Gender and Social Policy in A Global Context: Shahra Razavi and Shireen Hassim

Gender and Social Policy
in a Global Context
Uncovering the Gendered Structure
of ‘the Social’
Edited by
Shahra Razavi and Shireen Hassim
Gender and Social Policy in a Global Context
Social Policy in a Development Context
General Editors: Thandika Mkandawire and Huck-Ju Kwon, both of UNRISD
Social Policy in a Development Context is a new series which places social policy at the centre
of research while maintaining the United Nations Research Institute for Social Development
(UNRISD)’s unified approach to social development. The series provides a new and exciting
contribution to the literature in economic development and social policy. In economic devel-
opment, social policy has been recognized as an integral part of development, but the litera-
ture often falls short of elaborating social policy for a unified approach to economic and social
development. In social policy, analysis has concentrated mainly on European and North
American countries, and studies on developing countries often lack comparative rigour.
The bridge between economic development and social policy will not only contribute to the
academic research but also inform the policy debate at the international and national levels.
Titles include:
Olli Kangas and Joakim Palme (editors)
SOCIAL POLICY AND ECONOMIC DEVELOPMENT IN THE NORDIC COUNTRIES
Massoud Karshenas and Valentine M. Moghadam (editors)
SOCIAL POLICY IN THE MIDDLE EAST
Economic, Political and Gender Dynamics
Huck-Ju Kwon (editor)
TRANSFORMING THE DEVELOPMENTAL WELFARE STATE IN EAST ASIA
Maureen Mackintosh and Meri Koivusalo (editors)
COMMERCIALIZATION OF HEALTH CARE
Global and Local Dynamics and Policy Responses
Thandika Mikandawire (editor)
SOCIAL POLICY IN A DEVELOPMENT CONTEXT
Shahra Razavi and Shireen Hassim (editors)
GENDER AND SOCIAL POLICY IN A GLOBAL CONTEXT
Uncovering the Gendered Structure of ‘the Social’
Social Policy in a Development Context

Series Standing Order ISBN 978-1-4039-4295-1 (hardback) 978-1-4039-4296-8 (paperback)
(outside North America only)
You can receive future titles in this series as they are published by placing a standing order. Please contact
your bookseller or, in case of difficulty, write to us at the address below with your name and address, the
title of the series and the ISBN quoted above.
Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21
6XS, England
Gender and Social Policy
in a Global Context
Uncovering the Gendered Structure
of ‘the Social’
Edited by

© UNRISD 2006
Softcover reprint of the hardcover 1st edition 2006 978-1-4039-9630-5
All rights reserved. No reproduction, copy or transmission of this
publication may be made without written permission.
No paragraph of this publication may be reproduced, copied or transmitted
save with written permission or in accordance with the provisions of the
Copyright, Designs and Patents Act 1988, or under the terms of any licence
permitting limited copying issued by the Copyright Licensing Agency, 90
Tottenham Court Road, London W1T 4LP.
Any person who does any unauthorized act in relation to this publication
may be liable to criminal prosecution and civil claims for damages.
The authors have asserted their rights to be identified
as the authors of this work in accordance with the Copyright,
Designs and Patents Act 1988.
First published in 2006 by
PALGRAVE MACMILLAN
Houndmills, Basingstoke, Hampshire RG21 6XS and
175 Fifth Avenue, New York, N.Y. 10010
Companies and representatives throughout the world.
PALGRAVE MACMILLAN is the global academic imprint of the Palgrave
Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd.
Macmillan® is a registered trademark in the United States, United Kingdom
and other countries. Palgrave is a registered trademark in the European
Union and other countries.
ISBN 978-1-349-54534-6 ISBN 978-0-230-62528-0 (eBook)
DOI 10.1057/9780230625280
This book is printed on paper suitable for recycling and made from fully
managed and sustained forest sources.
A catalogue record for this book is available from the British Library.
Library of Congress Cataloging-in-Publication Data
Gender and social policy in a global context : uncovering the gendered structure of
‘the social’ / edited by Shahra Razavi and Shireen Hassim.
p. cm.––(Social policy in a development context)
Includes bibliographical references and index.
1. Women – Government policy – Cross-cultural studies. 2. Social policy –
Cross-cultural studies. 3. Sex role – Government policy – Cross-cultural studies.
4. Women – Social conditions – Cross-cultural studies. I. Razavi, Shahrashoub.
II. Hassim, Shireen. III. Series.
HV1444.G45 2006
305.40973––dc22 2005056620
10 9 8 7 6 5 4 3 2 1
15 14 13 12 11 10 09 08 07 06
Contents
List of Tables vii
List of Figures ix
List of Acronyms and Abbreviations x
Notes on Contributors xi
Preface and Acknowledgements xv
Foreword xvii
1 Gender and Social Policy in a Global Context: Uncovering

the Gendered Structure of ‘the Social’ 1
Shireen Hassim and Shahra Razavi
Part I Historical and Regional Trajectories in

Social Provisioning
2 Mothers at the Service of the New Poverty Agenda:
The PROGRESA/Oportunidades Programme in Mexico 43
Maxine Molyneux
3 Gender and Post-socialist Welfare States in Central

Eastern Europe: Family Policy Reforms in Poland and
the Czech Republic Compared 68
Silke Steinhilber
4 Maternalist Policies versus Women’s Economic

Citizenship? Gendered Social Policy in Iran 87
Valentine M. Moghadam
5 Gender Equality and Developmental Social Welfare in South Africa 109

Shireen Hassim
6 Social Policy Reforms and Gender in Japan and South Korea 130
Ito Peng
7 The Evolution of the Women-friendly State: Opportunities

and Constraints in the Swedish Welfare State 151
Barbara Hobson
8 The Adult-worker-model Family and Gender Equality: Principles

to Enable the Valuing and Sharing of Care 173
Jane Lewis and Susanna Giullari
v
vi Contents
Part II Labour Market Informality and the Search

for Social Security
9 Labour Market Informalization, Gender and
Social Protection: Reflections on Poor Urban
Households in Bolivia and Ecuador 193
Lourdes Benería and Maria S. Floro
10 Working People and Access to Social Protection 217
Francie Lund
Part III Gender Dimensions of Social-sector Restructuring

11 Gender and Health Sector Reform: Analytical Perspectives
on African Experience 237
Maureen Mackintosh and Paula Tibandebage
12 Health Sector Reform in China: Gender Equality and Social Justice 258
Jufen Wang
13 Secondary Education in the Indian State of Uttar Pradesh:

Gender Dimensions of State Policy and Practice 278
Jyotsna Jha and Ramya Subrahmanian
Part IV Financing Social Provisioning and

Counting in Women
14 Gendered Implications of Tax Reform in Latin America:
Argentina, Chile, Costa Rica and Jamaica 301
Evelyne Huber
15 Expectations versus Realities in Gender-responsive Budget

Initiatives 322
Debbie Budlender
Index 340
Tables
1.1 Time spent in unpaid and paid work by men and women in
two adult families with a child under 5 years old (average
hours per day). 11
1.2 Women’s hourly earnings as a percentage of men’s
hourly earnings 13
3.1 Entitlements and financing of family benefits, Poland, 2003 73
3.2 State social support benefits in the Czech Republic as
of 1995/96 74
3.3 Three family benefits, total spending and spending as a
percentage of GDP, the Czech Republic, 1996–2000 75
5.1 Government budget: size and distribution, South Africa 123
6.1 Ageing population (% population over 65) in Japan and
Korea, 1980–2030 134
6.2 Total average fertility rate in Japan and Korea, 1970–2030 134
6.3 Family structures in Japan and Korea 1980–2000 135
6.4 Divorce rate (crude divorce rate per thousand population),
1980–2002 135
6.5 Percentage of single parents in Japan 135
6.6 Percentage of single parents in Korea 136
6.7 Changes in the population coverage of the four major
social insurances in Korea 139
9.1 Employment pattern of husbands and wives in urban,
poor households, 2002 Bolivia and Ecuador Sample Survey,
by sex 203
9.2 Access to benefits by workers, 2002 Bolivia and Ecuador
Household Sample Survey, by type of benefit and by sex 204
9.3 Monthly earnings and earnings variability of
self-employed workers, 2002 Ecuador and Bolivia
Household Sample, by sex 205
9.4 Patterns of household decision-making in couple
households, 2002 Bolivia and Ecuador Household
Sample, by sex 207
9.5 Mean debt service ratio of borrowers, 2002 Ecuador Household
Sample, by sex and type of household headship 209
9.6 Loan characteristics by gender and credit source, 2002
Ecuador Household Sample 210
10.1 Gendered risks and vulnerabilities associated with different
stages of the life cycle 220
vii
viii List of Tables
11.1 Selected African countries: percentage of the labour force

that is female, and percentage of the male and female
labour force receiving wages or salaries 242
12.1 Shanghai social healthcare scheme, contributions by
percentage and age 267
12.2 Standard scale for the shared-responsibility stage of
urban insurance, Shanghai 268
12.3 Main aspects of the 2004 supplementary policy on
reducing medical care expenses, Shanghai 272
13.1 Gross Enrolment Ratios at secondary level (Grades IX to XII) 281
13.2 Crude drop-out rates and proportion of repeaters 281
13.3 Crude transition rates for different levels of school
education (1998/99 to 1999/2000) 282
13.4 Gender-wise pass percentages for Grades X and XII,
Uttar Pradesh, India 282
13.5 Gross Enrolment Ratio (GER) at different stages of
school education 283
13.6 Expenditure on school education by levels in
13.7 Percentage distribution between plan and non-plan
expenditures on elementary and secondary education,
13.8 Per student expenditure at elementary and secondary
education levels, Uttar Pradesh, India 286
13.9 Percentage distribution of secondary schools by
management in Uttar Pradesh, India 288
13.10 Number of examinees and pass percentages in different
types of schools (Grade X and Grade XII
examinations – 2002), Uttar Pradesh, India 289
13.11 Growth of schools in Uttar Pradesh, India 291
13.12 Annual growth rate for the number of institutions at
different school levels in Uttar Pradesh, India 292
14.1 Tax categories 302
14.2 Taxes as a percentage of GDP 303
14.3 Taxes as a percentage of total revenue 303
14.4 Population employed in the urban informal sector 314
Figures
1.1 Ratio of female to male gross primary school enrolment rates 3

1.2 Ratio of female to male gross secondary school enrolment rates 3
1.3 Ratio of female to male gross tertiary enrolment rates 4
1.4 Ratio of female to male adult (15+) literacy rates 4
1.5 Maternal mortality rates, 2000 5
1.6 Percentage of parliamentary seats held by women, 2004 5
1.7 Difference between male and female activity rates (economically
active population as a percentage of total population, 15+) 6
1.8 Informal employment as a percentage of total employment (15+) 12
1.9 Women’s hourly employment earnings as a percentage of men’s,
total employment (15+) 13
3.1 Family benefits as percentage of GDP, Poland, 1990–98 75
7.1 Women’s and men’s paid and unpaid work, weekly
hours 1990–91, 2000–01 165
7.2 Poverty rates of solo mothers 166
9.1a Areas of economic activity and labour use during economic growth 196
9.1b Areas of economic activity and labour use during severe
economic crisis 196
ix
Acronyms and Abbreviations
BpfA Beijing Platform for Action

CEDAW Convention on the Elimination of All Forms of Discrimination
Against Women
CGE Commission on Gender Equality (South Africa)
DAWN Development Through Active Women Networking Foundation (the
Philippines)
FOWODE Forum for Women in Democracy (Uganda)
GAD Gender and development
GBI Gender Budget Initiative
GEAR Growth, employment and redistribution (South Africa)
GRB Gender-responsive budget
IFI International financial institution
MTEF Medium-term expenditure framework
NCRFW National Commission on the Role of Filipino Women
NGO Non-governmental organisation
PER Public expenditure review
PRS Poverty reduction strategy
SNA System of National Accounts
TAF The Asia Foundation
TGNP Tanzania Gender Networking Programme
UNIFEM United Nations Development Fund for Women
WBI Women’s Budget Initiative
WNC Women’s National Coalition (South Africa)
ZWRCN Zimbabwe Women’s Resource Centre Network
x
Notes on the Contributors
Lourdes Benería is Professor of City and Regional Planning and Women’s Studies,
and Director of the International Studies in Planning Program at Cornell
University. Her work has focused on gender and development issues, paid/
unpaid work, globalization, structural adjustment and Latin American development,
among others. She is the author of many articles and books on these topics. Recent
publications include Gender, Development and Globalization; Economics as if all
People Mattered (2003), with its Spanish translation published in 2005, and Global
Tensions; Opportunities and Challenges in the World Economy, edited with Savitri
Bisnath (2003).
Debbie Budlender is a specialist researcher with the Community Agency for

Social Enquiry (CASE), a South African non-governmental organization working
in the area of social policy research. Over the last eight years Debbie has served as
consultant on gender-responsive budgeting to non-governmental organizations,
governments, parliamentarians and donors in approximately 30 countries in
Africa, Asia, Europe, Latin America and the Caribbean.
Maria S. Floro is an Associate Professor in the Economics Department at American

University, United States, where she teaches in the fields of gender, development
finance, and development economics. In 1987, she received her doctorate from
Stanford University. Her publications include books on Credit Markets and the New
Institutional Economics (1991), as well as Women’s Work in the World Economy
(1992), and several articles on topics including time allocation, informal sector,
credit, savings and asset ownership along gender lines. Her recent research focuses
on gender dimensions of financing strategies for development, and the intersec-
tion of gender, finance and home-based work among urban poor households in
both Latin America and Asia.
Susanna Giullari was a BA Fellow and a Research Officer employed on the

European Union-funded project when she completed her work with Jane
Lewis. She is now the Single Parent Action Network Policy/Research Officer.
She is currently working on a participatory policy project, funded by the
Big Lottery Fund, to raise the capacity of lone parents to influence policy devel-
opment in the UK, and on a EU-funded comparative research project investi-
gating the experiences of poverty and social exclusion of children living in
single-parent families.
Shireen Hassim is Associate Professor in Political Studies at the University of the

Witwatersrand, South Africa. She is author of Women’s Organizations and
Democracy in South Africa: Contesting Authority (2005) and co-editor, with Anne
Marie Goetz, of No Shortcuts to Power: Women in Politics and Policymaking in Africa
(2003).
xi
xii Notes on the Contributors
Barbara Hobson is a Professor of Sociology at Stockholm University and the

founder and a current editor of Social Politics: International Studies of Gender, State,
and Society. She has published numerous articles on gender and welfare regimes
concerning the themes of citizenship, global restructuring and most recently
birthstrikes and worklife balance. She has recently been awarded a European
Science Foundation grant to study men and worklife balance. Her current research
concerns issues of citizenship identity and diversity, within supra-contexts and
venues. She has been project leader for three major book projects within the
Comparative Gender Studies Program, of which she was Director for many years:
Recognition and Social Movements: Contested Identities, Agency and Power (2003),
Making Men Into Fathers: Men, Masculinities and the Social Politics of Fatherhood
(2001); and Gender and Citizenship in Transition (2000).
Evelyne Huber (formerly Evelyne Huber Stephens) is Morehead Alumni

Distinguished Professor of Political Science at the University of North Carolina at
Chapel Hill. She is co-author of Capitalist Development and Democracy (1992) and
Development and Crisis of the Welfare State (2001), and editor of Models of Capitalism:
Lessons for Latin America (2002). She is investigating the impact of transitions to
open economies and/or to democracy on systems of social protection in Latin
America and the Caribbean.
Jyotsna Jha was awarded a PhD in Economics of Education from Jawaharlal Nehru
University, New Delhi, in 1995. She has worked for both government and non-
governmental organizations, and for national and international agencies, for the
past several years in India. She has conducted research and evaluation studies
relating to various aspects of education policy, programming and financing. She
has also worked closely with teachers and administrators on matters of equity
in educational planning, research techniques, development of social learning
curriculum and educational management issues. She is the co-author of the recently
published book, Elementary Education for the Poorest and Other Deprived Groups (2005).
Jane Lewis was Barnett Professor of Social Policy at the University of Oxford and
is now Professor of Social Policy at the London School of Economics and Political
Science. Her recent publications include: Should We Worry about Family Change?
(The 2001 Joanne Goodman Lectures) (2003); The End of Marriage? Individualism
and Intimate Relations (2001); with Paul Bridgen, Elderly People and the Boundary
between Health and Social Care 1946–1991: Whose Responsibility? (2000); and with
Kath Kiernan and Hilary Land, Lone Mothers in Twentieth Century Britain (1998).
Francie Lund trained as a sociologist and community worker, and is Associate

Professor in the School of Development Studies at the University of KwaZulu-
Natal, South Africa, where she teaches social policy. She is also the director of the
Social Protection Programme in the international WIEGO network – Women in
the Informal Economy: Globalizing and Organizing. Her main research interests
are in the redistributive role of social security, and informal employment, and the
relationship of both to the dynamics of poverty and well-being. She has published
extensively, and most recently was a co-author of UNIFEM’s Progress of the World’s
Women 2005: Women, Work and Poverty.
Notes on the Contributors xiii
Maureen Mackintosh is Professor of Economics at the Open University, UK. Her

research interests are in the economics of markets in ‘social’ goods such as health
care and social care, both in the UK and in African contexts. Publications include:
The Market Shaping of Charges, Trust and Abuse: Health Care Transactions in Tanzania
(with Paula Tibandebage) (2005) and Commercialization of Health Care: Global and
Local Dynamics and Policy Responses, edited with Meri Koivusalo (2005).
Valentine M. Moghadam is currently Chief of Section in the Social and Human

Sciences Sector of UNESCO. Previously she was Director of Women’s Studies and
Professor of Sociology at Illinois State University. Among her books are Modernizing
Women: Gender and Social Change in the Middle East (2003, second edition) and
Globalizing Women: Transnational Feminist Networks (2005). She is co-editor, with
Massoud Karshenas, of Social Policy in the Middle East: Economic, Political, and
Gender Dynamics (2005).
Maxine Molyneux is Professor of Sociology at the Institute for the Study of the
Americas, University of London. Among her recent books are Women’s Movements
in International Perspective (2000), M. Molyneux and Razavi, S. (eds), Gender Justice,
Development and Rights (2002), M. Molyneux and S. Lazar, Doing the Rights Thing
(2003), M. Molyneux and N. Craske (eds), Gender and the Politics of Rights and
Democracy in Latin America (2002), and E. Dore and M. Molyneux (eds), The Hidden
Histories of Gender and the State in Latin America (2000).
Ito Peng is Associate Professor of Political Sociology at the Department of

Sociology, and the Director of Dr David Chu Chair Programme in Asia Pacific
Studies at the University of Toronto. She has published extensively on compara-
tive social policy in Japan, Korea, and Taiwan. Her current research projects
include political economy of social policy reforms in Japan and Korea, compari-
son of familialistic welfare states in Japan, Korea, Italy and Spain, and varieties of
capitalism in East Asia.
Shahra Razavi is Research Coordinator at the United Nations Research Institute

for Social Development (UNRISD) and has published widely on gender and liveli-
hoods. Her most recent publications include Agrarian Change, Gender and Land
Rights (2003), also published as a special double issue of the Journal of Agrarian
Change, Vol. 3, Nos. 1&2, January & April 2003; and Gender Justice, Development
and Rights, an edited volume with Maxine Molyneux (2002).
Silke Steinhilber is a researcher and consultant on women’s economic and social

rights in Central and Eastern Europe. She is a PhD candidate in the Political
Science Department of the New School for Social Research in New York. Formerly,
she worked for the International Labour Organisation in the Subregional Office for
Central and Eastern Europe in Budapest, Hungary.
Ramya Subrahmanian is a Research Fellow at the Institute of Development

Studies, University of Sussex working on issues of elementary education, child
labour, social exclusion and gender equality. She is a co-editor of Needs versus
Rights: Child Labour and the Right to Education in South Asia (2003) and Institutions,
xiv Notes on the Contributors
Relations and Outcomes: A Framework and Case Studies for Gender Aware-Planning
(1999).
Paula Tibandebage is Training Co-ordinator for Research on Poverty Alleviation

(REPOA) in Dar es Salaam, Tanzania. She is a political economist specialising in
social policy issues. Her recent work includes two co-authored articles: ‘Inclusion
by design: rethinking regulatory intervention in Tanzanian healthcare’, Journal of
Development Studies (2002), and ‘The market shaping of charges, trust and abuse:
Health care transactions in Tanzania’, Social Science and Medicine (2005).
Jufen Wang is Professor at the School of Social Development and Public Policy,
Fudan University, Shanghai, China, where she teaches courses on demography
and gender studies. She has published over 30 papers in various journals. Her main
publications include The Resolution of Pre-marital Pregnancy (2002) and Legal
Protection of Women’s Rights (2004). Her research interests are reproductive health,
women’s employment and social security issues.
Preface and Acknowledgements
This book is part of a series coming out of UNRISD’s timely and innovative
research programme on Social Policy in a Development Context. The programme
has launched a wide-ranging inquiry into models of social policy that are inclu-
sive, developmental and democratically grounded – especially for middle- and low-
income countries. Inclusive social policy, the programme maintains, is not a
luxury that only affluent countries can afford, but a necessity of social and eco-
nomic development. The programme has thus set itself the task of understanding
the social, political and economic conditions for the edification of a state-society
nexus that is developmental, inclusive and democratic.
A gender perspective on social policies in middle- and low-income countries, as

in the already institutionalized ‘welfare states’ until quite recently, has remained
on the margins of mainstream research. This book attempts to move the gender
analytical framework closer to the centre of social policy thinking by exposing
how the social institutions through which social policy is filtered are all indeli-
bly bearers of gender––be it families and communities, markets, informal
arrangements for care, or health and education systems, and the public sector.
Women’s unpaid care work continues to form the bedrock on which social pro-
tection is subsidized, with erosions in state provisioning impacting most
strongly on women. Despite women’s increasing participation in paid work,
labour markets continue to reproduce gender-based segmentations and inequal-
ities in wages and income, as well as in work-related social benefits and social
protection mechanisms. Attention to gender also reveals the extent to which
inequalities (of class, gender and region) are being intensified as a consequence
of shifts in the global economy, and processes of privatization and commercial-
ization taking place within countries.
As editors of this volume we would like to thank all the contributing authors
for their patience and persistence in responding to our numerous requests for
revisions. At UNRISD, Thandika Mkandawire was a major source of intellectual and
moral support; we also benefited from many useful discussions with Huck-Ju Kwon.
Our greatest debt goes to Alexandra Efthymiades for providing excellent research
assistance and for seeing the book through, with great patience and good humour.
We would also like to thank those who contributed to our work by agreeing to
review manuscripts, act as discussants, and provide advice at different points in
the life cycle of this project, especially Armando Barrientos, Ulla Björnberg,
Claudia Bonan Janotti, Sarah Cook, Monica Erwer, Éva Fodor, Tekaligne Godana,
Beth Goldblatt, James Heintz, Ping-Chun Hsiung, Elizabeth Jelin, Helen Schneider,
Stephanie Seguino, Ylva Sörman Nath, Ida Susser, Dzodzi Tsikata and Ann
Zammit.
xv
xvi Preface and Acknowledgements
We would also like to express our appreciation to the Centre for Global Gender
Studies at Göteborg University, especially Ann Schlyter, Monica Erwer and Monica
Lindberg-Falk, for organizing a memorable conference on the beautiful Marstrand
Island in Göteborg (Sweden) in May 2005.

Foreword
The UNRISD research programme on social policy consists of both regional and
thematic components. This volume presents the comparative research carried out
by a network of gender specialists, largely though not exclusively from developing
countries, exploring the gender dimensions of social policy. Although it was
always the intention of the social policy programme to address gender issues
across all regional and thematic components, we were also convinced that this
effort would have to be complemented by a focused analysis of social policy
through a gender ‘lens’. Without the latter, we feared, ‘mainstreaming’ gender
would risk diluting it.
From their diverse regional perspectives the contributions to this volume show
how social and economic rights are shaped by economic structures and reforms,
institutional designs and capacities, and political processes which are all deeply
gendered. It underlines the importance of thinking beyond states and markets in
social provisioning, including in the analysis the interactions between these and
other social institutions, especially the family and community that take on much
of the burden, particularly in developing countries where formal social provision-
ing is thin. Although there have been changes in the balance of work and care in
many societies, this book shows that in many contexts these changes have rein-
scribed rather than eroded gender inequalities. The volume illustrates why both
academic research and policy thinking need to factor-in gender hierarchies and
structures if they are to address some of the key challenges of contemporary
societies: the widespread informality and insecurity of paid work, and the crisis
of care.
UNRISD would like to take this opportunity to thank the Swedish International
Development Cooperation Agency (Sida) and the United Kingdom Department
for International Development (DFID) for their financial support of the project.
for its financial support of the project. As is the case with all other UNRISD pro-
jects, work on this project would not have been possible without the core funding
of the governments of Denmark, Finland, Mexico, Norway, Sweden, Switzerland
and the United Kingdom.
Thandika Mkandawire
Director, UNRISD
xvii
1
Gender and Social Policy in a Global
Context: Uncovering the Gendered
Structure of ‘the Social’
Shireen Hassim and Shahra Razavi
The past decade has witnessed a renewed interest in social policies, and some
governments have increased social spending to soften the impacts of economic
reform. These changes have come in the wake of widespread realization of the
failure of the neoliberal economic model to generate economic growth and
dynamism, and to reduce poverty. At the same time, processes of political liberaliza-
tion have opened spaces for social movements in many parts of the developing
world to articulate demands for more effective social policies that mitigate the
effects of market failures and reduce inequalities.
These contestations have coincided with a rediscovery of ‘the social’ in the
policy oriented literature, widely understood to embrace the cluster of social and
political institutions, norms, and relationships that define the boundaries of
market exchange, reduce transaction costs and enhance social and political stability.
Polanyi’s (1957) seminal work that showed the market to be a political and social
construct is widely cited today to explain the failure of the structural adjustment
packages that narrowly focused on ‘getting the prices right’, and to redirect atten-
tion to the institutional underpinnings necessary for successful market capitalism
(Ruggie 2003).1 However despite the movement away from the standard neoliberal
approach of the 1980s, and the increasing recognition given to institutions and
the state, there is little agreement on a number of critical issues. These include the
scope of social policy and the appropriate interface between social policy and
macroeconomic policy (Elson 2004; Mkandawire 2004; Tendler 2004); the role of
the state, not just as ‘regulator’ but also as a provider of social welfare; and the values
underpinning public policy, in particular core values of equality and redistribution
which seem to have been displaced by the discourse on poverty (Phillips 2001).
A gender perspective on social policies in the South, as in the North until quite
recently, has remained on the margins of these debates. This volume is an attempt
to move the gender analytical framework closer to the centre of social policy
thinking. From their different regional perspectives, the chapters in the volume
map out the complex ways in which social policies are filtered through social
institutions – families and communities; markets; care arrangements; health and
1
2 Shireen Hassim and Shahra Razavi
education systems; the public sector – that are ‘bearers of gender’. Moreover attention
to gender reveals the extent to which inequalities (of class, gender and region) are
being intensified as a consequence of shifts in the global economy, and processes
of privatization and commercialization taking place within countries. Women’s
unpaid care work continues to form the bedrock on which social protection is sub-
sidized, with erosions in state provisioning impacting most strongly on women.
Despite women’s increasing participation in paid work, labour markets continue
to reproduce gender-based segmentations and inequalities in wages/income,
work-related social benefits, and social security. As this book demonstrates, social
institutions are by no means homogeneous: economic, institutional and cultural
variations across countries and regions shape the nature of both risks faced and
forms of social protection available for women.
In this chapter, we root a theorization of gender and social policy in three key,
interrelated arenas: the nature of labour markets, the institutional basis for social
policy formulation (families, communities, markets and states) and the nature of
political contestation around social policy. In the first section, we lay out the
gendered nature of economic transformations in the late twentieth century,
drawing out the implications for gender equality of shifts in the nature of labour
markets and the relationships between paid and unpaid work. In the following
section, we link these changes in the structure of labour markets to a discussion
of the impacts of social sector restructuring. Here we examine the gender impli-
cations of commercialization and privatization of social services and income
supports as well as the policy turn to targeting and social insurance as a response
to the exclusionary effects of markets. The third section explores the institutional
basis for social policy formulation, examining more closely the assumptions
about gender roles and entitlements, especially in the key institutions of family
and community, and how these interface with the state. The relationship between
political democratization and the development of gender equitable social policy
is then examined. We are particularly concerned with exploring women’s
agency in relation to advocating for social policy change in ways that meet their
various needs.
To facilitate our discussion of the chapters, Figures 1.1 to 1.7 (pages 3 to 6)
capture some of the standard indicators of relative (or absolute) female status for
the countries included in this volume, ranked by per capita GDP (USD PPP
adjusted). These include indicators of educational status (gross enrolment rates at
the primary, secondary and tertiary levels, and adult literacy rates), health status
(maternal mortality rates), political status (percentage of parliamentary seats held
by women), and economic status (economic activity rates). While enrolment rates
in primary education (Figure 1.1) provide a generally more egalitarian picture
(India being an exception), inequality sets in at the secondary level (Figure 1.2)
especially for some low-income countries. Gross enrolment rates at the
tertiary level (Figure 1.3) provide a more complex picture with gender gaps being
in favour of males in some countries (such as Japan, South Korea, Bolivia, Tanzania),
and in favour of females in others (such as Sweden, Argentina, Jamaica), cutting across
3
Japan ($26,940) 1.00

Sweden ($26,050) 1.03
South Korea ($16,950) 1.00
Country (per capita GDP, $US, PPP adjusted)
Czech Rep. ($15,780) 0.99

Argentina ($10,880) 1.00
Poland ($10,560) 0.99
South Africa ($10,070) 0.96
Chile ($9,820) 0.98
Mexico ($8,970) 0.99
Costa Rica ($8,840) 0.98
Thailand ($7,010) 0.96
Iran ($6,690) 0.96
China ($4,580) 1.00
Jamaica ($3,980) 0.99
Ecuador ($3,580) 1.00
India ($2,670) 0.85
Bolivia ($2,460) 0.99
Tanzania ($580) 0.97
0.80 0.85 0.90 0.95 1.00 1.05
Figure 1.1 Ratio of female to male gross primary school enrolment rates
Note: Data are from 2001–02.
Source: World Development Indicators 2005, World Bank.
Japan ($26,940) 1.01

Sweden ($26,050) 1.21
South Korea ($16,950) 1.00
Czech Rep. ($15,780) 1.03

Argentina ($10,880) 1.06
Poland ($10,560) 0.97
Chile ($9,820) 1.02
Mexico ($8,970) 1.07
Costa Rica ($8,840) 1.08
Thailand ($7,010) 0.95
Iran ($6,690) 0.94
China ($4,580) 0.92
Jamaica ($3,980) 1.03
Ecuador ($3,580) 1.02
India ($2,670) 0.74
Bolivia ($2,460) 0.96
0.65 0.75 0.85 0.95 1.05 1.15 1.25
Figure 1.2 Ratio of female to male gross secondary school enrolment rates
Note: Data are from 1999–2002.
Japan ($26,940) 0.86

Sweden ($26,050) 1.54
South Korea ($16,950) 0.61
Czech Rep. ($15,780) 1.09
Argentina ($10,880) 1.48
Poland ($10,560) 1.43
Chile ($9,820) 0.93
Mexico ($8,970) 0.95
Costa Rica ($8,840) 1.16
Thailand ($7,010) 1.09
Iran ($6,690) 1.07
Jamaica ($3,980) 2.24
India ($2,670) 0.70
Bolivia ($2,460) 0.55
0.40 0.90 1.40 1.90 2.40
Figure 1.3 Ratio of female to male gross tertiary enrolment rates

Note: China and Ecuador are excluded due to lack of information. Data are from 2001–02, with the
exception of the figure for Bolivia which is from 1998.
South Korea ($16,950) 0.97

Argentina ($10,880) 1.00
Poland ($10,560) 1.00
Chile ($9,820) 1.00
Mexico ($8,970) 0.96
Costa Rica ($8,840) 1.00
Thailand ($7,010) 0.95
Iran ($6,690) 0.84
China ($4,580) 0.91
Jamaica ($3,980) 1.09
Ecuador ($3,580) 0.97
India ($2,670) 0.66
Bolivia ($2,460) 0.87
0.60 0.70 0.80 0.90 1.00 1.10
Figure 1.4 Ratio of female to male adult (15⫹) literacy rates

Note: Japan and Sweden are excluded because they have nearly universal adult literacy. The Czech
Republic has been excluded due to lack of information. Data are from 2000–02.
Source: World Development Indicators 2005, World Bank and Human Development Report 2004, UNDP.
income groups. Figure 1.5 captures maternal mortality, which seems to be more
sensitive to income levels, although again there is great diversity between coun-
tries reflecting differences in health systems and the accessibility of health care
(with Argentina and South Africa doing particularly badly for their income level).
5
1,600 1,500
1,400
1,200
1,000
800
600 540
420
400
230
200 130 87
56 76 44 43 83 31 13
82
9 20 10
2
0
In ($2 )
ua ($ 0)
m ($ 0)
($ 0)
($ 0)
($ 0)
($ 0)
ex ($ 0)
ut Ch ($8 0)
ric ($9 0)
Ar land 10, )
ze tina 10, )
So h R ($1 60)
re 15 )
ed $1 80)
)
26 0)
)
80
Ko . ($ 80
Ja n ($ 950
40
Ec dia ,46
Ja dor ,67
ca ,58
Th na ,98
nd ,58
os an ,01
69
o 84
7
Po a ( ,82
07
($ 5
liv ($5
ile ,9
Sw a ( ,7
n 6,0
,9
6,
ic 8,
ut ep 0,
6,
2
pa 2
Bo nia
n ($
ia
M ica
a
nz
ai
hi
la
Ir
e
R
C
ai
Ta
Af
ta
ge
h
c
h
C
So
C
Figure 1.5 Maternal mortality rates, 2000 (maternal deaths per 100,000 live births)
Source: Maternal Mortality in 2000: Estimates Developed by WHO, UNICEF, UNFPA, Geneva: WHO, 2004.
50.0% 45.3%
45.0%
40.0% 35.1%
35.0% 30.7%
29.8%
30.0%
25.0% 21.4% 22.6%
20.2% 20.2%
20.0% 18.5% 16.0% 17.0%
15.0% 11.7% 12.5%
8.8% 9.2%
10.0% 5.9% 7.1%
4.1%
5.0%
0.0%
In ($2 )
ua ($2 0)
m ($ 70)
C a ($ 80)
Ir $4 )
C aila $6, )
ta d ( 90)
8, )
ut Chi ($8 )
9, )
Ar lan 10 )
ze tina 10,5 )
)
0)
en 6, )
pa 26 )
26 )
)
0
( 80
M a ($ 10
o 840
Af e ($ 970
0
0
ed ($1 80
Ja ($ 950
($ 50
40
58
Th an ,58
Po a ( 82
ge d ( ,07
8
4
c 3,5
na 3,9
,0
ut ep 0,8
Sw rea 5,7
,0
,9
($
,
R $7
1
ia
$
$
Ko ($
an
(
ia
Ec ia
Ja or
os n
ic
ic
n
liv
d
nz
ai
hi
ric
ex
Bo
R
Ta
n
ch
h
h
So
So
C
Figure 1.6 Percentage of parliamentary seats held by women, 2004

Note: For countries with upper and lower houses of parliament, the data refer to the lower house only.
Source: Human Development Report 2004, UNDP.
Japan, 2003 ($26,940) 25.7%

Sweden, 2004 ($26,050) 4.0%
South Korea, 2003 ($16,950) 26.0%

Czech Rep., 2002 ($15,780) 18.3%
Argentina, 2003 ($10,880) 25.7%
Poland, 2003 ($10,560) 14.2%
South Africa, 2004 ($10,070) 15.8%
Chile, 2003 ($9,820) 35.5%
Mexico, 2004 ($8,970) 41.6%
Costa Rica, 2004 ($8,840) 38.9%
Thailand, 2004 ($7,010) 16.7%
Iran, 2000 ($6,690) 49.6%
Jamaica, 2003 ($3,980) 11.3%
Ecuador, 2003 ($3,580) 26.8%
India, 2000 ($2,670) 42.4%
Bolivia, 2000 ($2,460) 29.6%
Tanzania, 2000/1 ($580) 2.9%
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0%
Figure 1.7 Difference between male and female activity rates (economically active population
as a percentage of total population, 15⫹)
Note: Age ranges vary for some countries: Jamaica: 14⫹, South Africa: 15–65, Sweden: 16–64. For India and
Iran, estimates are projections based on EAPEP (Economically Active Population Estimates and
Projections) data, ILO.
Sources: LABORSTA Database, ILO; EAPEP Database, ILO; Labour Force Survey, September 2004, Pretoria:
Statistics South Africa, March 2005.
Figure 1.6 captures women’s relative access to parliamentary seats, which seem to
bear little relation to income levels, with Sweden and South Africa being the two
front-runners. Gender gaps in economic activity rates are shown in Figure 1.7, with
Sweden and Tanzania at the two ends of the income spectrum displaying the
narrowest gender gaps, while the widest gaps appear in Iran, India and Mexico
with income levels that are much higher than Tanzania’s.
The Economy as a Gendered Construction: Labour and Care
We make two central arguments in this section. First, we note that historically a
major preoccupation of social policy has been to respond to labour market risks
(such as unemployment, old age, maternity) through transfer payments condi-
tional on previous employment.2 However, women’s access to paid work does not
easily translate into their enjoyment of social protection mechanisms that flow
from paid employment. Gender-blind analysis of social policy has tended to
underestimate the extent to which labour markets are themselves gendered
(for example, sex-segregation of jobs, differential pay, gendered definitions of
‘skill’), features which in turn shape and constrain women’s access to such
Gender and Social Policy in a Global Context 7
benefits. Furthermore, the gender-segmented features of the labour market have

been exacerbated by liberalization policies that have weakened the link between
paid work and entitlements to social protection and provisioning. Second, for
gender analysts a central feature of women’s entry into paid work is the tension
that this produces with regard to their responsibilities to provide unpaid care. How
these tensions are experienced and understood diverge substantially across
contexts, and shape the particular balance of responsibilities between markets,
states, families and communities.
Paid labour and social protection: the historical legacy

Welfare-state development in all its regional variations has been marked by
numerous inequalities and exclusions (of gender and race). This is evident not
only in Europe and North America, but also in many developing countries, where
for various reasons the construction of effective and inclusive welfare systems
proved difficult. In the advanced industrialized countries of Western Europe, the
postwar social contract between capital and labour underpinning state social
regulation and provisioning was based on dominant normative assumptions
about gender difference, with breadwinning prescribed for men and caring/
homemaking for women. Many women of course were in the labour force, sometimes
continuously, but they tended to occupy the less protected niches as secondary
workers with limited access to social insurance benefits. Hence, for many women
who spent large parts of their lives outside the paid workforce, and even for those
who worked on an irregular basis, access to a pension or health insurance became
possible through their relations with a fully employed husband or father, as a
derived rather than individual entitlement.
Given this pattern, feminists have used the extent of women’s integration into the
paid labour market as a mechanism to distinguish between different welfare states.
At one end of the spectrum, Sweden demonstrated the most (although not com-
pletely) egalitarian form of welfare state, and was inclined to a ‘weak male bread-
winner’ or ‘dual breadwinner’ model (Lewis 1992).3 A formal emphasis on
egalitarianism also underpinned state socialism, which shared the Swedish model’s
aim of getting women, like men, into the workforce but without the democratic
political institutions of the Swedish state. In the state socialist countries, state policy
and rhetoric, at least initially, celebrated women’s liberation by defying bourgeois
models of the family and of femininity, by encouraging women’s presence in con-
tinuous full-time employment, and by taking public responsibility for the provision
of childcare. Gender equality was modelled on a male norm of paid work, and
labour-force entry was the route through which many women accessed a wide array
of social benefits and services. These egalitarian impulses notwithstanding, gender
inequalities in wages and labour market status (though far less accentuated than in
Western Europe) along with the burden of unpaid work in a shortage economy,
remained a palpable part of the legacy of state socialism for many women in the
region. A good portion of care work continued to be carried out within the family
and by women, while parental leave schemes were made available to women; only
under exceptional circumstances could fathers qualify for such leave.
Many developing countries, by contrast, have been characterized by weak state

institutions, partially commodified economies, low levels of formal waged labour
and a weak fiscal base. The colonial legacy, though experienced very differently,
left many countries with deepened social cleavages and lopsided economic
structures that postcolonial governments had a hard time shaking off. The colonial
impact was particularly devastating in the ‘enclave’ economies of southern Africa,
such as Zambia, where race, gender and location intersected to create highly dif-
ferentiated rights of access to state social provisioning. Most of the population was
excluded from welfare on the grounds of their rurality and reliance on subsistence
production (Mhone 2004). The combination of the pattern of male labour migra-
tion and colonial policy to keep women out of urban areas intensified women’s
responsibility for household reproduction and care in the subsistence-based rural
economies (Tsikata 2004). They tended to exhibit the most tenuous links with the
urban economy and any social benefits that came from it. After independence
state provision of important social services like health, housing and education
retained a bias towards the more developed urban areas, despite stated intentions
to make development broad-based.
Where the developmental state was successful in forging strong national indus-
tries, namely in parts of north-east Asia such as South Korea and Japan, both the
industrial structure and welfare provision remained stratified, with a dual labour
market structure that overwhelmingly privileged male workers in core industries.
The sectoral distribution of women’s employment, the size of the firms in which
they were employed, the occupations in which they were clustered and the nature
of their contracts combined to ensure that both the direct benefits of employment
and access to social welfare insurance were less advantageous for them. For example
corporate social welfare – which was an important source of welfare in South Korea
prior to the 1997 crisis – was far more generous in the large firms, whereas the bulk
of women workers were concentrated in small- and medium-sized firms that could
ill afford the same level of benefits (Cho et al. 2004).
In the case of Latin America, from the late nineteenth century some forms of
social protection and provisioning developed, even if to varying degrees, to create
the embryos of the future ‘social state’ in countries such as Argentina and Uruguay
(Filgueira forthcoming). These efforts were mainly concerned with the provision
of public health and primary education often as a means to ‘discipline and
homogenize’ (p. 5) the rural migrants and the European immigrants, while social
insurance was made available to privileged (and largely male) sectors of the labour
and armed forces. Such efforts were reinforced in the 1930s as corporatist arrange-
ments were put in place in countries like Argentina, Brazil and Mexico, in the
aftermath of the 1929 crisis, and established more securely in the postwar period.
This resulted in some expansion in entitlements, notably for organized labour
(Molyneux this volume). Social provisioning was expanded in the 1960s and
1970s so that by 1980 human development indicators in Latin America were on a
par with those of East Asia (ibid). However, the coverage that the social state
provided in Latin America varied enormously: social exclusion was far more exten-
sive in Central America, for example, while the southern cone countries tended
towards a more universalistic, even if highly stratified, model (Filgueira & Filgueira
2002).
In short, in deeply hierarchical and segmented societies, where some crucial
forms of social protection were linked to formal employment and where the latter
never included more than a small fraction of the population, vast sections of the
population, including the majority of women, were frequently excluded from
coverage. Yet it would be wrong to assume that women were absent from state
social provisioning and protection altogether. Not only did women make up a
significant proportion of social security beneficiaries as wives and daughters
(‘dependents’ in the language of social insurance) of male breadwinners, they were
also direct beneficiaries of some public services (health, education) as well as being
targets of so-called maternalist programmes aimed at mothers and their children
(Molyneux this volume). Women have also traditionally constituted a significant
share of state employees in the social sectors, as teachers, nurses and carers.
Labour market informalization: the demise of the

‘male breadwinner’ model?
The small size of the formal economy in most developing countries meant that job
security and work-related benefits remained privileges available to a relatively thin
stratum of workers, predominantly men. While these benefits could have been
extended over time to other sectors of the population (as in many successful welfare
states, and more recently in South Korea), since the early 1980s there has been a
global trend in the opposite direction. Paid work is becoming increasingly informal
and casual as workers have lost their work-related social benefits, although the extent
of convergence between patterns in North and South should not be exaggerated.
In Latin America in the 1980s and 1990s eight out of every ten new jobs created
were in the informal sector, while the 1999 manufacturing wage was only 3 per
cent higher than its 1980 level (Tokman 2002). Even in Costa Rica – the country
that has one of the strongest social states in the region – informal employment
accounts for nearly half of all employment (Heintz 2005). Those who work in the
informal economy are not covered by labour legislation for social protection and
earn less, on average, than those in the formal economy (ILO 2002).
Processes of labour informalization and casualization have been largely driven by
corporate interests, increasingly unhindered in their global search for ‘appropriate’
forms of labour, while trade unions have been weakened and sidelined and, with
few exceptions, unwilling to take informal workers on board. Hence, employers are
no longer forced to take responsibility for the social wage. In many developing
countries labour ‘flexibility’ was further imposed as a policy condition for debt
relief (via stabilization and structural adjustment programmes), on the ground that
labour markets were too rigid. This process of casualization has coincided with
another momentous change – the increasing feminization of the work force.
Gender and labour markets: continuity and change

With some striking exceptions (East and Central Europe and Central Asia), the last
twenty years have seen a surge in women’s labour-force participation worldwide,
with women’s activity rates nearing men’s in some countries. The forces propelling
women into the work force are complex and diverse. In the OECD countries, the
trend of going to work is most noticeable among mothers with young children –
the group that was most likely to drop out of employment in the post-second
world war period (Orloff 2002). However, workforce participation rates are crude
indicators of working women’s situations. Women in these countries, for example,
tend to work on a part-time basis far more frequently than men (ibid). While
taking on part-time work may involve an element of real choice in some cases, for
others it may be ‘involuntary’, because of either the lack of full-time employment
options or the need to accommodate care in the absence of other viable sources for
its provision, or both.
Feminists have been particularly interested in Nordic social democracies that
appear to have achieved relatively high levels of gender equality through female
labour-force participation and the redistributive mechanisms of social policy.
Sweden, for example, has been able to forge what Hobson (this volume) calls ‘par-
ticipation parity’ in the labour market with nearly equal numbers of women and
men in the workforce (Figure 1.7). Yet this achievement masks the manner in which
gender inequalities have been reinscribed in new ways. For example, a significant
portion of women – over a third – are working part time4 and more than 90 per cent
of all part-time workers in Sweden are women. As Hobson argues, the Swedish model
is thus in practice a ‘one and three-quarters’ model: men work full time and invest in
their careers, while women work part time in the public sector, where it is easier to
combine employment with having a family. Even though parental leave schemes
and ‘daddy leave’ quotas were designed to be gender-neutral, the outcomes are
not. Throughout the 1990s, between 10 and 12 per cent of parental leave days
were taken by fathers. In 2005 men’s share rose to 17 per cent – still a far cry from
the equal participation of men in unpaid work. As Lewis and Giullari (this volume)
remind us, although the balance of contributions women and men make to house-
holds in the form of cash and care is changing it remains gendered. In most coun-
tries, women have added paid work to their existing responsibilities for care, while
men have decreased the amount of paid work they do and increased their care-
work only slightly (see Table 1.1 for a snapshot of selected OECD countries).
Recent research finds that despite some improvements in the 1990s, levels of
gender segmentation in the labour market remain high throughout the world
(Anker et al. 2003). Women tend to congregate in relatively low-paid and low-
status work at the bottom of the occupational hierarchy, and also to have little job
security. Gender segmentations in labour markets are more difficult to capture in
developing countries with pervasive agrarian and informal sectors, which are not
sufficiently covered in most large-scale statistical surveys.
For many developing countries increasing poverty and the commodification of
the economy are changing the coping strategies of households and communities
in a multitude of ways, causing upheavals in gender and generational patterns of
work and responsibility. Research from sub-Saharan Africa, for example, shows
that in the context of economic crisis and reform, it is becoming increasingly
necessary for all household members – whether female or male, young or old – to
Table 1.1 Time spent in unpaid and paid work by men and women in two adult families
with a child under 5 years old (average hours per day).
Men Women (employed

(average for full-time Ratio: Women to
all men) in paid work) Men
Unpaid Paid Unpaid Paid Childcare All unpaid
Canada (1998) 4.0 6.3 5.1 5.9 1.4 1.3

United States (1995) 2.5 6.2 4.2 4.9 1.9 1.7
Denmark (1987) 2.3 7.2 4.0 5.4 2.0 1.7
Finland (1987) 2.8 6.1 5.6 3.9 2.8 2.0
Sweden (1991) 3.7 6.4 6.1 3.9 1.9 1.6
Italy (1989) 1.8 6.6 6.4 4.2 2.7 3.6
UK (1995) 3.1 6.3 7.4 3.5 1.4 2.4
Austria (1992) 2.2 6.9 5.8 4.7 2.2 2.6
Gemany (1992) 3.4 6.1 6.2 4.1 2.1 1.8
Netherlands (1985) 2.9 5.2 6.2 1.7 2.4 2.1
Australia (1997) 2.9 6.1 4.6 6.0 1.8 1.6
Source: OECD Employment Outlook, Paris: OECD, 2001, Table 4.5, p. 140.
take on paid work (Bryceson 1999). Much of this tends to be badly remunerated,
with women being overwhelmingly clustered in low-entry, low-return type
activities (Whitehead 2004).
According to the ILO (2002) informal employment5 comprises between one-half
and three-quarters of non-agricultural employment in developing countries. But
there are also important stratifications within the informal economy, based on
productivity and income as well as gender. The informal economy tends to be a
larger source of employment for women than for men in most countries (ILO
2002); Figure 1.8, based on a recent six-country study, shows strongly gendered
patterns of employment in the formal and informal economy (with the exception
of El Salvador where male and female rates are comparable). Women informal
workers tend to be over-represented in the more precarious and less remunerative
segments of informal work: they are more likely to work as own-account workers,
domestic workers, and unpaid contributing workers in family enterprises than are
men, while men are more likely to work as employers and wage workers (ILO
2002). This is confirmed by Beneria and Floro’s evidence (this volume) from
Bolivia and Ecuador where women tend to have relatively more precarious jobs
than men. Furthermore, as time allocation studies show, while informal work,
especially home-based work, may enable women to combine paid and unpaid
work (such as sewing garments on a piece-rate basis while looking over children
and doing housework), it entails an intensification of work which can seriously
affect women’s well-being (see Beneria and Floro’s chapter).
It is very difficult to find gender-disaggregated data on income and poverty.
The reliance on poverty lines and household expenditure data has profound impli-
cations for how gender issues are analysed. Measuring poverty on the basis of
100% 93.9% 95.1%

88.0% 89.5%
90%
78.4%
80%
69.8% 70.0% 68.7%
70%
60% 56.6%
48.6% 50.9%
50% 44.0%
40%
30%
20%
10%
0%
Costa Rica Egypt El Salvador Ghana India South Africa
(2003) (1998) (2003) (1998/9) (1999/2000) (2003)
Female Male
Figure 1.8 Informal employment as a percentage of total employment (15⫹)

Source: Heintz, James. ‘Summary of Country Case-Studies and Tabulations for UNIFEM’s Progress of the
World’s Women 2005’ Background paper prepared for Progress of the World’s Women 2005: Women, Work,
and Poverty. New York: UNIFEM.
household expenditure data effectively ignores the longstanding feminist

concerns about intra-household distribution, and it is very rare to find standard
surveys embarking on a quantitative exploration of intra-household poverty
(Razavi 1999; Whitehead and Lockwood 1999). Hence, evidence from smaller sur-
veys on men’s and women’s wages and earnings can be very useful to fill this
lacuna. Table 1.2 and Figure 1.9, which draw on the UNIFEM study, confirm that
women’s hourly earnings typically fall below those of men in identical employ-
ment categories; the gender gap in earnings is particularly pronounced among
own-account workers while it is narrowest in public wage employment.
The broad conclusion that one can draw from some of the available data is that
despite the convergence in men’s and women’s crude economic activity rates and the
erosion of the male-breadwinner model, gender segmentation is not disappearing
from the world of work. It is surprisingly robust, and places serious limits on women’s
access to income. It also has important implications for social policy – making
women’s access to social protection mechanisms and social services more con-
strained if these are provided on a commercial basis or on the basis of employment-
related contributions, as in the social insurance model. In theory it is possible to
extend the coverage of social insurance programmes to include informal workers.
Lund’s chapter shows that this is happening in a number of countries, including
Costa Rica, but only if their contributions are heavily subsidised by the state. In the
following section we examine the contours of social sector restructuring in the 1980s
and 1990s and the implications for women’s access to social services and transfers.
13
Table 1.2 Women’s hourly earnings as a percentage of men’s hourly earnings
Costa Egypt El Ghana South

