OECD (2004d), Benefits and Wages – OECD Indicators, OECD, Paris. OECD (2004e), Learning for Tomorrow’s World – First Results from PISA 2003, OECD, Paris. OECD (2004f), Ageing and Employment Policies – United Kingdom, OECD, Paris. OECD (2005a), Society at a Glance – OECD Social Indicators, OECD, Paris. OECD (2005b), Pensions at a Glance – Public Policies across OECD Countries, forthcoming, OECD, Paris. OECD (2005c), Long-term Care for Older People, forthcoming, OECD, Paris. OECD (2005d), Reconciling Work and Family Life – Social Policies for Working Families, forthcoming, OECD, Paris. Orzechowski, S. and P. Sepielli (2003), “Net Worth and Asset Ownership of Households: 1998 and 2000”, Current Population Reports, US Census Bureau, May, Washington DC. Pearson, M. and J.P. Martin (2004), “Should We Extend the Role of Private Social Expenditure?”, Social, Employment and Migration Working Paper No. 19, OECD, Paris. Percival, R. and A. Harding (2002), “The Costs of Children in Australia Today”, National Center for Social and Economic Modelling, Income and Wealth report, Issue 3, University of Canberra, October. Perry, B. (2002), “The Mismatch Between Income Measures and Direct Outcome Measures of Poverty”, Social Policy Journal Of New Zealand, Issue 19, December. Pettit, B. and J. Hook (2002), “The Structure of Women’s Employment in Comparative Perspective”, Luxembourg Income Study Working Paper No. 330, Center for Policy Research, Syracuse University, New York. Pettit, K.L.S., T. Kingsley and C.J. Coulton (2003), Neighborhoods and Health: Building Evidence for Local Policy, The Urban Institute, Washington DC. Piketty, T. (2003), “The Impact of Divorce on School Performance: Evidence from France, 1968-2002”, Centre for Economic Policy Research Discussion Paper Series No. 4146. Powell, L. (1998), “Part-Time Versus Full-Time Work and Child Care Costs: Evidence for Married Mothers”, Applied Economics, Vol. 30. Ringen, S. (1998), “The Family in Question”, Demos, London. Rodrik, D. (1997), Has Globalisation Gone Too Far?, Institute for International Economics, Washington DC. Romanow, R.J. (2002), Building on Values: The Future of Health Care in Canada, Commission on the Future of Health Care in Canada, Saskatoon. Rønsen, M. (2004), “Fertility and Public Policies – Evidence from Norway and Finland”, Demographic Research, Vol. 10, Max Planck Institute for Demographic Research, Rostock. Ruhm, C.J. (1998), “The Economic Consequences of Parental Leave Mandates: Lessons From Europe,”The Quarterly Journal of Economics, vol. 113(1), MIT Press, February. Ruhm, C.J. (2002), “Parental Employment and Child Cognitive Development”, NBER Working Paper No. 7666 (revised version).


195

REFERENCES

Salomon, L.M., W. Sokolowski (2001), “Volunteering in Cross-National Perspective: Evidence from 24 Countries”, Working Papers of the Johns Hopkins Comparative Nonprofit Sector Project, Baltimore. Skoufias, E. (2001), “PROGRESA and its Impact on the Human Capital and Welfare of Households in Rural Mexico: A Synthesis of the Results of an Evaluation by IFPRI”, International Food Policy Research Institute, Washington DC. Social Exclusion Unit (2004), Tackling Social Exclusion: Taking stock and looking to the future, Office of the Deputy Prime Minister, London. Spiess, C. K. and U. Schneider (2003), “Interactions between Care-giving and Paid Work Hours Among European Midwife Women, 1994 to 1996”, Ageing and Society, Vol. 23, Cambridge. Tanzi, V. (2000), “Globalization and the Future of Social Protection”, International Monetary Fund Working Paper, WP/00/12, Washington DC. Stall, R. and R. Lupton (2003), “Is Targeting Deprived Areas an Effective Means to Reach Poor People? An assessment of one rational for area-based funding programmes”, CASEpaper 70, Centre for Analysis of Social Exclusion, London School of Economics, London. UN Population Division (2003), World Population Policies, Department for Economic and Social Affairs, New York. US Census Bureau (2000), Statistical Abstract of the United States, Washington D.C. US General Accounting Office (2003), Nursing Home Quality: Prevalence of Serious Problems, While Declining, Reinforces Importance of Enhanced Oversight, GAO-03-361, Washington DC. WHO (2003), Investing in Mental Health, World Health Organisation, Geneva. Zedlewski S. R., Giannarelli S., Morton J. and Wheaton L. (2002), “Extreme Poverty Rising, Existing Government Programmes Could Do More”, Urban Institute, New Federalism, Series B, No. B-45, Washington DC, April.

196


OECD PUBLICATIONS, 2, rue André-Pascal, 75775 PARIS CEDEX 16 PRINTED IN FRANCE (81 2005 05 1 P) ISBN 92-64-00794-6 – No. 53885 2005

Extending Opportunities: How Active Social Policy Can Benefit Us All

All of Us Can Almost

Active Social Capital

Active Social Capital

God Hates Us All

Giving: how each of us can change the world

Reconceptualizing Social Policy (Introducing Social Policy)

All U Can Eat

All You Can Eat

God Hates Us All

God Hates Us All

All U can eat

This Book Is Overdue!: How Librarians and Cybrarians Can Save Us All

This Book is Overdue! How Librarians and Cyberarians can Save Us All

This book is overdue!: how librarians and cybrarians can save us all

When Johnny and Jane Come Marching Home: How All of Us Can Help Veterans

Nothing Can Keep Us Together

Power Rules: How Common Sense Can Rescue American Foreign Policy

Deadly Choices: How the Anti-Vaccine Movement Threatens Us All

The Death of Capital: How Creative Policy Can Restore Stability

The Power of Giving: How Giving Back Enriches Us All

Social Policy for Social Work

Deadly Choices: How the Anti-Vaccine Movement Threatens Us All

Secrecy in US Foreign Policy

Social Policy, Social Welfare and Scandal: How British Public Policy is Made

Dreaming reality: how dreaming keeps us sane or can drive us mad

All a man can ask

All a Man Can Ask

All a Man Can Be

All A Man Can Ask

All a Man Can Ask

Extending Opportunities: How Active Social Policy Can Benefit Us All

All of Us Can Almost

Active Social Capital

Active Social Capital

God Hates Us All

Giving: how each of us can change the world

Reconceptualizing Social Policy (Introducing Social Policy)

All U Can Eat

All You Can Eat

God Hates Us All

God Hates Us All

All U can eat

This Book Is Overdue!: How Librarians and Cybrarians Can Save Us All

This Book is Overdue! How Librarians and Cyberarians can Save Us All

This book is overdue!: how librarians and cybrarians can save us all

When Johnny and Jane Come Marching Home: How All of Us Can Help Veterans

Nothing Can Keep Us Together

Power Rules: How Common Sense Can Rescue American Foreign Policy

Deadly Choices: How the Anti-Vaccine Movement Threatens Us All

The Death of Capital: How Creative Policy Can Restore Stability

The Power of Giving: How Giving Back Enriches Us All

Social Policy for Social Work

Deadly Choices: How the Anti-Vaccine Movement Threatens Us All

Secrecy in US Foreign Policy

Social Policy, Social Welfare and Scandal: How British Public Policy is Made

Dreaming reality: how dreaming keeps us sane or can drive us mad

All a man can ask

All a Man Can Ask

All a Man Can Be

All A Man Can Ask

All a Man Can Ask

Recommend Documents