Rica Salvador Africa
Formal
Non-agricultural
Employers 56.6 n.a. 78.9 n.a. n.a.
Own-account 62.1 n.a. 45.1 54.6 n.a.
Private wage 84.6 151.9 87.5 n.a. 89.5
Public wage 87.7 107.6 116.2 84.2 95.6
Agricultural
Private wage 85.1 n.a. 105.4 n.a. 66.3
Informal
Non-agricultural
Employers 97.6 n.a. 83.8 n.a. 83.6
Own-account 50.6 n.a. 65.1 80.2 59.6
All wage n.a. 263.3 n.a. 69.8 n.a.
Private wage 79.5 317.1 75.4 n.a. 107.0
Public wage n.a. n.a. 135.2 88.0 99.2
Domestic 57.4 n.a. 56.2 n.a. 100.0
Agricultural
Own-account 53.3 n.a. 56.9 65.0 n.a.
Private wage n.a. n.a. 86.4 n.a. 98.5
Public wage n.a. n.a. 177.6 n.a. n.a.
Note: n.a. indicates that data were not available or that there were insufficient observations to derive
statistically significant estimates.
Source: Heintz, James. ‘Summary of Country Case-Studies and Tabulations for UNIFEM’s Progress of the
World’s Women 2005’. Background paper prepared for Progress of the World’s Women 2005: Women, Work,
and Poverty. New York: UNIFEM.
120%
100% 96%
80%
80% 72% 71%
60%
40%
20%
0%
Costa Rica (2003) El Salvador (2003) Ghana (1998/9) South Africa (2003)
Figure 1.9 Women’s hourly employment earnings as a percentage of men’s, total

employment (15⫹)
Source: Calculations based on estimates from Heintz, James. ‘Summary of Country Case-Studies and
Tabulations for UNIFEM’s Progress of the World’s Women 2005’. Background paper prepared for Progress of
the World’s Women 2005: Women, Work, and Poverty. New York: UNIFEM.
Gender and Social Sector Restructuring
There have been tremendous changes in the development policy mindset with
regard to the role of the social sectors in the last two decades. In this section we
address the implications of these shifts for the development of gender-equitable
social policies.
By the late 1980s it became increasingly evident that the poverty and social
disruptions caused by stabilization and adjustment were not the ‘transitional
phenomena’ or ‘frictional difficulties’ the international financial institutions (IFIs)
had initially assumed; they were pervasive, long-term and systemic. Studies of the
impacts of adjustment, and popular opposition to key adjustment-related
measures combined to bring into question dominant policy prescriptions. Global
policy pronouncements became less assertive about the imperative of cutting
social spending, more apologetic about the imposition of ‘user fees’, and began to
acknowledge that social policy could have a vital role to play in the development
process.
In the early 1990s the World Bank grudgingly accepted that adjustment
packages had paid too little attention to social privations and that in view of
empirical studies documenting the economic and social pay-offs from investments
in health and education it would be wise to prevent the ‘depreciation of human
capital’ during the adjustment process. Despite this shift, the dominant view was
that ‘fiscal restraint’ had to be strictly observed. The dilemma of how to respond
to social needs while remaining within the constraints of macroeconomic stabi-
lization was resolved by attempting to ‘target’ social expenditures to populations
most in need (Vivian 1995). Certain expenditures were thereby reallocated, for
example from secondary to primary education; and supplementary programmes,
‘safety nets’ and ‘emergency funds’ were developed for the poor.
A consensus seemed to be emerging on the importance of social provision and
social protection, especially in a globalized economy where vulnerabilities were
accentuated due to exposure to external risks. That this was not merely a rhetorical
shift becomes apparent from the increasing share of Official Development Assistance
(ODA) going to the social sectors (UNRISD 2005: Figure 8.1). Moreover, in several
countries, levels of public social spending were restored in response to popular
pressure and discontent (UNRISD 2005: Table 3.3). At the same time, the state’s
role in the development process as well as in ensuring effective governance was
increasingly recognized (World Bank 1997). But what precisely was the scope for
state action in the realm of social development and welfare policy? And how was
it to triangulate with the institutions of the market, family and civil society?
The 1997 Asian financial crisis prompted the G7 to request the World Bank to
formulate ‘social principles’ and ‘good practice of social policy’ as a guide to policy
makers worldwide. The new emphasis was reflected in the World Bank’s
2000/2001 World Development Report (Attacking Poverty), which identified ‘social
risk management’ (SRM) as the most sustainable basis for coping with risk and
reducing the vulnerability of the poor. In this framework, the state was expected
to provide ‘risk management instruments where the private sector fails’ in
addition to ‘social safety nets’ for risk-coping for the most vulnerable (Holzmann
and Jorgensen 2000: 18) – displaying continuities with the earlier generation of
minimal safety nets. As we will see later, the family and the community were seen
as central institutions in SRM, working in tandem with the market.
The welfare pluralist approach was adopted in a context in which the provision of
social services and social protection was being rapidly commercialized. Selectivity in
social policy thus went hand-in-hand with a trend towards multi-tierism in modes
of provision in several important areas – pensions in particular, but also healthcare
and education. While selectivity means narrowing the targets for support, multi-
tierism means reducing the state component and partially privatizing social protec-
tion. Market-based, individualized entitlements (such as private pensions and
health insurance) are thus seen as appropriate for those who can afford them, while
scarce public resources are to be channeled or ‘targeted’ to the poor in the form of
elusive safety nets and ‘basic’ public health and education services.
Hence, behind the apparent consensus about the importance of social policy
lurks the ‘traditional great divide in social policy debates’ (GASSP 2005) between
the so-called universal redistributive model with an emphasis on universal access
to social services and a focus on equity and redistribution, and the residual
approach which ‘sees social policies as a residual measure to address the plight of
the poorest and most vulnerable’ (p.3). The ascendance of the latter approach was
at least partly related to the particular policy emphasis on poverty alleviation in
the 1990s, now in full bloom with the Millennium Development Goals (MDG),
which sought to ensure that public resources were geared towards only the poor-
est (GASSP 2005). By contrast with developed countries, the elaboration of social
assistance and social security in developing countries has thus been rooted in the
new global discourse of poverty, which has in turn fed into and reinforced a
bifurcated view of social policy (markets for those who can afford them and ‘safety
nets’ for the poorest). Below we consider the key building blocks of the currently
dominant model of social policy – commercialization, targeting, and the new
discourse of welfare developmentalism – from a gender perspective.
Commercialization and familialization of welfare

Social-sector reforms (health, education, pensions) have, among other things,
entrenched the commodification of public services through the imposition of
‘user fees’ and other charges, enhanced the role of the private for-profit sector in
the provision of social protection and services (sometimes through the privatiza-
tion of what were previously state/public services), and shifted some of the unmet
need for welfare onto families (re-familialization). Adopting Mackintosh and
Koivusalo’s (2005) usage, we understand the concept of commercialization to
be ‘wider than the “private sector” of provision and finance, encompassing for
example commercial behaviour by publicly owned bodies … and broader than
“privatization”, which refers to the sale or transfer of state-owned assets into
private hands’ (p.2).
In many low-income countries healthcare commercialization has been a key
area of public concern and the model of reform promoted by the IFIs, widely
referred to as Health Sector Reform (HSR), has entrenched the shift toward greater
commodification of public services. In sub-Saharan Africa, as Mackintosh and
Tibandebage (this volume) show, HSR has been promoted in a context of severe
poverty and in the wake of economic crisis. In Tanzania, for example, user fees
were introduced in the early 1990s at a time when cuts in government spending
and removal of subsidies on basic goods were disproportionately affecting the
poor. In a context where a significant proportion of the population cannot afford
its basic needs, the imposition of fees for healthcare has been impoverishing.
The currently dominant model of health service provision has particularly prob-
lematic implications for women – as users of health services, as health service
workers, and as providers of informal care. As Mackintosh and Tibandebage show,
HSR in Africa seems to be based on unrealistic assumptions about poor women’s
ability to muster the economic resources needed to access services for themselves
and their dependents, about their unlimited time and capacity to provide unpaid
care when formal care remains out of reach, as well as problematic assumptions
about the interests of healthcare workers and clients (for example, seeing them as
antagonistic).
A common policy response to the exclusionary effects of ‘user fees’ has been the
promotion of mutual health insurance schemes as well as social insurance
schemes. Unlike social insurance schemes which are employment-based, mutual
health insurance (MHI) schemes are voluntary schemes to promote the inclusion
of the poor and vulnerable by pooling their risks and providing exemptions for
those unable to pay. However most community-based MHI schemes face the
problem of low participation rates and lack of financial sustainability; in Tanzania,
for example, many rural Community Health Fund (CHF) schemes have not been
able to extend their participation rates beyond 10 per cent of eligible households
(Tibandebage 2004). Inability to pay constitutes one of the main reasons for non-
enrolment, which is likely to affect women more severely given that they are more
cash-constrained than men, and are likely to have more health needs. It is also not
clear how such schemes can provide exemptions for the poor and ensure financial
sustainability in the absence of significant subsidies from the state – given the
difficulties of having cross-subsidies from the better off in small-scale voluntary
schemes in poor communities.
Enrolment in social insurance programmes is very often employment-based,
with mandatory affiliation. In the case of developing countries, coverage has
tended to be limited due the large size of the informal economy, and the high rate
of evasion of contributions, even by employers and employees in the formal sector
(Huber 2000). For these reasons even though social insurance schemes facilitate
resource mobilization via contributions, they may not be the most effective
vehicle for extending coverage to the majority of the population, particularly
women who, as we showed in the previous section, tend to be informally
employed. But in some countries efforts are being made to extend social protec-
tion mechanisms to formerly excluded categories of workers (Lund, this volume).
The full implementation of these schemes, however, remains a challenge given the
administrative difficulties of reaching dispersed workers with erratic incomes,
monitoring the implementation of their new rights, and making their membership
financially sustainable.
The health insurance scheme built by SEWA (Self Employed Women’s
Association) in India, discussed in Lund’s chapter, is an example of an employment-
based scheme that is successful in reaching women informal workers; this is done
through an integrated insurance scheme which also offers life and asset insurance.
The reasons for its success include exceptionally strong leadership, top-flight
expertise from insurance experts, a willingness to respond to members’ needs and
set realistic rates, as well as the fact that it has been built on existing patterns of
solidarity of SEWA as a trade union and as a set of cooperatives for women with
strong organizational capacities. Replicating such conditions will not be easy.
Health sector reform in China, discussed in Jufen’s chapter in this volume, is
taking place alongside a fundamental restructuring of the labour force marked by
massive unemployment in the state industrial sector and large-scale migration of
the rural workforce into export-oriented factories. This stands in marked contrast to
the ‘full-employment’ scenario of pre-reform China when industry was exclusively in
state hands and welfare was provided through the enterprise. To be sure, gender
segmentations in the workforce often translated into stratified rights to welfare in
the pre-reform era: men greatly outnumbered women as permanent state workers
and also predominated in senior grades, while collectives constituted a feminized
sector (Lee 2005). But the proposed schemes for health insurance do little to
reduce gender segmentations, and are in fact likely to exacerbate gender-based and
other forms of exclusion.
The ‘basic health insurance scheme for urban workers’, which is the principal
component of China’s health insurance scheme for the urban population, covers
those who are in the formal workforce and have permanent residence permits,
thereby leading to the exclusion of informal workers, migrant labourers and those
who are not part of the workforce. These exclusions are exacerbated by the
scheme’s highly individualistic design which does not even provide coverage for
the ‘dependents’ of the insured. Furthermore, as Jufen’s contribution to this
volume clearly shows, a social insurance model with gender-neutral design and
individualized accounts is likely to produce very unequal outcomes for men and
women in terms of access to benefits (relative to need) when it is filtered through
structural inequalities, especially inequalities in wages/income and years of
employment. If coverage in social insurance programmes remains employment-
based and individualized with little subsidy from the state, then women’s labour
market disadvantages are likely to feed into their weaker claims on healthcare.
In the education sector the logic of ‘targeting’ promulgated at the international
level has prioritized primary education, where the ‘rate of return’ is presumed to be
higher and where public expenditure is considered to be pro-poor. In countries such
as India, discussed in Jha and Subrahmanian’s chapter, this logic seems to have led
to a significant reallocation of public social expenditure from higher education to
primary education. While this redistribution and the accompanying reforms in
Indian education have been vital for rapidly boosting India’s abysmal education
enrolment rates (Figures 1.1–1.4), it has also led to the unfortunate neglect of
secondary education. The systemic interconnections between different parts of the

education system have been ignored: the availability of post-primary education, for
example, can strengthen the pipeline that channels students through the education
system by giving parents an incentive to send their children to primary school and
by providing the next cohort of teachers. Moreover, many of the benefits that girls
and women reap from education – access to employment, access to contraception,
ability to negotiate intra-household relations – materialize at the post-primary
level (UN Millennium Project Task Force on Gender Equality 2005: Chapter 3).
The reallocation of funds has been facilitated by donors’ financial support for
new investments in primary education, which require ‘counterpart’ government
funding. The concentration of public funds at the primary level has in turn facili-
tated an expanding role for commercial provision at the secondary level. This
raises questions about affordability and access for both boys and girls from less
privileged backgrounds. It poses particular problems for girls in a cultural context
where parents prioritize their sons’ education, especially when the direct costs of
schooling are high relative to household income.
Such mechanisms seem to have been at work in the Indian state of Uttar
Pradesh, where state directives and infrastructure grants for the provision of
private single-sex secondary schools for girls have been subverted at the local level, as
schools benefiting from such grants were not able to recruit enough girls to make
their enterprises profitable. While the larger concern for girls’ education at the
national and international levels made it a good political decision to have schemes
to improve the provisioning of girls’ schools, narrow political and electoral
considerations at the local level subverted the very basis of allowing subsidy for a
private enterprise. As Jha and Subrahmanian conclude, ‘In an environment where
girls’ education is not a highly valued choice, increased privatization [commercializa-
tion] is bound to act against their schooling participation’. In particular, the role
of the state in mobilizing demand and influencing change in the domestic
calculus that leads families to under-invest in adolescent girls’ schooling is under-
appreciated where commercialization is seen as the dominant policy choice.
An important set of reforms that swept through many middle-income countries
over the past decade was the reform of public pension programmes, where
multilateral financial agencies and private commercial interests were strongly rep-
resented, in addition to technocratic elites and domestic constituencies such as
trade unions and political parties. The outcomes have been diverse, and in some
countries domestic political coalitions were able to resist the privatization model
being imposed by the IFIs. While gender concerns do not seem to have surfaced in
the public debates on pension reform, the moves towards privatization have major
implications for women with typically lower labour market capabilities. The fact
that benefit levels in privatized and individualized systems correspond closely
to the overall contributions made by the insured person, means that women with
their typically lower wages and fewer years of employment are likely to gain lower
benefits (Huber and Stephens 2000b). The fact that fully privatized systems also
take into account life expectancy further works against women although this is
typically partially mitigated by allowing women to draw a pension at an earlier age
than men. In public systems women’s disadvantages are often mitigated by

generous minimum pensions, by the fact that life expectancy is not taken into
account, and that credit is sometimes given for periods of full-time care.
Targeting versus universalism: is it gender-coded?

Social policy regimes are hardly ever purely universal (providing benefits to the
entire population as a basic right) or purely based on targeting (providing benefits
to selected members of the population, usually on the basis of their ‘neediness’);
they tend to lie somewhere between the two poles along a continuum and
often in hybrid forms (Mkandawire 2005). As we noted earlier, targeting has
become particularly attractive in the context of tight fiscal policies and aid
dependence. In such contexts the relationship between social provisioning
and the political economy of resource mobilization tends to be weakened; indeed
targeting concentrates on the problem of disbursing given external resources, and
not on that of generating and disbursing domestic resources through taxation
(Mkandawire 2005).
The emphasis of poverty discourses on support only for the most vulnerable
groups in society has evaded some difficult distributional questions related to the
position of middle-income and formal-sector workers in developing countries,
many of whom relied on public-sector employment for their mobility and
educational opportunities. In both African countries and post-socialist countries
cutbacks in public-sector employment have led to the downward mobility of these
strata. As Haggard and Kaufman (1995: 11) argue, this has significant consequences
for resource allocation disputes because downward mobility can lead to anti-
democratic mobilization: it is difficult to emphasize solidaristic principles to
underpin social policies since ‘blue-collar and middle-class groups are unlikely to
support antipoverty subsidies unless they can share in the benefits, and they are
likely to oppose them if they entail a reduction in existing services’.
Targeting has been justified in terms of cost containment and efficiency – the
‘most efficient and commonsensical thing to do under the circumstances’
(Mkandawire 2005: 3). Ironically though, many developed and developing
country governments persist with targeting despite the ‘stubborn slew of empiri-
cal evidence suggesting that targeting is not effective in addressing issues of
poverty (as broadly understood)’ (ibid p.20). It involves heavy administrative costs
and demands the kind of administrative sophistication, information-gathering
capability and infrastructural capacity that is difficult to find in many developing
countries, with the result that significant numbers of poor people are often missed
from targeted programmes while some of the non-needy are included (ibid).
Targeted programmes may also divide communities and create tensions between
those who are included as beneficiaries and those who are not (Molyneux, this vol-
ume). Most strikingly, targeting tends to be associated with increasing levels of
inequality; this is well-captured in Korpi and Palme’s (1998) phrase ‘paradox of
redistribution’: ‘the more we target benefits at the poor and the more concerned
we are with creating equality via equal public transfers to all, the less likely we are
to reduce poverty and inequality’ (p.661).
The current jargon – ‘fiscal restraint’ and ‘efficient allocation of resources subject
to budget constraints’ – seems to suggest that the problems of targeting are
technical. By contrast, Mkandawire (2005) underlines the ideological and political
imperatives which determine the choice of instruments used to address poverty,
inequality and insecurity. As the literature on social assistance programmes pro-
viding unreciprocated aid to the ‘deserving’ poor has repeatedly shown, there is a
strong element of control that underpins targeted programmes which goes against
current notions of citizenship and empowerment. In measures such as community
targeting much discretionary power can be vested in local unaccountable admin-
istrators who then wield enormous powers over ‘matters of life and death’
(Mkandawire 2005).
It has long been argued in the context of US welfare policies that government
programmes were divided into two, gendered streams: those with the most
legitimacy guaranteed secure entitlements to some citizens in return for their
contributions, while other programmes provided unreciprocated aid to the
‘deserving’ poor (Fraser and Gordon 1994). The gender-coded contract-versus-
charity dichotomy persists today in many countries in the opposition between
social insurance and public assistance programmes. Welfare benefits targeted to
women as mothers, for example, tend to identify women and their children as
needy and poor, while shading out their contributions as workers/carers (whether
paid or unpaid) and their rights as citizens. Social insurance programmes, on the
other hand, confer entitlements on workers in return for their labour/financial
contributions, even if in fact benefits depart from actuarial principles and do not
reflect financial contributions (Fraser and Gordon 1994).
Current anti-poverty initiatives also invoke and depend on gendered assumptions
about identities, interests and responsibilities of citizens. In recent years women
have been heavily present in targeted social assistance programmes. This can, per-
haps, be seen as a response to feminist policy advocacy in many national and
global forums where women have been identified as ‘the poorest of the poor’ and
governments were requested to take appropriate action. One response has been in
the form of micro-credit initiatives, which have been directed at women on the
basis of their assumed thriftiness, efficiency and entrepreneurial skills.
Elsewhere social assistance is provided on the condition that clients conform to
certain normative understandings of ‘good motherhood’ or appropriate female
behaviour. This is most evident in the Mexican programme Oportunidades,
explored by Molyneux in this volume. This is a programme that aims to enhance
human development by focusing on children’s health, education and nutrition. It
is premised on strongly normative – but implicit – assumptions about the ‘natural’
role of mothers as full-time carers and nurturers of children. The programme
seems to be driven by the expectation that low-income mothers would accept ben-
efits for their children even if these are made available on demanding terms (the
unpaid work expected of them) as well as intrusive levels of control exercised by
programme managers. This may very well be the case, given the lack of any better
option as well as the side benefits that women may derive from involvement in
such programmes, such as greater self-confidence, an ability to access public
spaces and connect with other women beyond kinship and marriage networks.
But such objectives are not always built into the programme (through training and
group activities, for example), nor is it clear whether participation in such
programmes enhances women’s economic autonomy and security.
The dual gender stream in ‘developmental’ social policy

Given the poor record of neoliberal policies in reducing levels of poverty, an effort
has been made in recent years by some social policy analysts and international
organizations (or parts thereof) to re-think the basis of anti-poverty programmes
by focusing on the historical experiences of Northern welfare states. Some analysts
have attempted to shift the global discourse away from targeting by reasserting the
politics of social solidarity and universalism.6 Drawing on the historical experiences
of Nordic states the argument is made that social policy is not only an end in itself
but also a means of enhancing economic dynamism, development and growth – in
short, that social policy is developmental (Mkandawire 2004).
The concept of developmental social policy, which is sometimes used inter-
changeably with the notion of ‘productivist’ welfare state, has appeared in very
diverse policy contexts. The early ‘productivist’ logic in East Asian social policy
was premised on protecting and enhancing conditions for male workers in
selected core industries, concentrating public expenditure on education, and rely-
ing on the family for the provision of care. The concept of ‘developmental social
welfare’ also appears in various South African government documents, where it is
defined as a welfare system ‘which facilitates the development of human capacity
and self-reliance within a caring and enabling socio-economic environment’ (cited
in Hassim, this volume). Here it seems to have, perhaps unintentionally, instituted
a hierarchy of entitlements in which wage-earning programmes such as public
works are ideologically privileged over social assistance programmes. In yet
another policy context, that of the EU, the notion of developmental social policy
has been equated with ‘active labour’ and ‘work first’ strategies aimed at getting all
adults into the labour force.
One of the common concerns that seems to underpin the ‘productivist’ logic in
its different manifestations is the longstanding anxieties about the disincentives
that welfare ‘handouts’ can create for work effort, leading to apathy and
clientelism. Paid work, on the other hand, is seen as contributing to development
while providing a route out of poverty. There is clearly some truth to this logic:
economic and social policies must be able to create economic dynamism and full
employment – what East Asian developmentalist states were amply able to achieve
in the 1970s and 1980s, and what is so out of reach for many developing countries
(including South Africa) in the present era. But as we argued above, if poverty is to
be reduced then what is needed is decent employment because some forms of paid
work actually entrench poverty and erode people’s capabilities. But even with this
proviso we would argue that the ‘productivist’ logic remains problematic because
it does not take into account the relationship between paid and unpaid forms of
labour, and does not acknowledge that the latter is just as much at the heart of
provisioning of needs as the former.
The Confucian principle of ‘filial piety’, which underpinned the reluctance of

East Asian states to fund social care services, relied on women’s unpaid care work
within families. Similarly as Lewis and Giullari (this volume) extensively argue,
the EU ‘active labour’ agenda, which aims to get everyone, regardless of gender,
into the labour force but with scant attention to how individuals and households
would make arrangements for care work, is likely to have gender-differentiated
implications. As Elson puts it, ‘in order to be gender equitable, full-employment
policies must be complemented by entitlements for those in informal or part-time
paid work and entitlements for the providers of unpaid caring labour as citizens in
their own right’ (2004: 70).
Education, health and contributory pensions are often treated as productive
investments, while welfare grants tend to have a Cinderella-like status for finance
authorities, especially when they compensate women for their unpaid care work.
As Steinhilber’s contribution in this volume shows, this has been the fate of family
benefits in both Poland and the Czech Republic since the ‘transition’. Aggregate
expenditure on family benefits declined over 1990s: in Poland from 1.7 per cent of
GDP in 1990 to 1.06 per cent in 1998; in the Czech Republic from 1.6 per cent of
GDP in 1996 to 1.2 per cent in 2002, even while total social expenditure increased.
A similar reading emerges from Hassim’s analysis of the struggles of the Lund
Committee in South Africa in its efforts to protect and reform the child mainte-
nance grant against criticisms from many in government for fostering dependency
and having no developmental potential.
Comparative research on welfare states finds that service-heavy states (social
democratic states that typically fund and deliver welfare services such as health,
education, daycare and elderly care) tend to be more ‘woman friendly’ than transfer-
heavy Christian democratic ones which tend to fund but not deliver public
services (Huber and Stephens 2000a). The services considered include health and
education as well as daycare, which tend to be citizenship-based and which
facilitate women’s entry into the workforce (both by releasing women from
unpaid care work and also by creating jobs for women in the public care sector).
In the case of developing countries care services for children and the elderly are
far less developed (much of the care taking place through informal kinship net-
works as well as informal paid care) but there is no doubt that welfare services such
as health and education have the potential of reversing wider gender inequalities
and discriminations. As Mackintosh and Tibandebage argue, ‘health systems’
immense political and social importance, and the association of health care
reform with political and economic crisis, implies that health systems can be sites
of challenge to gender disadvantage’.
But there should also be a place for cash transfers or non-contributory income
supports such as child allowances, family benefits and pensions within ‘productive’
welfare – resisting the notion that these are ‘handouts’ for passive clients. Income
supports can play a crucial role in helping households to provide care, smooth out
consumption and build up the capabilities of all household members, as the
evidence cited by Lund on the South African old age pension clearly shows.
The provision of accessible and accountable health and education services, as

well as universal or near-universal entitlements to non-contributory pensions and
other income supports remains of utmost importance to correct market outcomes
and social arrangements that disadvantage low-income households, and low-income
women in particular. It also requires a sound system of taxation – as Huber’s
chapter argues and explores.
Looking at four countries – three in Latin America and one in the Caribbean –
Huber documents the mixed record that these countries have had over the
last two decades in reforming their tax systems, with the predominant trend
being a negative one. In three of the four countries overall tax receipts as a per-
centage of GDP has fallen (the exception being Costa Rica), and three of the
countries have come to rely more heavily on indirect taxes levied on domestic
goods and services (the exception being Jamaica). Both developments are bad
news for the state’s capacity to reduce poverty, inequality, and gender-based
disadvantages. With a lower tax revenue the state has less fiscal capacity to
provide the kind of social services that we know are critical for low-income
households and for low-income women in particular, and indirect taxes also
tend to be generally regressive. And yet while it is politically difficult, it is
nevertheless possible to construct overall progressive tax systems that are
effective in raising at least moderate levels of revenue to fund state social provi-
sioning, even at comparatively low levels of development, as demonstrated by
the example of Jamaica. As Huber concludes, the requirements for this are:
reliance on direct taxes with an adequate standard deduction and few special
deductions for higher income earners for an important share of the total, a
structure of indirect taxes that exempts only basic goods and services, and social
security taxes without a cap. However, there is no escaping the fact that taxing
high-income groups is politically costly, and that liberalization has made taxing
corporations more difficult.
Families, politics and state development in a global context
Debates about the relative balance of responsibilities between state and non-state
institutions in the provision of social services and social support are underpinned
by normative assumptions about the role of the state as well as about gender roles.
The residualist approach to social policy is not just an innocent statement describ-
ing the limited role played by the state in social provisioning, but one prescribing
how states ought to behave. As feminist critics have repeatedly argued, the shed-
ding of state responsibilities shifts even further the burden of social provisioning
to the unpaid providers of care within families, households, and communities. On
the other hand, defining the state ‘as the central locus of social policy’, as
Molyneux (forthcoming) notes, also carries normative assumptions, alluding to a
vision of a strong, capable and socially responsive state. While much of the debate
between neoliberals and their critics tends to be cast in terms of markets versus
states, in many countries, low-income populations have to creatively combine
social supports from a mix of formal and informal social institutions for their secu-
rity. As Molyneux goes on to argue, ‘social reproduction is in such cases secured by
a variety of social practices and institutions that exist independently or work in
conjunction with “state action”; yet this interface, if it is acknowledged at all, is
rarely analysed in the social policy literature’ (Molyneux forthcoming).
In this section we address the interface between the state and other social
institutions. Firstly, we discuss the implications for gender equality strategies of
the assumptions made by states about the role of the family in social provisioning,
and the ways in which state policies attempt to reform the family. Secondly, we
address the implications for developing gender-responsive social policies of
differing levels of capacity, reach and legitimacy of state institutions. Finally, we
consider the ability of women’s organizations to extract benefits for women out of
processes of political liberalization and democratization.
Assumptions about families and the politics of familialism

Existing welfare-state models are based on culturally and historically specific
conceptions of the divisions between public and private (and in particular on
relatively secularized public sectors), of the nuclear nature of the family, and of
fairly differentiated institutional spaces occupied by the care economy and paid
work. In these contexts, feminist critiques of Gösta Esping-Andersen’s (1990) test
of ‘decommodification’ (the extent to which citizens can attain a basic standard of
living independently of the market) have focused on adding the test of ‘defamil-
ialization’. This is defined as ‘the degree to which individual adults can uphold a
socially acceptable standard of living independently of family relationships, either
through paid work or through social security provision’ (Lister cited in Sainsbury
1996: 39). Hobson (1990) and Orloff (1993) have further refined this concept by
suggesting that feminist analysis should focus on access to paid work and capacity
to form and maintain an autonomous household. These critiques emerged in a
context where support to individuals was channelled through families as the unit
of entitlement, with women often required to conform to dominant stereotypes of
‘good’ wives and mothers. Entitlements based on citizenship are potentially
defamilializing, as they de-link entitlements from marital status and motherhood.
Yet both the test of decommodification and the test of defamililialization are
difficult to apply to less-industrialized developing country contexts, where com-
modification is weak and families and social networks, such as extended families
and religious groups, remain important cultural and survival resources. Feminist
social policy analysts by no means argue for a notion of individuals as atomized
and autonomous beings. Yet even the limited forms of defamilialization that are
proposed (for example, women’s capacity to maintain households autonomously
of their dependence on others) are difficult to apply in contexts where family and
kinship networks remain important to people’s livelihoods and security, and
where non-familial provision of social security is weak.
This kind of social embeddedness is not only a primary source of identity; it also
structures women’s economic entitlements by offering them some access to
resources such as land, housing and childcare even if only as a consequence of
their marital or maternal status. In extended families, the presence of multiple

contributors to the household’s security in high-risk environments acts as a
mechanism for spreading risk and mobilizing resources for investments and for
social reproduction (Whitehead and Kabeer 2001). In the midst of economic crisis,
when jobs disappear and the little state provision that there is becomes eroded,
these networks take on an even more critical role as the ultimate safety net.
Beneria and Floro’s contribution to this volume documents the ways in which
low-income households in Ecuador and Bolivia draw on kinship and family net-
works and systems of reciprocity and mutual support for protection, income-
maintenance and as a way of smoothing consumption. Formal social protection
mechanisms in these communities are conspicuous by their absence.
Yet, there are at least two issues that we need to consider when analysing the
ways in which families and households provide security and social provisioning in
many developing countries. First, feminist research amply shows the ways in
which women are often disadvantaged vis-à-vis men in the pursuit of livelihood
security, through both intra-household inequalities in the allocation of resources
and the sharing of burdens, as well as inequalities generated by biases in the wider
institutional arena. One example of such inequalities emerges from Beneria and
Floro’s chapter: not only do women in these low-income households continue to
shoulder a disproportionate share of the unpaid work involved in household
reproduction (alongside their increasing participation in paid work), but they also
have a higher debt burden than the men in their households as a consequence of
assuming greater responsibility for debt repayment and household maintenance.
Second, while classical liberal theory sees the family as a realm unto itself – distinct
from civil society and shielded from incursions by the state – the notion that states
and families operate as ‘separate spheres’ does not hold up to historical scrutiny
(Haney and Pollard 2003). Nor does such a framework assist our understanding of
contemporary reforms in state roles and obligations vis-à-vis social welfare, which
carry enormous implications for what is expected of families (and women in
particular). Indeed, comparative work on the family in contemporary societies,
especially post-socialist Eastern Europe, shows how ‘the familial’ may be deployed
to assist states’ reform of – and often retreat from – social life (Haney and Pollard
2003; Haney 2003).
There is growing policy interest in devolution of responsibility from the central
government to local governments (through decentralization) as well as non-state
actors, especially ‘communities’, according to the principle of subsidiarity – reliance
on the smallest possible unit that can perform a given social function effectively.
These ideas have always been very strong in religious-based conservative political
parties and movements, for example in Christian democracy, which has also placed
strong emphasis on traditional family values and arrangements. But the underlining
of women’s traditional roles associated with caring and the neglect of their eco-
nomic autonomy and security is also evident in some contemporary anti-poverty
programmes, as Molyneux’s analysis of the Oportunidades programme shows.
The current policy emphasis on subsidiarity tends to show very little interest in
the gendered workings of non-state institutions and the ways in which they can
reproduce and entrench gender inequalities. Who in the community does the car-
ing and on what terms? Who does the voluntary work? Who benefits from social
provisioning provided by religious-based voluntary institutions and on what
terms? Who in the household does the bulk of social provisioning and caring?
These questions are not new. They lie at the heart of debates to measure the extent
of social rights in welfare states and are central to contemporary processes that
have been loosely termed ‘reprivatization’. The care burden imposed by the AIDS
epidemic, as Mackintosh and Tibandebage note, has cruelly exposed the inadequacy
of the assumptions about women’s coping capacity and unlimited labour supply.
This is forcing onto the policy agenda questions about care and its provision, and
the costs to the carer of unpaid and voluntary work.7
Steinhilber’s chapter shows that despite the divergent reform paths taken in
Poland and the Czech Republic, in both countries the family was addressed as ‘the
most natural’ social institution to ensure care for its members; as increasing
responsibilities were assigned to families in the reform process, supporting families
in coping with the social fallouts from the transition was considered eminently
important. Gender relations in families, however, and the impact of family
benefits on them did not become an issue for public debate. Indeed, reform
discourses have invoked and aimed to recreate traditional gender arrangements
with motherhood as a full-time dedication. These idealized notions of the family
and of women’s roles are often in conflict with the reality of women’s day-to-day
lives and their continued attachment to the labour market (more so in the Czech
Republic than in Poland). Indeed, women face greater difficulties in the post-
reform era in combining employment and family responsibilities than they did
prior to the reforms. Ironically, despite the pro-family rhetoric, in both countries
aggregate expenditure for family benefits declined over the course of the 1990s. In
Poland, the share of family benefits in total social expenditure declined while total
social expenditure increased.
A more radical engineering of the family is evident in the Islamic Republic of Iran
where a revolutionary Islamist state has sought to restructure the family, and society
more broadly, along traditionalist lines by attempting to segregate all public
spaces, and to fully domesticate women.8 Moghadam argues in chapter 4 that state
‘developmentalism’ in the Middle East was undermined by its ‘neopatriarchal’
approach to women, gender, and the family, codified in family laws/personal status
codes that define women largely in terms of their filial, marital, and maternal roles,
place them under male guardianship, and deny them equality in access to family
wealth. Although state social expenditures did allow for social mobility and access
by some women to education and employment, on the whole, oil wealth and fam-
ily laws served to prevent the ‘commodification’ of women’s labour. On these
grounds she argues that processes of individuation and commodification need to
be supported as without these developments women are unlikely to be considered
as economic agents and right-bearing citizens. Hence, some disembedding of social
provisioning may be desirable as far as gender justice is concerned.
Individualist approaches to development have, however, been criticized for
assuming that poverty is the result of individual (or even broader cultural) deficits,
rather than structural features of dependent peripheral economies. Instead, from

both the left and the right of the political spectrum, there has been a tendency to
focus on the community as the locus of welfare and the site for social justice. Left
perspectives on community tend to emphasize the building of agency and
participation through local-level democracy that often draws on indigenous prac-
tices of collective responsibility, mutuality and reciprocity. While women provide
the backbone of caring and sharing within communities, notions of justice and
entitlements embedded in community institutions may, however, be oblivious to
gender equality.
Many community-based initiatives for addressing poverty have drawn on
women’s skills and capacities to develop programmes over which poor people have
control and to develop collectivist strategies that assist in broader democratization
of society. These kinds of collectivist programmes link community well-being to
political goals, often tied to socialist advocacy. In some cases, they are so success-
ful that populist governments co-opt them as part of the project of legitimation
(see Blondet 2002 on Peru). The political framework of collectivism, however, is
very different to communitarian approaches that emphasize sharing and caring
within communities rather than directing demands to the state (Midgely 1995:
91). These attempts to enhance community self-control can often be retrogressive
for women, particularly when linked to the entrenchment of traditional forms of
authority and cultural stereotyping of gender roles. These cultural contestations
can often be very muted yet the effect of powerful traditionalist interests and
visions can be far-reaching, as the case of Iran clearly demonstrates.
The resilience of these informal institutions, their ability to substitute for state
services, and their effectiveness at providing members with dignity and social
purpose, mean that these institutions must be recruited to the task of rebuilding
social cohesion in failing states (UNRISD 2005: 259). It may however prove very
difficult to insert gender-equality concerns (or broader social equality concerns) to
these processes, especially where traditional institutions have a patriarchal charac-
ter. But traditional institutions can be held accountable to basic constitutional
standards of social equality, if there is a strong, legitimate and capable state that is
committed to women’s rights.
As Lewis and Giullari show in chapter 8 , this kind of state was hard-won in the
North, and the commitments to gender equality have to be continually renegoti-
ated in the face of economic and political shifts. Yet these struggles have been
difficult for feminists to pursue in developing countries in the late twentieth
century. The closed, nation-state model of development that, for example, enabled
organized labour to extract some livelihood guarantees has been undermined by
processes of economic liberalization and the indebtedness of many countries. This
has impacted on political struggles to reduce the vulnerabilities of citizens to the
market. Finally, the institutional legacies of the state play a major role in determin-
ing the extent to which states are able to carry through developmental programmes
to reduce social inequalities.
Developmental state variation is marked by the different capacities – economic
and political, to be sure, but also infrastructural – of different states that shape the
extent to which collective struggles for greater protection are successful. The next
section examines the interplay between these factors more closely.
Reach, capacity and legitimacy of state institutions

There are enormous variations in state-society linkages and state capacity between
developmental states in Latin America, Asia and Africa. Postcolonial governments
in Africa were initially committed to the idea of an active state that would drive
development and poverty reduction. As Robert Bates points out, although African
governments adopted different developmental paths, ‘they were virtually all
activist’ (Bates 1994: 15). However, for at least three decades following indepen-
dence, most governments did not expand the institutions established by colonial-
ism (executive, civil service, police and army) in ways that consolidated
democracy or even their long-term ability to sustain a developmental focus. In
particular, institutions that would constrain executive power such as multi-party
elections, judicial independence and, outside the state, institutions that might
expand the legitimacy of the state and its capacity to represent diverse interests
(such as vibrant civil society) were either severely restricted or actively repressed.
By contrast, those institutions that were seen as either enhancing the capacity of
elites to manage or to remain in power, such as the bureaucracy, expanded rapidly.
Importantly, however, bureaucratic expansion was not tied to efficiency or to
citizen responsiveness.
Structural adjustment policies further weakened states’ commitments to social-
sector spending, undermining the purchase of nationalist governments on the
loyalties of poor people. Donald Rothschild (1994) argues that state autonomy was
weakened at a crucial stage and the state was unable to regulate society effectively
or to implement its ambitious developmental programmes. The inability of
postcolonial states to deliver on the developmental promises that accompanied
political independence, whether as a consequence of leadership deficiencies or
external impositions, weakened state legitimacy, with many groups in society dis-
engaging from making demands on the state and instead entrenching informal,
traditionally-based systems of governance and resource allocation. Citizens
bypassed the state as the locus of their demands, meeting their needs through a
combination of informal mechanisms and developing allegiances to local political
actors rather than the state per se. These developments reinforced the peculiar
dilemma that new democracies face, as Gordon White (1996) has pointed out: lack-
ing legitimacy, new democracies cannot become effective; lacking effectiveness they
cannot develop legitimacy.
To be sure, policy outcomes are shaped not just by political commitments to
invest in the reduction of social inequalities, but also by the capacity of institutions
to implement policy. Yet there is no simple relationship between public spending
and outcomes in terms of poverty reduction. A key intervening feature in the strug-
gles to develop effective systems of risk reduction for poor people is the institu-
tional character of the state. Democratization depends to a significant extent on
the consolidation of state power and the institutionalization of political contesta-
tion within elites and between elites and other social groups (Rueschemeyer et al.
1992). Gordon White argues that the effective linking of democratic and
developmental goals depends on the construction of ‘an effective developmental
state that is regulatory, competent and redistributive, and has the political authority
to manage social and political conflicts’ (quoted in Minogue 2002: 127).
The ‘good governance’ approach of the World Bank, which promoted democra-
tization through institutional development, recognizes the importance of ‘getting
the state right’. However, the emphasis is on strengthening the institutions that
are considered to be essential for capitalist development such as those responsible
for financial management, private property rights and rule of law. The aim of con-
tributing to more efficient government has not necessarily been linked to the
value of promoting government accountability downwards to citizens. Women
clearly have an interest in a responsive and accountable state, but one that is
responsive to their particular needs. There are gender-specific capacity failures in
all public institutions targeted for reform. Public expenditure management
systems fail to acknowledge women’s needs or distribute budgetary resources
equally. The civil service or judiciary may be dominated by men antipathetic to gen-
der equality. Women public-sector workers clustered at the bottom of state bureau-
cracies may be the first to be fired when cost-cutting efficiencies are introduced
(UNRISD 2005: Chapter 11).
As the Latin American and East Asian chapters suggest, competent public bureau-
cracies that are at least internally accountable can be made responsive to the needs
of women. In these regions states appear to be more able to act on political agree-
ments struck between political parties and other actors. Latin American states
achieved consolidation relatively early, often in association with demands for mass
incorporation into political institutions (Rueschemeyer et al. 1992). Although
there are variations in the social composition and democratic credentials of
states, on the whole Latin American states have been pervasive, relatively well-
institutionalized and with a strong history of interventionism. Grindle (1986: 13)
argues that ‘the Latin American state has played a considerable role in the mediation
of social conflict through co-optation, manipulation and coercion and this helps
account for the complexity and conflict that exists within the bureaucratic apparatus
of individual countries’.
The institutional context of democratization was also different in East Asian
states compared to sub-Saharan Africa. East Asian states inherited highly evolved
bureaucracies from a combination of Japanese models as well as from strong
networks between state and political and economic dynasties. This left a strong
foundation for the developmental states (indeed the authoritarian legacy allowed
developmental states to intervene in markets in productive ways). In these
contexts, the development of gender-equitable social policies was dependent far
more on winning political support and social consensus over the direction of
social policy than on state capacity to absorb women’s demands.
Weak state capacity also impacts on the kinds of strategies feminists can employ
in making social policy more gender-equitable. One of the most innovative of
these strategies, gender-responsive budgeting (GRB), has suffered from the com-
bined problems of political will and state capacity. This strategy is often adopted
because it is seen as the key tool for exposing and thereby re-directing government
policies and spending patterns to more gender equitable ends. As a ‘technical’ tool,
it is often regarded as a strategy that can bypass political and cultural obstacles to
equality. However, as Budlender argues in chapter 15 of this volume, GRBs can
often become a panacea for the more politically difficult work of redirecting
political will and institutional capacity. Budlender very clearly demonstrates the
differential outcomes of particular political contexts as well as the effects of the
lack of receptivity of politicians and state bureaucracies to change. Policy inertia
seems to dominate while opportunities for change (brought about by a combina-
tion of strong politicians, often female, with effective alliances in the face of
government receptivity to change) seem to be very rare moments.
Social policy as political struggle

Opportunities for creating states and political systems that are more responsive to
the needs of constituencies of poor people were greatly enhanced by the wave of
political liberalization that swept through developing countries from the late
1980s. Processes of political democratization raised expectations in civil society
that more redistributive policies would be followed, as indeed in many countries
they were. However, an important factor that needs to be taken into account when
considering the link between liberalization and redistribution is the strength of
political organization among poor and working people in developing countries. To
what extent was democratization forced from below rather than imposed from
above by a combination of local and foreign elite interests?
Even with the recognition of the need for a more activist state in global lending
institutions and the provision of more comprehensive protection for the poor, in
many African countries the impetus to provide social protection was externally
set, as part of the conditionalities of debt relief. The combination of this factor
with the weak tax base and small middle class in very poor countries had the effect
of removing social policy from the arena of national politics. As we suggested
above, these factors have consequences for the quality and financial sustainability
of social programmes. However, they also impact on the process of building a
social value consensus and on the political sustainability of social programmes
(Tendler 2004). As Tendler points out, in such situations the ‘national anchor’ for
social programmes is easily lost, and the anti-poverty agenda inadvertently can
become an ‘anti-labour’ agenda. Building programmes that provide protections
beyond the ‘poorest of the poor’ to include the organized working poor becomes
more difficult in the face of the combination of residualism promoted from above
by global lending institutions and populist arguments that employed workers
represent a labour aristocracy.
Similar political consequences are evident in the interventions of global lending
agencies in Eastern Europe. As Steinhilber shows, in Poland, which pursued a rapid
reform path, there was a wholesale dismantling of the welfare benefits system that
had existed under state socialism. The resulting residual, familial model, apparently
gender-neutral, downscaled ‘costly’ benefits and services that supported women’s
dual role as worker and mother. The absence of strong, local feminist lobbies or
allies in political parties allowed the adoption of a residualist welfare model that
seriously undermined women’s social rights.
A different dynamic was at play in East Asian developmental states, where the
process of democratization was more successfully pushed by local actors, and with
clear consequences for the expansion of social protection. Peng shows how male
bias in both the Korean and Japanese systems only softened as a consequence of
demographic shifts – declining birth rates and an ageing population – that in turn
became the touchstone for political competition. These demographic changes,
together with broader social changes, facilitated the erosion of traditional living
arrangements, increasing divorce rates and increasing numbers of single mothers,
accompanied by the increasing employment rates of women (including married
women). These social trends created a tension between caring needs on the one
hand (of children and the elderly) and the availability of women to provide
unpaid care on the other. Together these factors resulted in shifts in social policies
with increasing welfare expenditures. Care of the elderly shifted from being
means-tested to being rights based, while public supply of childcare was increased
and parental leave extended. All of this was facilitated by political regime shifts,
the extent to which social policy became an electoral issue and increasing
numbers of women in political office.
On the other hand, as Hassim’s chapter shows, similar processes of democratiza-
tion in South Africa have not had the same effects on the social welfare system.
There, the potential redistributive effects of regime shift and expansion of women’s
access to political office were mitigated by a dominant party system in which social
policy did not become part of electoral contestation, and by a labour market
characterized by high levels of unemployment rather than labour shortages.
Women’s movements in developing countries have an uneven record of
organizing for better and more appropriate forms of social protection. Many
women’s movements in developing countries have eschewed state-centred poli-
tics, questioning whether women’s citizenship could be expanded by states that
were fundamentally undemocratic in character. States are clearly not neutral in
designing and implementing social policies, but they are not self-evidently
patriarchal either. Gender relations are shaped by, and themselves shape, the
nature of the state and its relationships to other social institutions; in this sense
institutions are interpretive systems that give meaning to particular notions of
social positioning and citizenship. Yet unlike institutions such as the family and
community, states may be more permeable to women’s interests in contexts of
strong social conservatism. This may offer openings for women’s movements to
extend the reach of social programmes in ways that address women’s gendered
vulnerabilities. For example, in Latin America collectivization among poor women
did lead to successful pressures being brought to bear on ruling parties that were
seeking to extend their legitimacy among poor citizens. Populist governments in
the region sought to expand the basis of their legitimacy by expanding the welfare
net in ways that benefited both men and women.
It can be argued, therefore, that women have successfully made claims on the
state, very often by harnessing their maternal roles to political claims-making and
advocacy for better conditions and social support for women. Although maternal-
ist politics has had contradictory and different outcomes in different countries,
this form of claims-making shares an implicit acceptance that the rights women
were claiming should come in return for certain pre-given responsibilities tied to
traditionally-ascribed gender roles. This acceptance of traditional gender roles ren-
dered maternalist movements and demands controversial. The protean character
of maternalism also leant itself to subtle shifts from ‘a vision of motherhood in the
service of women to one serving the needs of paternalists’ (Koven and Mitchel
1993: 5). As such maternalism may reinforce the patriarchal gender order and
entrench women’s economic dependence on men – it can become ‘a cloak for
paternalism’ (ibid). For example, Islamist women activists in Iran have tended to
use maternalist arguments for greater social protection focusing on rewarding
women’s unpaid labour without addressing the underlying power relations of
gender. Consequently, Moghadam argues for a secular politics that focuses on
women’s access to the public sphere of paid labour and political participation.
From this platform, she argues, women are more likely to succeed in making family
relations more equitable.
Representation, political parties and social policy change

More recently, the international women’s movement (for example, in the Beijing
Platform for Action and WEDO’s global 50–50 campaign) is focusing on access to
formal political institutions as a key strategic lever to advance gender equality.
Such campaigns attempt to create mechanisms through which women can enter
into parliament in their own right rather than at the behest of benevolent political
parties. The success of this strategy rests on a number of factors, most importantly
the strength of gender equity lobbies in political parties and civil society, the way
in which political competition is organized and on the extent to which women
can make issues of social policy into electoral issues. There are many examples of
women’s movement activism, but fewer examples of successful party mobilization
in support of social policies.
On the face of it, certain kinds of one-party and authoritarian regimes have
paradoxically been permeable to some kinds of gender claims. These openings
were created not only for conservative maternalist political claims (as in the
authoritarian regimes in Latin America) but also for feminists. For example,
women in Uganda did benefit in many respects from the patronage of Yoweri
Museveni, gaining a large number of reserved seats in parliament. However, the
dependence of the women’s movement on the National Resistance Movement was
tested severely by the end of the 1990s, when women failed to gain support for
removing traditional obstacles to women’s ownership of land (Tripp 2002). The
Ugandan experience is a salutary reminder of the limits of party-political patronage
for gender activists.
The contributions by Peng (chapter 6) and Hobson (chapter 7) examine coun-
tries where gender equality was successfully inscribed in social policies. Hobson
argues that winning support from political parties and raising the electoral stakes of
gender is crucial to women’s movement success. However, there is enormous
variation in the political party landscape in developmental states. Where party

formation is relatively strong with deep histories and clearly developed ideologies,
the ability to make electoral gains depends on the extent to which there is politi-
cal commitment to redistribution. In Japan, for example, social policy emerged as
a key electoral issue in the 1990s, political parties distinguishing themselves on
the basis of their positions on social welfare. Women’s organizations had been
campaigning for social policy reform since the early 1980s, but it was only as a
result of broader political shifts that women in association with groups represent-
ing the elderly became an electoral constituency. Their prior organization around
social policy enabled them to offer policy alternatives to which political parties
had to respond.
By contrast, in many liberalizing African states, political parties are not as well
established around issues and social programmes, but tend rather to be vehicles for
personalistic power and ethnic ambitions. Trade unions are relatively weak, exacer-
bated by the large informal and agricultural sectors. Policy alternatives are rarely
the basis on which voting takes place, even in a country with a relatively diverse
and long tradition of party mobilization such as South Africa. In these cases,
the elite bias of political competition is reinforced and women’s organizations, like
other sectors of civil society, may see few incentives in advancing their claims
through the party system. In post-socialist states, for different historical reasons,
both political parties and women’s organizations are not deeply institutionalized
and the relationship between gender equality and social policies is not a central
political issue.
As Hassim’s chapter shows, the inclusion of women in the formal institutions of
the state, and references to ‘gender equality’ in policy documents, do not neces-
sarily lead to the redistribution of resources and power in ways that change the
structural basis of gender inequality. Inclusion can be an avenue for reinforcing
elite women’s access to the formal political system while not translating clearly
into policies that address the needs of larger constituencies of women. Although
many new women members of parliament have taken ‘poor women’ as their
constituency (at least in moral terms), it has been all too easy for them to focus on
anti-poverty programmes as the only form of pro-women intervention (Hassim
2006). Yet, as we have argued above, a focus on the ‘poorest of the poor’ is an
inadequate strategy for advancing gender equality as it does not build a sustain-
able system of social protection that takes account of women’s particular gender
interests, which can sometimes cut across social class. As Tsikata points out, we
need to be careful that the emphasis on the basic needs of poor women and their
children does not displace issues of gender relations and power (Tsikata 2000: 6).
On the other hand, without political rights and access to the public sphere,
women cannot even enter debates about social policy. Moghadam’s chapter
underscores the argument that economic and political rights are important
prerequisites for gender equality. ‘Neopatriarchy’ offers women few avenues for
political intervention and democratization of the formal institutions of state and
policy-making is self-evidently necessary as a key step in advancing democratic
social policy. Unlike South Africa, transitions from state socialism did not open
spaces for feminist interventions. Rather closer to the experiences of the Middle
East, post-socialist countries ‘retraditionalized’ the family, upholding the distinc-
tion between the public and the private spheres (Haney and Pollard 2003). The
family was seen as ‘a site of refuge amidst chaos and unpredictability … and served
as a model and metaphor for transition’ (p.7). In Poland, even the several brisk
changes from left-to right-wing governments during the 1990s did not result in
progressive policies for women, because of the dominant social conservatism
(Steinhilber, this volume). Women were poorly represented in government and
the bureaucracy and there were few openings for feminists to make policy inter-
ventions. However, in response to the scaling back of benefits women’s organiza-
tions are beginning to make social policy reform part of their political activism.
Similarly in countries such as Iran the family is idealized as the antithesis to
Western individualism, making feminist struggles for more egalitarian social
policies that recognize women’s rights extremely difficult to pursue.
Concluding remarks
We have noted that historically state social provisioning and protection was
premised on a normative male-breadwinner/female-carer model, even if in reality
many women were in the labour-force. That model has eroded over the past two
decades, with the global increase in women’s labour-force participation. However,
despite the convergence in men’s and women’s crude economic activity rates,
gender segmentation is not disappearing from the world of work – not even
in Sweden where men continue to work full time and invest in their careers,
while women work ‘part time’ in the public sector, where it is easier to combine
employment with having a family. The link between paid work and entitlements
to social benefits has been further weakened by processes of labour informalization
or casualization, which have been part and parcel of liberalization policies over the
past two decades. A greater proportion of women’s paid work (compared to men’s)
tends to be of the informal kind, and within the informal economy women tend
to dominate the more casual and less remunerative niches.
Labour market segmentation and casualization mean that increasing numbers
of people, particularly women, are likely to be excluded from access to social
services and income supports if these are provided on a commercial basis or on the
basis of labour ‘contributions’ as in the social insurance model – ironically the two
paths that have dominated social sector restructuring in the 1980s and 1990s. In
theory it is possible to extend the coverage of social insurance programmes to
include informal workers – as some countries in East Asia and Latin America have
attempted – but only if their contributions are heavily subsidised (by the state). An
effective means for reducing gender-based poverty and inequality would be public
provision of accessible and accountable social services (especially health and
education) as well as citizenship-based entitlements to basic income support
(pensions and family/child allowances).
Despite significant differences played by families and households in social
protection and provisioning (these being the ultimate safety nets in many poorer
developing countries), it is nevertheless intriguing that the provision of unpaid

care remains so feminized everywhere.9 This is a major factor feeding into
women’s disadvantages in the market economy. But it is also important to under-
line that care is central to human flourishing and to social and economic devel-
opment. However it is also an area that remains marginal to the concerns of
mainstream policy actors across ideological and political divides. Only under
exceptional circumstances is any explicit attention paid to women’s unpaid care
work. Otherwise policy actors either assume that families and communities (i.e.
women within those units) will continue to provide care, or as in the case of
conservative and religious-based movements and policies, reinforce women’s car-
ing duties as part and parcel of their ideological crusades to restrict women’s
choices.
But there are also serious policy questions about what to do about care: how in
particular to reconcile the needs and rights of those who require care with the
needs and rights of those who provide care (whether paid or unpaid), and how to
foster responsibility for sharing care between men and women? Provision of acces-
sible, affordable and high-quality care services (for children, the elderly and others
with intense needs) is of course a sine qua non. But it is impossible to fully
defamilialize care, as Lewis and Giullari emphatically argue. The bulk of unpaid
care work is carried out, largely by women and girls, within families, households,
and community and kinship networks. These are the institutional arenas that are
also likely to reproduce gender disadvantage. States often do mould certain kinds
of families and gender arrangements through legal instruments, often emphasiz-
ing women’s domesticity and restricting their economic citizenship. While it is
difficult for states to oblige men to share equally in care work, they can provide
incentives for them to do so. In the Nordic countries, for example, while ‘daddy
leave’ quotas may have had a small impact on how men and women divide their
unpaid care responsibilities, they nevertheless have enormous symbolic value. The
same countries have been less forthcoming in reforming the male employment
model, by reducing working time (a shorter working week and the regulation of
overtime, for example).
In the case of many developing countries, public or private care services for
children and the elderly remain underdeveloped and the great bulk of care is
assumed by women and girls as members of families and households, while some
of it (in middle-class households) is purchased through informal arrangements. In
these contexts, a priority must be the strengthening of states’ capabilities to
provide basic infrastructure (piped water, roads, electricity) and accessible and
affordable public health services, which are likely to reduce women’s unpaid work-
loads. The sharing of unpaid care work between women and men would require
different strategies from those used in advanced welfare states, given that the bulk
of paid work is unregulated in many of these countries.
The state is a key institution as an organizer if not necessarily a provider of social
protection and provisioning. It is clear that states that are well-institutionalized
are better able to translate political commitments into effective social policies and
delivery systems. Women thus have an interest in making states more responsive
and accountable to their citizens. Neoliberal approaches to state reform in

developing countries have, however, tended to undermine the capacity of states to
be responsive to the needs of their citizens. The renewed interest in the state (in
the ‘good governance’ paradigm) offers some opportunities for the creation of
gender-responsive states. But this would require that more attention be paid to
developing political accountability to citizens and that women be seen as part
of the ‘publics’ that need to be responded to and served.
Related to the point above, this volume suggests that we need ‘thicker’ under-
standings of democracy that go beyond supporting multi-partyism and the
numerical increase of women in national parliaments. Both of these are important
prerequisites for reducing inequalities, but they need to be buttressed by deeper
levels of political participation. This would include developing the capacity of
women’s organizations and civil societies in general to interpret and articulate the
needs of different constituencies of women in policy terms. It would also include
more strategic use of political parties as vehicles of representation by pushing for
social policies to become electoral issues.
Last, but not least, women have fought for the state to recognize their needs in
various ways (including maternalist demand-making) but not always in ways that
challenge the underlying power relations of gender. In some countries the absence
of strong feminist lobbies, or allies within political parties, has allowed the adop-
tion of a residualist welfare model that has seriously undermined women’s social
rights. Difficulties in clearly articulating women’s needs in social policy terms
seems to be as much a problem in the South as it is in the North.
Acknowledgements
We thank Alexandra Efthymiades for excellent research assistance, James Heintz for the
preparation of tables, and Francie Lund and Silke Steinhilber for extensive comments on an
earlier draft of this chapter.
Notes
1 One response has been the ‘good governance’ agenda promoted by the World Bank,
which rehabilitates the state by emphasizing the role of ‘lean’ state bureaucracies and
judiciaries in creating the conditions for market competition through the enforcement of
private property rights and contracts, and by ‘regulating’ private industry and commer-
cial social services (World Bank 1997, 2003; Fukuyama 2004).
2 The welfare state broadly defined not only includes transfer payments but also the public
funding and delivery of goods and services, such as education and health. However, the
quantitative literature on the development of the welfare state contains a one-sided focus
on transfer payments (and neglect of social services), which is related to the other major
lacunae in this research – the role of gender (Huber and Stephens 2000a).
3 The other categories were ‘strong male breadwinner states’ where Ireland and Britain were
placed, and ‘modified male breadwinner countries’ which included France (Lewis 1992).
4 Though long part-time, 20 to 30 hours per week.
5 ‘Informal sector’ covers informal enterprises. Informal employment is a broader category
that includes all workers who work without secure contracts, worker benefits or social
protection; so it can include those who work as informal wage workers for formal
enterprises or households (e.g. casual labourer, domestic worker) (ILO 2002).
6 ILO, UNRISD, UN/DESA, UNESCO, UNDP (and even some parts of the World Bank) are
among the organizations often mentioned as those promoting this global discursive shift
(see GASPP 2005).
7 There are some similar gender concerns about the burden and division of care work in the
more affluent countries in the context of ageing (Stark 2005).
8 What the Islamist state has attempted to do however should not be mistaken for social
reality. Processes of social change, in which women have been major change agents, have
subverted many of the Islamization measures promulgated by the state.
9 For developing countries, however, we need more systematic analysis of unpaid work
than is currently available through the existing time use surveys.
References
Anker, R., H. Melkas and A. Korten (2003) Gender-Based Occupational Segregation in the 1990s.
ILO Working Paper No. 16. Focus Programme on Promoting the Declaration on
Fundamental Principles and Rights at Work. Geneva: ILO.
Bates, R. (1994) ‘The Impulse to Reform in Africa’. In J. A. Widner (ed.) Economic Change and
Political Liberalization in Sub-saharan Africa. Baltimore: Johns Hopkins University Press.
Blondet, C. (2002) ‘ “The Devil’s Deal”: Women’s Political Participation and
Authoritarianism in Peru’. In M. Molyneux and S. Razavi (eds) Gender Justice, Development
and Rights. Oxford: Oxford University Press.
Bryceson, D. F. (1999) African Rural Labour, Income Diversification and Livelihood Approaches:
A Long-Term Development Perspective. ASC Working Paper No. 35. Leiden: African Studies
Centre.
Cho, Hyoung, Ann Zammit, Jinjoo Chung and In-Soon Kang (2004) ‘Korea’s Miracle and
Crisis: What Was in It for Women?’ In Shahra Razavi, Ruth Pearson and Caroline Danloy
(eds) Globalization, Export-Oriented Employment, and Social Policy: Gendered Connections.
Basingstoke: Palgrave.
Devereux, Stephen and R. Sabates-Wheeler (2004) Transformative Social Protection. Working
Paper No. 232, Brighton: IDS.
Elson, Diane (2004) ‘Social Policy and Macroeconomic Performance: Integrating “the
Economic” and “the Social” ’. In T. Mkandawire (ed.) Social Policy in a Development Context.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Princeton: NJ Princeton
University Press.
Filgueira, F. (forthcoming) ‘The Latin American Social States: Critical Junctures and Critical
Choices’. In Y. Bangura (ed.) Democracy and Social Policy. Basingstoke: Palgrave.
Filgueira, F. and Filgueira, C. (2002) ‘Models of Welfare and Models of Capitalism: The Limits
of Transferability’. In E. Huber (ed.) Models of Capitalism: Lessons for Latin America.
University Park PA: Pennsylvania University Press.
Fraser, N. and L. Gordon (1994) ‘Civil Citizenship Against Social Citizenship? On the
Ideology of Contract-Versus-Charity’. In van Steenbergen (ed.) The Condition of Citizenship.
London: Sage.
Fukuyama, F. (2004) ‘The Imperative of State-Building’. Journal of Democracy, Vol. 15, No. 2,
pp.17–31.
GASPP (Globalism and Social Policy Programme) (2005) Copenhagen Social Summit Ten Years
On: The Need for Effective Social Policies Nationally, Regionally and Globally. Policy Brief No 6,
Helsinki: GASPP/STAKES.
Grindle, M. (1986) State and Countryside. Baltimore: Johns Hopkins Press.
Haggard, S. and R. R. Kaufman (1995) The Political Economy of Democratic Transitions.
Princeton: NJ Princeton University Press.
Haney, L. (2003) ‘Welfare Reform with a Familial Face’. In L. Haney and L. Pollard (eds)
Families of a New World: Gender, Politics, and State Development in a Global Context. New
York: Routledge.
Haney, L. and L. Pollard (2003) ‘In a Family Way: Theorizing State and Familial Relations’. In
L. Haney and L. Pollard (eds) Families of a New World: Gender, Politics, and State Development
in a Global Context. New York: Routledge.
Hassim, S. (2006) ‘Women in African Parliaments: The Virtuous Circle of Representation’.
In G. Bauer and H. Britton (eds) Women in African Parliaments. Boulder: Lynne Rienner.
Heintz, J. (2005) The Growing Informality of Labour Markets. Paper presented at the
UNRISD/CGGS Conference on Gender and Social Policy, Marstrand, Sweden, 28–29 May.
Hobson, B. (1990) ‘No Exit No Voice – Women’s Economic Dependency and the Welfare
State’. Acta Sociologia 33, 235–250.
Holzmann, R. and S. Jorgensen (2000) Social Risk Management: A New Conceptual Framework
for Social Protection and Beyond. Social Protection Discussion Paper Series, No. 6,
Washington DC: World Bank.
Huber, E. and J. D. Stephens (2000a) ‘Partisan Governance, Women’s Employment, and the
Social Democratic Service State’. American Sociological Review, Vol. 65, No. 3, 323–42.
Huber, E. and J. D. Stephens (2000b) The Political Economy of Pension Reform in Latin America.
OPG No. 7, Geneva: UNRISD.
Huber, E. (2000) Social Policy in a Development Context: Notes on Social Security and Pensions
Systems. Presentation at the UNRISD Conference on Social Policy in a Development
Context, Stockholm, 23–24 September.
International Labour Organization (ILO) (2002) Women and Men in the Informal Economy:
A Statistical Picture. Geneva: ILO.
Korpi, W. and J. Palme (1998) ‘The Paradox of Redistribution and Strategies of Equality:
Welfare State Institutions, Inequality, and Poverty in the Western Countries’. American
Sociological Review, Vol. 63, 661–87.
Koven, S. and S. Michel (1993) ‘Introduction: Motherworlds’. In S. Koven and S. Michel (eds)
Mothers of a New World: Maternalist Politics and the Origins of Welfare States. New York:
Routledge.
Lee, C.K. (2005) Livelihood Struggles and Market Reform: (Un)making Chinese Labour After State
Socialism. OP No. 2, Geneva: UNRISD.
Lewis, Jane (1992) ‘Gender and the Development of Welfare Regimes’. Journal of European
Social Policy, Vol. 2, No. 3, 159–73.
Mackintosh, M. and M. Koivusalo (2005). ‘Health Systems and Commercialisation: In Search
of Good Sense’. In M. Mackintosh and M. Koivusalo (eds) Commercialization of Health Care:
Global and Local Dynamics and Policy Responses. Basingstoke, Palgrave.
Minogue, Martin (2002) ‘Power to the People? Good Governance and the Reshaping of the
State’. In U. Kothari, and M. Minogue (eds) Development Theory and Practice: Critical
Perspectives. Basingstoke: Palgrave.
Mhone, G. (2004) ‘Historical Trajectories of Social Policy in Post-Colonial Africa: The Case of
Zambia’. In T. Mkandawire (ed.) Social Policy in a Development Context. Basingstoke: Palgrave.
Midgley, J. (1995) Social Development: The Developmental Perspective in Social Welfare. London:
Sage.
Mkandawire, T. (2005) Targeting and Universalism in Poverty Reduction. Programme Paper
No. 23, Geneva: UNRISD.
Mkandawire, T. (2004) ‘Social Policy in a Development Context: Introduction’. In
T. Mkandawire (ed.) Social Policy in a Development Context. Basingstoke: Palgrave.
Molyneux, M. (forthcoming) Poverty Relief and the New Social Policy in Latin America: Mothers
at the Service of the State? Programme Paper, Geneva: UNRISD.
Orloff, A. S. (1993) ‘Gender and the Social Rights of Citizenship: The Comparative Analysis
of Gender Relations and Welfare States’. American Sociological Review, No. 58, 303–28.
Orloff, A. S. (2002) Women’s Employment and Welfare Regimes: Globalization, Export Orientation
and Social Policy in Europe and North America. SPD No.12, Geneva: UNRISD.
Phillips, A. (2001) ‘Feminism and Liberalism Revisited: Has Martha Nussbaum Got It Right?’
Constellations 8, No. 2, 249–66.
Polanyi, K. (1957) The Great Transformation. Boston: Beacon Press.

Razavi, S. (1999) ‘Gender, Poverty and Well-being: Introduction’. Development and Change,
Vol. 30, No. 3, 1–27.
Rothschild, D. (1994) ‘Structuring State-Society Relations in Africa: Toward an Enabling
Political Environment’. In J. A. Widner (ed.) Economic Change and Political Liberalization in
Sub-saharan Africa. Baltimore: Johns Hopkins University Press.
Rueschemeyer, D., E. Huber and J. D. Stephens (1992) Capitalist Development and Democracy.
Chicago: Universty of Chicago Press.
Ruggie, J. G. (2003) ‘Taking Embedded Liberalism Global: The Corporate Connection’. In
David Held and Mathias Koenig-Archibugi (eds) Taming Globalization: Frontiers of
Governance. Cambridge: Polity.
Sainsbury, D. (1996) Gender Equality and Welfare States. Cambridge: Cambridge University Press.
Stark, A. (2005) ‘Warm Hands in Cold Age – On the Need of a New World Order of Care’.
Feminist Economics, Vol. 11, No. 2, 7–36.
Tendler, J. (2004) ‘Why Social Policy is Condemned to a Residual Category of Safety Nets and
What to Do About it’. In T. Mkandawire (ed.) Social Policy in a Development Context.
Tibandebage, P. (2004) Health Sector Reforms and Gender Implications: A Case Study of Mutual
Health Insurance in Tanzania. Mimeo, Geneva: UNRISD.
Tokman, V. E. (2002) ‘Jobs and Solidarity: Challenges for Labour Market Policy in Latin
America’. In E. Huber (ed.) Models of Capitalism: Lessons for Latin America. University Park
PA: Pennsylvania State University Press.
Tripp, A.M. (2002) ‘The Politics of Women’s Rights and Cultural Diversity in Uganda’. In
M. Molyneux and S. Razavi (eds) Gender Justice, Development and Rights. Oxford: Oxford
University Press.
Tsikata, D. (2000) ‘Lip Service and Peanuts: The State and National Machinery for Women in
Africa’. Accra: Third World Network.
Tsikata, D. (2004) Economic Liberalization, the Informalization of Work and Urban Women’s
Livelihoods in Sub-Saharan Africa since the 1990s. Background paper for Gender Equality:
Striving for Justice in an Unequal World. Geneva: UNRISD.
UN Millennium Project Task Force on Gender Equality (2005) Taking Action: Achieving Gender
Equality and Empowering Women. London: Earthscan.
United Nations Research Institute for Social Development (UNRISD) (2005) Gender Equality:
Striving for Justice in an Unequal World. Geneva: UNRISD.
Vivian, J. (1995) ‘How Safe are “Social Safety Nets”? Adjustment and Social Sector
Restructuring in Developing Countries’. In J. Vivian (ed.) Adjustment and Social Sector
Restructuring. London: UNRISD/Frank Cass.
White, G. (1996) ‘Civil Society, Democratization and Development’. In R. Luckham and
G. White (eds) Democratization in the South: The Jagged Wave. Manchester: Manchester
University Press.
Whitehead, A. (2004) The Gendered Impacts of Liberalization Policies on African Agricultural
Economies and Rural Livelihoods. Background paper for Gender Equality: Striving for Justice in
an Unequal World, Geneva: UNRISD.
Whitehead, A. and M. Lockwood (1999) ‘Gendering Poverty: A Review of Six World Bank
African Poverty Assessments’. Development and Change, Vol. 30, No. 3.
Whitehead, A. and N. Kabeer (2001) Living with Uncertainty: Gender, Livelihoods and Pro-poor
Growth in Rural Sub-Saharan Africa. Working Paper 134, Brighton: IDS.
World Bank (1997) World Development Report 1997: The State in a Changing World.
World Bank (2000) World Development Report 2000/1: Attacking Poverty. Washington DC:
World Bank.
World Bank (2003) World Development Report 2004: Making Services Work for the Poor.
New York: Oxford University Press.
Part I
Historical and Regional Trajectories
in Social Provisioning
2
Mothers at the Service of the New
Poverty Agenda: The PROGRESA/
Oportunidades Programme in Mexico1
Maxine Molyneux
In Latin America as elsewhere in the world, gender bias and masculine prerogative
have prevailed in social policy as in social life more broadly, with entitlements
resting on culturally sanctioned and deeply rooted notions of gender difference
and patriarchal authority. These have generally accorded with idealized assump-
tions about the asymmetric social positions occupied by the sexes with male
breadwinners and female mother-dependents receiving benefits according to
these normative social roles. Such assumptions have proved remarkably universal
and enduring even where, as in Latin America, gender divisions have been
modified by women’s mass entry into the labour force and by equal rights
legislation.
This chapter considers the changes and continuities in social policy provision in
Latin America through a focus on the ways that women, in particular mothers, are
positioned within the new anti-poverty programmes that have followed structural
reform. It examines a flagship anti-poverty programme known as Oportunidades
(Opportunities) established in Mexico at the end of the last decade.2 Seen by some
commentators as a quintessentially neoliberal programme, and embodying many of
the main ideas of the ‘New Poverty Agenda’,3 Oportunidades represents a novel com-
bination of earlier social policy approaches with the contractarian, co-responsibility
models associated with new approaches to social welfare and poverty relief. This
chapter is in two main parts; the first provides the background context for the
emergence of the new approaches to poverty; the second describes and critically
examines the Mexican programme’s selective construction of social need.
Social policy in Latin America prior to the reforms
In Latin America, low tax revenues and weak commitments to redistributive

policies ruled out the development of effective, universal welfare systems. Only
five countries, Argentina, Uruguay, Costa Rica, Chile and Cuba4 developed a form
of welfare state and, with the exception of the latter, none achieved universality of
entitlement or coverage.5 Nonetheless, from the late nineteenth century, if to
43
44 Maxine Molyneux
widely different degrees, some forms of social provision began to evolve. These
were principally concentrated on the education and health sectors and, where
Bismarckian models were influential, as in Mexico and Chile, state pension
schemes, along with other forms of social insurance for privileged (predominantly
masculine) sectors of the labour and armed forces, accompanied the process of
state formation.
From the first decade of the twentieth century social rights increased as a result
of successful demands by organized labour and socialist parties for social reform,
with an incremental assumption of social responsibility by governments. In the
1920s and 1930s ‘improving the “race” ’ in order to secure the conditions for
development and head off threats of disorder became the leitmotif of the social
reform and eugenics movements. Many women were among the promoters of
‘social hygiene’ and its derivative, the science of puericultura (child development).
They energetically supported policy and legal changes which were maternalist in
orientation, demanding benefits and services for mothers and children. Mothers
were among the first to be recognized as social policy claimants whether as mar-
ried women or as ‘unfortunates’, that is, impoverished single mothers. However, it
was often stated in the discussion of these provisions that it was primarily in the
interest of their children that women might receive benefits of a financial, educa-
tional or medical kind. In other words it was in the construction of children’s
needs that their mothers received entitlements, that is, in order to better fulfil
their maternal responsibilities.
The era of nationalist state-centred development under corporatist populism
was inaugurated by the crisis of 1929 but was more securely established in the post-
war period. It brought some expansion in entitlements, notably for organized
labour, the natural constituency of corporatist regimes and a relatively privileged
sector for long afterwards. Social rights correspondingly expanded in Mexico,
Chile, Argentina and Brazil among others, and even as populist corporatism
waned, the technocratic developmentalism that replaced it continued to expand
the social sector.
By the end of the 1960s all but the poorest states had established the main
planks of social welfare, if at times in skeletal form. Health and education were
publicly funded, and social insurance systems covered some categories of formal-
sector workers. Regional policies were now influenced by the ECLAC (United
Nations Economic Commission for Latin America and the Caribbean) or CEPAL-
ISTA guidelines which drew on human capital theory to anchor social policy more
firmly in a discourse of development priorities, as Latin American states presided
over a rapid expansion of literacy programmes and primary education. At this
time too, the ‘basic needs’ approach was gaining support, leading to some greater
attention to ‘subsistence rights’6 through the provision of food to the poor, sani-
tary works, potable water and affordable housing. Positive growth rates, rapid
urbanization and social mobilization, all caused Latin American states, irrespective
of political inclination, to embark on programmes to meet rising social demands
and expectations. These decades saw Latin America leading the developing coun-
tries in terms of social expenditure and social coverage. There was a corresponding
Mothers at the Service of the New Poverty Agenda 45
improvement in human development indicators as life expectancy steadily

increased and infant mortality declined, to place Latin America by 1980 at the top
of the developing regions (Filgueira and Filgueira 2002).
While CEPALISTA developmentalism was associated with universalist principles
and despite the expansion of social provision from the 1960s, most of the region
still suffered from poor and skewed coverage and low-quality services. Social
policy provision in Latin America remained unevenly distributed between the
richer and poorer states and within these, between rural and urban populations, as
well as across sectors.7 The state sector was often underfunded, but was all too
often also blighted by poor administration with governments reactive to problems
as they arose, and susceptible to interest groups’ political demand-making.
Entitlements remained for the most part tied to formal employment with pensions
available only to a minority of workers, with some insurance schemes for disabil-
ity, unemployment and maternity. These arrangements did not cover the rural
sector or the large proportion (sometimes as much as 40 per cent of the active pop-
ulation) which was in the informal sector or in domestic service, often the largest
employer of urban women. In 1980 some 130 million people or 33 per cent of the
total population of Latin America lived below the UN-defined poverty line. Since
many of the poor were rural workers, or employed in the informal sector, they
were unable to qualify for social insurance systems.
In the lower-income countries such as Guatemala, Nicaragua, Bolivia, Ecuador
and Peru, and in some of the better-off states such as Brazil and Mexico, the situa-
tion was one of widespread poverty, sharp regional and ethnic inequalities, poor
state provision, a weak entitlement framework, minimal and far from efficient
safety nets. Organizational and managerial deficiencies added to the problem in
many countries, exemplified in multiple and overlapping welfare institutions,
legal complexity, statistical inadequacies, and lack of coordination between
departments (Mesa-Lago 1994: 76). For those living in and on the margins of
poverty, that is, up to half the population in half of the countries of the region in
the 1990s, the principal safety net was emergency relief (food aid, primary health),
family and kinship support, supplemented by the voluntary sector, comprising
NGOs, church-based relief and charitable organizations.
Gender and social policy

Social policy in Latin America, contrary to some readings, was not gender blind
but instead worked with gendered conceptions of social needs, ones which were
familial, patriarchal and paternalistic. While women gained access to education
and health, and entered the workforce, by broad consensus their primary duties
lay within the family. Liberal citizenship might extend to women in the public
realm, but in the private domain, a different order prevailed. Built into the earliest
forms of social provision were assumptions of female dependency on a male
breadwinner which positioned women as under the protection of ‘their’ men,
whether husbands of fathers. Motherhood, too, was assumed to be a natural and
inevitable destiny of women, one that brought them both rights and duties. Thus,
in Latin America, women enjoyed some universal rights, but other rights and
46 Maxine Molyneux
entitlements remained closely bound to the family, and were accessed by virtue of
their status as wives and mothers. If, in the public realm of work, women employees
gained recognition as mothers and were granted comparatively generous mater-
nity leaves, in the private realm, on marriage, they lost certain citizens’ rights.
Instead, like children, they gained a measure of protection, as those of dependent
status. Widows of soldiers, professionals and some formal-sector workers were able
on this basis to claim their deceased husbands’ pensions. Yet most working women
were located in low-paid jobs, in unorganized sectors of employment, and in work
that was considered supplementary to the male wage and lacked social protection.
In Mexico, the cataclysmic upheavals of the revolution did not greatly alter this
general picture. The constitution of 1917 and the civil code of 1928 accorded
women legal equality and gave them some new rights, such as the right to
divorce, but denied them full civil rights and political rights; universal female fran-
chise was not granted until 1953, later than in most other Latin American states.
Over time the law placed limits on men’s authority over their wives, but it was
only in 1974 that the code making women responsible for the domestic sphere,
and abandoning the home a specifically female offence, was repealed (Varley
2000).
As elsewhere in Latin America, Mexican women workers and women’s organiza-
tions pressed for the regulation of their working hours, in order to protect them
from over-exploitation. They gained popular support for these demands in a
context where there were widely expressed concerns that this was necessary to
safeguard their ‘maternal functions’.8 Paternalist sentiments were aroused by such
claims, with women and children positioned in this discourse as requiring protection;
many women protested that this was an excuse to deny them the right to equal
work and well-paid jobs. Mexico’s ‘perfect dictatorship’, a blend of one-party rule
and corporatism, combined revolutionary rhetoric, radical reform measures and
social conservatism. Catholicism retained a hold over the population and, despite
the secular disposition of the early revolutionaries, exerted a continuing influence
over social policy and women’s rights. In other ways too, women lost out in the
process of post-revolutionary state formation. Corporatism established a political
bond between the worker and the ruling party ensuring the loyalty of the more
powerful sectors of organized labour. Male-dominated trade unions were the prin-
cipal beneficiaries of corporatist social contracts that enrolled men in the service
of the state as workers and patriots, their compliance secured through negotiated
pacts over wages, working conditions and social security (Rosemblatt 2000).
Where women had acquired a significant presence in the workforce, they laboured
in poorly paid, less-organized sectors. Not only were they marginal to the contrac-
tual negotiations of the corporatist state, but they also occupied an ambiguous
place in wage negotiations since their very presence in the workforce could com-
promise the historic demand of organized labour for a ‘family wage’, one premised
on female dependency and the presence in the family of the full-time housewife
and mother.
Corporatist bargaining did however secure the passage of some welfare
measures, but what evolved was a selective, predominantly urban system of
welfare, with rural and indigenous women largely excluded from these and other
entitlements. Despite the agrarian reform’s revolutionary momentum, women
were more often than not denied rights to property in land, which through law or
custom remained a male prerogative, thus revealing the gender bias as well as the
limits of the ruling Party’s inclusive pretensions. If women gained entitlements as
workers, these too were not only restricted to a small section of the female
population, but were often unclaimable in practice.9 What little social protection
was available to women was more likely to be accessed through marriage and
family law which specified that it was a husband’s duty to provide for his wife and
children, and afforded women conjugal property rights.
While there occurred some limited individuation of women’s rights from the family
as a result of reforms spearheaded by feminist movements, these general features of
women’s social rights endured in Latin America. The restricted reach and scope of
social policy, the poor quality and difficulty of accessing many of the services and
benefits, meant that most low-income women could not and did not look to the state
for much in the way of support on their own behalf. They might be fortunate enough
to attain some minimum provision in education and health, and some support for
their children, but individual entitlements such as income support and pensions
were distant dreams for the majority. Security, such as it was, came from paid work
where it could be found, from marriage, kin and community, and from the church.
Reforming the social sector
The fragility and inefficiency of the social security systems prevailing across much
of Latin America were features that were sharply accentuated by the broader socio-
economic trends that set in from the mid-1970s. The oil shocks, the debt crisis and
subsequent recession of the 1980s combined with demographic pressures – and in
much of the region with political conflict – to erode the social sector at precisely
the moment when its expansion was most needed. This was a period which saw
more women entering the labour force, while households sought to cut consump-
tion, substituting market-purchased goods and services with reproductive labour,
largely provided by women (González de la Rocha 1994).
The human and social costs of the first phase of the structural reforms helped to
revive longstanding debates over social sector reform which now took place both
within the region and in international development policy arenas. Cornia et al.’s
UNICEF study, Adjustment with a Human Face (1987) is widely acknowledged as a
‘wake up call’ to international agencies to pay attention to the social costs of
adjustment, but it was also important for the policy recommendations that it
made. While presaging future policies, these recommendations also confirmed
trends that were already under way; these included more targeting as a means to
enhance the redistributive role of the state, employment-creating labour-intensive
public work schemes, subsidies of certain items, and, crucially, for the scope of
social policy to be widened during times of adjustment.
These ideas were already achieving wide acceptance in development institutions,
but would take time, political will, and resources to implement across the region.
48 Maxine Molyneux
Poverty was beginning to be acknowledged as a more enduring phenomenon than

initially supposed. If, as many argued, it was not transitional but a structural effect
of the New Economic Model, it would require closer attention to social assistance,
with anti-poverty measures and state programmes that were more extensive in
scope than Emergency Social Funds had allowed. By the end of the 1980s many
governments and international development agencies accepted that there was an
urgent need to address the ‘social deficit’ if the neoliberal reforms were not to be
violently rejected by the populations that had suffered so harshly from them. There
was also broad agreement that if the social sector was to be more efficient it had to
be brought into closer alignment with the market and with trends elsewhere
towards pluralizing service delivery. A new approach to social policy, in which
poverty was a crucial element, was evolving.
The central features of the new social policies as they evolved from the mid-1970s
conform to general descriptions of neoliberal policy assumptions, containing as they
do the familiar elements of targeting, privatization, multi-tierism and pluralization
of service providers, along with a greater reliance on the market for poverty relief,
most evidently in micro-credit programmes. But the way these elements were
interpreted and incorporated into social policy reflected not only broader shifts
within international development policy but also specific regional developments.
‘Neoliberalism’ in Latin America passed through at least two main phases. The first,
from the 1970s, generally seen as the period of ‘high neoliberalism’, coincided with
the policies of structural adjustment and stabilization adopted in the debt crisis.
These were the policies that were dubbed ‘market fundamentalism’ when govern-
ments, often under conditionalities imposed by the international financial institu-
tions, sharply decreased the realm of state action through privatization, imposed
tight fiscal controls, liberalized capital accounts, and opened up their economies.
Since then, there has been some revision of these policies, with results summed
up as ‘the post-Washington consensus’, sometimes hailed as ‘the end of neoliber-
alism’.10 The original adjustment package was modified in three ways that concern
us here: the state was partially rehabilitated in development policy and planning,
its role described as ‘facilitator’ by the World Bank and its efficiency to be
enhanced through good governance reforms; there was a clear recognition by the
international financial institutions and by Latin American governments that the
social deficit had to be addressed, so social policy was returned to the regional
agenda; and poverty relief became a central component of social policy. If in the
1980s policy attention focused on ‘getting the economy right’, in the 1990s there
were attempts to attend to the hitherto neglected social realm and to build appro-
priate institutions. Significant though these policy shifts might be, they did not
greatly alter the broader outline of macroeconomic policy; governments remained
committed to fiscal discipline and market-led growth.
The new social policy materializes
As Brock et al. (2001) argue, the evolution of development policy is not usefully
seen as following a uni-linear dynamic; the social policies and development
rhetorics that accompanied the post-adjustment phase are better described as

‘hybridized’, and seen as the result of a complex dynamic of power and agency,
involving a wider range of actors, interest groups and discourse coalitions than
top-down accounts usually allow. This is particularly true of social policy which is
less subject to influence by IFIs than some other areas of policy, as it has to be
approved by parliaments as well as by interest groups such as trade unions.
In Latin America by the early 1990s, reforms and policies affecting the social
sector were being developed, some of which had first been applied experimentally
from the 1970s, some of which were new. In the latter category were more decen-
tralized health and education provision, and the privatization of pensions; in the
former, a greater emphasis on participatory mechanisms in the delivery of social
welfare. All were intended to increase efficiency, accountability and quality
(Grindle 2000). As poverty moved up the scale of international priorities, 1990 saw
the launch of the World Bank’s New Poverty Agenda and towards the end of the
decade the Millennium Development Goals committed governments to halving
extreme poverty and hunger between 1990 and 2015.
Development policy analysts acknowledge that the ideas being advanced at this
time marked a significant departure from the structural adjustment objectives
pursued during the first phase of the reforms (Lipton and Maxwell 1992: 1).
Among the main changes were the importance of civil society, along with an
emphasis on certain concepts that became central to the New Poverty Agenda,
ones that were not novel in themselves but perhaps were so in combination. These
were the principles of participation, empowerment and co-responsibility.
Participation is the least novel of the three but it moved from the margins of
development practice in the 1970s to form part of mainstream thinking in the
1990s. Its combination of ethical (democratic) principle with efficiency arguments
gave it a wide appeal, seeming to trump bureaucratic centralism on both counts.
A wide range of participatory programmes were established, following the model
of demand-led assistance, first introduced in the Emergency Social Funds, which
were later renamed as Social Investment Funds and established on a more perma-
nent basis. Decentralization too, has proceeded in tandem with novel forms of
participatory policy fora. From a broader societal perspective, participation was
also argued to tackle the condition of social exclusion that frequently attends
poverty and deprivation and is seen as central to, if not synonymous with the
creation and maintenance of social capital, itself a policy concern of the 1990s.11
Empowerment, like participation with which it is linked (since participation is
one of the means to secure empowerment), moved into mainstream development
practice in the 1980s. Widely used in the practice of women’s organizations and by
NGOs, it has generally been understood as a process of transformation involving
both the acquisition of capabilities and changes in subjectivity that enable agency
to be exercised. Empowering the poor and the disadvantaged should result in their
gaining more voice and presence in decision-making arenas that affect their lives
and in developing the capabilities to enable them to escape poverty. In Latin
America support for such ‘empowerment approaches’ built on the critique of
traditional philanthropy that lived on in ‘assistentialist’ policies and which saw
50 Maxine Molyneux
‘beneficiaries’ as passive recipients of charity. The new policies instead situate

‘users’ as ‘stakeholders’ – with interests and responsibilities – who are ‘participants
in the policy process’. They are no longer ‘beneficiaries’ or ‘clients of the state’ but
empowered, active citizens capable of formulating their own needs and engaging
in the setting of priorities and the implementation of projects, whether commu-
nity development schemes, health and housing or micro-credit enterprises. As the
capabilities approach has gained wider acceptance in policy circles, empowerment
has come to mean that the poor are to be trained and educated to prepare them for
employment. Poverty ‘relief’ must not be considered a short-term palliative, but
through incorporating elements that enhance the capacities and choices of the
poor, they will be helped to develop the means to secure a route out of poverty.
A further dimension is added to the conceptual basis of the new approach by
‘social risk management’ as outlined in the 2000/1 World Development Report
Attacking Poverty wherein sustainable poverty alleviation entails measures to
increase the security of the poor, through developing their capacity to ‘cope, mit-
igate or reduce’ their risks. The risk-management approach has been adopted by a
wide range of multilateral lending institutions.
Third, and again, closely related to these previous concepts, is the principle of
beneficiary responsibility variously articulated in ideas of ‘co-management/
responsibility’ self-help or self-sufficiency, ideas that gained resonance in the
1980s when the state was identified as a major cause of development failure and
accused of nurturing a ‘dependency culture’. At the same time the World Bank,
concerned with cost-sharing and efficiency, formulated policies in which the
no-longer-passive recipients of state handouts became active participants in
meeting the costs of development. The growth of cost recovery, co-financing and
co-management schemes along with community participation and voluntary
work became a means to promote self-help in development and welfare projects.
As states moved towards targeted assistance programmes, attention focused on
how the poor could be encouraged to ‘help themselves’. This idea informed a
range of policies, from giving economic assistance (as in the case of micro-credit),
to providing basic education in nutrition and healthcare. These latter were
designed, as in the earlier ‘social hygiene’ movements of the 1920s and 1930s, to
‘modernize and civilize’ the poor, but also to equip them with the attitudinal
wherewithal to manage their own destinies, ‘free’ of state dependency but sub-
ordinated to the discipline of the market (Rose and Miller 1992).
Latin America’s new social policy
In Latin America the novel features of this ‘post-Washington consensus’ phase of

policy evolution lay in the specific regional interpretation of its key elements. This
was most evident in three areas: the changes in the locus and character of state
activities; the rise of parallel institutions to assist in the delivery of social welfare;
and the promotion of civil society partnership in development and poverty relief
programmes. How these elements combined with efforts to create a democratic
politics in post-authoritarian Latin America and resonated with historic demands
for reform are essential in understanding the ways in which social policy was
refashioned in the changed circumstances of the 1990s.
The Latin American experience of post-adjustment policy-making was marked
by an ad hoc accommodation to the precepts of the New Economic Model,
combining new and old policy elements, while at the same time generating some
original trends and initiatives that entered the global debate on poverty relief. In
practice policies range along a spectrum from social liberal variants with univer-
salist inclusionary principles to those based on targeted provision with a greater
role for privatized services. Brazil for example introduced a universal public
pension to which everyone over a certain age is entitled. Indeed the Brazilian
Constitution of 1988 was the first to combine proposals for an alliance between
state and civil society, with a commitment to universal social polices and raising
total public expenditure, along with new institutional structures, such as manage-
ment councils and public hearings, where state and civil society could ‘work
together to ensure that priority-setting matched the public and private interests
and secure accountability in the definition and delivery of social policies’ (Coelho
et al. 2002). In all of these variants, the Latin American region has seen the
establishment of a wide range of new agencies and institutional structures for
providing access to social services. Government-sponsored poverty programmes
have been established to complement the continuing work of Social Investment
Funds, and consultative processes and institutions have been put in place at local
and national levels. As elsewhere, the official policy discourses and forms of enti-
tlement that are being created in Latin America tend to place more emphasis on
individual responsibility, while social security is defined in official statements as
no longer residing solely with the state. It now involves the ‘co-management of
risk’: that is, the individual has to make responsible provision against risks
(through education and employment); the family too, must play its part (through
better care); while the market (through private interests) and the community
(through decentralization, ‘co-responsibility’ and the voluntary sector) are all
involved in the decentring of expectations of welfare from the state.
The specificity of the Latin American region not only stamped its mark on how
these ideas would materialize in policy but also how they would be received by
citizens. In a context of widespread distrust of the state and weak social protection,
the refiguring of state–society relations offered by the NSP received a mixed
response, not by any means all negative. The core ideas at least seemed to offer
some potential for advancing much needed reforms, if social and political condi-
tions allowed. Decentralization, ‘good governance’, accountability, participation
and urgent attention to poverty, resonated with the reform agendas of democratic
parties, movements, and civil society organizations that were working to democ-
ratize politics and society following years of military rule.12 From the 1980s calls to
‘deepen’ democracy, and to address the ‘social deficit’ of the adjustment years
converged with some of the Good Governance and state reform agendas. The
human rights movement in the 1990s was enjoying a particularly prominent
international role, and this impacted on Latin America at a time of considerable
receptivity to the new inclusions of women’s and children’s rights and indigenous
52 Maxine Molyneux
claims for recognition and justice (Molyneux and Lazar 2003). Women’s organiza-
tions of various kinds were particularly active in promoting women’s rights,
working simultaneously within communities and at state level to advance reforms
in the areas of violence against women, legal and political representation and
reproductive rights. They also helped to establish and sustain popular health
movements, leadership and legal literacy training for women throughout the
1980s and 1990s.13
Where was the state in this scenario? Analyses of neoliberal restructuring have
documented the scaling down of entitlements and of governments’ commitments
to universal provision showing the trend towards a greater reliance on the private
and third sectors for welfare delivery. These trends have been seen as evidence of
the shrinking of the state, its ‘hollowing out’, ‘evacuation’, even ‘disappearance’.
Does this accurately describe what has occurred in Latin America? The evidence
suggests a more complex picture, a less ‘zero sum’ situation. There has undoubtedly
occurred a decisive shift from the state-centric principles that previously governed
social policy, but this has not gone along with the ‘evacuation of the state’ from
social provision. This widely held view of post-1980 reforms does not allow us to
capture what is different about current social policy figurations of the state, or the
substantial changes that the state itself has undergone during the decades of
‘neoliberal hegemony’. In Latin America social reproduction in the domains of
health and educational provision remains by and large the responsibility of the
state across a range of countries, despite decades of creeping privatization and
underfunding. After the critical watershed years of the debt crisis when social
expenditure per capita fell to unprecedented levels,14 by 1991 it recovered the
levels registered at the beginning of the 1980s, and in recent years social expendi-
ture has generally risen across the region. While these figures must be treated with
caution, the role of the state in public welfare remains significant.
The state itself however has undergone reform in this process, led by efforts to
advance good-governance agendas designed to make state institutions more effi-
cient and accountable and by democratic reform parties and movements. This has
gone along with support for decentralization and deconcentration, with Latin
America taking the lead in the 1990s as the region that advanced furthest down
this path. Re-democratization involved a wave of constitutional reform across the
region, and decentralization was one of the democratic principles incorporated in
the new frameworks to redress a historic legacy of over-centralization. Across the
Latin American region from the 1980s, states have been engaged in a process
aimed at strengthening and reforming local government, while devolving a
greater share of the budget to locally administered state agencies. This process of
‘municipalization’ has brought the state back into the domain of welfare provi-
sion, albeit in a new guise. As we will see in the case of Mexico, poverty relief has
engaged states, both central and local, in a wide range of programmes involving
millions of dollars of public and international funding. None of this is to suggest
that the decentralization process in Latin America has overcome distributive prob-
lems or secured adequate citizen representation. Devolved resources remain sparse
and without plans to tackle regional economic regeneration, decentralization has
not generally produced a marked improvement in welfare coverage. Yet, while

moribund, corrupt and clientelized states abound in the region, states are still cen-
tral actors in the development and welfare domains. As Judith Tendler shows in the
case of Ceará with a population of 40 million in the Brazilian North East, the state
‘is doing more’ not less ‘and something quite different as well’ (Tendler 1997: 24).
These multiple changes in social welfare provision were bound to have conse-
quences for the large numbers of female poor. In recent years female poverty, as
distinct from the gender dimensions of poverty, acquired considerably more
policy attention. If during the period of SAPs women were the invisible army who
bore the costs of the adjustment to ensure household survival, the New Poverty
Agenda appeared to render women more visible. From the later 1980s women’s
poverty as well as their role in poverty relief programmes became increasingly
evident to policy communities. Feminist advocacy and research into the gendered
effects of adjustment played their part in securing this visibility: female poverty
was a central theme of all the International Women’s conferences and the Beijing
Platform for Action (PFA) called for it to be addressed as a matter of urgency. The
PFA proposed a number of priorities for assistance: the targeting of female-headed
households, greater participation of women in decision-making at community
and other levels, and the extension of credit to low-income women were among
them. The promotion of these ideas was also part of a broader effort by Latin
American women’s organizations to incorporate a gender analysis into regional
declarations and government policies. The ‘new social policy’ (NSP) therefore
evolved during the high point of global feminism and yet, as we shall see, its prac-
tical realization often meant that it existed in tension with the latter’s emphasis on
equality.
However, while the current focus on poverty has its novel features, it is also
marked by a continuum with earlier Women and Development approaches that
saw ‘integrating women’ as a way to secure broader development objectives, while
failing to tackle underlying causes of gender inequality. If this is a general rule that
all too often applies today, it is nonetheless important to emphasize the consider-
able diversity in the conception and implementation of these new programmes. As
we will see, the objectives of these programmes determine how women will be
involved and how they are affected. To illustrate this point, we now turn to
consider the Mexican anti-poverty programme, PROGRESA/Oportunidades.
PROGRESA/Oportunidades
Mexico is an upper-middle-income country and one of the most industrialized in

Latin America, but poverty is estimated to afflict half of the population, with a
fifth in extreme poverty, due to its highly skewed income distribution.15 Social
divisions inherited from the colonial period and deepened through urban bias
exist along regional, ethnic and gender lines, with 44 per cent of indigenous
Mexicans found in the poorest income quintile. Mexico’s state welfare system is
based on formal employment but coverage is restricted to just 55 per cent of the
population due to the character of its labour market (Laurell 2003: 324). Up to half
54 Maxine Molyneux
of the economically active population depends upon the informal sector for its
income, and has access to few benefits. Moreover the size of the informal sector
means that Mexico collects only 11 per cent of GDP in tax, well below the average
for Latin America (which is 18 per cent) and below that of relatively low-tax
countries such as the United States.16
The election of Partido Acción Nactional (PAN) leader Vicente Fox in 2000 ended
71 years of one-party rule by the Pártido Revolucionario Institucional (PRI), and
was accompanied by efforts to reform existing institutions along more democratic
and accountable lines. Fox pledged to make social justice a priority of his govern-
ment, recognizing that poverty was a ‘multidimensional phenomenon’, and
raising social expenditure by an average of almost 10 per cent per annum. The
main poverty relief programme directed at those in extreme poverty, PROGRESA
was modified and relaunched in 2002 under the name of Human Development
Opportunities (Desarrollo Humano Oportunidades) known today as Oportunidades.
The programme’s coverage, formerly restricted to the rural poor, was extended to
include urban and semi urban areas,17 and the number of those inscribed in the
programme was expanded from 2.6 million families (in 1999), the equivalent of
40 per cent of all rural families (Rocha Menocal 2001: 520), to 4.2 million families
in 2002 (of whom 2.9 were rural) (González de la Rocha 2003: 14). By 2005 it
covered five million households with an estimated 25 million beneficiaries.
Oportunidades is one of the most extensive programme of its kind in Latin America.
It is also considered to be the most successfully developed example of the region’s
NSP-inspired anti-poverty programmes.18 It has been judged to be particularly
effective in meeting its goals, and this is attributed to an unusually high degree of
presidential support and inter-ministerial collaboration along with an annual budget
(in 2004) of 25 billion pesos and a recent loan of $1 billion from the IDB. An
undoubted strength of the programme is that it is subject to regular evaluations,
including some by outside bodies, and has been responsive to suggestions for
improvements and modifications. Oportunidades is a targeted cash transfer pro-
gramme that attempts to combine short- and long-term objectives of sustainable
poverty reduction, as advanced by the social risk-management approach. As noted
earlier this approach aims to tackle poverty though helping the poor to ‘cope, mit-
igate or reduce’ their risk of falling into or being trapped in poverty. Oportunidades
aims to improve human development by focusing on children’s education, nutrition
and health. It is based on the assumption that poor households do not invest enough
in their human capital, and are thus caught in a vicious circle of intergenerational
transmission of poverty with children dropping out of school and destined to
suffer the long-term effects of deprivation.
Families selected for the programme are therefore helped, through cash transfers,
with the financial and opportunity costs of having children in school. The transfers
are primarily in the form of ‘scholarships’ for children to attend school,19 supple-
mented by additional cash to improve nutrition where required. The practical func-
tioning of the programme centres on mothers as the key to securing improvements
in the life chances of their children, born and unborn. It seeks to strengthen,
through workshops and monitoring, the mothers’ responsibilities for children’s
health and education and to improve the nutritional status of their children
(and of themselves, if they are pregnant or breastfeeding). Secondary outcomes
such as building the mothers’ capacities, empowerment, citizen participation and
strengthening community ties are part of the programme’s goals, but how these
are interpreted has varied over time and the quality of what is on offer under these
headings depends upon local authorities and cooperating professionals.
Oportunidades’ guiding principles are designed to differentiate it from assisten-
tialist programmes through an emphasis on the participants’ ‘active management’
of their risk through ‘co-responsibility’ (cogestion) or co-management. Responsibility
for health and education is to be recognised as not only the government’s but the
whole of society’s, and therefore should be assumed by the entire community.
However, the responsibility of the ‘entire community’ is perhaps better described
as being devolved to mothers who are those designated as being primarily respon-
sible for securing the programme’s outcomes. Co-responsibility is formalized
through a quasi-contractual understanding that in return for the entitlements
proffered by the programme, certain obligations are to be discharged by the two
parties, that is, the programme and the participating mother. This conditional
form of entitlement, although well established in other regions and emanating
from the United States, is of more recent origin in Latin America but it is being
widely adopted. In this case the responsible participants (mothers) receive their
stipend on condition that they fulfil the duties laid out by the programme
managers: this involves taking children for regular health checks, meeting targets
for ensuring their children’s attendance at school, attending workshops on health
and programme coordinators’ meetings, and contributing a set number of hours of
work to the community, typically for cleaning buildings or clearing rubbish.
Failure to comply with the requirements can lead to being struck off the
programme.
On the available evidence collected through regular evaluations, the programme
has been largely successful in its own terms.20 Its stipends have reduced household
poverty, and have improved school attendance of children, as well as the health
and nutritional levels of all those inscribed in the programmes.21 These are impor-
tant gains, and are extensively discussed elsewhere. Here we will focus on some of
the more contentious aspects of the programme as highlighted by the women
participants in interviews and evaluations before considering the question of the
programme’s gender impacts.
As far as the design of the programme is concerned there appear to be two main
criticisms made by participants which raise some general issues of principle. These
are the issue of targeting and that of co-responsibility. With regard to the first, one
large-scale multi-site assessment carried out by the World Bank and others22 found
that while none of the women who participated in the programmes doubted that
Oportunidades had helped them in their struggle against poverty, there were some
criticisms of the way targeting was applied. Despite the rigour of the selection
mechanisms, and despite the claim that the programme is intended to be seen as
a way to ‘[access] a social right in a situation of social inequality’ (Rivero 2002: 6),
the targeting process attracted the strongest criticism from participants in some
56 Maxine Molyneux
evaluation exercises. Along with a general sense that more information should be
made available on the programme and on the means-testing mechanism itself,
dissatisfaction was expressed over the selection process, which was felt to be
arbitrary, excluding people whose needs were considered just as pressing as those
included in the programme. Means-testing was felt to ‘generate a lack of trust,
social divisions and feelings of envy and exclusion’ among those not selected.
(González de la Rocha 2003: 17). These are common problems faced by targeted
social-protection programmes in contexts where poverty is extensive and deep,
although Oportunidades’ coverage is more extensive than most programmes, a
factor which has caused it to be dubbed by those in charge of it, ‘a near universal
programme’.
A second general complaint voiced by participants was that they felt ‘discrimi-
nated against’ by its demands on their time. They believed that they were ‘treated
badly or […] were asked to do things in ways that offended their dignity’ (Rivero
2002: 4). As they expressed it, because they were ‘paid by the government’ they
were expected to perform community work, such as cleaning schools and health
centres, while others in the community did not.23 In another evaluation women
complained of being made to do ‘absurd’ tasks just for the sake of keeping them
occupied. The requirement to do community work had been incorporated into the
earlier programme and was continued into the new post-PROGRESA design by the
Fox administration, but following recommendations by evaluators the amount of
work time contributed was reduced, and it is still an issue under consideration.
In light of such findings it is not surprising that there was, among some
communities, resistance to accepting the notion of ‘co-responsibility’. Rather, the
requirements of the programme were seen in terms of ‘obligations’ and partici-
pants felt that genuine co-responsibility would also oblige teachers to accept their
‘responsibility not to miss classes so much’. This ‘inequality of responsibility’
made some participants resentful of the way they were expected to meet targets set
for monitoring the health and education of their children and could be ejected from
the programme for failing to do so. Why, they asked, should a teacher’s salary
not be reduced if they fail to turn up to teach, since mothers were fined for not
meeting their targets? This latter point reflected a general criticism that there
were few reliable mechanisms of accountability where complaints regarding the
behaviour of officials or professionals could be processed. Nor were the partici-
pants given an active role in the design, management and evaluation of the pro-
gramme (González de la Rocha 2003). It is hard to square these findings with
the view that the programme was intended to function as a way of exercising
civil, political and social rights and as a means to achieve full citizenship
(Rivero 2002).24
Gender relations: now you see them, now you don’t
One of the claims made by Oportunidades programme managers is that it has

helped to empower the mothers and daughters who are its beneficiaries. It is to
this claim that we now turn. As far as gender is concerned it is clear that gender is
not only incorporated into, but is central to, the management and design of the
Oportunidades programme. There are four main aspects to this gender sensitivity:
first, the programme was one of the earliest in Latin America to give the financial
transfers (and principal responsibilities associated with them) to the female head
of participating households; second, the transfers associated with children’s school
attendance involved an element of affirmative action: stipends were 10 per cent
higher for girls than for boys at the onset of secondary school which is when the
risk of female drop-out is highest;25 and third, the programme’s healthcare bene-
fits for children were supplemented by a scheme which monitors the health of, and
provides support for, pregnant and breastfeeding mothers, and children under two
years of age. The fourth aspect of the project design which displays gender
sensitivity is the goal to promote the leadership and citizenship of the women
subscribed. These goals represent a combination of equality measures (for the girls)
and maternalist measures for their mothers, but with what outcomes and gender
impacts?
There is a paucity of detailed evidence on this question, and far from sufficient
to make an accurate estimate.26 Most information that is available comes from a
large-scale survey by Adato et al. (2000) and qualitative research by Escobar and
González de la Rocha (2004). These allow certain general points to be established.
In the first place, as is well known, improving the educational opportunities of girls
has strong potential to enhance their self-esteem and life chances, while at the
same time sending a message to households and to communities that girls are
‘worth investing in’. Secondly, stipends paid directly to mothers are widely
accepted to benefit their households through more equitable redistribution, but in
giving women direct control over cash resources, their standing in their commu-
nities as well as their leverage within the household can be enhanced. The evidence
from evaluations of the Mexican programme confirm these trends, although, as
one evaluation noted, while the mothers enjoyed some increased autonomy, this
did not necessarily translate into empowerment, since the latter depended on
more factors than control over a small money income (Escobar and Gonzalez de la
Rocha 2004). Women did however appear to feel that their self-esteem was
enhanced as a result of the stipends.27 They also appreciated the programme’s edu-
cation and training projects (including health and community leadership) where
these were well organized, but they wanted more access to education and training
(Adato et al. 2000).
More research into the gender impacts of the programme is needed to establish
if it is producing a redistribution of power and status within households and if so
to explain what effects this status reordering has on household livelihoods and
well-being. Transfers paid directly to women have the potential to generate con-
flict if men feel that they are entitled to control money resources and resent any
undermining of their authority. However, existing data indicate that no strong
relationship has been found linking the programme and the incidence of violence
in the home.28
These positive developments however might need to be qualified in the context
of more critical appraisals. While those available refer to the earlier years of the
58 Maxine Molyneux
programme they indicate issues that arose, some of which are ongoing, but also
the reflexive response of the programme to these problems. One evaluation of
Oportunidades by the Network of Rural promotoras and assessors (Red de
Promotoras y Asesoras Rurales) concluded that there was no significant improve-
ment in women’s position in their families,29 the stipend was insufficient to
overcome poverty, and the programme did not generate employment opportuni-
ties for school-leavers which would enable the cycle of poverty to be overcome.30
Escobar and Gonzalez de la Rocha (2004) further noted that the programme did
not take sufficient account of women’s income-generating and other activities
such as collective community work ( faenas) and that as a consequence women
could be overloaded with competing demands on their time.
In sum, Oportunidades has several positive features as a new human development
programme beyond its successes with regard to improving children’s health and
life chances. It has, over time, expanded its coverage and has sought to respond to
some of the gaps in its provision taking account of evaluations in the modifica-
tions of its programme. It has also made some headway in detaching poverty relief
from political patronage although much still remains to be achieved. It has also
helped low-income households to cope financially with the demands of school-
age children. It has remained, however, in essence a maternalist programme in
that it aims to fortify and normalize the responsibilities of motherhood as a way to
improve the life chances of children. Its human development rationale accurately
describes the programme’s goals as far as children are concerned, but its combined
focus on mothering and reproductive health has made it less likely to develop a
more differentiated set of capacities for the mothers. In effect Oportunidades creates
a dependency on a subsidy which confirms mothering as women’s primary social
role, one which may enhance their social status and self respect, but nonetheless
puts them at risk of remaining in poverty for the rest of their lives.
Female altruism at the service of the state?
Oportunidades exemplifies the maternalism at the heart of many of the new anti-
poverty programmes being established in Latin America, and its organizational
principles raise some important questions for gender analysis. Feminist theory
and practice has suggested that if gender equality is to be tackled in development
and welfare programmes these must have some potential to empower women and
enhance their capabilities in ways that enable them to challenge relations of
inequality and subordination and at the same time provide some scope for female
economic autonomy. The new anti-poverty programmes may successfully identify
some unmet needs within poor households and communities but attending to the
needs of the women (the mothers) who are central to the functioning of these
programmes is not their explicit aim, any more than is gender equality a key
objective. The social construction of need in these programmes is child-centred, as
is their overall organization. Women are incorporated into programme design
(i.e. are visible) but in a way that depends upon the gender divide for its success. Thus,
even as women might be empowered within these structures (through managing
the subsidy), such programmes in effect reinforce the social divisions through
which gender asymmetries are reproduced.
In the first place they depend upon women fulfilling their ‘traditional’ social
roles and responsibilities. Oportunidades does so by basing its programme on nor-
matively ascribed maternal responsibilities, in effect making transfers conditional
on ‘good motherhood’. Men are not incorporated in any serious way, and no effort
is made to promote the principle that men and women might share responsibility
for meeting project goals. These programmes unambiguously rest on normative
assumptions concerning ‘women’s roles’ so that the work women undertake in
ensuring that children’s needs are met is taken for granted as something that moth-
ers ‘do’. The social relations of reproduction therefore remain unproblematized,
and the work performed easily naturalized.
Latin American cultural constructions of femininity are strongly identified
with motherhood, and serving the needs of children and household is generally
considered a primary maternal responsibility. Motherhood is often offered as the
explanation for political or civic activism, and allied with moral virtue, altruism
and self-sacrifice. It is likewise assumed by programme managers and participants
alike that any actions that improve the well-being of children are not, as
Bradshaw and Quirós Víquez express it, a ‘burden’ for women, and any ‘costs’
they bear are ‘simply part of the mothering role’ (2003). If femininity is closely
bound up with an affective investment in a self-sacrificing or altruistic motherhood,
the ideological site for contesting the demands of maternalist programmes is not
one that is easily occupied. Beneficiaries who miss a clinic appointment or a
workshop because they were working, lay themselves open to the charge of
being ‘bad mothers’ who do not care for their children (Bradshaw and Quirós
Víquez 2003).
The Mexican programme seeks even greater commitment from mothers
through the regulation of their domestic responsibilities, situating them as the
principal managers of their families’ needs. This does however involve some status
reordering in the family in favour of mothers but within the traditional division of
labour. While much is said about the ‘individuation of the social’ in regard to neo-
liberal policies, this does not apply to the women in these programmes who are
bound ever more securely to the family. If there is a new element beyond the
purely technical administration of the project, it is that which is introduced by
some (albeit limited) sensitivity on the part of the designers to issues of gender
equality at least in the case of the girl children, if far less so in the case of their
mothers. One must conclude that gender equity considerations had some influ-
ence in the design of these programmes in recent years, sometimes as a result of
feminist advocacy through NGOs, sometimes as a result of the shift in public and
professional attitudes occasioned by the spread of feminist ideas.
It remains the case that the women in such programmes are primarily
positioned as a means to secure programme objectives; they are a conduit of policy,
in the sense that resources channelled through them are expected to translate into
greater improvements in the well-being of children and the family as a whole.31
Such benefits as are derived by the mothers themselves as a result of participation
60 Maxine Molyneux
in the programme are a by-product of servicing the needs of others. This is com-
pounded by the fact that there is little in the design of the programmes that
advances women’s economic autonomy or security. Training for the job market is
limited or non-existent, and there is scant if any, childcare provision for those
women who want or need it because they work, train or study. Poor women are
often involved in income-generating activities which, while precarious, may leave
them without much disposable time or flexibility.32 Indeed, while rarely acknowl-
edged in the case of women, participation in anti-poverty programmes can have
negative consequences in incurring opportunity costs by preventing or restricting
women’s freedom to engage in paid work (Bradshaw and Linneker 2003). In the
Oportunidades programme the assumption that mothers were available to carry out
‘their’ stipulated duties in respect of their children takes no account of the fact
that there was according to one evaluation, an ‘increasing dependency’ on
women’s earnings, even though these were often meagre.33
This is not to deny that many women might choose not to work and might not
perceive the programme’s demands as anything other than helpful in relieving
some of the pressure on them to obtain paid work – especially if little is available
to them. However given the importance of women’s lifelong economic precari-
ousness and their need to secure cash incomes, the relative lack of attention to this
issue is striking.34
Conclusions
Women have much to contribute to anti-poverty programmes. Their gendered

assets, dispositions and skills, their inclination towards involvement in household
survival and at community level, and their precarious relationship to the wage
economy, all help to make them appear a peculiarly suitable ally of anti-poverty
programmes. This is not least because they also represent an army of voluntary
labour, and can serve as potential generators or guardians of social capital. This
‘suitability’ can be understood as arising from the positive and negative aspects of
the gendered relations of poverty. If poverty is a multidimensional condition
involving deprivation and exclusion, then on the basis of indicators such as those
developed in the capabilities approach, women, lacking assets and with fewer
capabilities, might be considered to be more subject to deprivation and exclusion
than men. Yet, if this is often so, well-being is not only a question of material
goods but involves self respect, dignity, belonging and participation and it may be
the case that even disadvantaged women might have more access to forms of
social inclusion and well-being than men in similar material circumstances.
Women are vulnerable economically, chiefly because their labour-market situation
is precarious and interrupted by periods of childbearing and the demands of care-
giving. They often lead lives that are busy, engaged as they are in the informal and
care economies, both private and public, typically performing unpaid or poorly
paid work. Yet they can and frequently do, gain satisfaction, self-esteem, recogni-
tion and respect from motherhood and from activities that constitute a kind of
‘informal citizenship’ that takes their domestic activities from the isolation of the
family to public spaces, with some (albeit variable) development of their

capacities.
These gendered assets and dispositions are being increasingly recognized by the
international development agencies, but so far this has not brought significant
material benefits to the women involved. The costs many women bear through
juggling these multiple responsibilities in terms of weak labour-market links, lack
of support for carework and long-term security are rarely taken into account.
Prevailing policy assumptions still tend to naturalize women’s ‘roles’ and seek to
make use of them and influence how they are developed and managed subjec-
tively and situationally. Poverty relief is still treated all too often as a matter of an
unproblematized social need, abstracted from the social relations that produce it.
The classic assistentialist programmes that targeted women and children as high
risk and in poverty were based on this view, and were commonly associated with
paternalist notions of care and charity. They made little if any attempt to address
the conditions which placed their beneficiaries in these circumstances and
concentrated on short-term relief typically delivered in the form of food aid and
primary healthcare. The ideas of the New Social Policy try to go beyond this
through participation, gender awareness, capacity building, and by ‘responsibiliz-
ing’ the poor, yet in practice programmes still rely on outdated conceptions about
women’s social roles that take little account of their differentiated needs.
Programmes that give money or food to those in ‘vulnerable’ positions and fail to
strengthen households’ assets do little to reduce vulnerability (Bradshaw and
Quirós Viquez 2003). If women are to be provided with an opportunity for redefin-
ing the terms of their inclusion in their societies and in their polities, the
unequally valued forms of social participation for men and women that are
inscribed within the public and private spheres, and pervade the organization of
carework, the public sphere, paid work, and public institutional life need to be
challenged rather than deepened. The radical challenge to social policy from an
equality perspective is to de-ontologize ‘women’s roles’ and to help reconstruct
gender relations along more egalitarian lines in both the domestic and public
spheres. Marginalizing men from these responsibilities is not in their overall
interests any more than it is in children’s.
For women in poverty, current programmes appear to offer both risks and
opportunities. Women are placed to occupy a central role in the new poverty
agendas, and the evidence shows that many women in Latin America experience
well-being from participating in activities that are not tied to monetary reward.
But such participation is always conditional on the participant’s support of the
project, and on the gains, individual and collective, that it brings. Those involved
in voluntary work may be happy to give their time and effort but they still need
projects that enhance their capabilities through education or training, providing
links to employment, advancing credit for successful projects that enable them to
acquire their own assets. Above all, women need a reliable income source and sus-
tainable routes out of poverty, ones that are at the same time more realistic and
imaginative than the maternalist options that are currently in place. The limits of
programmes like Oportunidades are evident not only in their selective approach to
62 Maxine Molyneux
tackling social need, but in their narrow vision of how to overcome poverty;
stipends are no substitute for economic regeneration, and without attention to the
household livelihoods and long-term prospects of the poor, including women, such
programmes despite their good intentions, remain fundamentally trapped in
‘assistentialist’ paradigms.
Notes
1 This chapter is extracted from a comparative research paper entitled Poverty Relief and the
New Social Policy in Latin America: Mothers at the Service of the State? (Molyneux
forthcoming). I would like to thank Edurne Larracoechea for her invaluable research
assistance on this project. Thanks also to my colleagues at the Institute for the Study of
the Americas – Helga Baitenmann, Fiona Macaulay and Rachel Sieder – as well as to
Sarah Bradshaw, Jasmine Gideon, and María de la Paz López for their encouragement
and generous help with materials. Thanks, too to Elizabeth Jelin for her comments and
to the Oportunidades programme directors for their cooperation.
2 The analysis draws on conversations with Oportunidades Director Rogelio Gómez
Hermosillo and Concepión Stepa, Director of Planning and Evaluation, as well as on field-
visits to Huachinango, San Miguel de Tenango and Zacatlán de las Manzanas in July 2005.
3 This term was first used by the World Bank but it has acquired a wider currency since.
4 Colombia is a possible sixth according to some analysts. For overviews and analyis of
Latin American social policy see inter alia Abel and Lewis 1993 and 2002, Abel 1996,
Mesa-Lago 1994, Haagh and Helgo 2002, and Tulchin and Garland 2000.
5 It is significant that this group includes a socialist (Cuba), a market (Chile) and a mixed
economy (Costa Rica) model of welfare. See Mesa-Lago 2000 for elaboration of these
comparative observations.
6 See Eckstein and Wickham-Crowley (2003: 19) on subsistence rights.
7 Filgueira and Filgueira (2002) differentiate between countries characterized by ‘stratified
universalism’ (Uruguay, Chile, Argentina); ‘dual regimes’ (Brazil, Mexico) and ‘exclusionary
regimes’ (DR, Central America, except Costa Rica), Bolivia and Ecuador.
8 As early as 1906 in Uruguay bills were proposed to give rights to maternity leave, and leg-
islation to restrict women’s working hours were first introduced in Argentina in 1905.
9 Barriers to claim-making included administrative obstruction, low female educational
attainment, ignorance of rights; and for indigenous women, the lack of an identity card
would be sufficient to bar them from their entitlements.
10 By 2005 the talk was of the ‘post-post-Washington non-consensus’ reflecting the greater
pluralization of approaches to development in Latin America and in policy circles.
11 But one that was singularly gender blind. For a critical analysis see Molyneux 2002.
12 See Jelin and Hershberg 1996 for further discussion.
13 See Craske and Molyneux 2002 for case studies of women’s movement activism in the
1990s.
14 In real terms social spending per capita declined by 10 per cent between 1982 and 1986.
Even as it grew afterwards it remained 6 per cent below 1980 at the end of that decade,
and only recovered slowly in the 1990s (IDB 1996).
15 ECLAC and the Mexican government’s estimates broadly agree that 45 per cent of the
population live under conditions of poverty while other estimates put the figure as high
as 61 per cent in poverty and 26.5 per cent (25 million) in extreme poverty (Urquidi in
Middlebrook and Zepeda 2003). The top 20 per cent accounted for 59.3 per cent of
income in 1989; the bottom 20 per cent for 3.9 (Grindle 2000: 20).
16 In 1997 social expenditure accounted for 7.8 per cent of GDP. It has risen slightly since.
17 PROGRESA was preceded by PRONASOL also known as Solidarity, Mexico’s first large-
scale anti-poverty programme. Established in 1988 its conception of poverty relief was
quite different from PROGRESA and had party political objectives. It was designed by the
Salinas’ administration to offset the political consequences of the adjustment years and
revive the flagging political support of the PRI. According to Molinar and Weldon,
PRONASOL’s regional priorities were developed with three aims in mind: to reward PRI
loyalists, to reconvert PRD supporters and to punish PAN supporters (in Rocha Menocal
2001: 524). Such manoeuvres delivered the expected returns to the ruling party, but
the programme was discredited. When Zedillo assumed the presidency in 1994, he
replaced PRONASOL with PROGRESA claiming that his new anti-poverty programme
did not have a political agenda (Rocha Menocal 2001: 513). Although some political bias
continued, it was much reduced, the PRI lost the 2000 elections to the opposition
PAN party. Oportunidades has since sought to distance itself from this record of political
clientelism, with a public campaign message stating that social protection is a right and
allegiance is not due to any political party.
18 Some observers have preferred to ignore the Mexican contribution to the development
of the programme and have seen it as originating with the World Bank as it appears to
be in sympathy with its broader recommendations. PROGRESA was not imposed by the
World Bank and was intended to run only on federal funds with no direct funding from
the World Bank. Oportunidades is government-funded with loan support from the IDB as
above.
19 Monetary and educational grants are provided for each child under 22 years of age
who is enrolled in school between the third grade of primary and third grade of high
school.
20 The International Food Policy Research Institute website contains a number of evaluation
reports on different aspects of PROGRESA/Oportunidades. See in particular Skoufias and
McClafferty 2000, which covered the three years up until 1999, as a result of which the
programme was extended to rural areas. The results of a qualitative evaluation carried
out in six communities by Escobar and González de la Rocha were published in 2004 and
will be referred to here. Another large-scale evaluation is currently under way.
21 Escobar and González de la Rocha’s qualitative research confirmed Skoufias and
McCafferty’s findings that the largest positive impacts were on children in secondary school.
The latter’s survey finds a 10 per cent increase in enrolment for boys and 20 per cent
for girls along with an overall narrowing of the gender gap in education, particularly in
primary school.
22 Cited in Rivero 2002.
23 In the PROGRESA programme such work involved on average 29 hours per month.
24 Since the programme strives to separate social entitlements from political clientelism,
this is another sense in which citizenship is understood.
25 Grants rise with the age of the child and the sex difference in the grant starts with
secondary school, which is normally when girls drop out. In the third year of secondary
school monthly grants are about US$58 for boys and US$66 for girls.
26 An audit of the gender effects of the programme is currently under discussion.
27 Escobar and González de la Rocha 2004.
28 This is clearly contentious and requires further in-depth research. According to Adato
et al. 2000 and my own interviews with specialists in gender and poverty, there has been
considerable evidence of violence against women over control of the stipend in some
regions (Author’s interview, Oaxaca, July 2005).
29 Based on responses from 309 beneficiaries and 60 professionals in eight communities.
30 A new component has been added to the programme since these findings, the ‘Jovenes
con Oportunidades’ which provides youth training and work experience. However on a
field visit young people interviewed saw themselves as having no future in their
localities and dreamt of migrating. Without attention to rural livelihoods Oportunidades
risks educating the young for the US labour market.
31 BID data cited in Bradshaw and Quirós Viquez 2003.
64 Maxine Molyneux
32 See for example Gideon on Chile (2002), Bradshaw (2002) on Nicaragua.

33 Escobar and Gonzalez de la Rocha 2004.
34 A pilot project undertaken by the BID is working with mothers to engage in productive
activity; this is clearly a step forward and signals an awareness of this shortcoming.
References
Abel, C. (1996). Health Hygiene and Sanitation in Latin America c. 1870–1950. Institute of Latin
American Studies, University of London.
Abel, C. and Colin Lewis (eds) (1993). Welfare, Poverty and Development in Latin America.
Basingstoke: Macmillan.
Abel, C. and C. Lewis (eds) (2002). Exclusion and Engagement: Social Policy in Latin America 2,
Institute of Latin American Studies, University of London.
Adato, Michelle, B. de la Briere, D. Mindek and A. Quisumbing (2000). The Impact of Progresa
on Women’s Status and Intra-household Relations. Final Report. Washington, DC: International
Food Policy Research Institute, Washington D.C.
Angell, A. and C. Graham (1995). ‘Can Social Sector Reform Make Adjustment Sustainable
and Equitable?’ Journal of Latin American Studies, 27.
Arriagada, Irma (1998). ‘The Urban Female Labour Market in Latin America: The Myth and
The Reality’. Mujer y Desarrollo Series, No. 21, Santiago, Chile, ECLAC.
Barrientos, A. (2004). ‘Latin America: A Liberal-Informal Welfare Regime?’ In I. Gough et al.
(eds), Insecurity and Welfare Regimes in Asia, Africa and Latin America. Cambridge:
Cambridge University Press.
Bradshaw, S. (2002). Gendered Poverties and Power Relations: Looking inside Communities and
Households. Fundación Puntos de Encuentro, Managua.
Bradshaw, Sarah and Brian Linneker (2003). Challenging Women’s Poverty. London: Catholic
Institute of International Relations.
Bradshaw, Sarah and Ana Quirós Víquez (2003). ‘The Social Protection Network in Nicaragua:
A Gendered Analysis’. Unpublished draft.
Brock, K., A. Cornwall and J. Gaventa (2001). Power Knowledge and Political Spaces in the
Framing of Poverty Policy. Working Paper 143. Institute of Development Studies, Sussex.
Cardelle, A. (1997). ‘Health Care in the Time of Reform: Emerging Policies for Private–Public
Sector Collaboration on Health’. Miami, North South Centre, Vol. VI, No. 1.
Castameda, T. (2002). Tendencias de Largo Plazo en Tamamo, eficiencia y Focalización del gasto
Social en America Latina y el Caribe. Report prepared for the Social Protection Division of the
Banco Internacional de Desarrollo.
Chant, S. (1997). Woman-Headed Households; Diversity and Dynamics in the Developing World.
Basingstoke: Macmillan.
Coelho, S. et al. (2002). ‘Deliberative Fora and the Democratisation of Social Policies in
Brazil’. IDS Bulletin, Vol. 33, No. 2.
Cornelius, Wayne, A. Craig and J. Fox (eds) (1994). Transforming State-Society Relations in
Mexico: The National Solidarity Strategy. San Diego: La Jolla, California Centre for Latin
American Studies.
Cornia, Giovanni, R. Jolly and F. Stewart (eds) (1987). Adjustment with a Human Face. Oxford:
Clarendon Press.
Craske, N. and M. Molyneux (eds) (2002). Gender and the Politics of Rights and Democracy in
Latin America. Houndmills: Palgrave.
De Vylder, S. (2003). Gender Equality and Poverty Reduction Strategies. Discussion Paper
(unpublished).
Eckstein, S. and T. Wickham-Crowley (eds) (2003). Struggles for Social Rights in Latin America.
New York and London: Routledge.
Economic Commission for Latin America and the Caribbean (ECLAC) (1999). Social
Panorama of Latin America 1998. Santiago: ECLAC.
Economic Commission for Latin America and the Caribbean (ECLAC) (2000). The Challenge
of Gender Equity and Human Rights on the Threshold of the Twenty-First Century. Report to the
eighth session of the Regional Conference on Women in Latin America and the Caribbean.
Lima, Peru: ECLAC.
Enrique Rosas, Rocío (2002). ‘Women and Survival Strategies in Poor Urban Contexts: A Case
Study from Guadalajara, Mexico’. Journal of Development Studies, 18 (2–3): 81–108.
Escobar, A. (2002). ‘The PROGRESA Programme and social change in Rural Mexico’. In Louise
Haagh and C. Helgo (eds.), Social Policy Reform and Market Governance in Latin America.
Escobar Latapi, A. (2003). ‘The PROGRESA Programme and Social Change in Rural Mexico’.
In A. Escobar and M. González de la Rocha (eds), Evaluación Cualitativa del Programa de
desarollo humano Oportunidades. Mexico City: Serie, Documentos de investigacion 3,
SEDESOL.
Escobar Latapí, A. and M. González de la Rocha (2004). Evaluación Cualitative del Programe
Oportunidades. Mexico City: Centro de Investigación y Estudios Superiores en Antropología
Social.
Filgueira, Carlos and F. Filgueira (2002). ‘Models of Welfare and Models of Capitalism: The
Limits of Transferability’. In Evelyne Huber (ed.), Models of Capitalism. University Park PA:
Pennsylvania State University Press.
Gideon, Jasmine (2002). ‘Decentralization, Participation and Inclusion: Reassessing Primary
Healthcare Delivery in Chile’. In L. Haagh and C. Helgo (eds.), Social Policy Reform and
Market Governance in Latin America. Basingstroke: Palgrave Macmillan.
Goetz, A. M. and D. O’Brien (1995). ‘Governing for the Common Wealth: The World Bank’s
Approach to Poverty and Governance’. IDS Bulletin 26, 2.
González de la Rocha, M. (1994). The Resources of Poverty: Women and Survival in a Mexican
City. Oxford: Basil Blackwell.
González de la Rocha, M. (2001). ‘From the Resources of Poverty to the Poverty of Resources:
The Erosion of a Survival Model’. Latin American Perspectives 28: 72–100.
González de la Rocha, M. (2003) ‘Oportunidades y Capital Social’. Paper presented at the
Comisión Económica Para America Latina, Workshop on Social Capital and Anti Poverty
Programmes, Chile.
González Bombal, I. and R. Villar (eds) (2003). Organizaciones de la Sociedad Civil e Incidencia
en las Politicas Publicas. Buenos Aires: Libros de Zorzal.
Graham, C. (1991). From Emergency Employment to Social Investment: Alleviating Poverty in
Chile. Washington DC: Brookings Institute Press.
Graham, C. (1992). ‘The Politics of Protecting the Poor During Adjustment: Bolivia’s
Emergency Social Fund’. World Development, Vol. 20, No. 9.
Graham, C. (1994). Safety Nets, Politics and the Poor: Transitions to Market Economies.
Washington, DC: Brookings Institution Press.
Grindle, Merilee (2000). ‘The Social Agenda and the Politics of Reform in Latin America’. In
Tulchin and Garland (q.v.): 17–52.
Haagh, Louise and Camilla Helgo (eds) (2002) Social Policy Reform and Market Governance in
Latin America. Basingstoke: Palgrave Macmillan.
Hirst, P. and G. Thompson (2000). Globalization in Question. 2nd edition, Cambridge: Polity.
Inter-American Development Bank (IDB) (1996) Economic and Social at Progress in Latin
America. Report, Making Social Services Work, special edition, Washington DC: Johns
Hopkins University Press.
IFRI at http://www.ifpri.org
Jackson, C. (1998). ‘Rescuing Gender from the Poverty Trap’. In C. Jackson and R. Pearson
(eds), Feminist Visions of Development: Gender Analysis and Policy. London: Routledge.
Jelin, E. and E. Hershberg (eds) (1996) Constructing Democracy: Human Rights, Citizenhip, and
Society in Latin America. Boulder: Westview.
Kabeer, N. (1997). ‘Editorial. Tactics and Trade-Offs: Revisiting the Links Between Gender and
Poverty’. IDS Bulletin 28, 3: 1–25.
66 Maxine Molyneux
Kabeer, N. (2003). Gender Mainstreaming in Poverty Eradication and the MDGs. London:
Commonwealth Secretariat.
Kaufman, R. and G. Trejo (1997). ‘Regionalism, Regime Transformation and Pronasol: The
Policies of the National Solidarity Programme in Four Mexican States’. Journal of Latin
American Studies, Vol. 29, No.3: 117–45.
Laurell, A. C. (2003). ‘The Transformation of Social Policy in Mexico’. In K. J. Middlebrook
and E. Zepeda (eds), Confronting Development: Assessing Mexico’s Economic and Social Policy
Challenges. Stanford CA: Stanford University Press.
Linneker, Brian (2003). ‘Gender Comparisons of Capital Influences on the Well-Being of
Women and Households Experiencing Poverty in Nicaragua’. Unpublished report, mimeo.
Lipton M. and S. Maxwell (1992). The New Poverty Agenda: An Overview, IDS Discussion Paper
No. 306.
Middlebrook, Kelvin and Eduardo Zepeda (eds) (2003) Confronting Development: Assessing
Mexico’s Economic and Social Policy Challenges. Stanford CA: Stanford University Press.
Mesa-Lago, C. (1994). Changing Social Security in Latin America: Toward Alleviating the Social
Costs of Economic Reform. Boulder CO: Lynne Reinner.
Mesa Lago, Carmelo with Alberto Arevas de Mesa, Ivan Brenes, Verónica Montecinos and
Mark Samara (2000). Market, Socialist and Mixed Economies. Comparative Policy and
Performance: Chile, Cuba and Costa Rica. Baltimore: Johns Hopkins Press.
Mkandawire, T. (2004). ‘Disempowering New Democracies and the Persistence of Poverty’. In
Max Spoor (ed.), Globalisation, Poverty and Conflict. Dordsecht, London and Boston:
Kluwer.
Molinar Horcasitas, Juan and J. A. Weldon (1994). ‘Electoral Determinants and Consequences
of National Solidarity’. In Wayne A. Cornelius, Ann L. Craig and Jonathan Fox (eds),
Transforming State-Society Relations in Mexico: The National Solidarity Strategy. San Diego:
Center for US–Mexico Studies.
Molyneux, M. (2002). ‘Gender and the Silences of Social Capital: Lessons from Latin
America’. Development and Change, Vol. 33, No. 2: 167–88.
Molyneux, M. (forthcoming). Poverty Relief and the New Social Policy in Latin America: Mothers
at the Service of the State? Programme Paper, Geneva: UNRISD.
Molyneux, M. and S. Lazar (2003). Doing the Rights Thing: Rights-Based Development and Latin
American NGOs. London: Intermediate Technology Publishing Group.
Moser, C. (1996). Confronting Crisis: A Comparative Study of Household Responses to Poverty and
Vulnerability in Four Urban Communities. World Bank, Washington DC.
O’Donnell, G. (1993). ‘On the State, Democratization and Some Conceptual Problems: A
Latin American View with Glances at Some Postcommunist Countries’, World Development,
Vol. 21, No. 8: 1355–69.
OECD DAC (2001). Guidelines for Poverty Reduction, at http://www.oecd.org/
dataoecd/47/14/2672735. pdf.
Pastor, M. and C. Wise (1997). ‘State Policies, Distribution and Neoliberal Reform in Mexico’.
Journal of Latin American Studies Vol. 29: 419–56.
Piester, K. (1997). ‘Targeting the Poor: The Politics of Social Policy Reform in Mexico’.
In D. Chalmers, C. Vilas, K. Hite, S. Martin, K. Piester and M. Segarra (eds), The New Politics
of Inequality in Latin America. Oxford: Oxford University Press.
Red de Promotoras y Asesoras Rurales (2000). ‘Dinero del Diablo’. Unpublished report,
mimeo, May.
Rivero, Margot Aguilar (2002). ‘Oportunidades y Derechos Sociales: Un Proceso de construcción
social de Ciudadanía’. Unpublished report, mimeo.
Rocha Menocal, Alina (2001). ‘Do Old Habits Die Hard? A Statistical Exploration of the
Politicisation of Progresa’. Journal of Latin American Studies, Vol. 33: 513–38.
Rose, Nikolas and Peter Miller (1992). ‘Political Power Beyond the State: Problematics of
Government’. British Journal of Sociology Vol. 43 (2): 173–205.
Rosemblatt, K. (2000). ‘Domesticating Men: State Building and Class Compromise in Popular
Front Chile’. In E. Dore and M. Molyneux (eds), Hidden Histories of Gender and the State in
Latin America. Durham NC: Duke University Press.
Schell, P. (1999). ‘An Honourable Vocation for Ladies: The Work of the Mexico City Unión de
Damas Católicas Mexicanas, 1912–1926’. Journal of Women’s History, Vol. 10, No. 4: 78–103.
Schild, V. (2002). ‘Engendering the New Social Citizenship in Chile’. In M. Molyneux and
S. Razavi (eds), Gender Justice, Development and Rights. Oxford: Oxford University Press.
Skoufias, E., B. Davis and S. de la Vega (2001). ‘Targeting the Poor in Mexico: An Evaluation
of the Selection of Households into PROGRESA’, World Development, Vol. 29, Issue 10,
October: 1769–84.
Skoufias, Emmanuel and B. McClafferty (2000). ‘ “Is Progresa Working?” Summary of the
Results of an Evaluation by International Food Policy Research Institute (IFPRI) and Food
Consumption and Nutrition Division (FNCD)’. Discussion papers No. 118, Washington
DC, at http.//www.ifpri.org.
Stallings, B. and W. Peres (2000). Growth, Employment, and Equity: The Impact of the Economic
Reforms in Latin America and the Caribbean. ECLAC: Brookings Institute Press.
Tendler, J. (1997). Good Government in the Tropics. Baltimore: Johns Hopkins University Press.
Ticknell, A. and J. Peck (2003). ‘Making Global Rules: Globalisation and or Neoliberalisation?’
In J. Peck and H. Yeung (eds), Remaking the Global Economy: Economic Geographic
Perspectives. London: Sage.
Tulchin, J.S. and A. Garland (eds) (2000). Social Development in Latin America: The Politics of
Reform. Boulder CO: Lynne Reinner.
United Nations Development Programme (UNDP) (1995). Human Development Report: Gender
and Human Development. Oxford and New York: Oxford University Press.
Urquidi, Victor. (2003). ‘Mexico’s Development Challenges’. In Kevin Middlebrook and
E. Zepada (eds), Confronting Development: Assessing Mexico’s Economic and Social Policy
Challenges. Stanford CA: Stanford University Press.
Varley, A. (2000). ‘Women and the Home in Mexican Family Law’. In E. Dore and
M. Molyneux (eds), Hidden Histories of Gender and State in Latin America. Durham NC and
London: Duke University Press.
Weyland, K. (1995). ‘Social Movements and the State: The Politics of Health Reform in
Brazil’. World Development, Vol. 23, No.10: 1699–712.
Whitehead, Ann (2003). Failing Women, Sustaining Poverty: Gender in Poverty Reduction Strategy
Papers. Gender and Development Network Report, Oxford: OXFAM and Christian Aid.
Whitehead, L. (1993). ‘On Reforms of the State and “Regulation of the Market” ’. World
Development, Vol. 21, No. 8: 1371–93.
Wood, C. (2003). ‘Adjustment with a Woman’s Face: Gender and Macroeconomic Policy at
the World Bank’. In S. Eckstein and T. Wickham-Crowley (eds), Struggles for Social Rights in
Latin America. New York and London: Routledge.
World Bank (1997). World Development Report: The State in a Changing World. Oxford and
World Bank (2000). World Development Report 2000/2001: Attacking Poverty. Washington DC:
World Bank.
World Bank (2002). The Gender Dimension of Bank Assistance: An Evaluation of Results
Yaschine, I. (1999). ‘The Changing Anti-Poverty Agenda: What Can the Mexican Case Tell
us?’ IDS Bulletin, Vol. 30, No. 2: 47–60.
3
Gender and Post-socialist Welfare
States in Central Eastern Europe:
Family Policy Reforms in Poland and
the Czech Republic Compared
Silke Steinhilber
Introduction
Since 1989, post-socialist1 Central and Eastern European countries have under-
taken significant reforms of their social policies and social security systems. One of
the key challenges in so doing was balancing the tension between a tradition of –
and in some countries continued commitment to – extensive welfare provisioning
and income redistribution on the one hand, and the residualist social policy set-up
advocated by the currently dominant global neoliberal economic framework on
the other. Among the multiple challenges faced during the reforms, abolishing
gender inequality was given little, if any, priority.
Through an analysis of family benefits after 1989, including maternity and
childcare benefits, as well as transfers to families, this chapter illustrates that
reforms have nevertheless had important gender implications. I trace how and
why ‘costly’ benefits and services supporting women’s dual role as worker and
mother – a feature of the socialist past often considered ‘women-friendly’ –
changed dramatically. Thereby, I explore how the broader dynamics of post-socialist
social policy reforms have transformed the institutional and social environment
in which gender relations are reenacted and renegotiated.2
Both Poland and the Czech Republic faced the ‘competing demands of building
capitalism and democracy’ at the beginning of the 1990s (Orenstein 2001: 6). Yet
the two countries responded somewhat differently: Poland chose a rapid and rad-
ical economic transformation strategy – ‘shock therapy’. In the Czech Republic, on
the other hand, neoliberal and social democratic elements were combined into a
‘social liberal strategy of reform’ (Orenstein 2001: 7).
Both countries also followed different reform paths in the field of social policy.
During the 1990s Poland moved further than other post-socialist countries toward
a residualist and familial model of a welfare state, scaling down state involvement
68
Gender and Post-socialist Welfare States in Central Eastern Europe 69
in social protection, and promoting individual responsibility and – de-facto, if not

always explicitly – greater reliance on the family for the provision of well-being.3
Reforms in the Czech Republic also revoked important elements of the inherited
universalist tradition of welfare-state provisioning. On the whole, however, Czech
social policy discourses continued to refer more explicitly to social solidarity. As a
consequence, to date, a higher level of state involvement as well as a greater
commitment to income redistribution has been maintained.
Feminists have highlighted that it is not the degree of state intervention as such,
but the particular kind of policy configurations that shape state–gender relations
(e.g. Lewis 1993, 1997; Sainsbury 1999).4 As state-organized mediators between
paid employment and unpaid care work, family benefits are one key policy area in
this respect. Changes in family benefits thus have far-reaching consequences for
gender relations. The analysis of family benefit reforms in post-socialist Poland
and the Czech Republic here focuses on three aspects and their intersections:
(1) the influence of the state socialist legacy on present-day reforms, (2) the overall
economic and political shifts in family policy since 1989, including the gendered
values of key reform actors, as well as (3) the power dynamics between reform
actors, including, among other aspects, the relative strength of women’s voices
in the reform discourse. I deal with each of these issues below in a comparative
fashion.
The legacy of economic and social policies
Three main features characterized the gendered patterns of post-World War II

labour markets in state-socialist Central Europe.5 First, women were integrated
into the labour market in unprecedented numbers, in particular into the industrial
sector. This represented a significant change compared to pre-World War II
economic structures. As a result, by Western European standards, employment
rates of women were high throughout Central and Eastern Europe under
Communist party rule. Second, greater participation of women in employment led
to a partial breakdown of gender stereotypes in occupations. Women occupied
professions and posts from which they had largely been excluded prior to the take-
over of the Communist parties. However, both vertical and horizontal gender seg-
regation of employment opportunities prevailed. Women were not found in posts
of higher responsibility in numbers commensurate with their overall employment
participation and high levels of education. Lastly, a gender gap in wages survived
all state-socialist commitments to equality. Despite the Communist party’s
equality rhetoric and formal commitments, illustrated for example in the early
ratifications of international laws, such as ILO Convention 100 on equal pay for
equal work and work of equal value,6 women earned considerably less than men
in wide sectors of the economy.
Family policy in pre-1989 welfare states

The gendered economic structures were also reflected in state-socialist social
policy, and in family benefits specifically. At the end of the 1980s Poland and
70 Silke Steinhilber
Czechoslovakia7 had elaborate systems of social security benefits (Deacon 1992;

Castle-Kanerová 1992; Millard 1992). Support for families included childcare
benefits and leave schemes, complemented by a state-sponsored system of crèches
and kindergarten (which was, however, far less extensive in Poland than in
Czechoslovakia), and price subsidies for a number of essential products.8
The family benefit system worked differently for women and men. It supported a
high employment rate for women, by helping women workers to combine their
dual roles as employees and mothers. Benefits and rewards for motherhood enabled
the state to include women into the system of full, but not necessarily freely
chosen, employment. In contrast, men were not explicitly addressed as workers and
fathers, or provided support to combine employment with family responsibilities.
Men were even legally excluded from a variety of family benefits, especially from
childcare leave schemes. Only under exceptional circumstances, such as the death
of the mother, could men claim most of the family-related benefits.
State-socialist family policy thus reinforced a particular gender order within the
family and society: Pre-socialist gender role models and cultural stereotypes were
preserved, while women’s economic independence was strengthened (Siemienska
1983; Eermáková et al. 2000: 42). Family benefits embodied a state-socialist notion
of a ‘good working-class family’, reflecting persistent conservative and patriarchal
values of leading decision-makers at the time. The state valued women as unpaid
caretakers and homemakers, while incorporating them as paid workers into the
economy (Eermáková et al. 2000; Seibert 2001).
In both Poland and Czechoslovakia, public spending on family benefits prior to
1989 was high by international standards, amounting to 4.4 per cent of GDP in
Czechoslovakia, and 3 per cent in Poland at the end of the 1980s. Relatively, both
countries spent a greater share of family benefit expenditure on cash benefits than
on in-kind benefits (mainly childcare services) (UNICEF/MONEE 1999; Fajth
1994).
Eligibility to most family benefits was linked to employment, and their level
wage-based (Kucharová et al. 2003; Wóycicka et al. 2003). Yet, given the nearly full
employment, benefits were available virtually universally. Czechoslovakia offered
maternity leave and benefits, extended maternity leave9 and a special benefit for a
parent to care for a sick child. Poland provided for maternity leave and benefits,
childcare benefit (sick-child benefit) and child-raising leave for employed mothers.
Other benefits or services such as crèches and kindergarten were either state-
organized, or provided through the employing organization, sometimes on-site.
In addition to employment-based benefits, some family benefits were univer-
sally available and paid directly from the state budget. Czechoslovakia offered a
one-time birth grant and a maternity allowance. Poland had a child-raising
allowance for mothers, as well as a family allowance, which depended on the fam-
ily income, but not on employment. In 1974, Poland also introduced an alimony
fund to support single parents (de facto single mothers).
Maternity leave and benefits had long traditions in both countries: maternity
leave was introduced in the Czech Lands in 1948,10 as was the child allowance to
help families cover the costs related to raising children. In Poland, maternity leave
and benefits had been introduced as early as 1924. The benefit for caring for a sick
child was introduced in 1954.
Parental leave schemes and benefits were of a more recent origin. They were
introduced in response to declining birth rates, yet were subject to political con-
flict throughout. In Czechoslovakia, family benefits were debated from the second
half of the 1950s. A flat-rate birth grant was introduced in 1956, and an extended
two-year maternity leave in 1970, initially unpaid. At the same time, a maternity
allowance was introduced for mothers of two or more children who stayed at
home. In Poland, child-raising leave was first introduced in 1968, initially consist-
ing of a one-year leave without pay. In the 1970s, with a low demand for labor
because of economic difficulties, and with declining birth rates, the state encour-
aged women to stay home for longer time periods (Lohmann and Seibert 2003: 78).
Unpaid leave was extended to three years in 1972 (Wóycicka et al. 2003: 194). Paid
leave for two years was introduced only in 1980 in response to popular demand,
and as a concession of the Communist government to the opposition Solidarity
movement which had initiated a campaign on the issue. A great number of Polish
women took advantage of the paid parental leave because the benefit level
was decent, and discontent with the workplace environment widespread.
Moreover, the supply and quality of institutional childcare was unsatisfactory
(Cichocinska 1993).
Since childcare was typically a mother’s responsibility in both countries,
childcare facilities were crucial to women’s ability to combine employment and
family responsibilities. In comparison with many Western European countries,
childcare was relatively widely accessible, generous and comprehensive in
Czechoslovakia. However, in Poland demand for childcare places greatly exceeded
supply throughout the entire postwar period (Lohmann and Seibert 2003: 70).
Also, regional differences in the availability of childcare services were substantial,
especially between urban and rural settings. Criticism about the services was
widespread: they were often overcrowded, and the ratio between children and
employees was high.
In both countries, nursery care enrolment, i.e. for children below age three, was
always lower than kindergarten enrolment. Places were rare, the quality of nursery
services was low, and state incentives for home care for young children were
strong. Nursery care enrolment in Poland declined even further during the second
half of the 1980s. The number of factory-run nurseries decreased by a quarter, the
number of places by a third, and the number of children in nurseries by almost half,
from a peak in the mid-decade. Kindergarten attendance also declined noticeably
over the 1980s (Cichocinska 1993).
Economic, political and family benefit changes after 1990
Poland
Poland went through profound economic, political and social changes after 1989,
experiencing a triple challenge of transition to a market economy, integration into
a globalized world economy, and transformation for accession to the European
Union. Popular discontent with the economic reforms caused several swings from
left- to right-wing governments. Despite their divergent political orientations,
however, the different governments have on the whole followed a surprisingly
similar macroeconomic and social policy reform recipe. Macroeconomic austerity
measures dominated all reforms in the highly indebted country. In its broad
effects, gender equality policy – or the absence thereof – has also been surprisingly
similar under different governments, as the discussion below illustrates.
Following the postulates of the ‘Washington Consensus’, highly indebted
Poland initiated a rapid economic reform process which focused on macroeco-
nomic stabilization, price and trade liberalization, and privatization (Lavigne
1999; Orenstein 2001). Among the severe consequences was an initial steep
decline of GDP by about 20 per cent, a significant decline in real incomes, and a
drop in economic activity rates, as well as high unemployment and growing
poverty. During 1992 more than 25 per cent of all Polish families could not afford
to purchase the basic necessities of food and clothing (Seibert 2001: 42). The share
of the population below the poverty line hovered between 12 and 17 per cent in
the period 1995–2000 (Lohmann and Seibert 2003: 16).
Poland today is regarded a successful economic reformer in CEE. Relative
economic stabilization with fast economic growth ensued during the mid-1990s,
followed by a slowdown of the economy at the end of the decade and renewed
faster growth after 2003. Yet growth has had only a limited employment impact:
both women’s and men’s employment rates have been decreasing massively since
1990, and are low by international standards. Women’s employment rates are
significantly lower than men’s. Unemployment has remained at high levels of
between 15 and 20 per cent since the early 1990s, with women’s unemployment
consistently higher than men’s. Gender differences in wage levels are in the range
of 15–30 per cent, the highest gap being found in the highest wage group
(Chlon-Dominczak 2004). A joint EU/ Polish government analysis concluded that
‘there remain significant inequalities of opportunities between women and men
in the labor market’ (European Commission/Government of Poland 2001).
Three major social policy reforms were enacted in 1998/99: reform of the
education system, the national health insurance system, and the national pension
scheme, which was partially privatized. Family benefit reforms during the 1990s –
while no less profound in their impact – were designed and implemented in mul-
tiple steps: among the most significant was the decoupling of the entitlement to a
number of family benefits from employment. They were now paid from the state
budget – in some cases the central, in others the local. A majority of family benefits
became income-tested (with the exception of maternity benefits and short-term
childcare benefits) (see Table 3.1). Some benefits were cut, and the real value of
some further declined through inadequate adjustment for inflation (Wóycicka
et al. 2003).
The Czech Republic

In contrast to Poland, the Czech reform path has been described as a ‘hybrid
“social liberal” strategy for transformation’, referring to a more moderate reform
Table 3.1 Entitlements and financing of family benefits, Poland, 2003
Means-tested/
Basis for Financing not means
Benefit entitlement source tested?
Maternity leave Employment Social insurance Not means-

and benefit relationship revenue tested
Childcare leave and Employment Social insurance Not means-
benefit11 relationship revenue tested
Child-raising leave Employment State budget Means-tested
and allowance relationship revenues
Family allowance Not employment- State budget Means-tested
based revenues
Alimony fund Not employment- State budget Means-tested12
based revenues ⫹ ⫹ court order
repayments
Guaranteed Former employment State and local Means-tested
periodic benefit relationship13 budget revenues
Benefits for Not employment- State and local Means-tested
pregnant women based budget revenues
and women raising
children
Source: Wóycicka et al. 2003: 193–9.
approach, and the practice of political compromises caused by ambiguous election

results and negotiations of coalition partners (Orenstein 2001: 61). As in Poland,
economic reforms in the early 1990s focused on financial stabilization, price liber-
alization, and the privatization of former state-owned enterprises. However, the
relatively low level of indebtedness allowed the country to maintain many
features of the former welfare system, including universal housing subsidies and
health and welfare benefits (Kucharová et al. 2003).
Even more radical market-oriented Czech political forces, which advocated
welfare retrenchment and a focus on incentives rather than redistribution, saw
state provisions as buffers against the negative consequences of the economic
reforms. The centre-right Civic Democratic Party (ODS), for example, in govern-
ment between 1992 and 1997, employed social compensation mechanisms to
maintain public support (Orenstein 2001). On the other hand, the Social
Democratic (CSSD) minority government after 1998, despite having run on a pro-
welfare state electoral platform, did not have enough political clout to implement
major social policy reforms (True 2003: 15).
Sometimes referred to as the ‘Czech miracle’, the economic reform appeared to
be the most successful in the region during the first half of the 1990s. In 1996, for
example, unemployment was 3.5 per cent compared to 15.7 per cent in Poland.
Inflation was only 8.8 per cent compared to 19.9 per cent in Poland. But bank
failures and financial scandals ushered in a three-year recession during 1997–99,
accompanied by serious political turmoil. During this time, the Czech Republic
was lagging behind most transformation countries in annual GDP growth. By July
1998, unemployment had surpassed 6 per cent for the first time since 1989, with
wide regional differences. It was as high as 15.6 per cent in the hardest-hit locali-
ties. Almost 60 per cent of those unemployed were women and about a third
young people. Women’s unemployment has been consistently higher than men’s
throughout the decade of the 1990s, with particularly high rates of long-term
unemployment.
The fiscal deficit widened in 2001 to 3.2 per cent of GDP as compared to 1.6 per cent
in 1999. An austerity programme therefore foresaw cutbacks in state spending. In
2000 and 2001, the Czech economy recovered to above 3 per cent growth on aver-
age. The unemployment rate, however, has been in the 8.5 to 9 per cent range
since 1999 and above 9 per cent in 2002.
Czechoslovakia, as Poland, entered the 1990s with a comprehensive social
security system. Yet since the old system was inadequate to address some of the
new needs emerging from the transition, such as unemployment and poverty,
social security reforms were supported across the political landscape. Early
reforms included the introduction of a household subsistence minimum, which
serves as a poverty benchmark, and the formulation of cost-of-living increases
of benefits (Kucharová et al. 2003). Two new family benefits were introduced in
1990, the state compensatory allowance – intended to buffer the negative effect
of consumer price liberalization – and a parental allowance for at-home care of
children.14 Other early reform steps included the revision of the maternity
benefit formula, and the decoupling of the child allowance from employment.
Instead, it was made dependent on the number of children in a family, and
their age.
A major reorganization of social security happened in 1995–96. Benefits were
grouped into three subsystems: social insurance, social assistance, and state social
support. Most family benefits are provided as state social support, with parental
allowance, child allowance and social allowance forming 85 per cent of total state
social support spending (Kucharová et al. 2003). State social support benefits are
financed from general revenues and paid out through the district authorities.
Some of the benefits are income-tested, but income-testing is not applied in a very
restrictive manner so far (see Table 3.2).
Table 3.2 State social support benefits in the Czech Republic as of

1995/96
Non-income tested benefits Income-tested benefits
Parental allowance Child allowance

Maintenance allowance15 Social allowance
Foster care allowance Housing allowance
Birth grant Transport allowance (until 2004)
Funeral grant
Source: Kucharová et al. 2003: 116.

Comparative reform trends in both countries16
Several broad effects of the family benefit reforms became clear at the end of the
1990s. First, aggregate expenditure for family benefits in both countries declined
over the course of the 1990s:17 In Poland during the period 1990–98, from 1.7 to
1.06 per cent of GDP; and in the Czech Republic during the period 1996–2002,
from 1.6 to 1.2 per cent of GDP (Fultz and Steinhilber 2004). See Figure 3.1 and
Table 3.3.
It is noteworthy that in Poland, during most of the 1990s, the share of family
benefits in social expenditure declined, while total social expenditure increased.
This suggests that family benefits were more affected by cuts than other social
security benefits, for example pensions (Wóycicka et al. 2003: 207). Further
savings in family benefits were achieved by reducing benefit adjustments for
inflation. Consequently, the share of family benefits in household income
declined significantly (Wóycicka et al. 2003: 195).
A second effect was a declining number of beneficiaries and concentration of
benefits on low-income households. In the Czech Republic, the proportion of
families receiving the three most important family benefits fell by half. At the
same time, those still eligible for all three benefits experienced increases between
2.5 2.3
2.3 2.1
2.0
1.7 1.7
1.5
1.3 1.1 1.07
1.0 1.06
0.5
0
1990 1991 1992 1993 1994 1995 1996 1997 1998
Figure 3.1 Family benefits as percentage of GDP, Poland, 1990–98
Table 3.3 Three family benefits, total spending and spending as a percentage of GDP, the
Czech Republic, 1996–2000 (CZK, thousands)
Benefit 1996 1997 1998 1999 2000 2001 2002
Child allowance 12.19 12.50 12.49 12.47 12.75 12.80 13.35

Social allowance 6.24 6.22 6.27 6.25 6.20 6.00 6.27
Parent allowance 7.36 7.61 7.78 7.72 7.69 7.70 8.02
Total of three 25.79 26.33 25.54 26.44 26.64 26.50 27.64
benefits
Gross domestic 1 572.30 1 668.80 1 798.30 1 833.00 1 910.60 2 175.20 2 275.60
product (GDP)
GDP accounted for by 1.64% 1.58% 1.47% 1.44% 1.39% 1.22% 1.21%
the three benefits
Source: Government of the Czech Republic, quoted in Fultz and Steinhilber 2004: 260.
3 and 10 per cent of net family income (Fultz and Steinhilber 2004; Kucharová et al.
2003: 125).
In Poland, by 1998, family benefits were significantly more important for low-
income households, as compared to higher-income households. Family benefits
represented 12 per cent of household income for families in the lowest income
decile, but just 0.15 per cent for those in the top decile. The numbers of beneficia-
ries of employment-related benefits and of recipients of the family allowance
declined drastically, between 30 and 60 per cent, depending on the benefit scheme.
The decline is explained by high unemployment (in the case of employment-related
benefits), stricter benefit entitlements and the sharp decline in birth rates.
While on the whole fewer families receive benefits in Poland, the one benefit
which most directly serves women and their children, the alimony benefit, was also
the one with dramatically increased numbers of beneficiaries throughout the
1990s. Initially, the income-test for the benefit had been waived in 1989, prompt-
ing increased numbers of beneficiaries (Wóycicka et al. 2003). Yet, even after the
reintroduction of income-testing, the numbers of beneficiaries increased, as did
overall spending. Between 2001 and 2002 the fund saw a 6.1 per cent increase of
beneficiaries, and a 16 per cent spending increase (Ministry of Economy, Labour
and Social Policy 2003: 44). While this is an indication of the precarious situation
of single mothers in the country, the alimony fund was abolished in 2004,
sparking a wave of protest among women all over the country.18
Third, evidence suggests a gap between de jure entitlements to family benefits,
and their practical use and take-up (Heinen and Portet 2002; Steinhilber 2003).
There are many accounts of employers discriminating against women workers
with family responsibilities, and there is evidence that women in both countries
do not fully use their rights for fear of reprisals by employers, such as job-loss or
unfavourable reassignment (Fultz and Steinhilber 2004: 259).19
The politics of gender and family benefit reforms compared
While Poland and the Czech Republic undertook profound social policy reforms
since 1989, gender equality was not a high-priority topic in either country. For the
Czech Republic, Eermáková et al. conclude that at least until the end of 1997 –
when the Department for Equality of Men and Women was established – ‘women’s
issues were the least of governmental and state administration concerns’
(Eermáková et al. 2000: 31). Similarly, Wóycicka et al. argue that the family
policies implemented over the course of the 1990s range from ‘passive to harmful’
in their effect for gender equality in the reconciliation of parental and work
responsibilities (Wóycicka et al. 2003: 221).
Social policy reform debates in both countries did, however, focus widely on the
family and highlight the importance of combating family poverty. The family was
addressed as ‘the most natural social group which should (…) ensure the care for
its members and form their material and intellectual needs from childhood until
old age’ (Prusa 1994, quoted in Eermáková et al. 2000: 42). Since families were
regarded so important and were assigned increased responsibilities in the reform
process, supporting them to cope was considered eminently important. The gen-
dered nature of family relations, however, and the impact of family benefits on
them did not become an issue for public debate.
The proclaimed family focus of reforms matched popular demand. Before 1989,
support for families had been a constitutive part of state responsibilities, and was
regarded by citizens as at least partial compensation for the inadequacies of the
planned economy. The importance of family networks for day-to-day life in an
economy of scarcity has been highlighted (Veeerník 1995; Veeerník and Matßju
1999; Marksová-Tominová 2003; Seibert 2001; Cichocinska 1993; Giza-Poleszczuk
1992). In continuation, many citizens called for state-support to families as a matter
of social justice after 1989 (Steinhilber 2003; UNDP / Charles University Prague 2003).
In this climate, women’s needs were addressed indirectly and conceptualized in
stereotypical gender terms. Women benefited as members of other social groups,
such as members of low-income households or the unemployed. Echoing the
strong norm of motherhood and the importance of women’s role as caregivers in
both countries in the period after 1945, women today continue to be addressed as
mothers, possibly also as heads of households, but not as individuals or economic
actors (Kucharová et al. 2003: 134).
Changes in entitlement criteria have shifted the status of beneficiaries from
‘holders of personal rights to petitioners of the state’, required to prove their need
(Fultz and Steinhilber 2004: 260). Because of the pervasive practice of means-testing
of benefits and the contraction of the state in social security provision, this effect
is more pronounced in Poland than in the Czech Republic. Yet a renewed debate
about stricter means-testing, for example of the child allowance, in the Czech
Republic in 2004 (Kucharová 2004), illustrates that means-tested benefits are more
vulnerable to changing political conjunctures.
During the reforms, a debate about ‘good motherhood’ re-emerged in both
countries. Notably, good mothers were not referred to as mothers who care for
their children while also being employed. Instead, motherhood was proclaimed as
full-time dedication. In the Czech Republic, income restrictions associated with
the parental allowance exemplify this. While receiving the parental allowance, a
mother should not maintain her links to the labour market. A proposal of the gov-
erning Social Democrats to lift the income restrictions was countered by conserv-
atives who were concerned about Czech women becoming ‘bad mothers’ if the
restrictions were lifted. Women would then focus on making money instead of
caring for children (Marksová-Tominová 2003: 56).
In Poland, the strength of Catholic family values and the traditional image of
the ‘Polish Mother’ have also shaped the debate on ‘motherhood’ and family ben-
efits (Plakwicz 1992; Seibert 2001; Heinen and Portet 2002). Here, religious values,
the church and faith-based organizations play an important role in private and
political life (Inglehart and Norris 2003). Since 1989, abortion and women’s repro-
ductive rights have remained on the political agenda, and religious values have
influenced politics widely across party lines. The Church exerts particularly strong
influence on some areas of social policy, such as abortion, and remains a powerful
advocate for traditional family values and gender roles (Einhorn 1993; Heinen and
Matuchniak-Krasuska 1995; Zielinska 2000). The Church gained particular influence

during the right-wing government under Solidarity Electoral Action (AWS)
1997–2001.
The introduction of a new ‘benefit for pregnant women and women raising
children’ in 1993 is one piece of the Polish ‘motherhood’ debate; it also illustrates
its bitter ironies for women. The benefit was introduced jointly with the
infamously strict anti-abortion legislation. In practice, thus, the benefit offered
limited economic support to women who were deprived of their reproductive self-
determination. However, as expenditure dramatically exceeded estimations, both
the duration and the level of the benefit were soon cut, in two steps in 1994 and
2002 (Wóycicka et al. 2003: 199). While the anti-abortion legislation was still not
liberalized, the benefit level and duration were again significantly cut in 2002, this
time under a left-wing government. The example shows the fragility and superfi-
ciality of respect for women’s needs, which can easily be overturned by fiscal
priorities.
Government changes in Poland created a back-and-forth movement of family
policy-reform projects and proposals. The overarching goal of fiscal austerity, how-
ever, has guided both right- and left-wing governments. Right-wing governments
more clearly tried to instrumentalize family benefits to boost a traditional notion of
women as mothers and caregivers. It was also under right-wing governments that
concerns about declining birth rates and desires to control women’s reproductive
behaviour acquired particular importance in Polish social policy debates (Zielinska
2000). However, left-leaning governments have not been explicitly committed to
promoting gender equality either. Many opinion-leaders of the left have used a
less openly moralistic and conservative gender discourse, yet their eyes were set on
fiscal austerity rather than women’s rights as well.20
In the Czech Republic, in turn, shifts in political power and the weakness
of elected governments based on small (or no) majorities have led to a moderation
of reform proposals and political compromises, or to reform standstill (Orenstein
2001). Radical proposals, for example the introduction of stricter means-testing of
the child allowance, were softened in the political process. Moderate means-testing
may be seen as a small gain only, however, since social democrats initially had
strongly argued in favour of a universal child allowance (Marksová-Tominová
2003: 58). Yet observers argue that the practice of political compromise orches-
trated by relatively weak political powers has been one of the reasons for a wide
acceptance of the social policy reforms in the country (UNDP/ Charles University
2003: 63). The need for – and political practice of – compromise, however, has sus-
tained the sidelining of gender equality in social policy.
While not explicitly aiming at increased gender equality, both countries
increased formally equal treatment of women and men in access to family
benefits. Provisions of pre-existing laws that had barred fathers from using bene-
fits or had imposed stricter rules on them were eliminated (Fultz and Steinhilber
2004: 255). In Poland, equal treatment in the eligibility to childcare benefit and
child-raising allowance was guaranteed from 1995 and 1996, respectively. In the
Czech Republic, eligibility to the parental allowance was equalized in 1990, and
employment protection for men finally introduced in 2001. In an exceptional

move to direct a measure specifically at men, Poland extended two weeks of the
maternity leave and benefit to fathers.
Equal access to benefits, however, has not changed the entrenched gender roles,
nor has it affected the take-up of benefits. Available data, though incomplete,
suggest that fathers do not use family benefits in greater numbers than before.
In the Czech Republic during the period 1995–98, fathers represented roughly
0.75 per cent of recipients of childcare benefits, increasing to 0.9 per cent in 2003
(Kucharová et al. 2003: 130, OSI Network Women’s Program 2005b). In Poland,
statistics on the use of childcare benefits by men and women are not available; in
the Czech Republic, they are not included in the officially published statistics –
indicative for the lack of interest in gender trends on the side of public adminis-
tration in both countries.
To date, there has been no debate about ‘good fatherhood’ in either country, or
about fathers combining employment and family responsibilities. In both coun-
tries, measures to support parents in combining both responsibilities have
received only secondary attention during the social security reforms. The reform
of the Czech parental allowance is illustrative: designed as a universal benefit for a
full-time home carer, the parent was, in principle, allowed to work, but faced
severe earnings restrictions. Moreover, the child could be placed in childcare for
no more than three (later raised to five) days a month. This made continuing
education as well as job search – much less paid work – very difficult (Marksová-
Tominová 2003; Eermáková et al. 2000). The income restriction was frequently
violated, reflecting a gap between policy-makers’ intentions and women’s day-to-day
reality and needs. The income restrictions were finally lifted in 2003 (valid from
2004), after intense debates and lobbying efforts of women.21
Progressive voices in favour of gender equality have been institutionally weak
and often politically fragmented in both countries. For much of the 1990s,
women’s organizations struggled to become recognized political actors. Under
conditions of radical political change and in an ‘associational wasteland’ (Offe
1993), civil society organizations, among them women’s organizations, needed
to gain strength and stability first, while lack of funding and an inadequate
legal framework created severe obstacles (Eermáková et al. 2000). At the
same time, gender issues were marginalized politically and faced strongly nega-
tive public attitudes as issues associated with the past (Eermáková et al. 2000:
33–4). The purported women-friendliness of state socialism served to ‘delegitimate
women’s political activity in postcommunism’, as Gal and Kligman argue
(2000).
Women’s organizations, while becoming stronger and more numerous over the
1990s, were often concerned with topics other than family benefits, for example
the very low political representation of women after the first post-socialist
elections and employment, where women’s losses through the transition were
immediately felt strongly. In addition, in Poland reproductive rights motivated
women’s organizations strongly throughout the decade, crystallizing in major
debates in 1989 and 1992/1993. Though not successful in stopping the anti-abortion
legislation, the debate itself contributed to a strengthening of the organizational

networks of progressive organizations in the country.22 While social security was
not a central concern anyway, the ‘packaging’ of family benefit reforms caused
further problems for feminist social policy discussions: Because the benefit for
pregnant women and mothers was introduced in conjunction with the law on
abortion, progressive women’s activists had to find a way to argue for family ben-
efits without defending the abortion law. In turn, conservative and Catholic
women’s organizations clearly benefited from this. Only the discontinuation of
the alimony fund in 2004 instigated a short-lived broader coalition of women’s
organizations on a social security issue in Poland.
Women’s weak institutional representation at governmental and administrative
levels in both countries further contributed to the lack of attention to gender
issues during the reforms. It was not until the end of 1997 that a Department for
the Equality of Men and Women was established in the Czech Ministry of Labour
and Social Affairs. Therefore, a number of observers agree that institutional
progress came mainly in response to the requirements imposed on the country as
part of the EU accession negotiations, and argue that the state’s attitude toward
gender issues has been formal rather than indicative of a substantive commitment
to gender equality (Eermáková et al. 2000; Marksová-Tominová 2003).23
In Poland, no gender-equality institution existed during the period of the most
profound social reforms. The institutional representation of women and gender
equality experienced a number of turnabouts during the decade. An early proposal
for an Equal Status Act was initiated in 1992 by feminist groups and carried
forward by the Parliamentary Women’s Group. It proposed the creation of a
Commissioner for Equal Status.24 While the ensuing several-years-long debate did
not result in the enactment of the legislation, at least it helped the formation and
public presence of women’s organizations (Regulska 1998). Between 1992 and
1995, the post of the Plenipotentiary for Women and Family was vacant. After the
election victory of the right in 1997, the office was replaced by the Plenipotentiary
for Family Affairs, explicitly entrusted with the task of promoting a traditional fam-
ily model in line with the ‘family oriented policy of the state’ (ILO 1999). Again,
after a change in government in 2001, and in response to strong lobbying efforts by
women’s organizations, the Plenipotentiary for the Equal Status of Women and
Men was established. Vocal public resistance helped avert a weakening of the insti-
tution in 2004, yet it was immediately abolished by the new government in 2005.
While the institutional representation of gender equality within governmental
structures in both countries is ensured today, de facto, these institutions do not
focus on gender equality in family policy. This is so despite the fact that in prin-
ciple, social security is in their mandate. For example, the Polish
Plenipotentiary was entrusted with coordinating all activities to secure equal
opportunities for women and men. Yet, for example, it has not played a visible
role in the debates about the 2004 reform of the family benefit system.
Similarly, the Czech ministerial department has not dedicated itself to social
security issues either.
Concluding remarks
In the face of similar challenges to construct a market economy and provide social
protection to the citizens, post-socialist Czech Republic and Poland have opted for
different social policy reform paths. These differences have in turn had important
implications for gender equality, a topic which has received only scant, if any,
attention in both countries. During the 1990s Poland moved further than other
post-socialist countries in the direction of a residualist and familial model of a wel-
fare state by cutting, individualizing and familializing benefits. The central state’s
role in social provisioning was reduced massively by shifting responsibilities to the
local administration and individual citizens and their families. The Czech
Republic, in turn, where reforms often were justified in similar ways as in Poland,
has maintained a greater role for the state in the provision of social welfare, and
has upheld important elements of the inherited redistributive welfare structures.
The individualizing and familializing effect of Czech reforms has thus been more
moderate than in Poland.
In the international economic reform climate of the 1990s, characterized by a
clear focus on macroeconomic stability and fiscal austerity, budgetary concerns
have had a strong presence in social policy debates. The Czech Republic has seen
itself under weaker constraints in this respect than the highly indebted Poland. Yet
fiscal pressures as such do not explain where budgets are cut: the analysis of fam-
ily benefit reforms illustrates that economic reform ideologies and political
dynamics and power struggles interlink with family and gender ideologies of
reform actors. In the case of Poland, for example, family benefits suffered greater
losses than other social security benefits.
Despite the differences in reform approaches and their gendered results, family
benefit reforms in both countries show a number of important commonalities.
Most importantly, family benefits in both countries continue to embody a conser-
vative family model. Rarely during the post-1989 reforms were family benefits
referred to by policy-makers as instruments to advance gender equality. Moreover,
the discourse on family benefits has not included ‘non-standard’ families as legit-
imate family forms. Also, no incentives have been developed so far to stimulate
change in men’s gender roles, or to support the de-gendering of care work.
Policy-makers in both countries have upheld a longstanding tradition of seeing
family benefits as instruments to redress declining population growth – a practice
already well in use during socialist times, with no empirically clearly proven effect.
Pronatalism through family benefits, accompanied by strict anti-abortion policies, is
more pronounced in Poland, but is a topic on the rise in the Czech Republic as well.
While the analysis has illustrated the continuities with state socialist family policy,
it also shows a profound shift in the nature of family benefits and in their justifi-
cation since 1989. In the past, family benefits served the dual role of supporting
women in combining paid work and family responsibilities, while also forming
part of the state-sponsored system of income support and redistribution. Despite
women’s continued attachment to the labour market, even more so in the Czech
Republic than in Poland, women today no longer benefit from such a well-
furnished benefit structure. Therefore, today women face greater difficulties in
combining employment and family responsibilities and suffer from a widening
gap between entitlements in law and their practical value.
To different degrees, post-socialist governments have withdrawn from the
tradition of state-organized income-distribution and welfare provision, now
prioritizing the reduction of poverty through means-tested benefits while limiting
state intervention and emphasizing subsidiarity, i.e. the provision of support at
the lowest possible level. In fact, this often means reliance on women’s unpaid
family work. While similar in its direction, this change has been more brisk in
Poland than in the Czech Republic.
As a consequence of reforms, the avenue to access social rights for women has
become more narrow, and women’s need for state support is interpreted differently
today. A few benefits are still linked to employment – yet finding and keeping
employment has become more difficult under conditions of a market-economy.
Entitlement to other key family benefits is based on fulfilling a traditional gender
role as mother and caretaker under precarious economic circumstances. Particularly
those women who do not live in low-income households have therefore lost their
entitlement to benefits. This policy is questionable from the point of view of
gender equality: to enhance gender equality, support for women as a matter of
individual right is as important as support for combining the dual roles in the family
and the world of work. Both need to be accompanied by targeted interventions to
promote a shift in gender roles through policies specifically addressed at men.
Reforms in Poland and the Czech Republic have not moved toward this goal so far.
The legacy of past benefit systems, economic and political shifts since 1989, and
power dynamics between key reform actors and their gendered values and
ideologies constitute a multifaceted web in which gender is being negotiated in
the daily lives of women and men. In the name of ‘women’s interests’ and ‘needs’ –
but without their adequate political representation and participation in the
decision-making process – family benefit reforms have impacted women’s and
men’s employment and life choices. Reforms have created new challenges to
women’s economic independence, public presence and reproductive options,
while reform discourses have invoked a traditional gender structure, often in
contradiction to women’s day-to-day reality of full employment plus unpaid care
responsibilities. The durable conservatism with respect to gender roles and values
in both countries, together with the political marginalization of women, have
come to bear very forcefully on the reform processes since 1989. Despite the many
differences, it is here that both countries show many similarities – and are not very
different from their neighbouring countries, and some (fellow) EU member states.
Acknowledgement
The author is currently supported by the Berlin Programme for Equal Opportunities for
Women in Academia (Berlin Programm zur Förderung der Chancengleichheit für Frauen in
Forschung und Lehre).
Notes
1 The terms ‘real-existing socialism, ‘real socialism’ and ‘state socialism’ are used here
interchangeably, to distinguish the Central and Eastern European regimes from the
ideals of socialism. See True 2003: 181.
2 For earlier gender analyses of post-socialist social reforms, see, for example, Einhorn
1993, Funk and Mueller 1993, Aslanbeigui et al. 1994, Kligman 1994, Lobodzinska 1995.
3 The ironies of a discourse on greater individual responsibility which in fact hands
women the responsibility of picking up the slack have repeatedly been pointed out by
feminists.
4 See also Chapter 1 of this volume.
5 For a more extensive discussion, see Fodor 2004.
6 Poland ratified ILO Convention 100 on equal pay for equal work and work of equal
value (1953) in 1954, the Czech Republic in 1957.
7 Unless noted otherwise, data before 1993 used here relate to Czechoslovakia, thereafter
to the Czech Republic.
8 On the problems and contradictions within state-socialist social policy, see Castle-
Kanerová 1992, Millard 1992, Deacon 1992.
9 Before 1989, maternity and childcare leave together were regularly termed ‘maternity
leave’ or ‘extended maternity leave’. Here, ‘maternity leave and benefits’ is used in the
sense of ILO Convention 183 as related to the protection of mother and child in relation
to the child’s birth. ‘Childcare leave’/ ‘parental leave’ is granted for a temporary absence
from employment to raise a child. Childcare benefits may or may not be linked to leave.
Conversely, leave may or may not come with a cash benefit.
10 Maternity benefits to insured women existed from 1928, yet the law was not
consistently implemented until after World War II (Pavlik 1985).
11 Short-term sick-child leave.
12 Means-testing of alimony benefits was waïved in 1989 and reinstated in 1999.
13 The guaranteed periodic benefit is payable to a single parent who has exhausted her/his
right to unemployment benefits.
14 The maternity benefit is available to mothers for a maximum of 28 weeks (37 weeks in
case of multiple births). It is insurance-based, paid from sickness insurance, with a
replacement rate of 90 per cent of the net income, up to a ceiling (1990). While
maternity benefits are provided to fathers only under exceptional circumstances (death
or disability of the mother), fathers can receive the parental benefit for full-time care
directly following the birth of a child (maternity leave is not mandatory for women), but
the parental benefit is lower than the maternity benefit.
15 Allowance for the families of soldiers.
16 The following section is based on Fultz and Steinhilber 2004.
17 Note that different levels of aggregation of family income data by household size
complicate cross-country comparisons.
18 Further reforms foresaw some additional support for single mothers, boosting a unified
social assistance benefit.
19 New data from the Czech Republic indicates an increased take-up of parental leave by
women: 60 per cent now return to the labour market only when their child is three.
Worse labour market conditions, an increased interest in full-time care for children at
home, a small increase in fertility (second children) and the shortage of childcare facili-
ties may account for this (V. Kucharová, private communication, June 2005).
20 A much-quoted example is the strict anti-abortion legislation which was not relaxed
under the left-wing government either.
21 Restrictions about childcare use have been maintained. Only from 2006 children can
attend kindergarten for four hours per day without any impact on the parental allowance.
22 Periods of right-wing governments also contributed to the consolidation of conservative
and Catholic women’s organizations (M. Fuszara, personal conversation June 2005).
23 Interestingly, the ministerial department in charge of equal opportunities for women

and men was placed in the Department for the Preparation of the Integration of the
Czech Republic into the European Union.
24 The bill is the longest-discussed draft in the history of the Polish parliament. It has been
amended and debated in every four-year term since 1993 (OSI Network Women’s
Program 2005a).
Bibliography
Aslanbeigui, Nahid, Steven Pressman and Gale Summerfield (1994) Women in the Age of
Economic Transformation. Gender Impact of Reforms in Post-socialist and Developing Countries.
New York and London: Routledge.
Castle-Kanerová, Mita (1992) ‘Social Policy in Czechoslovakia’. In B. Deacon (ed.) The New
Eastern Europe: Social Policy Past, Present and Future. London: Sage.
Eermáková, Maria, Hana Hasková, Alena Krizková, Marcela Linková, Hana Mariková and
Martina Musilová (2000) Relations and Changes of Gender Differences in the Czech Society in
the 90s. Prague: Institute of Sociology of Academy of Science of the Czech Republic.
Chlon-Dominczak, Agnieszka (2004) ‘Gender Dimensions of Social Security Reforms in
Poland’. Presentation at the seminar, The Gender Dimensions of Social Security Reform in
Central and Eastern Europe: Case Studies of Poland, Hungary and the Czech Republic.
organized by the International Labour Office and Central European University, Budapest,
23 February 2004.
Cichocinska, M. (1993) ‘Gender Aspects of Dismantling the Command Economy in Eastern
Europe: The Case of Poland’. In V. Moghadam (ed.) Democratic Reform and the Position of
Women in Transitional Economies. Oxford: Clarendon Press.
Deacon, Bob (1992) ‘Eastern European Welfare: Past, Present and Future in Comparative
Context’. In B. Deacon et al. (eds) The New Eastern Europe. Social Policy Past, Present and
Future. London: Sage.
Einhorn, Barbara (1993) Cinderella Goes to Market. New York: Verso.
European Commission / Government of Poland. (2001) Joint Assessment of Employment
Priorities in Poland. http://europa.eu.int/comm/employment_social/employment_
analysis/japs/poland_en.pdf, accessed on 24 April 2005.
Fajth, Gaspar (1994) Family Support Policies in Transitional Economies: Challenges and
Constraints. Innocenti Occasional Papers, Economic Policy Series, No. 43. Florence: UNICEF.
Fodor, Eva (2004) ‘Women at Work: The Status of Women in the Labor Markets of the Czech
Republic, Hungary, and Poland’. Occasional Paper written for the preparation of the report
Gender Equality: Striving for Justice in an Unequal World. Geneva: UNRISD.
Fultz, Elaine and Silke Steinhilber (2003) ‘The Gender Dimensions of Social Security Reforms
in Central and Eastern Europe’. In E. Fultz, M. Ruck and S. Steinhilber (eds) The Gender
Dimensions of Social Security Reforms in Central and Eastern Europe. Case Studies of the Czech
Republic, Hungary and Poland. Budapest: ILO, Subregional Office for Central and Eastern
Europe.
Fultz, Elaine and Silke Steinhilber (2004) ‘Social Security Reform and Gender Equality: Recent
Experience in Central Europe’. International Labour Review, Vol. 143, No. 3: 249–73.
Funk, Nanette and Magda Mueller (eds) (1993) Gender Politics and Post-Communism.
Reflections from Eastern Europe and the Former Soviet Union. New York and London:
Routledge.
Gal, Susan and Gail Kligman (2000) The Politics of Gender after Socialism. Princeton NJ:
Princteon University Press.
Giza-Poleszczuk, A. (1992) ‘Familism and System Transformation’. Polish Sociological Bulletin,
Vol. 3, No. 4: 315–27.
Heinen, Jacqueline and A. Matuchniak-Krasuska (1995) ‘Abortion in Poland: A Vicious Circle
or a Good Use of Rhetoric’. Women’s Studies International Forum, Vol. 18, No. 1: 27–33.
Heinen, Jacqueline and Stephane Portet (2002) ‘Political and Social Citizenship: An
Examination of the Case of Poland’. In M. Molyneux and S. Razavi (eds) Gender Justice,
Development and Rights. Oxford: Oxford University Press.
Ingelhart, Ronald and Pipa Norris (2003) Rising Tide: Gender Equality and Cultural Change
around the World. Cambridge: Cambridge University Press.
International Labour Office (ILO) (1984) ‘Family Benefit’. In Introduction to Social Security,
Chapter 11. Geneva, Switzerland, 1984.
International Labour Office (ILO) (1999) Information communication by the ILO national
correspondent in Poland on the Bill on Equal Status of women and men and general pre-
sentation of the situation of women. Internal Communication. Warsaw, 22 April 1999.
Kligman, Gail (1994) ‘The Social Legacy of Communism: Women, Children, and the
Feminization of Poverty’. In J. Millar and S. Wolchik (eds) The Social Legacy of Communism.
Washington DC and New York: Woodrow Wilson Center and Cambridge University Press.
Kucharová, Vera (2004) ‘Gender Dimensions of Social Security Reforms in the Czech
Republic’. Presentation at the seminar, The Gender Dimensions of Social Security Reform
in Central and Eastern Europe: Case Studies of Poland, Hungary and the Czech Republic.
Organized by International Labor Office and Central European University. Budapest,
23 February 2004.
Kucharová, Vera, M. Kotynková and Ladislav Prusa (2003) ‘The Gender Dimensions of Social
Security Reform in the Czech Republic’. In E. Fultz, M. Ruck and S. Steinhilber (eds) The
Gender Dimensions of Social Security Reforms in Central and Eastern Europe. Case Studies of the
Czech Republic, Hungary and Poland. Budapest: ILO, Subregional Office for Central and
Eastern Europe.
Lavigne, Marie (1999) The Economics of Transition – From Socialist Economy to Market Economy.
New York: St. Martin Press.
Lewis, Jane (ed.) (1993) Women and Social Policies in Europe. Work, Family and the State.
Aldershot Edward Elgar.
Lewis, Jane (1997) ‘Gender and Welfare Regimes: Further Thoughts’. Social Politics:
International Studies in Gender, State and Society, Vol. 4, No. 2: 160–77.
Lobodzinska, Barbara (1995) Family, Women and Employment in Central-Eastern Europe.
Westport, CN: Greenwood Press.
Lohmann, Kinga and Anita Seibert (eds) (2003) Gender Assessment of the Impact of EU
Accession on the Status of Women in the Labour Market in CEE. National Study: Poland.
Warsaw: Karat Coalition.
Marksová-Tominová, Michaela (ed.) (2003) Gender Assessment of the Impact of EU Accession on
the Status of Women in the Labour Market in CEE. National Study: Czech Republic. Karat
Coalition. Prague: Gender Studies Center.
Marody, M. and A. Giza-Poleszczuk (2000) ‘Changing Images of Identity in Poland: From the
Self-Sacrificing to the Self-Investing Woman?’ In S. Gal and G. Kligman (eds) Reproducing
Gender: Politics, Publics, and Everyday Life after Socialism. Princeton NJ: Princeton University
Press.
Marx Ferree, Myra (1993) ‘The Rise and Fall of ‘Mommy Politics’: Feminism and Unification
in (East) Germany’. Feminist Studies, Vol. 19, No. 1: 89–115.
Millard, F. (1992) ‘Social Policy in Poland’. In B. Deacon (ed.), The New Eastern Europe: Social
Policy Past, Present and Future. London: Sage.
Ministry of Economy, Labour and Social Policy (2003) Poland 2003: Report on Labour Market
and Social Security. Warsaw: Ministry of Economy, Labour and Social Policy.
Offe, Claus (1993) ‘The Politics of Social Policy in East European Transitions. Antecedents,
Agents, and Agenda of Reform’. ZeS, Arbeitspapier No. 2/93, Bremen.
Open Society Institute (OSI) Network Women’s Program (2005a) Your Sex Matters in Poland.
Equal Opportunities for Women and Men. Monitoring Law and Practice in New Member
States and Accession Countries of the European Union. Fact Sheet. http://www.
soros.org/initiatives/women/articles_publications/publications/equal_20050502/
poland.pdf, accessed on 1 July 2005.
Open Society Institute (OSI) Network Women’s Program (2005b) Your Sex Matters in the Czech
Republic. Equal Opportunities for Women and Men. Monitoring Law and Practice in New
Member States and Accession Countries of the European Union. Fact Sheet.
http://www.soros.org/initiatives/women/articles_publications/publications/equal_200505
02/czechrep.pdf, accessed on 1 July 2005.
Orenstein, Mitchell (2001) Out of the Red: Building Capitalism and Democracy in Postcommunist
Europe. Ann Arbor: University of Michigan Press.
Pavlik, Z. (1985) ‘Czechoslovakia’. In V. Bodrova and R. Anker (eds) Working Women in
Socialist Countries: The Fertility Connection, Geneva: ILO.
Plakwicz, J. (1992) ‘Between the Church and the State: Polish Women’s Experience’. In
C. Corrin, Super Women and the Double Burden. London: Scarlet Press.
Regulska, Joanna (1998) ‘The ‘Political’ and its Meaning for Women: Transition Politics in
Poland’. In J. Pickles and A. Smith (eds) Theorizing Transition: The Political Economy of Post-
Communist Transformation. New York: Routledge.
Research Institute for Labour and Social Affairs (RILSA)/ United Nations Development
Programme (UNDP) (1999) Human Development Report: Czech Republic 1999. Prague: RILSA.
Sainsbury, Diane (ed.) (1999) Gender and Welfare State Regimes. Oxford: Oxford University
Press.
Seibert, Anita (2001) ‘From Matka Polka to New Polish Woman: Women and Restructuring in
Poland’. Unpublished PhD thesis, Monash University, Clayton, Victoria, Australia.
Siemienska, Renata (1983) ‘Women and the Family in Poland’. In E. Lupri (ed.) The
Challenging Position of Women in Family and Society. Leiden: E. J. Brill.
Steinhilber, Silke (2003) ‘Women’s Views on Social Security Reform: Qualitative Survey’. In
E. Fultz, M. Ruck and S. Steinhilber (eds) The Gender Dimensions of Social Security Reforms in
Central and Eastern Europe: Case Studies of the Czech Republic, Hungary and Poland. Budapest:
ILO. Subregional Office for Central and Eastern Europe.
True, Jacqui (2003) Gender, Globalization and Postsocialism: The Czech Republic after
Communism. New York: Columbia University Press.
United Nations Children’s Fund/Monitoring Eastern Europe Project (UNICEF/MONEE)
(1999) Women in Transition. Florence: UNICEF/ MONEE.
United Nations Development Programme (UNDP)/ Charles University Prague (2003) Human
Development Report. Where Do We Come From, What Are We, and Where Are We Going?
Prague: UNDP and Charles University.
Veeerník, Jiri (1995) ‘Household Incomes and Social Policies: The Czech Republic in the
Period 1989–1995’. SOCO Project Paper No. 25. Vienna.
Veeerník, Jiri and Matfju, Petr (1999) Ten Years of Rebuilding Capitalism: Czech Society After
1989. Prague: Academia.
Wolchik, S. and A. Meyer (eds) (1985) Women, State and Party in Eastern Europe. Durham NC:
Duke University Press.
Wóycicka, Irena, Bozena Balcerzak-Paradowska, Agnieszka Chlon-Dominczak, Irena
Kotowska, A. Olejniczuk-Merta and Irena Topinska (2003) ‘The Gender Dimensions of
Social Security Reform in Poland’. In E. Fultz, M. Ruck and S. Steinhilber (eds) The Gender
Dimensions of Social Security Reforms in Central and Eastern Europe: Case Studies of the Czech
Republic, Hungary and Poland. Budapest: ILO, Subregional Office for Central and Eastern
Europe.
Zielinska, E. (2000) ‘Between Ideology, Politics, and Common Sense: The Discourse of
Reproductive Rights in Poland’. In S. Gal and G. Kligman (eds) Reproducing Gender: Politics,
Publics, and Everyday Life after Socialism. Princeton NJ: Princeton University Press.
4
Maternalist Policies versus
Women’s Economic Citizenship?
Gendered Social Policy in Iran
Valentine M. Moghadam
That social policy is gendered is now a truism in the feminist literature. But what
are the different patterns of (gendered) social policy formation and implementa-
tion across regions in the world economy? In the Middle East, for example, what
are the state policies, practices and institutions that directly influence the welfare
and security of various groups within a particular society, especially as far as
women are concerned? And if social policy is partly about protecting against risks
and contingencies while also providing social equity, do all social policies serve to
expand women’s citizenship? In this chapter I examine the gendered nature of
social policy in Iran, with a focus on the problems associated with women’s
economic citizenship that have resulted from Iran’s economic structure and its
cultural/ideological institutions. An oil-dependent economy and gender relations
codified in Muslim family law have had implications for women’s access to
employment and economic resources, as well as overall citizenship.
The literature on social policy is overwhelmingly concentrated on the experi-
ences and models of the welfare state in advanced capitalist countries, and in par-
ticular Europe and North America. The feminist component of this literature
consists largely of historical investigations of maternalist politics and policies in
the United States,1 or comparative studies of gender and welfare regimes in the
core countries,2 or studies of the distribution of care and the problem of employed
women’s double burden.3 Gosta Esping-Andersen’s typology of the three worlds of
welfare capitalism and T. H. Marshall’s theory of the evolution of civil, political,
and social rights of citizenship often conceptually frame these investigations.4 The
emerging literature on social policy in a development context suggests that East
Asia, among other regions, might represent a different experience and model
(Kwon 2005). The present chapter expands the empirical understanding of
gendered social policy in a development context, while also proposing a distinc-
tive experience and model based on the characteristics and consequences of an
oil-dependent economy and Muslim family law in one Middle Eastern country.
Elsewhere I have argued that in contrast to the rights-based model of the Nordic
welfare state or the developmentalist East Asian experience, social welfare
87
88 Valentine M. Moghadam
provisioning in the post-colonial Middle East and North Africa (MENA) was tied to
the imperatives of state-building and legitimation, was financed by the regional
oil economy, and reflected and reinforced a ‘patriarchal gender contract’ premised
on women’s family roles.5 The MENA state’s ‘developmentalism’ was undermined
by its ‘neopatriarchal’ approach to women, gender, and the family, codified in
family laws/personal status codes that define women largely in terms of their fil-
ial, marital and maternal roles, place them under male guardianship, and deny
them equality in access to family wealth. Although state expansion and social
expenditures did allow for social mobility and access by some women to education
and employment, on the whole oil wealth and family laws served to prevent
women’s ‘labour commodification’ in the MENA region, in contrast to other
regions in the world economy. Through most of the period between 1950 and
1990, the MENA model allowed for generous if non-inclusive social provisioning
and high wages, and provided the (rather small) salariat with generous benefits. But
in the subsequent period, the family wage was undermined by economic crisis and
then by the neoliberal prescription, while the reinforcement of Muslim family law
became the focus of Islamist groups (Moghadam 1998; 2003, ch. 2).
These are early days in the study of social policy in the Middle East. The grow-
ing literature on poverty, the limited work on pensions and social security reform
in Arab countries, and the more extensive literature on socio-economic develop-
ment lack attention to the gender dynamics of social policies.6 The focus of the
emerging literature on women’s citizenship in MENA is on the problems of
women’s civil and political rights of citizenship, and deals only minimally with
social rights associated with employment, or with economic citizenship.7 On the
other hand, family law is a social policy, and a highly gendered one that has
implications for women’s access to employment and income, although the large
and important literature on family law has not explored the links between family
law, women’s economic activity, and social development.8
In the sections that follow, I begin by elaborating on this chapter’s theoretical
framework, including the concepts of economic citizenship, maternalist politics,
the rentier and neopatriarchal state, Muslim family law, and the patriarchal gender
contract. I then turn to the gender/social policy regime in Iran, starting with a
brief historical overview and drawing attention to continuities and discontinuities
before and after the 1979 Islamic revolution. I argue that the state remained
rentier and neopatriarchal after the revolution, but a key difference was the
strengthening of Muslim family law and a state-driven maternalist discourse and
politics. The paper ends with reflections on the future of Iran’s gender/social
policy regime.
Conceptual overview
The central premise of a feminist approach to social policy is that the sexual division
of labour – which is influenced by the organization of economic production –
shapes policy decisions and results in different outcomes for women and
men. Marxist-feminist research has shown how women’s and men’s differential
roles in production and reproduction, and especially women’s greater responsibilities
Maternalist Policies versus Women’s Economic Citizenship? 89
for reproductive/domestic work, has rendered women a ‘reserve army of labour’

that could be marginalized, integrated or exploited according to the demands of
accumulation (Jenson 1986). Feminist social policy research has posited that early
welfare policies in Europe and the United States reflected the sexual division of
labour and were predicated upon the male breadwinner/female homemaker ideal
(Orloff 1993; Sainsbury 1996); as such, the welfare state is patriarchal (Pateman
1988). Other feminist scholars have emphasized the ‘women-friendly’ nature of
welfare regimes. The relationship between gender and social policy, therefore,
should be seen as complex and interactive. As Orloff notes, if gender relations
‘profoundly shape the character of welfare states’, it is also true that ‘the institu-
tions of social provision – the set of social assistance and social insurance programs
and universal citizenship entitlements to which we refer as “the welfare state” – affect
gender relations’ (Orloff 1996: 51).
The concept of women’s economic citizenship used in this chapter is adopted
from Alice Kessler-Harris (2001), who defines it as the right to hold a job of one’s
choice and the economic resources necessary to sustain an average household. In
her study of women’s quest for economic citizenship in the United States, Kessler-
Harris demonstrates that early social policy, including labour legislation,
envisioned men as primary breadwinners supporting families and viewed women
as wives and mothers. To the extent that women had economic rights, these
consisted largely of the right to be supported by a male breadwinner and protected
by the state in their mothering role (via the protective maternity provisions of
labour law). She also shows how the originally masculine rhetoric behind social
security gradually was forced to change when millions of American women joined
the labour force during World War II and in the decades afterwards. Kessler-Harris’
study confirms that across cultural and economic systems, many countries have
upheld marriage, motherhood and family as key social institutions. As in the
United States in the early part of the twentieth century, a patriarchal gender
contract holds sway in the MENA region, although it is beginning to unravel
under the force of new economic realities, demographic changes and women’s
activism.
Institutions of social provision may be women-friendly (or feminist) or they
may be predicated on the male-breadwinner/female-homemaker model (or patri-
archal). In the core countries, these institutions have evolved in accordance with
economic imperatives, demographic changes and organized women’s demands;
many welfare agencies are staffed by women at various levels who in turn help to
initiate policy debates and policy reforms (e.g. on parental leaves, ‘daddy days’,
flexi-time, etc.). The Nordic model of the welfare state may rightly be called
women-friendly (Hernes 1987; Hobson 1998).
The rentier state has a different logic. The concept of the rentier state, or rentier
economy, is applied to a country’s reliance on substantial external rent, such as the
sale of oil or transit charges (see Richards and Waterbury 1996). The consequences
of rentierism are said to be far-reaching. First, only a small fraction of the population
is directly involved in the creation of wealth. Second, the work–reward nexus is
diminished; rewards come from clientelism or patronage or proximity to the state.
Distributive policies or welfare provisioning amount to handouts from the state
rather than entitlements accrued to gainfully-employed tax-paying citizens. Third,

because the state is not dependent on taxation, there is far less demand for politi-
cal participation. (Thus: ‘no taxation, no representation.’) Saudi Arabia exhibits an
extreme form of rentierism, but the concept has been applied to Pahlavi-era Iran
and other MENA states. The rentier state has access to huge economic resources,
and is able to finance generous social policies for key segments of the population,
but it is also vulnerable to external shocks.
In the Middle East, the rentier state is also ‘neopatriarchal’ (Sharabi 1988),
meaning that the state rests on both modern and traditional institutions and
social relationships. The neopatriarchal state upholds the ‘patriarchal gender con-
tract’ (Moghadam 1998), by which I mean a gender ideology, a set of cultural
norms, and a social relationship predicated on the male-breadwinner/ female-
homemaker role. In the MENA region, the patriarchal gender contract includes the
provision – which is inscribed in the Muslim family laws that prevail in the region –
that men are responsible for the ‘maintenance’ (nafaqa) of their wives, and wives
are required to show ‘obedience’ (tamkin) to their husbands. Muslim family law
and the family wage ensure that most women have access to social welfare pro-
grams through their fathers or husbands. This is in some sense similar to what Ann
Orloff (1993) has described as ‘paternalist’ social policies in the early years of the
welfare state in Europe and North America.
Muslim family law is usually known as a code that governs marriage, divorce,
maintenance, paternity, and custody of children (An-Naim 2002). Critics are
typically concerned about its impact on civil and political rights as understood
globally or as enshrined in the Convention on the Elimination of All Forms of
Discrimination against Women (CEDAW). Less is known about the socioeconomic
consequences of Muslim family law, the way that it contravenes the International
Covenant on Economic, Social and Cultural Rights (the ICESCR, to which most
MENA countries are signatories), and its impact on women’s economic citizen-
ship. Derived from the Sharia, or Islamic law, Muslim family law defines the rights
and responsibilities of spouses and kin members, and aims to provide comple-
mentary justice for women and men. For example, because Muslim men have the
right to unilateral divorce and women do not, Muslim family law mandates a sum
of money from the groom to the bride (mahr in Arabic, mehrieh in Persian), which
is sometimes paid in full and sometimes deferred until the event of divorce. The
highly formal Islamic marriage contract stipulates the amount, which is usually
agreed to by both families. Within this legal and cultural framework, however,
women are seen primarily as dependants and men as breadwinners. As a result, the
discourse of motherhood is pervasive in Muslim societies, and few if any policies
are in place to actively encourage the sort of adult-worker family model that is
characteristic of many European countries, the US, or China.
Feminist concepts of ‘republican motherhood’ and of ‘maternalist policies’ are
relevant to the Middle East and especially the Iranian case discussed in this paper.
Linda Kerber (1980) used the term ‘republican motherhood’ to denote the ideals of
domesticity that legitimized women’s civic activity, especially moral reform,
throughout most of the nineteenth century. This is a concept that travels easily to
the Islamic Republic of Iran in the 1980s and 1990s, where women sought
to improve their legal status and social positions within the acceptable discursive
frameworks of revolutionary motherhood and religious duty (see Gheytanchi
2001). Maternalist policies refer to mother and child welfare policies that emerged
in the late nineteenth and early twentieth centuries, especially in the United States
and Britain, and which channelled social benefits directly to women as mothers.
There is some consensus that American middle-class women reformers – cut off as
they were from institutionalized politics – used the language of motherhood to
legitimize their political activism and to advocate better conditions and policies
for poor, working-class, or immigrant women. There is less consensus as to
whether these policies were ‘feminist’ or if they served to reinforce the patriarchal
social order. Some studies show that mother–child policies were palliative, or
temporary, or intended to strengthen rather than undermine the patriarchal family
(Lewis 1980; Ladd-Taylor 1994). Skocpol (1992) has argued that maternalist social
policies – which she defines as policies by women for women – characterized the
early welfare state in the United States.9 Koven and Michel (1993) distinguish
between radical and conservative maternalists, and demonstrate that maternalism
is a political discourse that is sometimes ‘a cloak for paternalism’ and sometimes
deployed by women for women. They write:
We apply the term to ideologies that exalted women’s capacity to mother and
applied to society as a whole the values they attached to that role: care, nurtu-
rance, and morality. Maternalism was the central and defining core of some
women’s vision of themselves and of politics.
Koven and Michel 1993: 4
They show, too, that at the heart of maternalism lies the paradox of entering the
public political arena by reinforcing the traditional female sphere of children,
family, nurturance and care.10
These feminist insights – along with a political economy approach – frame the
present study of gender and social policy in Iran, as well as the Middle East as a
whole. The regional oil economy combined with Muslim family law produced a
patriarchal gender contract that rendered the majority of MENA women economic
dependents. The dominant gender ideology has held that men are the providers of
women and children, and women are primarily mothers and wives. Until rela-
tively recently – that is, until the post-oil-boom downturn – the family wage and
the family law reinforced this situation. The beneficiaries of social security systems
thus have been largely men and their dependants; women’s welfare has been
presumed to be attained within the family or through the provisions of Muslim
family law; and the vast majority of Middle Eastern women have relied on family
support systems for protection vis-à-vis risk, contingencies and disasters.
This has been the case also in Iran, where, especially after the Islamic revolution
of 1979, women’s family roles were emphasized and female labour-force participa-
tion declined. In the late 1980s, Islamic women bureaucrats began to campaign for
improvements in women’s legal status and social conditions, and they did so
within a conservative religious/maternalist discursive frame. Accordingly, the
Islamic Republic of Iran adopted ‘maternalist’ policies for women that were
predicated on the patriarchal gender contract rather than on feminist concepts of
individual and collective women’s rights.
In the more recent neoliberal era, non-governmental organizations of all types,
including women’s organizations, have taken on an increasingly important role in
Iran, as elsewhere in the world economy. NGOs appear to have a three-fold role:
(a) they fulfil certain neoliberal requirements in the context of the withdrawal of
the state in welfare provisioning;
(b) they are part of the democratization process, or political liberalization, and
help to constitute the emerging civil societies across the region; and
(c) they play a key role in the women’s movement, with cultural, political, and
economic/employment implications for women and for gender relations.
Iranian women’s NGOs have been playing an increasingly prominent role in

delivering social services such as healthcare, literacy, legal counselling and the like
to women and girls, as well as helping to build civil society and generate employment
for women at a time when public-sector employment opportunities are diminish-
ing and the private sector remains predominantly male. As such, the women’s
NGOs serve to counter the adverse effects of the political economy and the family
law on women’s economic citizenship, albeit in a limited way.
The state, family law, and women’s economic citizenship
In the neopatriarchal state, unlike liberal or social democratic societies, religion is

bound to power and state authority, and the family, rather than the individual,
constitutes the universal building block of the community. As such, the neopatri-
archal state and the patriarchal family reflect and reinforce each other. The
concept describes an ‘ideal type’, and there have been variations in the region as
well as important reforms in the status of women initiated by modernizing or
revolutionary regimes. However, these reforms often have been undone, and
Middle Eastern states remain authoritarian and bound to cultural/ ideological
institutions such as Muslim family law. This is why those MENA states that have
signed on to the Convention on the Elimination of All Forms of Discrimination
Against Women (CEDAW) have done so with reservations that undermine the
spirit of CEDAW. As of March 2005, the Islamic Republic of Iran still had not
signed CEDAW.
Muslim family law – also known as the Personal Status Code in North Africa and
the Levant – is a social policy whose political significance and consequences for
women’s economic participation may be specific to the MENA region. It is
intended to regulate family life according to the norms of the Sharia, or Islamic
law; it is meant to ensure the rights and responsibilities of family members, espe-
cially those of the head of the family; and it aims to guarantee security to the wife
in the event of divorce or widowhood. The content of family law varies to some
degree across countries, depending on the legal school in place.11 But the broad
features of family law are similar across Muslim countries, as they commonly
reflect, and reinforce, a patrilineal logic.
The Islamic marriage contract requires the consent of the bride, and in some
countries women may insert stipulations into the contract, such as the condition
that she be the only wife (An-Naim 2002). However, marriage gives the husband
the right of access to his wife’s body, and marital rape is not recognized
(Welchman 2001). Children acquire citizenship and religious status through their
fathers, not their mothers. Muslim women are not permitted to marry non-
Muslim men, and non-Muslim widows cannot inherit from Muslim husbands.
A key aspect of Muslim family law is that women are required to obtain the per-
mission of father, husband, or other male guardian to marry, seek employment,
start a business, or travel. This is the source of a common legal inconsistency in the
Middle East, because placing women under the guardianship of a male contradicts
constitutional guarantees of equality of all citizens and labour laws which ensure
rights for men and women workers. As Pateman (1988) argued in relation to the
‘sexual contract’ in early modern Europe, Muslim family law denies daughters and
wives their full civil rights.
The unequal inheritance aspect of Muslim family law and civil codes compro-
mises women’s economic independence but is a sensitive issue; it is seen on one
level as a divine imperative revealed in the Koran and on another level as an impor-
tant part of the patriarchal gender contract whereby women are provided for by
their fathers, husbands or brothers. Islamic law gives women the right to own and
dispose of property, but they inherit less property than men do. Sons inherit twice
as much as daughters, but they are also expected to look after their parents in old
age. In a polygamous context, a deceased man’s inheritance and his pension are
divided among his widows, children and other relatives that he may have been sup-
porting; this could leave the widows with insignificant pensions. Thus, even though
Islamic norms require that fathers and husbands financially support their daughters
and wives, it is also the case that divorced, widowed or abandoned women without
access to jobs or a steady source of income, especially among the low-income social
groups, are often left in a state of impoverishment. Indeed, Loewe (2000) found that
among Egypt’s poorest are divorced, abandoned and widowed women.
In Iran (and elsewhere), a husband has the legal right to forbid his wife or
unmarried daughter to seek employment or continue in a job. As such, Muslim
family law contravenes the ICESCR, which is the main framework of economic
citizenship. The ICESCR prescribes the right of people to a freely chosen job;
equitable and equal wages for work of equal value; dignified working conditions
for workers and their families; professional training; equal opportunities for
promotion; protection for families, especially for children; maternity protection;
protection of boys, girls and teenagers against economic exploitation. Inasmuch as
Muslim family law and norms in Iran and elsewhere prevent women from apply-
ing for or staying in a job without the permission of father or husband, and in
some countries certain occupations and professions are off-limits to women (such
as the profession of judge in Iran), this denies women the right to enjoy the
ICESCR’s provision for ‘a freely chosen job’.
The mahr – or dower – is a sum of money promised by the groom to the wife
before marriage. Typically, the wife receives a portion of it at the time of marriage,
and the balance is deferred. But whereas this practice was initially meant to pro-
vide a sort of social insurance for the wife in the case of divorce or widowhood, in
the modern era it symbolizes women’s economic dependence.12 Moreover, legal
statutes, based on Islamic law, assert that men owe their wives material support or
maintenance and cannot command them to contribute to the family economy.
Indeed, one of the grounds for which women can legitimately petition for divorce
is non-support by the husband. This legal requirement, too, may have been salu-
tary in early Islamic history, but in the modern era it has functioned to perpetuate
the patriarchal gender contract.
In some ways, Muslim family law may be seen not only as a pre-modern or pre-
feminist code for the regulation of family relations, but also as a way of retaining
family support systems in the place of a fully functioning welfare state predicated
on concepts of citizen contributions and entitlements. The welfare of wives and
children remains the responsibility of the father/husband. When a woman seeks a
divorce or is divorced, her maintenance comes not in the form of transfers from
the state but in the form of the mahr that is owed to her by her husband, or (in the
Islamic Republic of Iran) the ojrat-ol-mesl, which is the monetary value of the
domestic work she has performed over the years. I discuss this new element of
Iran’s gender/social policy regime in the section below, but let us first examine the
historical context.
The gender/social policy regime in Iran
Women have a long history of economic activity in Iran, particularly in the carpet-
weaving and textiles sectors. In the late Qajar period (the turn of the twentieth
century), women workers were largely concentrated in home-based and small
workshop enterprises. Iran joined the International Labour Organization in 1914,
and in 1920–21 the ILO informed the government of its concern over working
conditions in the carpet industry in Kerman. This led to a series of decrees
designed to improve working conditions with respect to length of the work-day,
minimum age of employment for girls and boys, safety, sanitation and health.
According to Floor (1985: 88), two articles of the decree of 17 December 1923 read
in part: ‘boys and girls shall work in separate workshops; mixed workshops are
absolutely forbidden’, and ‘foremen (those who dictate the pattern to the workers)
shall not enter the girls’ workshop, where forewomen shall be employed.’ Iran’s
Civil Pension Fund was established in 1923, although there are no data to deter-
mine how many women, if any, were enrolled at that time. In the 1930s, under
Reza Shah, women’s social participation expanded as more schooling was made
available and the growing state sector absorbed some of the educated female
labour force (Moghadam 2000). In other areas, however, women remained
disadvantaged, as they could not vote and were subordinated within the family.
Moreover, they were all but excluded from Iran’s rapidly expanding oil industry,
which at the time was not only a labour-intensive industry but also completely
male-dominated.
After Reza Shah’s forced abdication in 1941, the coalition government of Ahmad
Qavam sought to implement a wide-ranging social programme. In 1946 it
promised distribution of state lands, women’s suffrage, provincial assemblies,
elimination of unemployment, construction of rural clinics, schools and irrigation
projects. Qavam established a Supreme Economic Council that was charged with
implementing a Five Year Plan, setting of minimum wage, protecting national
industry, drafting a plan to distribute crown lands, and supporting the peasantry.
He recognized trade unions and appointed a Supreme Labour Council to introduce
unemployment insurance, wage scales, and negotiating committees between man-
agement and labour. His government, however, did not survive to pursue these
radical policies (Messkoub 2005). Thus by the mid-1960s illiteracy rates remained
high in Iran: just half of the urban population was literate compared with only
15 per cent of rural population. The female literacy rate of 17 per cent was less
than half that of men (40 per cent). The situation improved somewhat by the early
1970s – after nearly a decade of a rural literacy campaign spearheaded by the
Literacy Corps.
During this period, Iran enjoyed unprecedented economic growth, largely as a
result of domestic investments fuelled by rising oil prices. Oil revenues accrued to
the state, and its ever-growing budget paid for development projects, prestige
projects (such as expensive monuments and royal celebrations), military procure-
ments, infrastructural development and social programs. Skilled workers and
high-level manpower were imported from South Korea and the United States,
among other countries, but Iran had its own large workforce to man the oil indus-
try, manufacturing and the booming construction sector. Oil revenues allowed for
high wages and low taxes. Karshenas has studied manufacturing wage levels
comparatively and has established that wages were higher in the Middle East,
including Iran, than in southeast Asia and Mexico. At the same time, income and
other taxes were among the lowest in the world (Karshenas 1997, 2001). High wages
and generous social policies for government employees – financed not by income
taxes or employee contributions but by oil revenues and other rents – also were
meant to strengthen and expand the state, establish national integration, and build
a social base of support for the state.
As the Shah of Iran sought to establish national unity and mobilize support for
his rule, the government established an array of social policies and welfare benefits.
Government employees received pensions, subsidized staples, and interest-free
loans for housing; a workers’ shareholding program applied to those in large
establishments; the urban poor could use free schools, free health clinics, and
cheap food; and peasants benefited from the land reform, a literacy campaign and
a healthcare campaign. For women, Shah Mohammad Reza Pahlavi extended the
vote to women and modernized the family law – largely due to advocacy by a group
of elite women in and around the state or with connections to the royal family.
The oil-boom era

During the state-building, oil-boom era, social policies included guaranteed public
sector employment for graduates, labour legislation that favoured workers, free
education, social insurance for government workers and those in large enterprises,
subsidies (utilities, petrol, bread), and paid maternity leaves for women
professionals. Various state enterprises and ministries had their own policies and
benefits packages. The most generous were offered by the armed forces and the
National Iranian Oil Company, while the Ministry of Roads had a hospital and
seaside resort for its employees.13 Social security benefits (sickness, disability,
death, pensions) were available to all regular employees, including women. State-
owned factories employing a certain number of women (usually 50 or 100) were
required to establish nurseries, and working mothers were also entitled to nursing
breaks. Social welfare services for the poor, including poor women, were chan-
nelled in part through institutions such as the Pahlavi Foundation and the
Women’s Organization of Iran.
During the 1960s and 1970s, the female labour force grew, and by 1976 women
comprised 12–14 per cent of the total labour force. Women were found in the civil
service (largely in teaching, healthcare and social services, with some participation
also in public administration), in large-scale factory employment and of course in
traditional small-scale activities. Though the oil and petrochemicals industries
were male-dominated, a small proportion of women worked in clerical positions
in the National Iranian Oil Company. Those in the modern sector were beneficia-
ries of labour legislation and social insurance. In the 1970s women retired on
average two to three years before men but with similar annual pensions; women
comprised around 15 per cent of all pension recipients (Messkoub 2005).
Class, urban, and gender biases were inscribed in the social policies, and
especially in the social security policies and provisions of labour law. For the most
part, the principal beneficiaries were men, members of the urban middle class or
the ‘labour aristocracy’. Social security and labour legislation did not extend to
domestic workers/servants, irregular or casual workers, unpaid family workers, or
peasants, rural workers and small farmers. Given the size of the rural population in
Iran, this meant that the majority of citizens relied on family support systems – a
situation that helped to reproduce the patriarchal family through large, extended
families with a preference for sons, and the unpaid labour of female members.
However, this also was a time when elite women in Iran began to organize and
mobilize for improvements in women’s social conditions and their legal status.
Much of that work was carried out by the Women’s Organization of Iran (WOI). By
the late 1970s, the WOI had over 400 branches and 120 women’s centres through-
out the country that offered daycare centres, legal literacy, counselling, family
planning, health, nutrition, and vocational training (Afkhami 2003; Afary 2005).
The Family Protection Law of 1967 and 1975 provided unprecedented rights for
Iranian women in the areas of divorce and child custody; for example, a wife’s
permission was required should her husband seek a second wife. However, a man
still had the right to forbid his wife from working outside the home, if he found
her job ‘dishonourable’; and she could not travel outside the country without her
husband’s written permission.
Given the huge increases in the national income in the 1970s, more invest-
ments should have been made in creating a literate, educated and skilled labour
force, both male and female. Public services did expand to a large extent, but Iran
did not achieve full literacy or eradicate child and maternal mortality. The 1979
revolution also set back socioeconomic gains by at least a decade. Thus by 1990,
Iran’s educational attainment rates were far below those of other countries at com-
parable levels of development or income. In particular, illiteracy was high among
adult women, and gender gaps were very wide. The mean years of schooling of
adult men and women (aged 25 and over) was 4.6 and 3.1, respectively. But in far
poorer Vietnam, the corresponding figures were 5.8 and 3.4 years. The Philippines
did much better, at 7.8 years for men and 7.0 for women.14
The Islamic era

The 1978–79 revolution initially had the limited objective of ending the dictator-
ship, but Islamization set in and transformed gender relations as well as the
country’s social policies, in particular the family law (which in Iran’s case is not a
separate, integrated law but consists of a series of articles found in the Civil Code).
The Islamic Republic’s very ambitious constitution required the government to
provide full employment to its citizens, including (presumably) women citizens.
But this constitutional guarantee has been undermined by
(a) poor economic conditions, inadequate domestic and foreign investments, and
subsequently low levels of job creation;
(b) a preference for investments in capital-intensive, male-intensive sectors such
as oil, gas and nuclear energy;
(c) a constitutional clause extolling the virtues of motherhood; and
(d) the ubiquitous Islamic criteria, including the fact that under the Islamic
Republic’s Sharia-based provisions on women and the family, women cannot
seek jobs without the approval of their fathers or husbands.
Iran’s Islamic Revolution had a number of adverse outcomes, especially for

women. First, women’s employment patterns and rates changed dramatically.
Most Iranian women had been working in the fields of teaching and healthcare,
and this remained the case after the revolution. But numerous women lost their
jobs due to their political, ideological or religious backgrounds and were replaced
by ‘Islamic’ women. The proportion of women in public services remained stable,
but a number of occupations and professions were deemed inappropriate to
women, such as jobs in the judiciary and in some engineering and medical fields.
Female judges and many high-ranking lawyers lost their jobs, but the greatest job
loss was experienced by working-class women in medium-sized or large factories
(Moghadam 1988). During the 1980s the Islamic state relied heavily on oil exports
for its revenues; these were needed to fight the war with Iraq and to finance an
array of subsidies for the population. As a result, the Iranian economy did not
experience the diversification that the revolutionary coalition and the new
constitution called for. In the 1990s, women’s share of the roughly one million
civil service jobs steadily increased, but their overall share of the labour force
remained small, because of men’s far higher involvement in the oil industry, state-
owned industrial enterprises, hotels and the private sector. In 1996 women’s share
of all manufacturing jobs (formal and home-based alike) was 14 per cent and their
overall share of paid employment was 12 per cent (Moghadam 2003, chs. 2, 7).
After the revolution, a massive ideological campaign was launched to tie women
to their family roles, and the new Islamic family law restored men’s rights to
polygamy, unilateral divorce, custody of children after divorce, and control over
their wives’ mobility. In matters of inheritance, women became severely disad-
vantaged. For example, a man could inherit all of his wife’s wealth, but she was
entitled only to one-fourth (if he had no child) or one-eighth (if he had children)
of his movable property and of the value of his estate. This pertains to cases of
permanent marriage; partners in a temporary marriage receive no inheritance.15 In
polygamous marriages, wives must divide among themselves the allotted one-
fourth or one-eighth. Article 1117 of the Civil Code stipulates that a man has the
right to prevent his wife from taking employment ‘if he deems such employment
would be at variance with their family interests and values’.
Although women’s educational attainment rates have risen steadily, the result of
both Iran’s political economy and its cultural/ideological institutions has been low
levels of female salaried employment, high rates of inactivity, disproportionately
high female unemployment, and a huge gender-based income gap. Iran’s labour
force has been predominantly masculine. The burgeoning private sector remains a
largely male domain, with work conditions that are not appealing to most women.
Gender bias has been rampant in employment and in the country’s social insurance
programs. Even though occupational restrictions were lifted, the profession of
judge remains exclusively male. Male and female workers alike have paid 7 per cent
of their wages toward their retirement insurance, but in the event of a woman’s
death, her family could not collect retirement benefits. All working women are
required to appear in hejab, which is at minimum a large scarf covering all the hair
and neck, and a long-sleeved and loose-fitting smock or manteau that covers the
body’s contours. In some government agencies, more conservative attire (chador)
is required of the women employees. Of course, professional women in the civil
service have been beneficiaries of social policies, and the time of writing they rep-
resented about 36 per cent of Iran’s total civil service employees. However, they
constitute only a small proportion of the total female economically active popula-
tion, and though they constitute only 12 per cent of the total salaried labour
force, their unemployment rate in 1996 was 12.5 per cent (IRI 1997; 1999: 99).16
Clearly, Iranian women’s economic citizenship is seriously circumscribed.
This may be one reason for the growth of women’s NGOs in Iran. When job
opportunities in the formal sector are foreclosed, women form their own businesses
or associations. This is an area that remains under-researched in Iran, although
Povey (2004) has undertaken some preliminary work on women’s NGOs. Women’s
NGOs include publishing houses, environmental protection groups, research
centres rural women’s centres, charitable agencies and professional associations.
The women’s press in Iran has been proliferating, and consists of magazines and
women’s studies journals such as Zanan, Hoghough-e Zanan and Farzaneh, and a
publishing house, Roshangaran Press. The Women’s Cultural Centre has produced
calendars, compendiums and a feminist journal, and maintains a website.
The WID-oriented NGOs engage in advocacy for enhanced women’s rights, sometimes
by using the Islamic Republic’s own maternalist discourse against it (see discussion
below on ojrat-ol-mesl). Forming women’s NGOs may be seen, therefore, as a strat-
egy to counter women’s economic dependence and high unemployment rates, as
a way of enhancing women’s social participation, and as a form of advocacy for
women.
Lack of job opportunities also may be behind another phenomenon, similarly
under-researched and largely anecdotal at this time – the expansion of the practice
mehrieh, which had been in decline in the final years of the Pahlavi era, especially
among the educated, westernized elite. By all accounts, mehrieh has become more
rather than less entrenched in Iranian society. The Islamic state’s ideological pressure
is surely one reason, but what I am suggesting here is that young Iranian women
and their parents may very well be insisting on mehrieh due to economic and
employment uncertainties.
Maternalist politics in the Islamic Republic

One of the defining features of the Islamic Republic of Iran has been its attempt to
institutionalize women’s domesticity and maternalism. A constitutional clause
extolling family, motherhood and the raising of ‘committed Muslims’; the abroga-
tion of the Pahlavi state’s Family Protection Act and suspension of its family-
planning programme; the encouragement of part-time work for married women
who chose to stay in the labour force; the exaltation of the ‘mother of martyrs’ –
these are examples of how the Islamic Republic of Iran has sought to create a
Revolutionary Motherhood or Republican Motherhood of a new type. Initially, the
model of the new Islamic woman was Fatemeh, daughter of the prophet
Mohammad and wife of Imam Ali, the selfless mother of martyrs who remained
steadfast in the faith. Elsewhere I have written about the elements of the Islamic
republican discourse on women, or ‘the three D’s’: – the danger that unveiled and
unmarried women represented, their essential difference, and domesticity as the ideal
state for women (Moghadam 1988). At the same time, in order not to completely
alienate women and to maintain the support of the population of Islamist women,
the regime under Ayatollah Khomeini instituted some measures that favoured its
female supporters. For example, it did not ban women from employment; there
were always several token women members of parliament; martyrs’ widows were
granted custody of their children and allowed to retain custody even after remar-
riage; and the regime introduced a new marriage contract in 1987 that allowed for
the stipulation of conditions by wives. After Khomeini’s death, the family-planning
policy was reinstated – which Iranian women enthusiastically embraced.17
But throughout the 1980s and 1990s Iran’s social policies and legal frameworks
ensured the persistence of the patriarchal gender contract, women’s economic
dependence, and inequalities in economic citizenship. The Civil Code stipulates
that a husband is responsible for the upkeep of the family, and Article 1107 even
states that ‘the wife’s living expense includes shelter, clothing, food and furniture
that are commensurate with her status, which may even include a maid in case
she is used to having one or she needs one due to illness or incapacity.’ As one
commentator has explained, with approval and pride: the Iranian legal system
does not obligate women to do any household activity. Even the wife’s nursing
of her own child ‘is not one of her unwavering responsibilities’. He states, citing
Article 1176 of the Civil Code: ‘The mother is not obliged to suckle her own
baby unless the baby cannot be fed by any other means’ (Mehrpour 1995: 60).
Here the commentator appears unaware of the larger implication of such a
‘right’ on the part of the wife and mother – which is to exempt her from any
labour whatsoever other than sexual services to her husband and childbearing
and therefore to deny her any economic citizenship. In the event of divorce, he
continues,
when the wife is free from any blame or fault, [the husband] is obliged to pay a
lump sum to his wife, which payment is in addition to his obligations on mar-
riage portion [mehrieh], recompense of the wife’s past services in married life
[ojrat-ol-mesl, of which more, below], and alimony payment during the speci-
fied time after divorce when she is religiously prohibited from a new marriage
Mehrpour 1995: 61; my emphasis
The Islamic Republic may have exalted mothers in its official discourse, but the
Civil Code stipulates a fractional share of a deceased husband’s wealth for the wife.
The stated justifications for women’s lesser shares are ‘the marriage portion and
provisions of women’s living expenses by the husband’ (Mehrpour 1995: 58).
The Civil Code also cites the Qur’anic verse ‘God has ordained that amongst the
children, the son’s share is double the daughter’s’ to establish unequal inheritance
for sons and daughters. Nonetheless, advocates of the Islamic Republic’s legal and
social policy frameworks insist that
men and women are entitled to equal rights in possessing and executing the
privileges of their legal capacity, and as such there is no distinction between men
and women. All the means and tools by which men can obtain economic
resources are legally at the disposal of women as well. … The only limitations on
the employment of women concern jobs which defy family values and interests.
Mehrpour 1995: 61
Many feminists, however, have noted that despite its ‘motherist’ discourse, the
Islamic Republic of Iran has in practice devalued motherhood through its policies
and laws. Lawyer Mehrangiz Kaar has highlighted the contradiction between the
Islamic state’s formal praise for mothers and Iranian mothers’ lack of legal protection,
such as the right to custody of their children, the right to open a bank account for
their children, and control over their own bodies (Kaar 1996). Her articles in the
women’s magazine Zanan – which came to exemplify the Islamic feminist per-
spective in the 1990s – concentrated on women’s real status under Iranian civil law
and family law and called on women to engage in ijtihad (independent reasoning
and interpretation) to claim their rights on behalf of their status as mothers
(Gheytanchi 2001: 566). Ojrat ol-mesl was one such example of women’s ijtihad to
further their social rights. Another was to argue that in cases of divorce, the courts
should calculate the wife’s deferred dower according to an index updated for
inflation (Poya 1999: 101–2).
Ojrat-ol-mesl, or the law mandating wages for housework, was passed by the
Majles (parliament) in late 1992. The regime’s Islamist women supporters played
an important role in the adoption of this policy – largely because of their dismay
over the ease by which men could divorce their otherwise good Muslim wives. As
believing women, their protests and recommendations were delivered in a
religious idiom. As Hoodfar explains:
Islamist women activists argued that women, like all other Muslims, are
entitled to the fruit of their labour on the grounds that Islam is against exploita-
tion; that in Islamic tradition wives have no duties to their husbands beyond
being faithful and are not required to work in their husbands’ homes, to the
extent that women are not even obliged to breast-feed their children without
payment from their husbands. Therefore, because all women do, in fact, work
in their husbands’ homes, they are entitled to the fruit of their labour!’
Hoodfar 2000: 311
Such women may have been advocating a maternalist policy in Skocpol’s sense.
However, not all Iranian feminists supported the new policy, and many felt
ambivalent about the use of the discourses of republican motherhood and the
rights of mothers under Islam. Although the law clearly gave women additional
benefits and increased the ‘cost’ of divorce to men, ojrat-ol-mesl was functional for
the Islamic Republic in that it carried no financial implications for the govern-
ment, it reinforced women’s maternal roles, and there was no onus on the
government to provide jobs or social assistance to divorced women.18 Moreover, as
we have seen above, these entitlements are due only to wives who are deemed to
be not at fault in the case of divorce. The legal scholar An-Naim (2002: 110)
emphasizes the point: ‘A woman is entitled to half of her husband’s assets if the
court finds the divorce was initiated by the husband and was not caused by any
failure on the wife’s part.’ This would appear to make ojrat-ol-mesl a conservative
maternalist policy – in Koven and Michel’s sense – that reinforces the patriarchal
gender contract rather than expands women’s economic citizenship. Moreover,
ojrat-ol-mesl ‘is difficult to apply in practice, partly because of the difficulty in
assessing wages for housework’ (An-Naim 2002: 110). Anecdotal evidence suggests
that it is far from an equal sharing of marital assets. Unless the husband has put
property or bank accounts in his wife’s name, he is entitled to keep all ‘his’ assets
except for that portion which is calculated as the ojrat-ol-mesl – or the wife’s
entitlement for the years of household and childcare.19
And what of other maternalist policies in the Islamic Republic of Iran, especially
for working women? In the early 1990s, and in response to its supporters’ concerns
about the status of women, the government of Hashemi Rafsanjani established the
Bureau of Women’s Affairs, which worked with a number of Iranian WID experts
and helped to establish WID or women’s units in various ministries. It also had
connections with a nationwide network of women volunteers who engaged in a

variety of charitable and social service work among the poor (Moghadam 1998,
ch. 7). Continuing its pattern of pro-family policies, the Islamic Republic
instituted the following (IRI 1999):
● Payment of a marriage allowance and family support allowance to insured male
or female employees;
● Early retirement for women aged 45 with a minimum service period of 20 years;
● Reduction in the working hours of women in government employment;
● Introduction of a family support allowance for widows and women with
disabled spouses;
● Payment of pensions to survivors of deceased female government employees.
At least two of these measures are similar to the family-friendly policies found in
continental Europe, but the overall thrust of the measures is to reinforce women’s
family roles. Given women’s overall lower incomes, an average retired woman’s
pension can be rather low, and certainly insufficient to support a family. In fact,
the policies have been designed on the assumption of a salaried male household
head as the real breadwinner.
The Islamic Republic’s 1990 labour law also includes maternalist measures that
are similar to the maternal protection policies of many countries. The labour law
bans women and all workers under 18 from working at night. For pregnant
women, the law proscribes the assignment of heavy work; requires employers to
transfer pregnant women to lighter work at the same pay when their regular work
is deemed heavy or otherwise harmful to their condition; and prohibits women
from carrying heavy loads during pregnancy and for 10 weeks after giving birth.
The law entitles women to 90 days of maternity leave, at least half of which must
be taken after childbirth, for the first three children. New mothers are given time
off to nurse their babies in the workplace crèche (IRI 1990; Centre for Women’s
Participation 1999). The crèches in which new mothers may put their babies and
nurse them in the course of the work-day are usually found in factories.20
However, because factory work is generally deemed inappropriate to women,
relatively few women are involved in the manufacturing sector and there have
been no campaigns to encourage working-class women’s employment. The leaves,
nursing breaks, crèches and childcare centres, therefore, are enjoyed by only a
fraction of the economically active female population.
Healthcare services have expanded in the Islamic Republic, and special
programmes for pregnant women are in place, such as immunization against
diphtheria and tetanus. Maternal mortality rates have fallen, and couples are
required to attend a seminar in family planning prior to receiving their marriage
certificate. This has been accomplished under the public healthcare system through
the network of Rural Health Centres and Urban Health Centres, and since 1990 the
Ministry of Health and Medical Education has employed the services of women
Community Health Volunteers.21 According to a 1999 government publication,
about 80 percent of urban mothers and 73 percent of rural mothers have been
attended to more than twice during their pregnancy. Trained professionals
assist at 86 percent of births in I.R. Iran, although the number of deliveries

attended by untrained people is six times higher in rural areas than it is in the
cities
IRI 1999: 103
These policies and outcomes may reflect social welfare goals, concerns for the
health of mother and child, and the advancement of women. But conservative
legislators and jurists have blocked bills that would raise the age of marriage to 16,
give women the same inheritance rights as men, or award a temporary stipend to
widows disadvantaged by inheritance laws from their late husbands’ estates
(Shahidian 2003: 222). A bill that would ban sexual integration at hospitals and
public health facilities was cancelled only due to a public outcry. In frustration,
Mehrangiz Kaar wrote in Zanan (cited in Gheytanchi 2001: 573):
In the beginning of the Persian new year, I am listening – in the midst of

political battles in Iran – to understand who I am and what are my rights which
I have lost under the name of motherhood. Women are asking one question:
under the contemporary legal system of Iran, are women citizens or possessed
objects? I hope you understand that when a victim – at the peak of her suffering –
speaks of injustices and objects, a historical event proceeds. Take her seriously!
Nobel laureate Shirin Ebadi similarly has observed a discrepancy between the
stated objectives of some of the laws and policies of the Islamic Republic of Iran,
and their real objectives and implications:
Priority is given to the betterment of women’s status throughout the country,

given their supreme dignity and the Muslim woman’s fundamental role in
strengthening the foundations of the family, as well as strengthening social,
scientific, and artistic projects.
A closer look at the place of women as viewed by cultural policymakers will
reveal their emphasis on family values; a woman’s independence, her social
situation, and the discriminations leveled against her are never at issue.
Policymakers view women as wives and mothers, who need cultural reinforce-
ment and guidance to better fulfill their domestic roles.22
If the Islamic Republic of Iran’s ‘motherist’ policies sought to valorize motherhood

and domesticity through such redistributive/allocative measures as pensions for
mothers of martyrs, compensation for housework and childrearing in the event of
divorce, and expansion of the mehrieh, most Iranian women seem to be more
interested in the reform of family law and in increasing their economic and
political participation. Indeed, like MENA feminists in general, Iranian women’s
rights activists, while not discounting the economic value or cultural capital of
mehrieh and nafaqa, prefer a discourse emphasizing civil, political and social rights
that are in line with international standards (Moghdam 2003, ch. 8). It is for these
reasons that the Center for Women’s Participation drafted in 2004 a parliamentary
bill that would have the Islamic Republic of Iran sign and ratify CEDAW.
Conclusions
This chapter has demonstrated how Muslim family law, as a social policy, impedes
women’s economic citizenship, and why it is that Iranian women activists (like
activists in other MENA countries) have prioritized reform of family law as a
necessary condition for the expansion of women’s rights and welfare. Whereas the
reform of family law is not in the interests of the rentier neopatriarchal state, it is
necessary for the strategic gender interests of Iranian women. I have argued that
Muslim family law today is an anachronistic social policy that reinforces the
patriarchal gender contract, undermines women’s economic citizenship, and does
little to provide for women’s welfare.
For these reasons, Iranian women have embarked on a quest for economic
citizenship. Although economic and cultural barriers continue to prevent large-
scale labour incorporation of women, middle-class women who are in the labour
force have been pushing for reform of those barriers – such as Muslim family law –
while also engaging in other strategies, such as forming or joining NGOs, in order
to enhance their social participation, employment opportunities and access to
economic resources.
This study also has confirmed the contradictory nature of maternalist discourses
and policies, especially as they have been practised in the Islamic Republic of Iran.
Following Koven and Michel, we note that maternalism in Iran has been a ‘cloak
for paternalism’ more than a radical discourse and vision emanating from the
women’s movement. Whereas Islamic women advocates have deployed a religious-
motherist discourse and were responsible for the adoption of ojrat-ol-mesl, secular
feminists have stressed the need to make Iran’s legal frameworks, including its
family law, more consistent with international standards and norms. Similarly,
because of the official discourse of women’s domesticity and the ideal maternal
role, secular feminists have steered clear of a maternalist discourse that seeks
rewards for women’s unpaid labour or housework. Rather, they emphasize the
need for more employment opportunities and equality in social insurance. Unlike
Europe, where ‘the highly unequal distribution of unpaid work and especially car-
ing labour underlies much of the inequality found in employment and reproduced
in social security’ (ILR 2000: 115), the issue in Middle Eastern countries like Iran is
the highly unequal distribution of paid work, along with explicitly discriminatory
family laws, that underlies the gender inequalities found in the region.
Indeed, in almost all Middle Eastern countries, maternalist politics and
discourses are associated with the neopatriarchal state and with Islamist move-
ments, including their female members. The masculine nature of the labour force
and the predominance of the discourse of motherhood is surely why MENA
feminists tend not to engage in a maternalist politics or advocate for policies
pertaining to women’s care work. Their priorities, instead, are reform of family law
in order that family relations are made more equitable and that women’s economic
and political participation is enhanced (Moghadam 2002a). Many non-Islamist
women have distanced themselves from religious-maternalist discourses and
campaigns, or have used them only instrumentally.23
There is no doubt that many Iranian families ignore Muslim family law and
provide daughters with the same inheritance as sons. A husband also may put
property in his wife’s name. (This was common in pre-revolutionary Iran, especially
among the elite.) But the fact is that the laws disadvantage women and apparently
so does much social practice. This, after all, is why Islamic women eventually cam-
paigned against the most egregiously discriminatory aspects of Muslim family law;
and why they lobbied for a mechanism to provide divorced women with some
share of the husband’s assets. One may also note that, having faced the limits of
the religious-maternalist discourse, Islamic women in and around the Centre for
Women’s Participation have deemed it necessary to lobby for Iran’s accession to
the Women’s Convention. How the state will respond remains to be seen.
Notes
1 This includes research by Linda Gordon, Theda Skocpol, Sonya Michel and Seth Koven,
Sonya Michel and Robin Rosen, and Molly Ladd-Taylor – who tend to have different
approaches to and definitions of ‘maternalist’ in the social policy or welfare state
context.
2 Representative of this body of research are studies by Gisela Bock, Helga Hernes, Barbara
Hobson, Ann Shola Orloff and Diane Sainsbury.
3 See, for example, studies by Nancy Folbre (1994), Arlie Hochschild (1989), Jane Jenson
(1986) and Jane Lewis (1980).
4 Esping-Anderson (1990) and Marshall (1964). Marshall’s theory remains powerful and
appealing, inasmuch as it integrates development and social policy frameworks and
offers a historical perspective.
5 The argument is made in a larger study of gender and social policy in the Middle East,
with a three-country focus on Iran, Jordan and Tunisia (Karshenas and Moghadam
2005). The present chapter draws on some of the conceptual discussion there.
6 See, for example, Richards and Waterbury 1996; Tzannatos and Kaur 2002.
7 Exceptions are Moghadam (1998) and Amira Sonbol (2003). See also the essay by Barbara
Swirski (in Joseph 2000), on differential access to social rights and the state’s welfare
benefits by Jewish and Palestinian women citizens in Israel.
8 On family law see, for example, Esposito and DeLong-Bas 2001; An-Naim 2002; Joseph 2000.
9 Among other things, Skocpol was keen to dispute the notion of the United States as a
relative ‘laggard’ among the core countries in terms of social welfare provisioning, and
to make a case for the revival of generous social policies in her country. What was dis-
tinctive about the early American policies, she argues, was the extent to which women
reformers were involved in their design and implementation. This has been challenged
by other scholars, who take issue with Skocpol’s definition of maternalist policies as
woman-centered, largely progressive, feminist, and cross-class. See the debate between
Skocpol and the historian Linda Gordon in the journal (now defunct) Contention, Vol. 2,
No. 3, Spring 1993: 139–89.
10 It is worth noting that maternalist discourses and politics have a long history in the
women’s movement, and have been especially evident in women’s peace and human
rights movements.
11 There are four Sunni schools (Hanafi, Hanbali, Maliki, Shafii) and one Shia school
(Jaafari) of Islamic jurisprudence. For details on Islamic family law across countries and
figh schools, see An-Naim (2002).
12 Mahr also commercializes the marriage contract. See Fatemeh Moghadam (1994).
13 Information from members of the author’s family, retired civil servants from the ministries
of agriculture, roads and transportation, and culture. Of course, the vast majority of
employees of the armed forces, the Oil Company, and the Ministry of Roads were men.
14 The figures are from UNESCO (1994), and the full table may be found in Moghadam
(1998), ch. 2.
15 Temporary marriage, known as sigheh, is an Iranian Shia phenomenon.
16 According to UNDP estimates, women’s earned income in 2002 was $2,835, compared
with men’s earned income of $9,946. See UNDP (2004), Table 24.
17 Family planning was revived and aggressively implemented in the wake of the economic
and social difficulties created by the eight-year war with Iraq, growing poverty due to
declining oil revenues and the state budget, and an exceedingly high birth rate during
the 1980s. In 1993 the parliament passed the Family Planning and Population Control
Act. This was the beginning of the ‘rationalization’ of social policy – within limits – in
the Islamic Republic. By the new century, Iranian (and Tunisian) women had, at 2.2
births per woman, the lowest fertility rates in the region, followed by Lebanon and
Turkey.
18 Yesim Arat (2000: 278–79) reports that a similar debate took place in Turkey, when
women’s groups proposed that an amendment to the civil code allow women to select
the option of sharing assets acquired during the marriage. Some feminists opposed this
measure, fearing that women would continue to see housewifery as an attractive
prospect, decline to work outside the home, and continue their economic dependence.
19 In early 2005, the sister of a friend of the author was seeking, without success, a larger
share of her husband’s assets after 22 years of marriage during which time the husband
had been frequently abusive. The amount of ojrat-ol-mesl that was due her was consider-
ably less than the average price of a home/apartment in Tehran.
20 In 1994, the author visited a crèche and childcare centre in a large textiles and garments
factory outside Tehran, and found them to be of high quality. See details in Moghadam
(1998), ch. 7.
21 In the early 1990s, the Iranian government mobilized around 20,000 women volunteers
to help implement its new family planning policy, but later refused them permission to
set up an independent organization. See Hoodfar (1998).
22 Shirin Ebadi, Women’s Rights in the Laws of the Islamic Republic of Iran (2002), reproduced
in the Iranian on-line feminist journal Badjens at http://www.badjens.com/ebadi.html
(accessed 20 December 2004).
23 Personal communication, Latifa Jbabdi of Morocco’s Union de l’Action Feminine,
Helsinki, 10 September 2004. Jbabdi was describing the feminist strategy of mobilizing
support for the reform of Morocco’s very conservative family code, the Mudawanna. She
told the author: ‘We had to break the taboo [of tampering with the family law, seen as
divine], and so we studied the Qur’an. We found that many aspects of the family law
contradicted the Qur’an. We are not Islamist feminists, but our country is a Muslim
country, and so we used religious language.’ The Iranian feminist lawyer Mehrangiz
Kaar, a staunch secularist, nonetheless urged ijtihad as a way of motivating believing
women to advocate change in the Islamic Republic.
References
Afary, Janet (2005) ‘Mahnaz Afkhami: A Memoir’ (book review essay). Journal of Middle East
Women’s Studies, Vol. 1, No. 1 (Winter): 147–57.
Afkhami, Gholam Reza (ed.) (2003) Jame-eh, Dowlat va Jonbesh-e Zanan-e Iran, 1342–1357:
Mosahebeh ba Mahnaz Afkhami (Society State and Women’s Movement in Iran: 1963–1978:
Interview with Mahnaz Afkhami). Bethesda: Foundation for Iranian Studies.
An-Naim, Abdullahi (2002) Islamic Family Law in a Changing World. London: Zed Books.
Arat, Yesim (2000) ‘Gender and Citizenship in Turkey’. In Suad Joseph (ed.) Gender and
Citizenship in the Middle East. Syracuse University Press.
Center for Women’s Participation (1999) Special Laws and Regulations for Women in the Islamic
Republic of Iran. [in Persian]. Tehran: Office of the President, Centre for Women’s Participation.
Esping-Anderson, Gøsta (1990) Three Worlds of Welfare Capitalism. Princeton NJ: Princeton
University Press.
Esposito, John, with Natana J. DeLong-Bas (2001) Women in Muslim Family Law. Syracuse NY:
Syracuse University Press.
Floor, Willem (1985) Labour Unions, Law and Conditions in Iran (1900–1941). University of
Durham Center for Middle Eastern and Islamic Studies, Occasional Papers Series No. 26.
Folbre, Nancy (1994) Who Pays for the Kids? Gender and the Structures of Constraint. London
and New York: Routledge.
Gheytanchi, Elham (2001) ‘Civil Society in Iran: Politics of Motherhood and the Public
Sphere’. International Sociology, Vol. 16 (4): 557–76.
Hernes, Helga (1987) Welfare State and Woman Power: Essays in State Feminism. Oslo:
University of Oslo Press.
Hobson, Barbara (1998) ‘Women’s Collective Agency, Power Resources, and Citizenship
Rights’. In Michael Hanagan and Charles Tilly (eds) Recasting Citizenship. Totawa NJ:
Rowman & Littlefield.
Hochschild, Arlie (1989) The Second Shift: Working Parents and the Revolution at Home. New
York: Viking.
Hoodfar, Homa (1998) ‘Volunteer Health Workers in Iran: Social Activists?’ Montpelier:
WLUML Occasional Paper no. 10.
Hoodfar, Homa (2000) ‘Iranian Women at the Intersection of Citizenship and the Family
Code: The Perils of Islamic Criteria’. In Suad Joseph (ed.) Gender and Citizenship in the
Middle East. Syracuse University Press.
ILO (1990) Yearbook of Labour Statistics: Retrospective Edition on Population Censuses
1945–1989. Geneva: ILO.
ILR (2000) ‘Introduction: Social Policy and Social Protection’. International Labour Review,
Vol. 139, No. 2: 113–117.
Islamic Republic of Iran [IRI] (1999) Human Development Report of the Islamic Republic of Iran,
1999. Tehran: Plan and Budget Organization and United Nations.
Islamic Republic of Iran [IRI] (1990) [1369]. Labour Law and Social Security of the Islamic
Republic of Iran. Tehran: Korshid Publications (in Persian).
Islamic Republic of Iran [IRI] (1997) Statistical Yearbook 1375 [1996]. Tehran: Central
Statistical Office.
Jenson, Jane (1986) ‘Gender and Reproduction: Or, Babies and the State’. Studies in Political
Economy, No. 20 (Summer): 9–46.
Joseph, Suad (ed.) (2000) Gender and Citizenship in the Middle East. Syracuse University Press.
Kaar, Mehrangiz (1996) ‘Women and Personal Status Law in Iran: An Interview with
Mehrangiz Kaar’. Middle East Report, No. 198 (January–February): 36–8.
Karshenas, Massoud (1997) ‘Macroeconomic Policy, Structural Change, and Employment in
the Middle East and North Africa’. In Azizur Rahman Khan and M. Muqtada (eds)
Overcoming Unemployment. London: Macmillan.
Karshenas, Massoud and Valentine M. Moghadam (eds) (2005) Social Policy in the Middle East:
Economic, Political, and Gender Dynamics. Basingstoke: Palgrave Macmillan.
Kerber, Linda (1980) Women of the Republic: Intellect and Ideology in Revolutionary America.
Chapel Hill: University of North Carolina Press (1st edn).
Kessler-Harris, Alice (2001) In Pursuit of Equity: Women, Men, and the Quest for Economic
Citizenship in 20th Century America. New York: Oxford University Press.
Koven, Seth and Sonya Michel (1993) ‘Introduction’. In Seth Koven and Sonya Michel (eds)
Mothers of a New World: Maternalist Politics and the Origins of Welfare States. New York: Routledge.
Kwon, Huck-ju (ed.) (2005) The Developmental Welfare State and Policy Reforms in East Asia.
Basingstoke: Palgrave Macmillan.
Ladd-Taylor, Molly (1994) Mother-Work. Women, Child Welfare, and the States, 1890–1930.
Chicago and Urbana: University of Illinois Press.
Lewis, Jane (1980) The Politics of Motherhood: Child and Maternal Welfare in England,
1900–1936. London: Croom Helm.
Loewe, Marjus (2000) ‘Social Security in Egypt: An Analysis and Agenda for Policy Reform’.
Background Paper for the 7th Annual Conference of the ERF, Amman (26–29 October).
Marshall, T. H. (1964) Citizenship and Social Class. Cambridge: Cambridge University Press.
Mehrpour, Hosein (1995) ‘A Brief Review of the Woman’s Economic Rights in Iranian Legal
Systems’. Farzaneh, Vol. 2, No. 7 (Fall/Winter): 49–64.
Messkoub, Mahmood (2005) ‘Constitutionalism, Modernization, and Islamization:
A Political Economy of Social Policy in Iran’. In M. Karshenas and V. Moghadam (eds)
Social Policy in the Middle East. Basingstoke: Palgrave Macmillan.
Meyer, Mary K. and Elisabeth Prugl (eds) (1999) Gender Politics in Global Governance. Lanham
MD: Rowman & Littlefield.
Michel, Sonya and Rianne Mahon (eds) (2002) Child Care Policy at the Crossroads: Gender and
Welfare State Restructuring. New York and London: Routledge.
Moghadam, Fatemeh E. 1994 ‘Commoditization of Sexuality and Female Labor Participation
in Islam: Implications for Iran 1960-90’. In Mahnaz Afkhami and Erika Friedi (eds) In the Eye
of the Storm: Women in Post-Revolutionary Iran. Syracuse, NY: Syracuse University Press.
Moghadam, Valentine M. (1988) ‘Women, Work and Ideology in the Islamic Republic’.
International Journal of Middle East Studies, Vol. 20 (2) (May): 221–43.
Moghadam, Valentine M. (1998) Women, Work and Economic Reform in the Middle East and
North Africa. Boulder CO: Lynne Rienner Publishers.
Moghadam, Valentine M. (2000) ‘Hidden from History? Women Workers in Modern Iran’.
Iranian Studies, Vol. 33, No. 3–4 (Summer/Fall): 377–401.
Moghadam, Valentine M. (2003) Modernizing Women: Gender and Social Change in the Middle
East. Boulder CO: Lynne Rienner Publishers (2nd edn).
Orloff, Ann Shola (1993) ‘Gender and the Social Rights of Citizenship: The Comparative
Analysis of Gender Relations and Welfare States’. American Sociological Review, Vol. 58
(June): 303–28.
Orloff, Ann Shola (1996) ‘Gender in the Welfare State’. Annual Review of Sociology, Vol. 22:
51–78.
Pateman, Carole (1988) ‘The Patriarchal Welfare State’. In Amy Gutmann (ed.) Democracy and
the State. Princeton NJ: Princeton University Press.
Povey, Elaheh Rostami (2004) ‘Trade Unions and Women’s NGOs: Diverse Civil Society
Organizations in Iran’. Development in Practice, Vol. 14, Nos. 1/2 (February): 254–66.
Poya, Maryam (1999). Women, Work, and Islam. London: Zed.
Richards, Alan and John Waterbury (1996) A Political Economy of the Middle East. Boulder CO:
Westview (2nd edition).
Sainsbury, Diana (1996) Gender, Equality and Welfare States. Cambridge: Cambridge
University Press.
Shahidian, Hammed (2003) Women in Iran: Gender Politics in the Islamic Republic. Westport
CT: Greenwood Press.
Sharabi, Hisham (1988) Neopatriarchy: A Theory of Distorted Change in Arab Society. New York:
Oxford University Press.
Skocpol, Theda (1992) Protecting Soldiers and Mothers: The Political Origins of Social Policy in the
United States. Cambridge MA: Harvard University Press.
Sonbol, Amira al-Azhary (2003) Women of Jordan: Islam, Labour, and the Law. Syracuse NY:
Syracuse University Press.
Tzannatos, Zafiris and Iqbal Kaur (2002) ‘Welfare State Policies in the Middle East and North
Africa’. In Ethan P. Kapstein and Branko Milanovic (eds) When Markets Fail: Social Policy
and Economic Reform. New York: Russell Sage Foundation.
UNESCO (1994) Education for All: Status and Trends 1994. Paris: UNESCO.
UNDP (2004) Human Development Report 2004. New York: OUP.
Welchman, Lynn (2001) ‘Capacity, Consent and Under-Age Marriage in Muslim Family Law’.
The International Survey of Family Law, 2001 Edition. Cambridge: Cambridge: University
Press.
5
Gender Equality and Developmental
Social Welfare in South Africa
Shireen Hassim
Introduction
Mainstream welfare-state literature identifies four welfare-state regimes: liberal,

social democratic, corporatist and state-socialist. More recently, a fifth category
has emerged: the developmental welfare state. As with research on Esping-
Andersen’s categorization of ‘three worlds of welfare capitalism’, early research on
developmental social policy is revealing enormous variants within this category as
well (Mkandawire 2004). Feminist studies of developmental welfare are poorly
developed, and thus far we have very few analyses of how gender is conceptualized.
Yet, like other welfare state regimes, developmental states are regulatory in nature:
they ascribe meaning to the social category of gender and create a normative
framework within which needs are adjudicated and considered to be worthy of
attention. While welfare states in the north traditionally redefined the relation-
ship between work and family, developmental states operating in the context of
informalization of labour markets redraw the boundaries between and responsi-
bilities of state, community, families and individuals. Analyses of developmental
states have focused on the types of need that ought to be prioritized, and the
extent to which these needs could be satisfied within a range of fiscal and global
constraints. However, there has been less attention to the ways in which different
types of developmental states interpret needs and particularly to the gendered
nature of needs interpretation.
In seeking to address the challenges of high levels of poverty, the African
National Congress (ANC) government has explicitly adopted the approach of
developmental welfare – indeed South Africa is regularly described as ‘probably the
developing world’s largest and most generous welfare state’ (Bernstein 2005).
Unusually, South African policy frameworks are underpinned by very strong for-
mal commitments to gender equality, as social rights and gender equality have
both been conceived as integral to citizenship in democratic South Africa. The
constitutional obligations to gender equality are enacted by the creation of an
institutional framework (the national machinery for women) to ensure the inclu-
sion of gender equality concerns in policy formulation. In the early 1990s,
women’s organizations, led by the Women’s National Coalition, argued for the
109
110 Shireen Hassim
greater representation of women in elected political bodies (Hassim 2002). As a

result of the successes of these strategies, the country has consistently been among
the highest performers in the world in terms of the numbers of women elected to
political office. These developments created the expectation that social and
economic policies would follow suit in placing the expansion of women’s social
rights at the core of government priorities. However, South Africa’s political
representation performance in the ten years since the inception of democracy
significantly outstrips its performance in improving women’s economic position.
On the Gender and Development Index (GDI) of the United Nations, South Africa
ranks 90th out of 144 countries. Clearly, political presence does not necessarily
mean that poor women’s interests will be adequately addressed in economic and
social policy.
South Africa provides a useful case study for the examination of the gendered
nature of developmental welfare. In this chapter, I consider the impact of
processes of democratization on the structure and ideology of welfare institutions
in the post-apartheid South Africa. I analyse the extent to which these processes
have expanded women’s citizenship entitlements. Has democratization shifted
the nature of state regulation and conceptions of need and entitlement? How are
interpretations of need concretely manifested in policies and programmes? I argue
that while women have made enormous strides in gaining recognition for their
particular political disadvantages, there has been slower translation of political
rights into social rights. The welfare system remains constrained by narrow con-
ceptions of the role of the state and by distrust of rights-based demands on state
resources. These have impacted on the extent to which social inequalities of gen-
der are eroded by the democratic state.
Democratizing social welfare
The democratic government inherited a state that from a social policy perspective
was unique in Africa. Social assistance was already inscribed as state responsibility,
albeit in a racially discriminatory form, prior to the creation of apartheid. State
pensions, the major area of government spending on welfare, have always been a
non-contributory form of social assistance and although means tested were
awarded regardless of work status and with no reciprocal obligations on recipients.
Pensions were supplemented by a range of other non-contributory, means-tested
social assistance mechanisms. In general, as Lund et al. (1996: 97) have characterized
it, the South African system is a ‘strange combination of mainly British welfare tra-
dition and apartheid policy’ where the distinction between the ‘deserving’ and the
‘undeserving’ poor is marked by race (Van der Berg 1997). Race was regarded as
shaping lifestyle patterns as well as entitlements. Under apartheid, African exclu-
sion from welfare was justified on the grounds that ‘people accustomed to modern
lifestyles and consumption patterns had greater need of social protection than those
in rural subsistence agriculture, who were not proletarianized and were thus
presumed to be better placed to meet traditional subsistence needs’ (Van der Berg
1997: 486).
Gender Equality and Developmental Social Welfare in South Africa 111
The provision of pensions and other grants (primarily child maintenance and
disability) expanded in the period between the two world wars partly in response
to unionized white working-class struggles and as a mechanism for addressing the
‘poor white’ problem. Inclusion of poor whites (mainly Afrikaners) was a corner-
stone of the Afrikaner nationalist movement and the creation of safety nets
targeted at this group legitimated ideological processes of building a racially
exclusive state system. Africans – even those who were workers in the urban
formal economy – were excluded from welfare provisions on the grounds that the
burdens of social reproduction would be carried by extended families. Indians
were regarded as temporary residents in South Africa; a repatriation policy was still
adhered to formally. On the basis of recommendations of the Carnegie
Commission of Enquiry, which sat in 1929, a state welfare department was
established for the first time in 1937.
African urbanization and the rise of urban slums created enormous anxieties for
the state during this period, although the response was not the comprehensive
welfare system that developed to support poor whites. Over time some benefits,
such as pensions and grants for the blind and disabled, were extended to African
and Indian people (men and women) although benefits were pegged at different
levels on the basis of racist assumptions about the basic needs of different groups.1
For example, African pensions were set at one-tenth the amount for whites and
the means test employed for Africans was far more stringent (Van der Berg 1997:
487). Nevertheless, the acceptance of the idea that all groups were entitled to some
benefits is startling given the denial of political and civil rights to all but whites.
During the apartheid period, the racial spread of welfare benefits was retained but
the gap in the levels of benefits between whites and Africans widened considerably.
Apartheid also began to be inscribed not just in the ideological premises of policy
but also in the ways in which services were delivered to citizens. Administratively,
there was a shift in the 1950s from a single Welfare Department in government for
the whole population to separate departments for different race groups, although
all departments were governed by the same welfare legislation (Lund 1988: 24).
Lund (1988) points out that not only did the apartheid government force private
voluntary agencies to stop offering services on a racially inclusive basis, it also
insisted that voluntary welfare bodies not have racially mixed management com-
mittees. State subsidies to private welfare agencies stipulated racially discriminatory
salaries for social workers.2
The 1950s were also a period of heightened political mobilization against the
unfolding of the apartheid state, with women being among the most vociferous
critics of attempts to control the movement of African people and to limit African
women’s access to the labour market through the imposition of pass controls. An
independent, non-racial national women’s movement, the Federation of South
African Women (FSAW) emerged that aligned itself with the ANC and linked
women’s struggles for emancipation with those of the national liberation
movement. The FSAW played a central role in redefining women’s political roles
away from being ‘the tea ladies of the struggle’ to articulating concrete political
demands, within a radical motherist frame (Wells 1992). The ideology of
112 Shireen Hassim
motherhood and the political language of ‘motherism’ became firmly anchored in

both the women’s movement and the national liberation movement, a defining
trope in nationalist discourses on gender. The strength of this approach increased
in the face of apartheid’s denial of rights to family for Africans, and the state’s
programmes to reduce African women’s fertility, introducing Depo Provera and
IUDs through the state health system during the 1960s.
Motherism had a powerful impact on the language of social policy in South
Africa that persists into present policy discourses. In the anti-apartheid movement,
motherism was articulated as a radical and empowering discourse and as a means
through which women could gain recognition for their gendered responsibilities.
The Women’s Charter of Demands, adopted by the FSAW in 1954, echoed policy
reforms being promoted by proponents of the welfare state in Britain. The Charter
demanded that the state provide four months’ maternity leave on full pay for
working mothers, together with maternity homes, antenatal clinics, child welfare
centres, crèches, nursery schools and birth-control clinics. Its list of radical
demands included subsidized housing and food and provision of a range of basic
services, a minimum wage and the banning of nuclear and atomic bombs. These
very concrete demands went beyond a general political demand for the extension
of political citizenship (Walker 1981) and reflected the importance placed by
women on the creation of an inclusive welfare state.
Under apartheid, the main grant for child and family care was the state mainte-
nance grant (made up of a parent allowance and a child allowance), which was
awarded on a means-tested basis to certain categories of women and their chil-
dren. As with other grants, the state maintenance grant was awarded on a racially
differentiated basis with whites receiving the highest amount followed by Indian
and Coloured people at the same level, followed by African people at the lowest
end of the scale. It was also unevenly administered under the racial and home-
land-segregated welfare delivery system: some administrations did not award the
grant and some awarded only the child allowance component. Although African
families constituted the majority of poor households, most African families did
not benefit from the grant (only two per 1000 African children received the grant);
they were largely excluded through a range of administrative measures. For
example, the homelands and ‘independent’ states such as the Transkei did not
administer the grant, rendering vast swathes of the African population without
access to social welfare. The application of a means test resulted in many white
people being filtered out of the system due to their higher overall income. By con-
trast, grants to Indians and Coloureds increased dramatically in the 1970s and
1980s as the state embarked on a reform programme that would offer limited citi-
zenship rights to these groups. The majority of the beneficiaries were Coloured
and Indian families – 48 per 1000 children and 40 per 1000 children respectively.
In the mid-1980s, welfare policy was revised to erode the ‘welfare statism’ that
was seen to have permeated policy. The Department of Constitutional Planning’s
Directorate of Social Planning (an interesting location for welfare policy that
reflected the increasing links between meeting social welfare needs and the
political and military objectives of the government (Patel 1992: 41)) issued a
report recommending further racial segregation of welfare, the privatization of

welfare provision (‘the state would act as a safety net only where individuals,
communities and the private sector were unable to take on new roles and respon-
sibilities’ (Lund 1988: 25)) and the devolution of welfare provision to provincial
and local authorities. By 1990 the government was well on the road to downscaling
its commitment to the provision of social welfare even though reforms aimed at
deracialization increased expenditure on African people dramatically.
The provision of welfare was not only inflected with assumptions about race but
also with very fixed, Eurocentric views about gender roles. Throughout the
apartheid era the welfare system retained the view that ‘people would live in two-
generational nuclear families with a male head of household; that there would be
full employment in the formal waged economy and that women would be at
home’ (Lund et al. 1996: 97). In reality a wide variety of household forms coex-
isted both between and within population groups, and access to the labour market
was racially differentiated. Influx controls and migrancy resulted in urban workers
(both female and male), having attachments and responsibilities to more than one
household. While unemployment was virtually non-existent for whites due to
preferential access to the labour market, it was highly prevalent among Africans.
The result was that whites could access a range of supplementary economic bene-
fits in the private sphere (for example, medical and unemployment insurance,
workers’ compensation and retirement provisions) that were unavailable to Africans.
In part because of high levels of unemployment, pensions came to assume signif-
icant importance in poor households as a source of household security. Although
pensions are allocated to individuals, they are consumed as a household asset thus
having redistributive implications. Case and Deaton (1996: 23) found that pensions
reached almost three times as many women as men. Their study found that
23.7 per cent of African households received an old age pension and that 66.4 per cent
of pensions went to households in rural areas, the location of the poorest house-
holds (cited in Lund et al. 1996: 102). Women draw a pension at age 60, men at
age 65; this has added to the evidence that pensions are a gender-sensitive mech-
anism of redistribution. However, as I will argue below, there are systemic barriers
to women’s ability to access social assistance that may limit the gender sensitivity
of all grants.
In the late 1980s, the government adopted the principle of welfare privatization.
The National Party government argued that welfare had to be understood as a
partnership between the state, the private business sector and voluntary religious
and community associations. It therefore encouraged ‘community and individual
responsibility for meeting needs through market mechanisms, emphasizing
volunteerism, mutual aid, reciprocity between providers and consumers, fees for
service and private practice in welfare provision’ (Patel 1992: 45). In the process,
no doubt the government hoped to deflect political opposition to the racialized
allocation of public resources. However, among the highly mobilized communities
in African townships, as well as Indian and Coloured women who were the major
recipients of welfare, the idea of privatization of welfare was anathema. Similarly,
cash-strapped and overburdened providers of social services assumed that the
114 Shireen Hassim
advent of democracy would result in an expansion rather than contraction

of state responsibilities. Leila Patel’s survey among non-state, grassroots providers
of social services in the early 1990s found that 75 per cent opposed privatization of
social welfare and advocated the expansion of state responsibility in financing
social services (Patel 1992: 102).
Despite considerable discussion in anti-apartheid organizations during the late
1980s about policy alternatives that would be democratic and redistributive, and
despite the social-development initiatives of grassroots organizations such as the
civics movement that dealt with problems of crime, alcoholism and social conflict
in street committees and ‘people’s courts’,3 there seems to have been little discus-
sion about what role welfare specifically would play in meeting basic needs and
redistributing resources. Discussions of the relationship between gender equality
and social policy were even more rare, as feminists were generally marginal to
political and policy debates. Discussions of economic policy occurred in isolation
from debates about social policy, even though as late as 1990 Francie Lund was
raising the question of the affordability of the general British model of the welfare
state in South Africa and calling for a holistic discussion of social policy. It is there-
fore not surprising that the period since 1994 has not been characterized by an
unambiguous redistributive policy vision for social welfare. Rather than locating
social policy strongly and unambiguously within a redistributive framework, in the
democratic period the bulk of government efforts has focused on linking welfare to a
policy of developmentalism. In the next section I examine the gendered assump-
tions behind this approach, examining the implications of the developmental
approach for the ability of women to access their citizenship entitlements.
Although a highly advanced social security system was in place by 1994, there
were gaps in this system: pensions covered by far the bulk of the welfare budget,
while other grants (disability, parents’ allowances and child support) were
relatively small. The rapid inclusion of women into government also resulted in
the women’s movement being able to put through a range of new legislation, such
as the Domestic Violence Act and the Maintenance Act, both of which required sub-
stantial increases in budgetary allocation if they were to be effectively implemented
(Vetten 2005).
The notion of developmental social welfare
The concept of developmental social welfare was first outlined in the

Reconstruction and Development Programme in 1994 and is embedded in the
White Paper on Social Welfare, gazetted in February 1996 and adopted in 1997.
The approach is defined in the White Paper as being to create a welfare system
‘which facilitates the development of human capacity and self-reliance within a
caring and enabling socio-economic environment’ (Government of South Africa
1997: ch. 2). Recognizing that economic growth in itself will not enhance the
social and economic well-being of citizens, the White Paper argues for ‘the equitable
allocation and distribution of resources … . Social development and economic
development are therefore interdependent and mutually reinforcing’. The link
between social and economic development is not mere rhetoric. A key plank in the
developmental social welfare platform is the creation of employment; this is a
responsibility of macroeconomic policy in general but also specifically of the
Department of Social Development (previously Welfare and Population
Development) through the Community Based Public Works Programme, funded
out of the social security and welfare budget. Public works programmes are seen to
have the twin benefits of addressing the infrastructural needs of the country as
well as reducing poverty and long-term dependence on state assistance. They are
therefore seen as developmental in their impact, and considerable resources have
been directed into these programmes.
The White Paper identifies a wide and impressive set of guiding principles, 17 in
all, including democracy, human rights, justice, transparency and accountability.
An examination of the list of principles suggests the policy is based on wide-ranging
and expansive conceptions of citizenship. In the particular formulations of the
guiding principles, it is compatible with the capabilities approach developed by
Amartya Sen, which focuses less on the elaboration of the entitlements of citizen-
ship and much more on whether all members of society are capable of achieving
an enhanced quality of life. For example, non-discrimination is specified in
addition to equity; where equity refers to the distribution of material resources,
non-discrimination focuses on ‘tolerance, mutual respect, diversity, and the inclu-
sion of all groups in society’. The policy is based on the principle that the quality
of life of all people should be raised through a redistribution of resources and ser-
vices. In yoking the cultural concept of ubuntu (humanity) the White Paper signals
the importance of cultural norms and values, particularly the principle of caring
and mutual interdependence, to the project of development. Elsewhere, in the
‘Agenda for Action’, the policy emphasizes the need for government programmes
to ensure the realization of citizens’ ‘dignity, safety and creativity’.
This expansive notion of development initially found resonance and widespread
support in civil society. The White Paper was developed under the new ‘rules of
the game’ with regard to public decision-making introduced by the ANC govern-
ment, in which policies would be developed through extensive consultations
between government and civil society, and where public policies in a range of
areas – the economy, health and welfare, trade and so on – were assumed to be
synergistic. The White Paper was developed in a context of general optimism
about the ability of the state to lead a process of transformation and a faith in the
democratic process.
As a policy document produced out of compromise between different interests,
the White Paper inevitably embodies numerous tensions that are left to particular
programmes to deal with. From a feminist perspective, the most noteworthy tension
is between the capabilities approach and neoliberal caveats. Social assistance pro-
grammes were to be based on a principle of affordability and sustainability – that is,
they were to be ‘financially viable, cost efficient and effective’. These neutral terms
obscure the extent to which welfare budgets are vulnerable to the imperatives of
‘fiscal responsibility’. The plan of action for the White Paper recognized the fiscal,
economic and infrastructural constraints on government’s ability to implement
116 Shireen Hassim
the principles of developmental social welfare. Social security provisions were thus
to be ‘phased in’ on the principle of the progressive realization of benefits (this
wording conforms to the constitutional provisions on social rights) as well as
sustainability. The issue of how rapidly this would be done and what the targets
for inclusion would be are not specified in the White Paper and, like the question
of affordability, would be intensely contested in the welfare debates that ensued.
Despite the emphasis on participation, poor women were relatively poorly orga-
nized and had very little voice in national-level policy debates, making it unlikely
from the outset that they would have much power over budgetary choices.
Although developmental welfare seeks to provide a third way between inequality-
enhancing residualist systems and expensive social democratic systems,
commitments to social justice are relatively weak. There is an inbuilt normative
choice in the emphasis on public works programmes as opposed to expansion of
the scale of welfare benefits that sets up a two-tier system of benefits, with people
on work-related programmes treated as ‘deserving’ poor and those on welfare (and
particularly mothers drawing the child support grant), as either passive and depen-
dent subjects or cunning exploiters of the system. Reflecting on the notion of
developmental social welfare in 2003 in the context of a debate on the Basic
Income Grant, Ravi Naidoo of the trade-union research agency NALEDI argues
that the ‘trendy concept of developmental social welfare has failed miserably’ and
that it has embedded notions of ‘an undeserving poor’ and ‘promoted ways for the
“able-bodied” to pull themselves up by their own bootstraps’ (Naidoo 2003).
The White Paper names the importance of the non-state welfare sector (religious
organizations and non-governmental organizations) as well as informal special
support systems, including community care, in meeting social service needs.
Again, this is posed in the White Paper in the spirit of ‘national collective respon-
sibility’, but it leaves open the question of the precise balance between the different
sectors of welfare provisioning. In this respect the emphasis on the cultural value
of caring might perhaps be seen as loading the dice against women, who bear the
practical burdens of care-work within families and communities. As Lund et al.
(1996: 115) note, ‘a double equation is at work which assumes that community
care is equal to care by families which is equal to unpaid care by mostly women’.
Women’s caring burdens have dramatically increased as the HIV/AIDS infection
rates have assumed pandemic proportions. Indeed, early evidence from the
pandemic is showing that it is not only women who are carrying an even greater
burden of caring but that children are increasingly having to take on these roles.
The shift away from the language of rights and entitlements in the White Paper
would seem to dilute the particular (and greater) responsibility of the state in
meeting social security needs through the redistribution of public resources.
A final area of tension in the White Paper relates to the ways in which ‘the
family’ is invoked in the document. Under the subheading ‘The Family’,
the White Paper simply states: ‘The family is the basic unit of society. Family life
will be strengthened and promoted through family-oriented policies and pro-
grammes’. It is not difficult to understand why ‘strengthening family life’ is a
desirable goal for many in South Africa, given the ways in which the migrant
labour policies of the colonial and apartheid state, and the dominance of
residentially based domestic work denied basic human comfort and intimacy to so
many Africans. The White Paper leaves open, perhaps intentionally, the definition
of what constitutes family (while stipulating that social security provisions should
include homosexuals). These specificities are left to particular policies to articulate
and while in some instances (such as the Lund Committee on the implementation
of a child support grant) opportunities were taken to shift away from nuclear,
male-headed family forms as the norm, this approach is not guaranteed in the
overarching policy framework, leaving perhaps too much to the political will and
ideological perspectives of particular policy-makers.
The increasing centralization of macroeconomic decision-making from 1996
undermined the assumptions of consultative, participatory decision-making
assumed by the White Paper. For women’s organizations, the ability to leverage
the symbolic power and legislative representation of women into policy outcomes
was severely undermined. Government’s assertion of fiscal restraint introduced a
new discourse into policy-making: the debate because less concerned with what
was desirable and more concerned with what was possible. Affordability was often
assessed in narrow fiscal terms and by prioritizing gross inequalities rather than a
concern with the long-term costs of failing to address pervasive systemic inequal-
ities. The formal provisions of the constitution proved inappropriate in dealing
with the ways in which government prioritizes spending. Although the right to
social security is entrenched in the socioeconomic-rights clause in the constitu-
tion (section 27) the implementation of this right is by no means automatic, nor
does it guarantee that the extent of social security provided will be adequate to
ensure a decent standard of living. The important proviso to the right to social
security is a qualifying clause in section 27 of the Bill of Rights, which defines the
state’s obligations as limited to ‘available resources’ (Government of South Africa
1996, section 27).
The immediate consequences of the emphasis on affordability were seen in the
process of overhauling the system of child and family benefits, which the
Department of Welfare instituted in 1996 and which coincided with the public
debate on the White Paper on Social Welfare. The changes to the welfare provisions
for children began in a transitional context of translating broad policy formula-
tions into concrete programmes, and the deliberations of the Lund Committee,
which spearheaded the changes, reveal the tensions between pursuing equity (by
ensuring that poor African families would benefit from social assistance) and
affordability.
By July 1996, poor mothers were paid a total monthly grant of R565 (made up
of a parent allowance of R430 and a child allowance of R135) for a maximum of
two children up to the age of 18 (that is, to a maximum of R700). For those fami-
lies receiving it, the grant played a major poverty-alleviation role in raising overall
household income above the household subsistence level. However, as noted
above, the grant reached very few African households, with Indian and Coloured
families being the primary beneficiaries. Attempts to entrench the racial equity of
the grant produced the perverse consequence that while more poor families
118 Shireen Hassim
received the grant the monetary value of the grant was drastically reduced. The
Committee had to work within the government’s decision not to increase overall
spending. At the time, government had already cut defence spending significantly
(virtually halved between the 1990/91 and 1996/97 fiscal years) and increased
social security spending by 120 per cent over the same period (Gelb 2004). There
was a strong perception in the Committee that no further increases were likely and
that the demands on the existing budget were likely to increase under the impact
of the HIV/AIDS epidemic. The Committee’s recommendations were therefore
based on a fiscally constrained scenario in which the already inadequate welfare
budget faced further cuts from the central government.
A number of concerns have surfaced with regard to the effective implementa-
tion of the child support grant. Firstly, women’s organizations have criticized the
two-tier means test to establish eligibility for the grant. The means test requires
that the primary caregiver prove that he or she is a member of a household with a
combined income below R9,600 (approximately $1,300) per annum for urban
households and R13,200 (approximately $1,800) per annum for rural dwellings or
those in informal areas. Other elements of the means test include a requirement
that the primary caregiver shows that she/he is actively seeking employment. The
Lund Committee was itself not convinced of the benefits of means testing, favouring
instead a universal grant. The Committee noted that the means test imposed
administrative costs on the system as it required extensive assessment and
monitoring and imposed costs on parents, which could act as a negative incentive,
particularly for foster parents. The Committee also warned that means testing
could be a biased process. ‘Communities, especially in poorer areas, are so greatly
under the domain of traditional leaders with extensive powers of patronage that
caution should be exercised in this approach. The track record of civic associations
in impartial decision-making is likewise uneven’ (Lund Committee 1996: 29).
However, Department of Welfare officials insisted on the retention of this mecha-
nism, concerned that there might otherwise be overwhelming numbers of people
applying for the grant and that many of them might be ‘unworthy’, at least in
economic terms. The Committee stressed that should a means test be adopted, it
should be a simple economic test and should not ‘in any way depend on a defini-
tion of a family’ (Lund Committee 1996: 92). However, the old means test appears
to still be in place. Recent research shows that many welfare officials informally
add on new ‘tests’ of eligibility, such as proof of tax being paid, and letters certified
by police officers (Goldblatt 2005). Anecdotal evidence suggests that welfare offi-
cials turn mothers who are overtly lesbian away. The South African NGO Coalition
(SANGOCO) has called for the existing means test to be replaced with one that is
based on the income of the primary caregiver, arguing that this will be easier to
administer. As Liebenberg points out, ‘this is particularly important in view of
the fact that the child support grant requires a doubling of the capacity of the
welfare system to process grants. The present system is already over-burdened with
huge backlogs in poverty-stricken areas’ (Liebenberg 1998: 11).
These problems are exacerbated by poor management and delivery systems,
and in some cases corrupt practices at the provincial level, which initially led to
under-spending of welfare budgets for three consecutive years, with the child
support grant showing the slowest take-up rate (EPRI 2002). In an assessment of
the effectiveness of the grant commissioned by the Department of Welfare and
conducted in 2000, the Community Agency for Social Enquiry (CASE) found that
the capacity and competence of the Department of Welfare to administer the grant
was limited by a range of problems: insufficient information, outdated application
forms, lack of coordination between the Departments of Welfare, Health and
Home Affairs, poor departmental cooperation with NGOs and community organi-
zations and by ‘indifferent and even hostile attitudes on the part of Welfare staff’
(CASE 2000: 120). The government has sought to deal with some of the adminis-
trative problems related to delivering grants by privatizing grant payments and
creating a common agency for grants payments that would take back control over
grants payments from the provincial to the national level of government. This has
by no means made the grants more accessible; indeed many pensioners have com-
plained that queues remain as long as ever and that the private companies are
using the payment process as an opportunity to sell insurance policies and even
short-term loans to pensioners. The provinces with the best resources (Gauteng
and Western Cape) have been most successful in reaching delivery targets, while
those with the greatest need (Eastern Cape and Limpopo) also have the least effective
delivery systems and have been the least successful in reaching their targets
(Guthrie 2002).
Most street children and children in child-headed households cannot access the
grant because they do not have the necessary identity documents to apply, or they
do not know the procedures for application. The Alliance for Children’s
Entitlement to Social Security (ACESS) estimates that 75.8 per cent of South
African children live below a poverty line of R400 per month. Many NGOs have
recommended increases in the level of the grant, but noted that ‘this would
require a political decision involving a trade-off with other grants and budgetary
items’ (Guthrie 2002: 122). The government has been responsive to these recom-
mendations, improving information about the grants and increasing the amount
and age limit of the grant. By 2003, close to 3 million poor children were
receiving the grant (Woolard 2003) and by December 2004 this number had
increased to 5.4 million, despite the administrative difficulties described above
(Taylor 2004).
A major discursive shift introduced by the Lund Committee was the move away
from the ‘familist’, male-worker model of social policy. The Committee chose to
adopt a ‘follow the child’ policy, with emphasis on the child rather than the carer –
in other words, the grant would be paid to the primary caregiver on behalf of the
child.4 Means-testing would therefore no longer comprise any component of
moral assessment of the worthiness of carer (at least in theory); nor would carers
who were not mothers be excluded from the grant. The Committee adopted the
policy in recognition of the fluidity and diversity of household forms. This was a
significant shift away from the normative assumption that the nuclear family was
the desirable model, which was inscribed in apartheid-era social welfare; and,
importantly, it addressed the rights of children born into polygamous families.
120 Shireen Hassim
At least in formal terms, the effect was to de-link childcare from normative
assumptions about family forms or gender responsibilities and particularly to de-
link care work from mothering. As the Committee noted, ‘it resolves the problem
of how to define the family in such a complex and multi-cultured society. It says
that children, however many in a household, of whatever status, are important
and need to be protected’ (Lund Committee 1996: 91).
Of course, these shifts in the core assumptions of the child support grant do not
as yet translate into shifts in the views of welfare officials and the media. In reality,
moral discourses continue to infect social security provision, crowding out rights-
based arguments for social security. In debates about the child support grant, there
was a strong view among many in government that welfare grants reinforced
apartheid privileges, had no developmental potential and should be phased out in
favour of greater attention to programmes such as community-based public works
(Francie Lund, personal communication). Even some ANC women members of
parliament took a conservative view of welfare as reinforcing a ‘culture of entitle-
ment’, with welfare grants seen as handouts that reinforced dependency on the
state. As one MP argued, ‘women should look at developing themselves’
(Parliament of South Africa 2000). Fraser-Moleketi herself accused poor people of
not doing enough: ‘communities had to change the thinking of those who held
out their hands for help but kept their sleeves down, a sign that they were not
willing to work’ (Koopman 1999). The child support grant continues to be blamed
for increasing teenage pregnancy (that is, that young girls are getting pregnant in
order to access cash), that women are spending the money on ‘clothing and
lipstick’ and even an allegation by the Minister of Social Development that mothers
have ‘rented’ out their children to others so that they can claim social assistance.
Newspaper reports castigate ‘runaway mothers’ who ‘claim the child support grant
meant to feed their offspring’ (Rhupiso 2003). Even the Chairperson of the
Commission on Gender Equality had to be gently reminded by feminist activists
not to fall into the trap of stereotyping women receiving the child support grants
as undeserving.
The Basic Income Grant
The Lund Committee’s brief was a narrow one, and the debates which accompanied
the release of its recommendations and subsequent problems with the implemen-
tation of the child support grant as well as major inefficiencies in the delivery of
all social grants have reopened larger questions about the overall system of social
assistance. A central concern among many welfare activists was the narrow reach
of social security. However redistributive in effect, by their nature pensions only
reach a limited number of people. Although by 2004 uptake of the child support
grant outstripped government expectations, at the moment poor children over
14 do not receive support. Up to 60 per cent of the poor – mainly those
between the ages of 14 and 60 – are not getting any social security at all. It is
estimated that 11.8 million of the poorest 23.8 million South Africans live in
households that receive no social assistance (Lund 2004). The trickle-down effects
of the Growth, Employment and Redistribution strategy (GEAR) are not material-
izing; levels of growth remained well below the rate required to address underlying
social needs. Unemployment has continued to rise and employment creation
through the public works programme was disappointing. It has been estimated
that the Community Based Public Works Programme only created between 13,000
and 33,000 jobs per annum between 1996 and 2001 (McCord 2003). By the end of
the 1990s, the Congress of South African Trade Unions (COSATU) as well as a
range of civil society organizations were arguing strongly for a more inclusive sys-
tem of social security that would have poverty-reducing effects.
Poverty is, unsurprisingly, a deeply gendered phenomenon. Statistics from
South Africa’s labour force study in 2003 showed that women on the whole had
lower incomes, higher unemployment rates and less access to assets than men
(Seidman-Makgetla 2004: 2). African women make up 42 per cent of the workforce
but only 30 per cent of the employed population. Those women who are
employed find themselves in the worst-paid sectors of the labour market, notably
in domestic and retail work. In 2003, 96 per cent of domestic workers were black
(that is, African, Indian and Coloured) women and 93 per cent of these workers
earned under R1,000 (approximately US$180) per month (Seidman-Makgetla
2004: 7). African rural women are the poorest category of citizens: in 1997, 65 per
cent of African female-headed households in rural areas were poor compared to
54 per cent of male-headed households (Seidman-Makgetla 2004: 40). At 29.4 per
cent, the mortality rate among African women was more than twice that of white
women in 1994 (11.5 per cent) (Seidman-Makgetla 2004: 41). It has been estimated
that 53 per cent of South Africans, including 60 per cent of the country’s children,
live in households with the lowest per capita consumption (Samson 2002: 71). On
the United Nations’ Gender and Development Index, even though women and
men had comparable school enrolment and adult literacy ratios, men earned twice
as much as women (Budlender 2004: 10). These gender vulnerabilities are com-
pounded by the HIV/AIDS pandemic. African women are most vulnerable to HIV
infection, more women than men are HIV-positive and women are likely to
become more infected at a younger age than men (Albertyn 2003). The pandemic
imposes additional burdens on women in their roles as primary carers of family
members who are HIV positive. These caring tasks, moreover, have to be performed
in the context of poor basic services such as the availability of clean water,
electricity and modern sanitation.
To what extent could social policy address these gendered dimensions of
poverty? There have been few concrete proposals put forward by the women’s
movement. Instead, many women’s organizations have allied themselves to the
proposal for a Basic Income Grant (BIG), put forward by COSATU as a universal
poverty-reduction mechanism. COSATU envisaged that a relatively small universal
grant (R100 per month) would be introduced for all individuals (including
children). Although neither business nor government were enthusiastic about a
BIG,5 COSATU kept up pressure for the grant both within the tripartite alliance
and in the media. In July 2001 the BIG Coalition involving 12 organizations was
formed. By this time, government had appointed a Committee of Inquiry into a
122 Shireen Hassim
Comprehensive System of Social Security for South Africa (Taylor Committee) in

March 2000, chaired by feminist activist and leading member of DAWN, Viviene
Taylor.
The Committee’s brief included, among other tasks, a review of existing grants,
including social assistance and social insurance mechanisms and, after a process of
consultation, to make specific recommendations for implementation. Apart from
practical administrative and fiscal considerations, the Committee’s recommenda-
tions were also to take account of the ‘adequacy of adherence to principles of
social solidarity’ (Taylor Committee 2002: 10). The Committee released a report in
March 2002 entitled Transforming the Present – Protecting the Future, in which it
made a range of recommendations, including mechanisms to improve the institu-
tional arrangements for grant payments, streamlining existing grants and laying
out a financial framework for comprehensive social protection. Political attention,
however, focused most intensely on the Committee’s recommendation with
regard to the BIG. The Committee recommended the ‘gradual development of a
comprehensive and integrated income support’. It noted that the ‘conditions for
an immediate implementation of a Basic Income Grant do not exist. In particular
there is a need to first put in place appropriate capacity and institutional arrange-
ments to ensure effective implementation’ (Taylor Committee 2002: 63). The
Taylor Committee argued that existing social assistance schemes reduced poverty
by 23 per cent while the BIG would reduce the gap by as much as 74 per cent, with
over six million people raised above the poverty line. The Committee recom-
mended that a BIG be phased in by 2006. Although this was a somewhat guarded
recommendation, it was taken as endorsement of the views of COSATU and the
BIG Coalition, who then used it as the basis for further activism, now arguing for
implementation of the Taylor Committee recommendations.
Although the Taylor Committee report was received with acclaim by civil society
and even supported in principle by the Minister of Social Development, the ANC
voted against it at the party’s National Congress in December 2002. Senior ANC
member and government spokesperson Joel Netshitenzhe, reporting on Cabinet
debate on the Taylor report, commented that Cabinet had a different philosophy to
that adopted by the Taylor Committee. Cabinet’s view was that ‘only the disabled or
sick should receive handouts, while able-bodied adults should enjoy the opportu-
nity, the dignity and the rewards of work’ (cited in Matisonn and Seekings 2002).
The government has countered criticism of the existing welfare system by
pointing to steady increases in budgetary allocations to welfare since 2000. Servaas
van der Berg has shown that ‘the first years after the political transition saw a large
and significant shift of social spending away from the more affluent to the
formerly disadvantaged members of the population, and that most social spend-
ing is relatively well targeted to reach those most in need of it’ (Van der Berg 2001:
157). As Stephen Gelb (2004) shows in his analysis of social-sector spending
(Table 5.1), there have been striking increases in budgetary allocations to welfare,
even compared with other social-sector spending.
The most recent research on the economic and social impact of social grants,
conducted on behalf of the Department of Social Development by the Economic
Table 5.1 Government budget: size and distribution, South Africa
1990/1 1995/6 1998/9 2001/2 2002/3 2003/4 2004/5 2005/6
Education 18 21 22 20 20 20 20 19
Health 9 10 11 11 11 11 11 11
Housing, other 13 5 3 4 4 5 5 5
social services
Social services 46 46 48 48 49 50 51 51
(total)
Protection 20 17 16 17 17 17 16 16
services
Economic 14 11 9 11 12 13 13 13
services
Interest 12 19 20 17 15 13 13 13
Other 8 7 8 7 6 7 6 6
Total 100 100 100 100 100 100 100 100
Source: Stephen Gelb, Inequality in South Africa: Nature, Causes, Responses ( Johannesburg: The Edge
Institute, 2004).
Policy Research Institute (which has been associated with the BIG Coalition), shows
that rising expenditure has begun to impact positively on poverty, labour-market
participation and productivity. Minister of Social Development Zola Skweyiya
argues that ‘income support is more than a safety net for the poorest and most
vulnerable people in our society. It is also a trampoline that enables many people
in these households to jump over the barriers of economic and social exclusion’
(Ministry of Social Development 2004). The study showed a 66.6 per cent reduction in
poverty, when the destitution poverty line (a measure of relative destitution based
on the household expenditures of households in the lowest 20 per cent of the
income distribution) is used as a benchmark. The study found that the progressive
extension of the child support grant up to the age of 14 would reduce the poverty
gap by 57 per cent. Receipt of the child support grant correlates to increasing
enrolment of children in primary school, with Lund (2004) reporting a remarkable
8 per cent increase in enrolment in grant-receiving households. The EPRI study also
found that the major social grants are significantly and positively associated with
greater household expenditure on food and consequently better nutritional
outcomes. The impact of existing social grants also seems to be addressing other
areas that are of concern to the proponents of a Basic Income Grant. The study
found that ‘social grants provide potential labour market participants with the
resources and security necessary to invest in high risk/high reward job search, [and
that] living in a household receiving social grants is correlated with a higher success
rate in finding employment’ (Ministry of Social Development 2004).
The likely impacts of a Basic Income Grant on poor women
Nor surprisingly, the most prominent debates about the BIG have related to afford-
ability and administrative capacity. The impact of the Basic Income Grant on
124 Shireen Hassim
women has not been directly addressed by the major participants in the BIG
Coalition, but the proposal raises a number of issues for feminist analysts. A range
of women’s organizations have argued that the BIG should be supported by
women on the grounds that women carry the major burden of poverty, and that
‘women are often more responsible than men in using income collectively and
developmentally’ (Goldblatt 2003: 2). The universal nature of the grant may
remove the stigma attached to other social assistance grants, and the absence of a
means test will reduce the degree of moral regulation of women. However, there
are a number of assumptions embedded in arguments for the BIG that bear closer
attention. These relate to assumptions about intra-household behaviour, the
impact of a direct cash grant on women’s autonomy and the consequences of a
universal grant for enhancing women’s citizenship entitlements and shifting from
notions of female dependency. I will deal with these very briefly.
There has been little attention among BIG proponents to the internal power
dynamics within households. It is assumed that members of a household will pool
their grants and that these will then be used for the benefit of the household as a
whole. This assumes also that the interests of the household are unitary and that
there will be consensus as to the spending priorities. There is inadequate research
into internal household dynamics in South Africa and much of the discussion is
thus highly speculative. However, existing research suggests that households are
not monolithic entities in which incomes will be pooled (Posel 2001). Some of
these concerns have been raised in relation to the BIG. As Beth Goldblatt asks, ‘is
it possible or likely that lack of trust will prevent women and men from pooling
the BIG or deciding together how it should be spent?’ (Goldblatt 2003: 2). Francie
Lund (2004) has pointed out that there is a further assumption that people live in
just one household, alluding to the fact that household forms in South Africa are
fluid and diverse. Eva Harman has shown how the distribution of social grants is
‘mediated by social relations, historical dynamics and material conditions’
(Harman 2003: 3). While her research shows that grants are pooled within some
poor households, this does not mean that distribution within the households is
fair and grants may indeed increase dependency of non-grant recipients on recipi-
ents. Merely receiving a grant, such as a pension, does not mean that the recipient
will have the power to decide how the grant will be used. The elderly, for example,
may receive pensions and thereby secure their place and some degree of care from
their families, but may give up the grant to younger (possibly male) and more
powerful members of the household. From this point of view, a direct grant to
each member of the household may reduce such dependencies, but lack of power
in the private sphere may lead to grants being appropriated. Work on the spread of
HIV shows the extent to which women lack power over decision-making within
households – even where the risk of not asserting their voice is extreme illness and
even death (Albertyn 2003). Some have argued that a BIG will reduce the extent of
domestic violence as women will have access to cash and therefore independence.
This seems a far-fetched expectation.
An area in which there is more likely to be a gendered impact lies in the poten-
tial shift that would result in rural households away from remittances to reliance
on a BIG. Although migrant remittances are crucial to the survival of rural house-
holds, they are also part of a cycle of power. Dorrit Posel’s research on the varied
patterns of distribution of migrant remittances within rural households suggests
that we cannot assume that migrants remit to rural households out of altruism
(Posel 2001). Posel found that remittances are being directed not at households
but at particular individuals in households, and that migrants may be more
responsive to the needs of some household members than to others. Some
claimants within households – such as old people and children – are seen as valid
claimants whilst others (notably young women) are seen as invalid.
A universal grant may therefore have the effect of uncoupling support for rural
households from the vagaries of the remittance system and might offer rural
women greater autonomy and enable them to overcome other barriers to eco-
nomic security. Although women have yet to gain equal rights to ownership and
access to communal land, they are able to purchase other forms of land through
accessing a state subsidy (Walker 2001). The major barrier to this thus far has been
lack of cash (purchasers are required to make a cash deposit before the state will
issue a subsidy). An income grant might make it possible for women in households
where there is more than one woman to pool their income and purchase the land
they are currently working. A basic monthly income might make it easier for
women to access low levels of credit to increase their productivity. Evidence sug-
gests that there is a definite gender pattern in household spending. Lund’s work on
pensions shows that the grant is most generally put to common household use
and that, moreover, pension money that goes to women is spent better (that is, on
food, health and education) than that which goes to men (Lund 2004). Directing
cash grants to women is therefore likely to enhance the quality of life of poor
women and children. However, these assumptions have not yet been tested by
empirical research and cannot be said to decisively provide support for the idea
that a small cash grant will be empowering for women.
The paucity of gendered analysis in social assistance debates is partly attributable
to the decline in feminist activism since the advent of democracy. Gender politics
in the last ten years has centred on increasing women’s representation in various
institutions, while the original link between representation and equality outcomes
appears to have been broken. In the apartheid era, a clear line was drawn between
struggles for formal equality and those for substantive equality. Formal equality –
the achievement of equal rights and opportunities – was regarded as an inadequate
conceptualization of liberation. The achievement of formal political and civil
rights, while an important gain in itself, was understood as a weak form of equal-
ity that would have little impact on the lives of poor women. What was needed
was substantive equality, understood as the transformation of the economic con-
ditions that produce gender equality.6 The Women’s Charter for Effective Equality,
adopted by the Women’s National Coalition in 1994, articulates a notion of equal-
ity that is closer to the vision of substantive equality, with a very clear emphasis on
the structural and systemic underpinnings of women’s subordinate status.
Definitions of equality of course need to be grounded in the particular context
in which claims are being made. I would argue that a strong notion of equality,
126 Shireen Hassim
one that would provide some guidance about appropriate policy choices in South
Africa, would rest on the extent to which formal discrimination in law and policy
is reduced,7 the extent to which overall poverty is reduced, the degree to which
women have autonomy and are able to make choices free of the constraints of care
work within families and communities, as well as free of the pressure to remain in
oppressive and violent relationships (Orloff 1993: 319),8 and the extent to which
women feel safe in society. This notion of equality has specific implications for
social policy, as it would require that resources be directed in such a way that they
serve not only to address the needs of the poorest women, but also become part of
an incremental process of enhancing women’s autonomy and full participation in
political and economic processes.
In South Africa and in other developing-country contexts the conceptualization
of equality primarily in terms of labour-market participation and autonomy from
family responsibilities is flawed. Firstly, the issue of access to labour markets is
different in the context of high and chronic joblessness. Secondly, the caring
needs of middle-class women have not resulted in demands for publicly funded
systems of family support. The availability of cheap childcare and domestic work
for the middle class meant that there was virtually no demand for high quality
state-provided childcare. The economic and political processes that accompanied
demands for greater public responsibility in providing care in Northern welfare
states (Lewis and Giullari, chapter 8 of this volume) do not exist in South Africa.
In South Africa the supply (of private childcare) has outstripped demand and
consequently the bulk of care work is likely to remain within the private sphere.
Conclusions
This chapter has shown that although the transition to democracy has led to the
elaboration of a wide-ranging set of civil, political and social rights, the gendered
patterns of poverty and inequality have not been significantly reduced. The dera-
cialization of social assistance has included some poor women in the social security
safety nets, but the gap in coverage excluded the majority of poor women between
the ages of 14 and 60. The extent to which poor women have been able to access
their citizenship rights has been limited by the faltering political will to address
poverty in a comprehensive manner, by an overarching macroeconomic frame-
work that prioritizes fiscal restraint over redistribution and by an administrative
system and infrastructure that is unable to fulfil basic tasks of service delivery.
Despite some progress in using social grants as part of a developmental strategy
and despite the increasing monetary expenditure on the social sector, the type of
welfare system being built in South Africa retains many of the features of a residual
system, reacting only to the worst effects of market or family failures and provid-
ing assistance to social groups seen as ‘deserving’. Women access support only
through their children until they reach the age of 60. Provision of benefits is not
generous, benefits are not universal but income-tested, and access to benefits is
difficult. Unless the basis of entitlements changes in ways that recognize women’s
ntitlements as citizenship rights, poor women will continue to be excluded from
the system of social entitlements. Equally importantly, unless the increased

representation includes debate and activism about the meanings of gender equal-
ity in the South African context, the likelihood is the parity in representation will
increase the access of women elites rather than have the outcome of increased
gender equality.
Acknowledgement
I would like to thank Beth Goldblatt for many hours of debate about gender and welfare in
South Africa, and for her critical reading of earlier versions of this article. Thanks also to
Francie Lund for her thoughtful comments and corrections of an earlier draft.
Notes
1 Notably, however, family allowances for large low-income families, instituted in 1947,
did not include Africans.
2 Lund (1988) points out that welfare agencies accepted these changes because they were
highly dependent on state subsidies. However, some did make private arrangements to
equalize salaries of their social workers.
3 See Patel 1992, chapter 3, for a useful discussion of these initiatives.
4 However, Goldblatt (2005) argues that the removal of the parental allowance (almost
exclusively accessed by mothers) that was part of the state maintenance grant denies
women access to money that they were in the past able to claim as their own entitlement.
5 Paradoxically, the Democratic Party, which in many respects can be seen as the party of
business, has supported the introduction of BIG.
6 Elsewhere Catherine Albertyn and I have argued that even this conceptualization of
freedom is limited, as it fails to address the social and cultural dimensions of inequality.
See Albertyn and Hassim (2003).
7 Of course, certain forms of gender discrimination are both inevitable and desirable, such
as in relation to maternity and reproductive health benefits.
8 Orloff (1993) takes this argument much further in suggesting that social policies should
aim at decommodification of gender relations by enabling women to form and maintain
autonomous households. I am hesitant to apply this notion to women in the South
African context, given the particular cultural attachments and support systems that
women value within family and communities. It could also be argued that the high num-
ber of women-headed households in South Africa suggests that women are indeed free to
form autonomous households, but this has patently not empowered women to become
full and equal citizens.
Bibliography
Albertyn, C. (2003) ‘Contesting Democracy: HIV/AIDS and the Achievement of Gender
Equality in South Africa’. Feminist Studies, Vol. 29, No. 3: 595–615.
Albertyn, C. and S. Hassim (2003) ‘The Boundaries of Democracy: Gender, HIV/AIDS and
Culture’. Development Update 4(3): 138–64.
Bernstein, Ann (2005) ‘Let the Market Rule in Welfare’. Business Day, 4 May.
Budlender, Debbie (2004) ‘Women and Poverty’, at www.genderstats.org.za/poverty.html,
accessed 30 August 2005.
CASE (2000) A Social Impact Study on the Phasing-in of the Child Support Grant. Johannesburg:
CASE.
Case, A. and A. Deaton (1996) Large Cash Transfers to the Elderly in South Africa. Research
Program in Development Studies, Princeton NJ: Princeton University.
128 Shireen Hassim
EPRI (2002) ‘Social Security Reform and the Basic Income Grant for South Africa’. EPRI Policy
Report, 31.
Gelb, Stephen (2004) Inequality in South Africa: Nature, Causes, Responses. Johannesburg: Edge
Institute.
Goldblatt, B. (2003) ‘Teen Pregnancy and the Abuse of the Child Support Grant: Addressing
the Myths and Stereotypes’. Agenda, No. 79: 59–81.
Goldblatt, B. (2005) ‘Gender and Social Assistance in South Africa’s First Decade of
Democracy: A Case Study of the Child Support Grant’. Politikon, November.
Government of South Africa (1996) Constitution of South Africa.
Government of South Africa (1997) White Paper on Social Welfare.
Guthrie, T. (2002) ‘Family Social Security Benefits in South Africa’. Social Dynamics 28(2):
122–45.
Harman, Eva (2003) ‘Social Grants and their Social Circulations’. Paper presented at the
Second Annual USBIG Conference of the Eastern Economic Association, 21–23 February.
Hassim, S. (2002) ‘A Conspiracy of Women: The Women’s Movement in South Africa’s
Transition to Democracy’. Social Research, 69(3), Fall: 693–732.
Koopman, Andre (1999) ‘Poor Urged to Roll Up Their Sleeves’. Cape Times, 25 May.
Liebenberg, S. (1998) ‘Social Security and Human Rights: Current Issues’. Economic and Social
Rights in South Africa Review, 1(2): 1–2.
Lund Committee (1996) Report of the Lund Committee on Child and Family Support. Pretoria:
Government Printers.
Lund, F. J. (1993) ‘State Social Benefits in South Africa’. International Social Security Review,
46(3): 5–25.
Lund, Francie (1998) ‘State Restructuring of Welfare’. Transformation 6.
Lund, Francie (2002) ‘Crowding in Care, Security and Micro-enterprise Formation: Revisiting
the Role of the State in Poverty Reduction, and in Development’. Journal of International
Development, 14(6): 1–14.
Lund, Francie (2004) ‘Gender and Social Security: What We Know, What We Think We Know
and What We Really Don’t Know’. Paper presented at Centre for Applied Legal Studies
Workshop in Gender and Social Security, December.
Lund, Francie, with Sally Ardington and Mary Harber (1996)’Welfare’. In Debbie Budlender (ed.)
The Women’s Budget. Cape Town: IDASA.
Matisonn, Heidi and J. Seekings (2002) ‘Welfare in Wonderland? The Politics of Basic Income
Grant in South Africa, 1996–2002’. Paper presented to the BIEN 9th International Congress,
Geneva, 12–14 September.
McCord, A. (2003) ‘Public Works as a Component of Social Protection in South Africa’. Paper
presented to the Development Policy Research Unit/ Trade and Industry Policy Secretariat
Forum, The Challenge of Growth and Poverty: The South African Economy Since
Democracy, 8–10 September.
Ministry of Social Development (2004) ‘Press Release on Income Support Findings’,
5 December.
Mkandawire, Thandika (2004) ‘Social Policy in a Development Context: Introduction’. In
T. Mkandawire (ed.) Social Policy in a Development Context. Basingstoke: Palgrave.
Orloff, Ann Shola (1993) ‘Gender and the Social Rights of Citizenship: The Comparative
Analysis of Gender Relations and Welfare States’. American Sociological Review, 58(3):
303–28.
Naidoo, Ravi (2003) ‘The Basic Income Grant: Poverty, Politics and Policy-making’. Harold
Wolpe Memorial Lecture, University of Natal, August.
Parliament of South Africa (2000) Minutes of Portfolio Committee on Welfare, 13 October
2000.
Patel, Leila (1992) Restructuring Social Welfare: Options for South Africa. Johannesburg: Ravan
Press.
Posel, Dorrit (2001) ‘How do Households Work? Migration, the Household and Remittance
Behaviour in South Africa’. Social Dynamics 21(1): 165–89.
Rhupiso, Victor (2003) ‘Dumped with Daddy’. Sunday Times Metro, 23 February.
Samson, Michael J. (2002) ‘The Social, Economic and Fiscal Impact of Comprehensive Social
Security Reform for South Africa’. Social Dynamics, 28(2): 69–97.
Seidman-Makgetla, Neva (2004) ‘Women and the Economy in South Africa’, at www.
genderstats.org.za/economy.html, acccessed 30 August 2005.
Taylor Committee (2002) Transforming the Present – Protecting the Future. Report of the
Committee of Inquiry into a Comprehensive System of Social Security for South Africa. Pretoria:
Department of Social Development.
Van der Berg, Servaas (1997) ‘South African Social Security under Apartheid and Beyond’.
Development Southern Africa, 14(4): 481–503.
Van der Berg, Servaas (2001) ‘Redistribution Through the Budget: Public Expenditure
Increase in South Africa, 1993–1997’. Social Dynamics, 21(1): 140–64.
Vetten, Lisa (2005) ‘ “Show me the Money”: A Review of Budgets Allocated Towards the
Implementation of the Domestic Violence Act’. Politikon, November.
Walker, Cherryl (1981) Women and Resistance in South Africa. London: Onyx.
Walker, Cherryl (2001) ‘Piety in the Sky? Gender Policy and Land Reform in South Africa’.
Paper presented at the UNRISD Conference on Agrarian Change, Gender and Land Rights,
Geneva, 6–7 November 2001. Geneva: United Nations Research Institute for Social
Development.
Wells, Julia C. (1992) We Now Demand! The History of Women’s Resistance to Pass Laws in South
Africa. Johannesburg: University of Witwatersrand Press.
Woolard, I. (2003) ‘Social Assistance Grants, Poverty and Economic Growth in South Africa’.
Research paper, Development Policy Research Unit, University of Cape Town.
6
Social Policy Reforms and Gender in
Japan and South Korea
Ito Peng
Introduction
Feminists are sharply divided into two camps on the issue of family and gender
policies in Japan and South Korea (henceforth Korea). On the one hand are those
who claim that despite the remarkable developments in the social, economic and
political spheres, very little has changed in terms of women’s status (Moon 2002).
Proponents of this ‘no real change’ perspective point to these countries’ poor
records in the Human Development Report (HDR) indices on gender equality
(Gender Development Index or GDI, and Gender Empowerment Measure or GEM)
as evidence of the persistence of gender inequality. Indeed, both Japan and Korea
are amongst the lowest in the international rankings, and Japan has fallen even
further behind in recent years despite the government’s claimed efforts to advance
gender equality.1 Based on these figures, there is good reason for scepticism about
these developmentalist economies’ abilities to ensure gender equality. As feminist
studies have shown, economic growth in many East Asian countries has occurred
simultaneously with gender inequality in both public and private spheres
(Seguino 1997; Hsuing 1996).
On the other hand, there are those who argue that despite their poor standings
on international measures, noticeable changes have been made in terms of gender
equality in Japan and Korea (Osawa 2000, 2003; Lee and Park 2003). Rather than
focusing solely on outcome measures as the means to evaluate progress towards
gender equality, more attention needs to be paid to the processes of policy change
and to how gender issues are being incorporated into social policies.2 Evidence
suggests that a wide range of policy measures have been introduced in both Japan
and Korea since 1990 to address the issues of family and gender, and attempts have
been made in the direction of gender mainstreaming.
This chapter cautiously adopts the latter approach, examining family and gender
policy reforms in Japan and Korea in the 1990s with an aim to explaining why
these reforms have been made, and what they mean in terms of progress towards
gender equality. I argue that some important changes have been made in social
welfare policies to support family and gender equality, and that the reasons for the
130
Social Policy Reforms and Gender in Japan and South Korea 131
changes are both sociological and political. In both countries changes resulting
from demographic ageing and low fertility, greater participation of women in the
labour market and in politics, and political regime shifts have created pressures on
the government to push for gender equality. However, there remain significant
structural and institutional barriers to women’s full participation in the labour
market and in politics, and a significant gap exists between individual ideas and
policy claims about gender equality on the one hand, and actual practice on
the other. The reasons for policy changes in relation to the family and gender are
post-industrial changes and political regime shifts. The two most important post-
industrial changes are firstly, demographic ageing and declining birth rates and,
secondly, changes in family and gender relations.3 In Japan and Korea the rapid
ageing of the population combined with declining fertility has sparked fears of an
increased dependency rate, labour shortages, greater social security burden, econ-
omic slowdown, and ultimately population decline. To exacerbate the situation,
the increase in women’s labour-market participation rates and widening genera-
tional divide in the ideas about family and gender relations have brought issues of
child and elderly care, and the role and capacity of the family in providing indi-
vidual welfare, to the centre of policy debates. These have undoubtedly created
imperatives for policy change.
Both countries also experienced political regime shifts in the 1990s, which
disrupted the earlier structure of social policy-making, creating opportunities for
new actors to engage in policy-making process and for new policy ideas to be
pushed forward. In Japan political realignments following the collapse of the con-
servative Liberal Democratic Party’s (LDP) political hegemony in 1993 created a
number of opportunities for social policy reforms. In Korea the Asian economic
crisis of 1997 created openings for political regime shift and enabled subsequent
governments to initiate welfare expansion. In both Japan and Korea, important
changes in family and gender policies happened therefore in the context of polit-
ical regime shifts. I argue that while post-industrial pressures may have forced
Japanese and Korean people and governments to re-think social policies, changes
in domestic political dynamics provided important openings for social and politi-
cal actors to carry through the reform agenda.
This view of social policy changes in Japan and Korea stands against the earlier
developmentalist perspective of East Asian welfare states which sees social poli-
cies as subjugate responses to economic development. In highlighting the
importance of the post-industrial and political factors, this chapter challenges
such a static view of East Asian welfare states, and offers an alternative perspective
that explains how structure and politics can facilitate welfare state change.
Following a brief discussion of the limits of the developmentalist view of
East Asian welfare states, I compare domestic post-industrial pressures confronting
Japan and Korea. I will then outline social policy reforms in these countries
since the 1990s, and explain how structural and political changes affected wel-
fare state restructuring in these countries. Finally, in the last section I will reflect
on the implications of family and gender policy reforms for women in these
countries.
132 Ito Peng
Welfare state development in Japan and Korea: beyond the

‘productivist’ welfare regime
The comparative study of East Asian welfare states has come a long way since the
early 1990s, when scholars began to undertake more systematic studies of
countries such as Japan and Korea. Building on the idea of the developmental
state, many first-generation East Asian welfare-state scholars focused on the role of
the state and the bureaucracy in shaping economic and social policies (Goodman
and Peng 1996; Goodman, White and Kwon 1998). Growing out of this literature,
the economic productivist approach sought to explain welfare developments in
East Asia in terms of the state’s prioritization of economic development, and hence
the subordination of social policies to economic and industrial policies. Holliday
(2000), for example, argues that in Japan and Korea while the state underpins the
market and families with some universal programmes such as health insurance
and pensions, only those linked to the ‘productive’ elements of society, typically
the male workers in core industries, are accorded minimal social rights.4
Alternatively, others examined the Confucian underpinnings of the East Asian
welfare regime. Jones (1990, 1993) has pointed out the pervasive influence of
the Confucian tradition underlying the success of East Asian welfare states. The
Confucian principles of filial piety and industry are thought to contribute to a
strong familial sense of welfare obligation and self-help. The notions of gender
role segregation and women’s subordination to men and to the family are also
used by the state as the rationale for its reliance on women’s unpaid work at home
and in the community (Palley 1992; Sung 2003). These factors in turn shape the
normative foundation of the dual labour-market structure that overwhelmingly
privileges male workers in core industries, and leads to strong assumptions about
women’s welfare responsibilities.
These perspectives, however, are being challenged. First, social policy has
become much more politicized in the last couple of decades. In Japan, social policy
and welfare reform have become an important policy issue and political parties
have became more vocal in advocating particular social policy choices. In Korea
while the democratization process helped to mainstream social policy, the Asian
economic crisis of 1997 further highlighted the political salience of social policy in
electoral politics. Simply put, the political currency of social policy in both Japan
and Korea rapidly appreciated in the 1990s.
Second, the productivist welfare-state argument also faces some difficulties in
explaining the evident expansion of social welfare programmes aimed at the ‘non-
productive’ sectors of society in the 1990s. If the productivist welfare-state logic pre-
vails, social policies in these countries should be more focused on the ‘productive’
sectors of the society as the economy weakens, and one would expect preferential
allocation of resources to ‘productive’ sectors and greater reliance on the family’s wel-
fare responsibilities, as was the case for Japan in the 1980s. In the 1990s, however, the
exact opposite dynamics were observed. In Korea significant social welfare expansion
followed after 1997; in Japan, ongoing economic malaise notwithstanding, welfare
expansion continued. What is more, not only was there expansion in social services,
most of the social policy reforms were targeted to ‘non-productive’ sectors of these
societies such as the elderly, children and the unemployed.5 In Japan, social care for
the elderly and children expanded alongside reforms to strengthen gender equality
and to protect women and children in vulnerable situations. In Korea, universaliza-
tion of healthcare benefits and pensions took place together with radical reforms of
unemployment insurance, social welfare, and family and gender policies. The focus
on ‘non-productive’ sectors of society is further confounded by the new emphases on
gender equality, and work and family harmonization, concepts that are clearly at
odds with Confucian principles of gender role segregation.
Of course, one may argue that these reforms could simply be counter-cyclical mea-
sures, and that state policies remain basically ‘productivist’ in nature. True, job cre-
ation through public investment can be an effective counter-cyclical measure in times
of economic crisis. The promotion of active labour-market policies in both countries
can be seen as essentially ‘productivist’ as well. However, this begs two questions.
First, why social welfare, and why at this time. In the past, the typical counter-cyclical
measure for both the Japanese and Korean governments was to work with employers
to sustain high rates of employment, and if necessary, governments used public
construction and other forms of public-works programmes to kick-start the economy.
Indeed, that was precisely what the Japanese government did in the early 1990s, until
they ran out of steam with little success and a massive government deficit. These mea-
sures made ‘productivist’ sense because they targeted the productive sectors of society
and provided male employment. Implicit in the productivist logic was a highly gen-
dered notion of economic growth: productivism through protection and enhance-
ment of male workers in productive sectors. Promoting gender equality in the labour
market was not part of the productivist counter-cyclical logic. Even if we see the cur-
rent welfare-state expansion as part of the same productivist counter-cyclical logic,
we have to explain why there has been a shift in policy understandings of both the
causes of economic malaise and solutions proposed to counter it. What are the
underlying causes of this shift in policy thinking, and what are its implications?
Second, the idea that East Asian welfare states are somehow uniquely produc-
tivist may be misplaced. As illustrated by the historical experiences of Nordic wel-
fare states, social policies have always been implicitly linked to economic
development. Indeed, as Mkandawire (2004) points out, social policy is basically
developmental. What is important, then, may be not the productivist nature of
social policy reforms, but rather the mechanisms and motivations underlying pol-
icy changes. In other words, we need to move beyond identification and classifi-
cation of East Asian welfare regimes to analysing causal processes of change.
Welfare-state expansions in Japan and Korea therefore require more in-depth
examinations which go beyond description.
Post-industrial pressures
Demographic shifts
Japan and Korea have been experiencing similar demographic changes over the
last few decades. In both countries population is ageing and fertility is declining
134 Ito Peng
Table 6.1 Ageing population (% population over 65) in

Japan and Korea, 1980–2030
1980 2000 2020 2030
Japan 9.1 17.2 26.9 29.6

Korea 3.8 7.2 15.1 23.1
Sources: NIPSSR, Jinko Dotai Tokei (Population Statistics), 2004;

KNSO (2005) Future Household Projection.
Table 6.2 Total average fertility rate in Japan and Korea, 1970–2030
1970 1980 1990 2000 2003 2010 2020 2030
Japan 2.13 1.75 1.54 1.36 1.29 1.32 1.38 1.38

Korea 4.53 2.83 1.59 1.47 1.19 1.21 1.24 1.28
Sources: NIPSSR, Population Statistics, 2004; KNSO (2005) Future Household Projection.
(Tables 6.1 and 6.2). Although Korea is still demographically young compared to
Japan and other OECD countries,6 the pace of ageing is rapid. As shown in
Table 6.1, the elderly population in Korea will constitute about 23 per cent of the
total population by 2030. The main cause of demographic ageing in both
countries is the low fertility rate. Given the current total fertility rate, demographic
ageing in Korea will occur within an extremely compressed time frame. The issue
for Korean policy-makers therefore is not simply that of declining fertility and an
ageing population, but more importantly, the speed at which these demographic
changes are taking place. There is concern that Korea will not be prepared for
changes of this magnitude in terms of its social infrastructure.
Changes in family and gender structures

The process of defamilialization (i.e. decline of the traditional family structure) is also
quite evident in Japan and Korea. As shown in Table 6.3, the proportion of three- (or
more) generation households has declined, while that of single-person households
has increased since 1980, suggesting an erosion of ‘traditional’ living arrangements.
Divorce rates and the proportions of single-mother families are also on the increase.
In Japan, the divorce rate nearly doubled between 1980 and 2000, while in Korea, it
increased by nearly five-fold (Table 6.4). The recent increase in divorce in Korea has
been attributed to family breakdown after the economic crisis. Studies suggest that
the increase in divorce rates after 1997 are largely due to economic hardship (Lee and
Park 2003), while the continued increase in divorce rates after 1997 can be also attrib-
uted to changes in people’s ideas about marriage and to conflicts between family
members (Park 2005). In both countries there is a growing bipolar pattern with rela-
tively high divorce rates among those married for five years or less (i.e. the younger
demographic cohort) and among those married for 20 years or more (i.e. postwar
‘baby boomers’). The increases in numbers and proportions of single-mother families
also closely parallel the increase in divorce (see Tables 6.5 and 6.6).
135
Table 6.3 Family structures in Japan and

Korea 1980–2000
1980 2000
Japan
Nuclear 60.3 58.4
Three-generation 19.7 13.6
Single-person 18.1 24.1
Korea
Nuclear 76.8 75.0
Three-generation 17 8.4
Single-person 4.8 15.5
Sources: MHWL (2001) White Paper on Health,

Welfare, and Labour; KNSO (2003) Future Household
Projection.
Table 6.4 Divorce rate (crude divorce rate per thousand population), 1980–2002
1980 1990 1992 1995 1999 2000 2001 2002
Japan 1.22 1.28 1.45 1.60 2.00 2.10 2.27 2.30

Korea 0.6 1.1 1.2 1.5 2.5 2.5 2.8 3.0
Sources: NIPSSR, Population Statistics, 2004; Yeong-Ran Park, 2005.
Table 6.5 Percentage of single parents in Japan

(Units: thousands of people; %)
% of
single
parent *
Total Widowed Divorced Unmarried Others families
1983 Total 885.4 326.2 443.2 38.3 77.7 (1985) 5.5

Single mother 718.1 259.3 352.5 38.3 67.9 4.7
Single father 167.3 66.9 90.7 n/a 9.8 0.8
1993 Total 947.2 245.2 605.1 37.5 38.0 (1995) 6.3

Single mother 789.9 194.5 507.6 37.5 33.4 5.5
Single father 157.3 50.7 98.5 n/a 4.6 0.8
2003 Total 1,399.2 180.6 1107.4 70.5 37.9 (2000) 7.7

Single mother 1,225.4 147.2 978.5 70.5 27.3 6.8
Single father 173.8 33.4 128.9 n/a 10.6 0.9
* Proportion of single parent families in relation to all families with children under 18 years of age.
Source: MHWL (2005) Report of the Heisei-15th year National Survey of Single Mother and Other Families (Heisei 15-nen do
Zenkoku Boshisetai tou chousa kekka houkoku); NIPSSR, Population Statistics, 2004.
136 Ito Peng
Table 6.6 Percentage of single parents in Korea

(Units: number of persons; %)
Marital Status of Household-Heads % of

Single
Married Death of parent
Total (Separated) spouse Divorce Unmarried families
1995 Total 959,972 216,067 526,320 123,969 93,616 7.4

Single father 172,398 50,666 68,022 51,080 2,630 1.3
Single mother 787,574 165,401 458,298 72,889 90,986 6.1
2000 Total 1,123,854 252,917 502,284 245,987 122,666 7.9

Single father 219,997 58,227 64,058 92,810 4,902 1.5
Single mother 903,857 194,690 438,226 153,177 117,764 6.3
Source: KNSO, Report on Population Census (1996, 2001).
Surveys of public attitudes towards marriage and divorce also suggest that
divorce is becoming more socially acceptable in Japan and Korea (Cabinet Office –
Japan 2005; Hong 2005; Kim 2003b). Moreover, there are also some changes in the
intergenerational contract, the most significant being the weakening of the
parental expectation of receiving support from children in old age, and a corres-
ponding decline in children’s expectation of supporting parents in their old age
(Cabinet Office, Japan 2005; Retherford et al. 2001).
Finally, employment structures in both countries have changed over the last
couple of decades. There has been a steady informalization of work throughout
the 1990s. In Japan, the proportion of people in non-standard forms of work
increased from 20.9 per cent to 30.4 per cent between 1995 and 2003 (MHWL
2004), while in Korea, the proportion of those employed in non-standard work
rose from 45.5 per cent in 1990 to 52.1 per cent in 2000 (KNSO 2005).7 In both
countries the majority of non-standard workers are women. The employment rate
of married women in Japan rose from 48 per cent in 1980 to 56 per cent in 2000,
while that in Korea increased from 40 per cent to 48 per cent over the same
period.8 In both cases, while the informalization of work has pulled more women
into the labour market, it has also widened the gap between child and elderly care
needs and the availability of unpaid care provided by women.
Social policy responses
The two governments’ main response to these social and demographic changes
has been to ‘activate’ the labour-market participation of women and the younger
elderly, while encouraging childbirth. This approach has been preferred to develop-
ing policies to encourage immigration, for example. In Korea, the postwar family
planning policy was officially terminated in 1996, and a number of woman-
friendly and gender-positive policies have been introduced to encourage women’s
social, political and labour-market participation. In Japan a series of woman-friendly
policies were introduced in the 1990s to help families harmonize work and family
responsibilities. These include: the Gold Plan in 1989, the Angel Plan in 1994,
introduction of parental leave and flexible hours for workers with childcare
responsibilities in 1996, and the Long-term Care Insurance in 1997.
Social expenditure as a proportion of GDP in Japan grew from 11 per cent to
17 per cent between 1990 and 2000, and in Korea from approximately 3 per cent
to 7 per cent between 1990 and 1999 (OECD 2004). Although a large portion of
these expenditures is accounted for by pension and health insurance, social wel-
fare expenditures also grew. In Japan, social welfare expenditure doubled from
¥4.8 trillion in 1990 to ¥10.9 trillion in 2000 (NIPSSR 2002), most of which was
accounted for by the expansion in long-term elderly care and childcare. In Korea,
social welfare expansion was even more remarkable: government expenditure on
the Employment Insurance Program increased from 4.7 million Won in 1996 to
306,172 million Won in 1999; the budget for childcare rose from 41,876 million
Won to 436,903 million Won between 1991 and 2002; and elderly welfare
increased from 37,861 million Won in 1990 to 407,767 million Won in 2003
(Lee and Park 2003).
Japan – social care expansion

In Japan elderly care reforms began with the Gold Plan in 1989. It expanded the
provision of home and institutional care for the elderly on a needs basis.9 In 1994
the Plan was revised with further increases in the target provisions and budget,
until 1997 when it was replaced by the Long-term Care Insurance (LTCI) Scheme
(implemented in 2000). As a form of universal mandatory social insurance, LTCI
enabled the government to collect insurance premiums from all adults over the
age of 40. It provides domiciliary and institutional care services to people over the age
of 65, and those between 40 and 64 who need care as a result of premature ageing.
The LTCI was an important reform because it shifted the principle of elderly care
from a means-tested to a rights-based program. Also, unlike the German model,
which is funded solely by insurance premiums, the Japanese LTCI is funded
equally by the insurance premium and the general tax revenue, and therefore
ensures some degree of redistribution.
Policies towards families with dependent children in Japan also benefited from
an expansionary thrust in the 1990s. The Angel Plan (1994) increased social care
and support for families with children, particularly public childcare. Under the
Angel Plan the number of public spaces for children under the age of two
expanded, as well as the number of multi-functional childcare centres, spaces for
extended-hour childcare, temporary childcare centres, local childrearing support
centres, and after-school programmes. Parental leave policy was also introduced in
1994 to allow either one of the parents to take up to one year of childcare leave.
A 40 per cent income replacement was added to this in 1998 to encourage more
parents to take the leave. Moreover, flexible working hours, child allowance, and
other forms of support for young families have also been introduced since 1996 to
enable parents to harmonize family and work responsibilities.
On the broader gender policy front, the Council of Gender Equality was
established within the Cabinet Office in 1996, and in 1999 the Basic Law for
Gender Equal Society (Danjo Kyodosankaku Kihon Ho) was introduced to promote
138 Ito Peng
gender equality more systematically. The law included provision for the
institutionalization of gender-equality principles throughout the national and
local governments, and the establishment of a monitoring system to measure the
level of change. The Council also broadened its influence throughout the 1990s.
For example, its policy proposals have been incorporated in the recent social
policy reforms (Cabinet Office, Japan 2005; Osawa 2003),10 including reforms to
divide the employee pension between husbands and wives, allowing divorced
women the right to claim half of the husband’s pension, and the extension of the
employee pension to part-time workers. Similarly, the Council has also introduced
a proposal to abolish the special tax break for married women, which was criti-
cized by feminists for discouraging married women’s full-time employment and
privileging middle-class male breadwinners. It has also been successful in pushing
for family policy reforms through the Low Fertility Measures Plus One Program
(Shoshika Taisaku Plus One), which includes targets of abolishing waiting lists for
public daycare, increasing the parental leave take-up rate to 10 per cent for fathers
and 80 per cent for mothers, a minimum five-day paternity leave, and shorter
working hours for workers with small children.
The policy-making process has also become more open to women’s groups. The
LTCI legislation was one of the first cases of direct engagement by women’s groups
in the social policy-making process (Eto 2001; Peng 2002a). The Council of Gender
Equality’s core group, for example, consists of 25 members, 13 of whom are min-
isters of various government ministries, while 12 others are ‘experts’ made up of
university professors and representatives of business and NGOs, most of whom are
women. In addition, the Council also set up a community advisory committee
called Egalité Network comprised of representatives from 88 different civil society
groups. Women politicians are also taking a more active role in policy reform. The
Domestic Violence Law (1999) and Child Abuse Prevention Law (2004), for
example, both involved a special parliamentary bill (gi’in rippo) put forward by a
cross-party coalition of women MPs (Gelb 2004; Iida 2004). This method of legal
reform involves MPs making a draft bill proposal directly in the parliament, and
bypassing the bureaucracy.
Korea – social security expansion

The Korean welfare state grew steadily throughout the first half of the 1990s, and
then took a sudden expansionary leap after 1997. Almost every aspect of the social
security system expanded. The coverage for (un)employment insurance increased in
1998 in response to the rise in unemployment resulting from economic restructur-
ing. (Un)employment benefits were extended to firms with five or more employees
(whereas before they only covered firms with 30 or more employees), and job secu-
rity and training benefits were also extended to firms with five or more employees
(prior to this change only firms with 70 or more employees were covered) (Kwon
2001: 225). In 2000, the coverage was further broadened to all employees, including
part-time government workers, and the length of benefit was also increased from a
maximum of 60 to 210 days, depending on individual work history (Sung 2003).
Table 6.7 shows this expansion in social insurance coverage in Korea since 1997.
Table 6.7 Changes in the population coverage of the Four Major Social Insurances
in Korea
1997 2000
EI WCI PI HI EI WCI PI HI
Firm size ⫺5 X X X O O O O O
(number of regular 5–9 X X O O O O O O
employees) 10–29 O O O O O O O O
30⫹ X O O O O O O O
Daily workers X X X O 䊐䊐 O O
Temporary workers X X X O 䊐䊐 O O
Self-employed X X X O X O O O
Unpaid family workers X X X O X O O O
O ⫽ covered, 䊐 ⫽ partially covered, X ⫽ uncovered. EI ⫽ employment insurance; WCI ⫽ workers

compensation insurance; PI ⫽ pension insurance; HI ⫽ health insurance.
Source: Lee, Hye-Kyung (2005).
Coverage for other social insurance schemes also expanded after 1997. The
National Pension Plan (NPP) was extended to all workers, including daily, tempo-
rary, self-employed and family workers (Kim 2003). It is, however, important to
point out that the outcome in relation to women is still small. Although the total
number of people insured by NPP has more than doubled between 1990
(4,651,678) and 2002 (12,248,483), women make up less than a third of the total
(29.5 per cent in 1990, and 30.1 per cent in 2001). The percentage of women who
are covered by NPP increased from 6.4 per cent of the total population in 1990 to
15.6 per cent in 2002, but this is still significantly lower than the coverage given to
men (15.2 per cent and 36.7 per cent respectively) (KWDI 2004). The main reasons
for this are women’s lower employment rates and the newness of the NPP. The
National Health Insurance coverage, on the other hand, has expanded since
the end of the 1970s, and currently nearly the entire population is covered.
However, the level of co-payment (user fees) for NHI is still relatively high, making
it difficult for low-income families to utilize healthcare.
The Livelihood Protection Programme, which provided basic income support to
the very poor, was also reformed in 1998 to become the National Basic Livelihood
Security Programme (NBLS), resulting in a 50 per cent increase in the number of
recipients within a year (from 1 million to 1.5 million) (Sung 2003). A work-fare
component called the ‘self-reliance policy’ was added to the NBLS in 1999 to
engage recipients in public-works programmes, while providing them with medical
assistance, childcare and school fees, and counselling services. In 2001, approxi-
mately 46,000 people were involved in this programme, 70 per cent of whom were
women, and half of those were middle-aged single mothers (Lee and Park 2003:
16). The expenditures for NBLS nearly quadrupled from 900 billion Won in 1997
to 3,400 billion Won in 2002 (Lee 2005).
The Mother-Child Welfare Programme was introduced in 1999 to provide
housing support, job training, and employment and childcare assistance to
140 Ito Peng
single-mother families. This programme was renamed Mother-Father-Child

Welfare Programme in 2002 to include single-father families and other working
poor families. The Infant-Child Care Programme (introduced in 1991), which
provides public childcare for working parents, also grew. The national budget for
childcare increased more than ten-fold during the 1990s, and the number of
childcare centres expanded from 9,085 in 1995 to 22,147 in 2002 (Lee and Park
2003).11 The Maternity Protection reform was also introduced in 2001, extending
maternity leave from 60 days to 90 days, with a paid childcare leave benefit of
300,000 Won per month. The budget for public elderly care also increased more
than ten-fold between 1990 and 2001 (Lee and Park 2003).
Finally, the Ministry of Gender Equality was established in 2001 (renamed to
Ministry of Gender and Family in June 2005) to advance the gender-equality
agenda and to implement and manage gender-equality activities and related poli-
cies (Park 2005). Many leaders of the women’s movement and feminist NGOs have
moved into the Ministry in an attempt to affect policy change from within.12 As a
result, the Ministry’s profile has risen. It has expanded its social policy portfolios,
including childcare, women’s social participation, skills training, and domestic
and sexual violence. In addition, the implementation of the Mother-Father-Child
Act and the Basic Act on Healthy Family has been added to the Ministry’s portfolio.
The Basic Act on Women’s Development in 2002 has also made gender analysis a
mandatory procedure for all government ministries and local governments starting
in 2005 (Park 2005).
Explaining social policy reforms in the 1990s: post-industrial

pressures mediated by changes in domestic politics
The changes described above suggest that the 1990s was indeed a watershed
decade for both Japan and Korea. Faced with similar social and demographic
imperatives, both countries had little choice but to undergo welfare-state restruc-
turing. However, translating these imperatives into policy reforms requires political
will. As illustrated by the experiences of European welfare states, despite similar
post-industrial pressures, social policy responses have diverged, largely because of
different domestic political conditions and policy options.
Similar post-industrial pressures in North America, for example, did not translate
into social policy expansion in either Canada or the US. It would be therefore
premature to draw direct relationships between post-industrial pressures and
welfare-state expansion in Japan and Korea. It is important to trace mechanisms
through which these pressures have been translated into policy changes. For this
we need to turn our attention to the domestic political dynamics. In Japan and
Korea, the structural changes of the 1990s also coincided with political regime
shifts: in Japan with the end of the conservative Liberal Democratic Party’s (LDP)
political hegemony in 1993, and in Korea with the transfer of power from the
conservative party to the opposition government in 1998. In both cases, the shifts
disrupted the political status quo and opened opportunities for change.
Japan
In Japan the elderly population had already reached 12 per cent by 1990, and the
basic social security system, introduced in the early 1960s, was beginning to
experience structural pressures associated with the maturing of the system. Not
surprisingly, then, a great deal of the increase in social security expenditure in the
1990s did result from rising pension and healthcare costs owing to population age-
ing. The important policy shift in the 1990s was therefore not so much associated
with the increase in health and pension benefits, which would have been
expected, but rather with the unexpected expansion of social care services. The
causes of social care expansion were in fact functional as well as political. While
post-industrial changes did affect people’s thinking about social welfare, not much
happened until the 1990s. Social care reforms in Japan are therefore a good
example of post-industrial imperatives facilitated by political regime shift.
In Japan, the so-called ‘crisis of the family’ had been brewing since the 1980s
when women’s groups began to agitate for more social care. This idea, however,
did not make much impact on social policy debates in the 1980s, until the fertil-
ity rate declined below the historical low of 1.58 in 1989. By the early 1990s the
idea of family crisis was promoted not only by the government but also by the
media and civil society groups. The image of the crisis ranged from the notion of
the family as a ‘caring hell’ (kaigo jigoku) for women (Higuchi 1997; Koreika Shakai
wo Yokusuru Josei no Kai 1998; NFDR 1997, 2001), to young people being depicted
as ‘Parasite Singles’ (Iwakami 2000; Yamada 1999), and the fear of labour shortage
and population decline (MHW 1997).13 Underlying these disparate characteriza-
tions was a shared sense of crisis about the adverse effects of social and structural
changes, especially fertility decline and rapid ageing, on the Japanese family and
society.
Work–family tensions for women were evident. Not only was the married
women’s employment rate increasing, but more women were working outside of
the home. The proportion of women in self-employed work or family businesses
declined from 39.0 to 18.4 per cent between 1975 and 1999, while the proportion
of women working outside the home rose from 59.8 to 81.4 per cent over the same
period (NFDR 2001). The majority of women were also experiencing stress and
psychological burdens as a result of having to care for their elderly relatives while
working (Ministry of Labour 1998). Opinion polls showed a decline in support for
the idea of the male-breadwinner/female-housewife model, and an increase in
support for women’s lifetime employment (Cabinet Office, Japan 2005). Women
in their twenties and thirties were less willing to accept the traditional division of
labour within marriage and were more willing to consider divorce as an option
(Retherford et al. 2001).
This shared sense of the family being in crisis due to social and economic
changes in Japan emerged in the context of a political regime shift. The collapse of
the LDP’s political dominance in 1993 led to a chaotic political scramble as politi-
cians and parties tried to realign themselves in a new and uncertain political situ-
ation. Within a few years, no less than a dozen new political parties appeared and
disappeared, while politicians jumped from one party to another in an effort to
142 Ito Peng
gain advantage. Political cleavages were redrawn as political parties tried to

capture largely middle-class votes along new issue lines (Pempel 1998; Peng and
Wong 2004). Social policy emerged as an important political issue. The first
coalition government headed by the Social Democratic Party of Japan (SDPJ) and
Komeito (Clean Government Party) ran on the pro-social welfare ticket to
distinguish itself from the conservative LDP and to gain electoral support from
women and the elderly. Public opinion polls also confirmed that social welfare was
a winning issue.14 As social welfare became increasingly prominent in political
contestation, other political parties, including the LDP and the newly formed
Democratic Party of Japan (DPJ) also joined the bandwagon in support of social
policy reforms. Special research teams to study and develop proposals for family
support policies were established within the Ministry of Health and Welfare in
response to the low fertility rate, changes in family structure, and to the shifts in
the labour market. Many of the proposals that would not normally pass the
conservative LDP parliament, including the expansion of public childcare,
parental leave, flexible working arrangements and new child protection legisla-
tion, were incorporated into the Angel Plan in 1994, under the SDPJ-led coalition
government.
Certainly political competition was an important factor for social policy reform
in Japan, particularly after 1993. Civil society groups such as the Women’s
Committee for the Improvement of the Ageing Society (WCIAS) had been cam-
paigning for social care reforms since the early 1980s, but the then ruling LDP paid
little attention to them. Politics changed dramatically once the power shifted from
the LDP to coalition governments. Women’s groups and other civil society groups
saw opportunities to engage in policy-making. In the case of the LTCI, for exam-
ple, the Ministry of Health and Welfare began to work closely with WCIAS to push
for policy reform against the LDP opposition. In this case, the Ministry’s interests
in finding alternatives to the problems of elderly healthcare outside of the medical
insurance system coincided with women’s interests in social care. The SDPJ-led
coalition government also began to practice more inclusive policy-making by
bringing new stakeholders such as NGOs into the policy-making process (Iida
2004). When the LDP-led coalition government took power in 1996, it tried to
stall the draft LTCI bill and prevent it from going through parliament. However,
that attempt eventually failed as WCIAS and other civil society groups represent-
ing elderly people turned the issue into a political campaign. They formed a
national coalition of civil society groups to lobby for the LTCI legislation.
Campaigning under the banner of ‘Ten thousand people in support of the Long-
term Care Insurance’, the civil society coalition held public forums and seminars
on elderly care issues, and lobbied local and national politicians to support the bill
(see Eto 2001; Peng 2002b).
New social policies such as LTCI were important for the coalition governments
not only because they provided a new policy focus for parties like SDPJ and
Komeito, but also because they helped new political parties such as DPJ to garner
support from women and the elderly (Democratic Party of Japan 2002).15 These
policies also helped lay out new mechanisms for civil society engagement in
policy-making. Furthermore changes in the electoral system from the single-

district multi-seat system to a combination of single district single-seat and
proportional representation systems made linkages between civil society groups
and political parties more crucial. New political parties such as Sakigake and the
Japan New Party found it important to build closer ties with identifiable and orga-
nized civil society groups that could mobilize votes. These political parties have
targeted emergent voluntary-sector groups to forge political links.16 Since the
introduction of the Non-Profit Organization (NPO) law in 1998, which enabled
NPOs to incorporate, all the major political parties have established an NPO
liaison committee to help develop links with civil society organizations.
New political parties and civil society groups also adopted fresh strategies to
push for policy changes. For example, rather than going through the traditional
route, as was the case for WCIAS in the development of the LTCI, new civil society
groups also began to use the process of special parliamentary bills (gi’in rippo) to
bypass the lengthy and bureaucrat-dominated draft proposal bill. The NPO law,
Domestic Violence law (2000), and Child Abuse Prevention law (2004) were all
passed by special parliamentary bills. This policy strategy – rarely used until 1998
because it forces politicians to take strong policy advocacy roles, often in opposi-
tion to bureaucrats – proved highly effective when civil society groups wanted to
pass new legislation quickly and when they were able to mobilize themselves
to gather enough political support from members of parliament. They lobbied
women politicians to press for policy change through gi’in rippo.
In sum, in Japan, while the structural and ideational impacts of demographic
and societal changes were important, the political regime shift of 1993 provided
important openings for change. The regime shift disrupted the political status quo
and created new policy competition amongst politicians and political parties over
social welfare issues. Civil society groups also took the opportunities to engage in
public policy-making, and they actively participated in shaping the contents and
directions of social policy reforms. New legislation such as the LTCI and NPO laws,
in turn, further helped institutionalize civil society’s advocacy roles in social
policy-making.
Korea – welfare state expansion

In Korea, the transition to democratic politics in 1987 was the first turning-point
for the country’s social policy development. Democratization brought about polit-
ical realignment, increased political competition, and rejuvenated civil society.
Although the 1987 election resulted in the victory of Roh Tae Woo, backed by the
military regime, it nevertheless established the notion of electoral democracy and
set a new political pattern. As Wong (2004) points out, in Korea social welfare
emerged as a winning issue from the beginning of the political realignment
process because it effectively cut across the existing personal/regional cleavages
dominant in Korea’s democratic politics. Strong economic growth also helped
expand interest among middle class voters in social policy. Under the Kim Young
Sam government (1993–98) a number of political reforms were introduced.
Provincial and local elections were reinstituted in 1993, followed by the purge of
144 Ito Peng
the former military elites (the Hanahoe) and the jailing of the coup leaders under
the military regimes. Like post-1993 Japan, in Korea new political parties emerged
and new political alliances were forged and broken, as politicians vied for political
advantage (Kim 2000). In the meantime, civil society groups also reorganized, and
shifted their focus from the broader democracy movement to specific social policy
advocacy. As Seong (2000) points out, the civil society movement in Korea after
1987 entered the mainstream as new civic organizations, representing diverse
popular concerns such as environmental protection, social welfare, women’s
issues, educational reforms and crime control, began to gain hegemony over
special-interest groups such as employers’ associations and labour unions. They
pushed for welfare-state expansion and the mainstreaming of social policies in the
1990s.
The suddenness of the Asian economic crisis in 1997 caught everyone off guard:
overnight, foreign investments were pulled out, stripping Korea of its assets and
reserves; stocks plunged; the economy halted; and finally, in what President Kim
Dae-Jung called ‘the country’s second humiliation’ (the first being its colonization
by Japan from 1910 to 1945), Korea had to turn to the IMF for an emergency bail-out.
The IMF conditions were stringent: currency devaluation, economic liberalization
and labour-market restructuring. Ironically, it turned out to be a watershed for
political regime shift in Korea. In the 1997 presidential election, public discontent
with the Kim Young Sam government’s ineffectiveness in handling the crisis
turned into a swing vote for the opposition candidate Kim Dae-Jung who cam-
paigned on the platform of social policy reform on the one hand, and modernization
and responsibility on the other.
Kim Dae-Jung’s electoral coalition was far from united. Indeed, from the
beginning significant tensions existed between the pro-welfare and pro-business
interests within his camp. The imperatives of the economic crisis, however, offered
him the opportunity to create a government–employer–labour tripartite council
and to force the ‘grand consensus’, a two-pronged, if not contradictory, strategy of
economic liberalization and welfare-state expansion. The labour movement
agreed to radical market liberalization as directed by the IMF, in exchange for the
expansion of the social safety net. Clearly many of the social policy reforms in
Korea took place against the backdrop of social and economic imperatives follow-
ing the economic crisis in 1997. However, we should not understate the impact of
the economic crisis on public support for change and on civil society groups’
willingness to back a reformist leader like Kim Dae-Jung. Certainly, pro-welfare
civil society groups constituted Kim Dae-Jung’s support base. Once the welfare-state
reforms were introduced they became the basis from which civil society began to
push for further expansion.
Even after the economic crisis the expansionary reforms continued. With the
economic crisis behind, the social policy reforms were increasingly informed by
the idea of changes in demographic patterns and family and gender relations.
Policy debates on elderly care and childcare expansion since 1998, for example,
mirror the Japanese family policy debates. Despite its relatively young population,
demographic ageing and low fertility are now the priority issues in the Korean
government’s social policy agenda (Park 2005; Lee 1999b; Cho 2000).17 Long-term
Care Insurance pilot projects (based on the Japanese model) will begin in 2007.
Similar attempts are being made in terms of the expansion of public childcare services.
The policy focus has also shifted from the economic crisis to the persistence of
low-income households, particularly those headed by women (Lee and Park 2003),
with much of the recent family and gender policy reforms being informed by
studies of poverty among women. Lee (1999) found that women were particularly
hard hit by the economic crisis, with two-thirds of the poor and four-fifths of poor
elderly people being women. Government studies also show the negative implica-
tions of the ‘dismantled’ families not only as they relate to financial hardships, but
also in relation to problems associated with childrearing, caregiving, and emotional
status. Finally, the combination of defamilialization and increased employment
among married women has also been identified as creating problems for elderly
care (Sung 2003; Lee and Park 2003). In summary, the mainstreaming of social
welfare in Korea since 1987 was propelled by the economic crisis that brought
about a shift from pro-business to a pro-welfare government. Once the crisis was
over, social policies remained informed by the idea of the family crisis caused by
the changes in demographic structure, and family and gender relations. Like
Japan, social policy reforms occurred in the context of structural and political
changes.
Implications of Social Policy Reforms for Women
Have any of these policy reforms made a difference to the lives of women in Japan
and Korea? The Gender Empowerment Measure and Gender Development Indexes
indicate very little change. However, at the ground level more subtle changes seem
to be taking place. For example, the number of women parliamentarians in Japan
increased from 26 (3.4 per cent) to 79 (10.8 per cent) between 1980 and 2001,
while their proportion at the prefectural and local government levels rose from
1.2 per cent in 1980 to 7.9 per cent in 2003. More evident changes in women’s
political representation are observed at the city level. In large cities such as Tokyo
and Osaka, the proportion of women politicians increased from 8.9 per cent in
1990 to 21.5 per cent in 2003. There has been an increase in the representation of
women in national policy advisory councils as well: in 1980, less than half of these
councils (46.2 per cent) had a woman member, and women represented a mere
4.1 per cent of all the council members; by 2003, however, the proportion had
changed to 98 per cent and 26.8 per cent respectively (Cabinet Office, Japan 2005).
In Korea, similar changes can be seen. In the 1992 parliamentary election there
were 35 women amongst a total of 1,106 candidates from local and national con-
stituencies (3.2 per cent); of this number, three were elected from a total of 296
seats (1.0 per cent). In the 2004 election, the percentages of women candidates
and electees increased to 10.4 and 13.0 respectively. Part of the reason for the
increase in elected women politicians is the introduction of the proportional
representation system through electoral reform. Similarly women’s representa-
tion in government advisory councils also increased from 9 per cent in 1990 to
146 Ito Peng
27.3 per cent in 2002, and the proportion of women officials in the government
bureaucracy also increased from 23.9 per cent in 1990 to 34.0 per cent in 2003
(KWDI 2004). In both Japan and Korea it appears that some effort has been made
to raise women’s representation in public and political office over the last couple
of decades.
In addition to women’s entry into the government and bureaucracy, there has
been a noticeable change in women’s participation in civil society groups as well.
In Japan, the number of registered NGO groups increased from 1,176 to 14,657
between 1998 and 2003. A national survey of NGOs in 2001 found women made
up nearly half (48.5 per cent) of all workers while elderly and retired people made
up about a third (31.4 per cent) (Cabinet Office, Japan 2005). In Korea, women’s
organizations have emerged as the main advocacy groups in the social policy-
making process (Lee 2005). The number of women’s organizations in Korea
increased from 62 in 1994 to 93 in 2001 (KWDI 2004). Studies of civil societies in
both countries show NGOs are becoming increasingly politically active, and
governments too have become more open to civil society groups’ engagement in
public policy-making (Lee and Peng 2005; Shipper 2005). What these trends
suggest is a possible change in the governance structure vis-à-vis public policy-
making. As social policies in both countries become more decentralized and as
civil society becomes more closely engaged in policy-making, public policies may
be further re-shaped.
Finally, we cannot understate the importance of ideational shifts in these coun-
tries. Public opinion polls in both countries suggest noticeable changes in men’s
and women’s attitudes towards gender relations. In Japan, the proportion of those
who claim that a woman should continue to work even if she has a child has risen
from 26 per cent to 38 per cent for women and 19 to 37 per cent for men between
1992 and 2002. Similarly, during the same period, the proportion of respondents
who endorse the male-breadwinner/female-housewife model has declined from
55.6 to 43.3 per cent for women, and 65.7 to 51.3 per cent for men; while the pro-
portion who reject the model has increased from 38.3 to 51.1 per cent for women
and 28.6 to 42.1 per cent for men (Cabinet Office, Japan 2005). In Korea, similar
attitudinal shifts with respect to women’s employment are observed.
Yet, these gains made at political, community and ideational levels are offset by
significant structural resistance, particularly within the labour market. In both
Japan and Korea employers are reluctant to adopt new gender policies, and the
legislation is inadequately enforced. In Japan, despite the public campaign to
promote parental leave, the rate of take-up has increased only marginally from
46 per cent for women in 1996 to 64 per cent in 2002, and in the case of men it
actually declined from 0.8 per cent in 1996 to 0.3 per cent in 2002 (MHWL 2004).
Workers are reluctant to take full parental leave because of negative pressures from
employers and work colleagues, and the fear of being made redundant (Cabinet
Office, Japan 2005). Workers will be reluctant to exercise their rights if they fear
negative consequences and if the legislation does not fully and actively enforce
their rights. In both Japan and Korea, despite the increase in public childcare
spaces since 1990, a significant number of children are still on the waiting list. In
both countries, but particularly in Korea, the large and progressive informalization of
employment also undermines women’s ability to achieve economic security. Despite
the extension of social insurance coverage to non-standard workers, a significant por-
tion of them remain uncovered, and more importantly, the social security systems in
both countries remain a poor defence against insecure employment and inadequate
income. These suggest that until the labour market and employment practices
change, the journey towards gender equality will indeed be slow and painful.
Acknowledgement
I would like to thank Kwansei Gakuin University for a travel grant to enable me to conduct
interviews in Japan and Korea during the academic year of 2001/2, and Social Sciences and
Humanities Research Council of Canada (SSHRCC) for the research grant (2003–7) which has
enabled me to continue and develop this research.
Notes
1 The UNDP report on the Gender-related Development Index (GDI), ranked Japan 12th
and Korea 29th amongst 174 countries in 2002. The same report also ranked Japan 38th
and Korea 68th in the Gender Empowerment Measures (GEM).
2 Interview with Mari Osawa, Professor of Gender and Social Policy, Institute of Social
Sciences, University of Tokyo, and also a member of the Gender Equality Council and
the Pension Reform Advisory Council, 17 February 2004.
3 Post-industrial changes also include exogenous factors such as economic globalization
and internationalization. Though clearly important, it will be beyond the scope of this
paper to deal with them adequately.
4 It should be pointed out here that in both Japan and Korea the male workers in core
industries, such as those represented by large business conglomerates, only represent
about 20 per cent of the total labour force.
5 In Japan, the economy stayed stagnant throughout the 1990s. Hope of economic recov-
ery in 1996 was, however, dashed by the Asian economic crisis of 1997. Korea was one
of the worst-hit countries in East Asia, and required financial rescue by the IMF.
6 The current OECD average is approximately 13 per cent.
7 Non-standard work normally means salaried and waged work that is part-time, tempo-
rary, contract or day work. In Japan, the majority of non-standard workers are part-
timers, while in Korea, most of them are contract and day workers.
8 These figures are for both standard and non-standard employment. In both cases the
majority of married women are non-standard workers.
9 The Gold Plan set targets for service expansion in elderly care for the next ten years.
Some examples of expansion are: an increase in the numbers of home helpers from
31,400 to 170,000, day services for the elderly from 4,300 to 17,000, beds in short-stay
centres from 1,000 to 60,000, and elderly healthcare institutions from 28,000 to 280,000
between 1989 and 1999.
10 Interviews with: Mari Osawa, 17 February 2004; Jozuka, director of the Gender Equality
Commission, Home Ministry, 20 February 2004; Akaishi, co-director of Femin, a national
women’s organization in Tokyo, 24 February 2004.
11 The percentage of children under six attending childcare increased from 7.4 per cent to
21.5 per cent between 1995 and 2002 (Lee and Park 2003). In Japan, it increased from
27 per cent to 33 per cent between 1995 and 2000.
12 Interviews with: Hyekyung Lee, Professor, Yonsei University, Korea, 1 February 2002;
Dr Wha-Soon Byun, Dr Yeon-Ran Park, and Dr Dayoung Song, Korea Women’s
148 Ito Peng
Development Institute, 4 February 2002; Ae-Ryung Kim, Director of Policy, Ministry of

Gender Equality, and Yang-Ji Ryu, Senior Policy Advisor, Ministry of Gender Equality,
5 February 2002.
13 The ‘parasite singles’ refers to young and not-so-young single adults who continue to
cohabit with and be financially and emotionally dependent on their ageing parents.
14 Polls show that next to economic recovery, social welfare – particularly welfare for the
aged – was the top concern for people over the age of 20. At the same time, public mis-
trust towards politicians, bureaucrats and governments soared to over 80 per cent in
1992, and remained there for the rest of the 1990s (Mainichi Shimbun National Opinion
Surveys 1992–2000). The PMO’s own public-opinion survey conducted in May 1994 also
found that social welfare and elderly care issues outweighed all other issues, including
the economy. For example, to the multiple-answer question asking respondents what
they would like the government to invest in, 61 per cent stated improved healthcare,
social welfare and pension systems, 47 per cent indicated care for the elderly and dis-
abled, 44.7 per cent stated a better tax system, and 37.6 per cent wanted better economic
policies (Japan-PMO 1994).
15 Interview with Mari Osawa, member of the Gender Equal Society Commission,
16 February 2004.
16 This has been clearly an important political factor in the development of NPO law in
1998. See Pekkanen (2000) for further discussion.
17 Interview with the Dr Kyung Tae Moon, Deputy Minister of Health and Welfare,
21 November 2003.
Bibliography
Cabinet Office, Japan (2005) Danjo Kyodo Sankaku Hakusho, 2004 (2004 White Paper on Gender
Equality), Tokyo: Cabinet Office.
Cho, Namhoon (2000) ‘Social Impact of Korea’s Low Fertility Rate’. Chislin IIbo, 1 October.
Eto, Murase Mikio (2001) ‘Public Involvement in Social Policy Reform: Seen from the
Perspective of Japan’s Elderly-Care Insurance Scheme’. Journal of Social Policy, 30(1): 17–36.
Democratic Party of Japan (2002) Tayona Lifestyle wo Ikiru Jidai no Jiritsu to Anshin no Seisaku
(Policies for Independence and Security in the Age of Different Lifestyles). Tokyo:
Minshuto Seisaku Chosa Kai.
Gelb, Joyce (2004) Gender Policies in Japan. New York: Palgrave.
Goodman, Roger and Ito Peng (1996) ‘East Asian Social Welfare States’. In Gosta Esping-
Andersen (ed.) Welfare States in Transition. London: Sage.
Goodman, Roger, Gordon White and Huck-ju Kwon (1998) The East Asian Welfare Model:
Welfare Orientalism and the State. London: Routledge.
Higuchi, Keiko (1997) ‘Kaigo no Shakai-ka wo Koshite Susumeta’ (How We Promoted
Socialization of Care). In Higuchi, Keiko (ed.) Chiiki de Miru, Minna de Mitoru: Joseiga
Susumeru Kaigo no Shakaika (Everyone in the Community Takes Care of the Elderly: Women
Promoting Socialization of Care). Kyoto: Minerva Shobo.
Holliday, Ian (2000) ‘Productivist Welfare Capitalism: Social Policy in East Asia’. Political
Studies, Vol. 48: 706–23.
Hong, Joo-Yun (2005) ‘Poll Shows Shift in Korea’s Priorities’. Joong Ang Daily, 23 November.
Hsiung, Ping-Chan (1996) Living Rooms as Factories: Class, Gender and the Satellite Factory
System in Taiwan. Philadelphia: Temple University Press.
Iida, Keisuke (2004) ‘Human Rights and Sexual Abuse: The Impact of International Human
Rights Law on Japan’. Human Rights Quarterly, 26(2): 428–53.
Iwakami, Mami (2000) ‘Shoshi-Koreika Shakai no Oyako Kankei: 30dai wo shoten ni shite
kangaeru (Parent-Child Relationship in the Low Fertility and Ageing Society: with a Focus
on People in their 30s)’. Seikatsu Kyodo Kumiai Kenkyu, Vol. 2: 19–26.
Jones, Catherine Finer (1990) ‘Hong Kong Singapore, South Korea and Taiwan: Oikonomic
Welfare States’. Government and Opposition, 25 (Autumn): 446–62.
Jones, Catherine Finer (1993) ‘The Pacific Challenge: Confucian Welfare States’. In Catherine
Jones (ed.) New Perspectives on the Welfare State in Europe. London: Routledge.
Kim, Byung-Kook (2000) ‘The Politics of Crisis and a Crisis of Politics: The Presidency of Kim
Dae-Jung’. In Kongdan Oh (ed.) Korea Briefing, 1997–1999: Challenges and Change at the
Turn of the Century. Armonk, New York: M.E. Sharpe, 2000.
Kim, Sang Kyun (2003) ‘A Systemic Reform of Korean National Pension Scheme’. Paper
presented at the Canada-Korea Social Policy Research Symposium, Seoul, 22–23 November.
Kim, Young-hoon (2003) ‘Women Delay Marriage, Report Shows’. Joong Ang Daily, 3 July.
Korean National Statistics Office (KNSO) (1996) Report on Population Census. Seoul, KNSO.
Korean National Statistics Office (KNSO) (2003) Future Household Projection. Seoul, KNSO.
Korean Women’s Development Institute (KWDI) (2004) Statistical Chart on Women. Seoul:
KWDI.
Koreika Shakai wo Yokusuru Josei no Kai (ed.) (1998) Tomodaore ka Tomodachi Shakai he
(Together We All Collapse or Together We Support Together). Tokyo: Akashi Shoten.
Kwon, Huck-ju (2001) ‘Globalization, Unemployment and Policy Responses in Korea:
Repositioning the State?’ Global Social Policy, 1(2): 213–34
Lee, Hyekyung (1999a) ‘Globalization and the Emerging Welfare State – the Experience of
South Korea’. International Journal of Social Welfare, 8: 23–37.
Lee, Hyekyung (2005) ‘Civil Society and Welfare Reforms in Post-Crisis South Korea’. Paper
presented at Canada-Korea Social Policy Research Cooperation Symposium, Munk Centre
for International Studies, University of Toronto, Canada, 26–27 January.
Lee, Heykyung and Yeong-Ran Park (2003) ‘Families in Transition and the Family Welfare
Policies in Korea’. Paper presented at the Canada–Korea Social Policy Research Symposium,
Seoul, 22–23 November.
Lee, Hyekyung and Ito Peng (2005) ‘Civil Society and Social Policy Reform in Japan and
Korea’. Paper presented at Association of Asian Studies Annual Congress, 31 March–4 April,
Chicago.
Lee, Hyun-Seung (1999b) ‘Korea’s Aging Population’. Monthly Chosun, September.
Mainichi Shimbun (2000) Mainichi Shimbun Seron Chosa (Mainichi Shimbun National Opinion
Survey). Tokyo: Mainichi Shimbun.
Ministry of Health and Welfare (MHW) (1997) Heisei 9-nendo Kosei Hakusho. (1997 White
Paper on Health and Welfare), Tokyo: MHW.
Ministry of Health, Welfare and Labour (MHWL) (2001) Heisei 13-nendo Kosei Rodo Hakusho.
(2001 White Paper on Health, Welfare, and Labour). Tokyo: MHWL.
Ministry of Labour (1998) Heisei-9 Hataraku Josei no Jijo (Situation of Working Women, 1997).
Tokyo: 21-seiki Shokugyo Zaidan.
Mkandawire, Thandika (ed.) (2004) Social Policy in a Development Context. New York: Palgrave
MacMillan.
Moon, Seung Sook (2002) ‘Carving out Space: Civil Society and the Women’s Movement in
South Korea’. Journal of Asian Studies, 61(2): 473–500.
National Institute for Population and Social Security Research (NIPSSR) (2002) Jinko Dotai
Tokei (Population Statistics). Tokyo: NIPSSR.
National Institute for Population and Social Security Research (NIPSSR) (2004) Population
Statistics, 2004. Tokyo: NIPSSR.
150 Ito Peng
Nihon Fujin Dantai Rengokai (NFDR) (Association of Women’s Organization of Japan).

(1997) Josei Hakusho, 1997 (White Paper on Women, 1997). Tokyo: Horupu.
Nihon Fujin Dantai Rengokai (NFDR) (Association of Women’s Organization of Japan).
(2001) Josei Hakusho, 2001 (White Paper on Women, 2001). Tokyo: Horupu.
OECD (2004) Social Expenditure Database. Paris: OECD.
Osawa, Mari (2000) ‘Government Approaches to Gender Equality in the mid-1990s’. Social
Science Japan Journal, 3(1): 2–19.
Osawa, Mari (2003) ‘Japanese Government Approaches to Gender Equality Since the mid-
1990s’. Occasional Paper Series 9, Wayne State University College of Urban, Labor and
Metropolitan Affairs, February.
Palley, H. (1992) ‘Social Policy and the Elderly in South Korea: Confucianism, Modernization
and Development’. Asian Survey, 32: 787–801.
Park, Yeong-Ran (2005) ‘Gender Dimensions of Family Policy in Korea’: Paper presented at
Canada–Korea Social Policy Research Cooperation Symposium, Munk Centre for
International Studies, University of Toronto, Canada, 26–27 January.
Pekkanen, Robert (2000) ‘Japan’s New Politics: The Case of the NPO Law’. Journal of Japanese
Studies, 26(1): 111–48.
Pempel, T.J. (1998) Regime Shift: Comparative Dynamics of the Japanese Political Economy. Ithaca
NY: Cornell University Press.
Peng, Ito (2002a) ‘Japan: The Interaction of Gender and Demography in a Low-Birth Rate
and Aging Society’. In Rianne Mahon and Sonya Michel (eds.), Child Care at the Crossroads:
Gender, Entitlements, and Welfare State Restructuring. New York: Routledge.
Peng, Ito (2002b) ‘Social Care in Crisis: Gender, Demography and Welfare State Restructuring
in Japan’. Social Politics, 9(3): 411–43.
Peng, Ito and Joseph Wong (2004) ‘Growing Out of the Developmental State: East Asian
Welfare Reform in the 1990s’. Paper presented at International Sociological Association,
Research Committee 19, Conference, 2–4 September, Paris, France.
Prime Minister’s Office Japan (1994) Seron Chosa (Public Opinion Survey). Tokyo: PMO, May.
Retherford, Robert D., Naohiro Ogawa and Rikiya Matsukura (2001) ‘Late Marriage and Less
Marriage in Japan’. Population and Development Review, 27(1): 65–102.
Seguino, Stephanie (1997) ‘Gender Wage, Inequality and Export-led Growth in South Korea’.
Journal of Development Studies, Vol. 34: 102–32.
Seong, Kyoung-Ryung (2000) ‘Civil Society and Democratic Consolidation in South Korea:
Great Achievements and Remaining Problems’. In Larry Diamond and Byung-Kook Kim
(eds) Consolidating Democracy in South Korea. Boulder CO: Lynne Rienner.
Shipper, Apichai W. (2005) ‘Shaping the Japanese State: The Role of Foreigner Support
Groups in Redefining Membership Rules and State Responsibilities’. Paper presented at
Association of Asian Studies Annual Congress, 31 March–4 April, Chicago.
Sung, Sirin (2003) ‘Women Reconciling Paid and Unpaid Work in a Confucian Welfare State:
The Case of South Korea’. Social Policy and Administration, 37(4): 342–60.
United Nations Development Programme (UNDP) (2002) Human Development Report.
Available at http://hdr.undp.org/statistics/data/indic/indic_218_1_1.html, accessed on
3 August 2005.
Yamada, Masahiro (1999) Parasite Singles no Jidai (The Age of Parasite Singles). Tokyo: Chikuma
Shinsh.
Wong, Joseph (2004) Health Democracies: Welfare Politics in Taiwan and South Korea. Ithaca NY
and London: Cornell University Press.
7
The Evolution of the Women-friendly
State: Opportunities and Constraints
in the Swedish Welfare State
Barbara Hobson
Introduction
Sweden is an interesting case in which to consider gender and social policy,

looking through the lens of institutions, social actors and gendered discourses on
equality. It has been one of the most interventionist states, mitigating the power
of capital and markets to control workers and working conditions. It scores high-
est on the dimensions of social rights that decommodify workers, that is, policies
that weaken a worker’s dependence on the market (Esping-Andersen 1990; Korpi
and Palme 1998). In the literature on comparative welfare regimes, it is presented
as the society where social citizenship and the pursuit of equality were shaped by
class-based politics; the cultural narratives that resonate in the society have been
those of working-class struggles (Acker 1992; Åmark 1992). Social rights in this
framework are the outcome of struggles over the distribution of society’s resources
where unionized workers and employers have been the main protagonists (Korpi
1980). The redistributive effects of wage and social policies are evident in that
Sweden scores very low on measures of social stratification and has one of the
most compressed wage structures.
Sweden is also cast as the exemplar of the women-friendly society with policies
that enable women to combine employment with having and caring for families.
It is considered the paradigm of the weak breadwinner society (Lewis 1992), in
which women are least dependent on a male breadwinner; this is reflected in the
low levels of women’s poverty after divorce. Women’s ability to form independent
households without falling into poverty has been referred to as defamilialization
(Hobson 1990, 1994; Orloff 1993). This process involves women’s integration into
paid work but it is also dependent upon state provision of housing allowances and
income-maintenance benefits, which acts as a buffer for solo mothers, who after
divorce tend to have on average a third less income than married mothers
(Hobson and Takahashi 1997). In coining the phrase ‘women-friendly state’, Helga
Hernes (1987) underscored not only social citizenship rights, but also participa-
tory citizenship rights, and in particular women’s agency in policy-making. She
151
152 Barbara Hobson
maintained that there were dynamic and interactive processes at work in women-
friendly states: policies that have facilitated participatory citizenship through
incentives to enter the labour market produce active recruitment of women into
unions and politics. These processes laid the ground for mobilizing women’s
constituencies to make claims on the state for more power in the shaping of
welfare institutions (Siim 1994).
Alongside this view that highlights the importance of Swedish social policies in
promoting participatory citizenship in the spheres of family, paid work and poli-
tics (Hernes 1987; Bergqvist 1999), one also finds less optimistic and critical voices
toward women’s inclusion into state institutions and policy-making. For instance,
some feminists argue that the Swedish welfare state has produced a system of
segregating men and women into different responsibilities with different power
positions (Hirdman 1989). Equality has meant privileging male norms. In the
same vein, others maintain that Sweden’s women-friendly policies are top-down
policies not framed in women’s interests but rather in the interest of family and
children (Eduards 1991; Ohlander 1992) and that there is feminism without femi-
nists (Gelb 1989). Taking the most pessimistic view of feminism in Sweden, Elman
(1995) claims that there is neither feminism nor feminists as seen in the failure to
address radical feminist issues, such as violence and sexual harassment.
This contested view of Sweden and gender equality reflects ideological rifts over
what feminism is, debates over gender difference and gender equal strategies, over
autonomous feminism versus institutionalized feminism. This is a dead-end
debate. My purpose in this paper is to broaden the conceptual terrain in gender
and social policy away from the ideological debates over autonomous and institu-
tional feminism in Sweden toward more process-oriented approaches that
highlight the mechanisms of power and agency in shaping gender-equality
outcomes. My approach assumes that gender strategies are contextually bound
and connected to social movements. This is to take seriously the notion of
capabilities (Sen 1992; and Lewis and Giullari in chapter 8 of this volume), and to
consider what kinds of policy frameworks and institutions promote individual
choices and participation in different spheres of life – the economic, social and
political. It is also important to consider how enhanced individual capabilities and
agency facilitate collective agency, that is, the ability of organizations and con-
stituencies of women to make claims, to have a voice and influence in creating
gender-sensitive social citizenship rights (Hobson and Lister 2002). Within the
comparative research on gender and welfare states there is an assumption that
societies with high rates of labour-force participation and defamilialization poli-
cies reduce agency inequalities (Korpi 2000); they increase women’s capabilities to
be full citizens. These in turn lead to greater participatory citizenship and collec-
tive political presence, thus creating women publics. The example of the former
socialist states demonstrates that high rates of labour-force participation and weak
economic dependency do not in themselves lead to the creation of women’s polit-
ical presence. Hence we need a more nuanced analysis of participation parity and
agency inequality in order to unravel a complex set of processes: How and on what
basis women are included in policy-making spheres shapes not only outcomes in
The Evolution of the Women-Friendly State 153
citizenship rights, but also political identities and the meanings of emancipation.
What kinds of claims are made and which women actors come to speak for
women, is historically and contextually grounded, and rooted in previous interac-
tions between social actors and institutions.
In this chapter I trace the role of women’s collectivities in shaping the women-
friendly Swedish welfare state. I address three sets of questions. Firstly, how have
feminist political actors (individuals and organizations) and their allies in political
parties and trade unions shaped the contours of Sweden’s ‘women-friendly’ wel-
fare state? Secondly, how have alliances and coalitions with political parties and
unions acted as constraints for women’s claims for equal participation in political,
economic and social spheres? And finally, which groups have been left out or
marginalized in the participatory model of gendered citizenship? Linked to the
question of which groups were not recognized, we need to consider which dis-
courses became recognized as embodying gender equality and which actors
became the privileged spokespersons for feminism.
Embedded in these questions are mechanisms of power and agency. Two that
I have used in my own work (Hobson 2003) are highly relevant to social
movements involving women’s movements and claims-making: brokering and
boundary-making. Brokering is a concept in social-movement theory that I have
defined as the pursuit of allies who gain advantages from their linkages with a
movement or collectivity, who can be sympathetic, lukewarm, resistant or even
antagonistic to the claims a group is making (Hobson 2003). In the case of Swedish
women’s movements, these allies have been leading men in the Social Democratic
Party and Liberal Party as well as trade union representatives. Boundary-making is
a mechanism that reveals the processes by which particular actors (and often their
organizations) become the certified spokespersons for the group, the recognized
actors in public discourse who are interviewed in the media and/or serve on
government commissions (Hobson 2003). These actors have the power to shape
which issues and claims are heard and legitimated, and which are shaded out and
marginalized. Because Swedish feminism early on became institutionalized and
linked to the Social Democracy Party and its ideologies, it is possible to view
the clear path-dependencies in discourse and policy around gender equality,
seen through the privileging of discourses that coupled paid work to women’s
emancipation.
In the following sections I provide a contextualized account of these processes
in the following periods: the first phase of Social Democratic governance in the
1930s, the second phase of welfare-state expansion in 1970s, and the third phase
of global restructuring and the emergence of feminist politics and feminist
political parties in the 1990s.
Foundations
The long duration and hegemony of the Swedish Social Democratic Party is well
known; it has been the only party with the possibility of winning a parliamentary
majority: from 1932 to 1988 the Social Democrats’ share of the vote was between
154 Barbara Hobson
42 and 50 per cent. In the policy realm, Sweden is a society where politics and
markets have been coupled. Up until the mid-1980s, there was a highly institu-
tionalized corporatist bargaining system composed of employer representatives,
unions and members of government. According to one historian of Swedish
politics, the political organization and political culture have inhibited the devel-
opment of social movements based on identities other than class, such as gender,
ethnicity and race (Åmark 1992).
Whereas most studies of Sweden and gender equality begin in the ‘golden age’
of welfare-state expansion in the 1960s and 1970s, in order to understand women’s
agency in the development of women-friendly policies, it is useful to go back to
the first phase of welfare-state formation in the 1930s and the feminist activism
that emerged in the same period. This was a period of political opportunity for
women’s groups as the Social Democratic Party was courting women as voters. It
was also a time in which there was a flowering of women’s organizations and
cross-party alliances.
Political opportunity
After a serious electoral defeat in 1928, the Social Democratic strategy shifted to a
mass-membership party, which downplayed the rhetoric of class struggle and
revolution. This election, called the Cossack Election – the label that conservatives
used in the campaign rhetoric – resulted in a decrease in Social Democratic
representation from 86 to 75 seats. Talk of revolution was replaced by security
(trygghet) (Olsson 1990: 94). Leaders of the party sought to broaden their base and
realized that it was clearly in their interest to get women to become party members
(Karlsson 1990: 323). However, there was no attempt to place women on the
voting lists nor were women’s issues notably present in the party platform.
It is important to keep in mind that the surge in Swedish feminist mobilization
came not during suffrage but in the decade after, the 1930s, a period in which
many Western feminist movements were in retreat. In Sweden large numbers of
women joined organizations for the first time. Sweden’s Social Democratic
Women’s Union increased their membership four-fold.1 Almost twice the number
of women were mobilized compared to those who fought for suffrage. Women’s
organizations that had been divided by class, education, political loyalties and
ideological differences (Lindholm 1991), in the 1930s began cooperating across
party lines. In 1939 a feminist wrote in Morgonbris (the Social Democratic women’s
journal), ‘I have experienced a shift in the women’s movement from a strictly
bourgeois phenomenon, led by middle-class women with great opposition from
working-class women, to a movement that has become everyone’s concern’
(Morgonbris 1939: 5).
This cooperation among women’s groups was facilitated by the shift in rhetoric
within the Social Democratic Party, which opened up new discursive and political
opportunities, as seen in a metaphor that became a master-frame in Social
Democracy, the People’s Home (Folkhemmet). Immortalized in the famous speech
of Prime Minister Per Albon Hansson, the People’s Home was compared to a home
in which there ‘are no privileged or deprived members, no pets and no stepchildren.
In the good home there is equality, solicitude, cooperation and helpfulness’

(quoted in Tingsten 1973: 265). This symbol of the caring state, which would
accrue in meaning over many decades of Social Democratic hegemony, reveals
how boundary-making could operate on two levels: firstly, the marginalization of
radicals in the party, and secondly as a frame within which Social Democracy could
mobilize women, gain influence in their own party, and become the recognized
spokespersons for feminism.
Women’s inclusion in social democracy: participation

as the mainframe
Feminists seeking to mobilize women faced a situation in which a significant
proportion of women did not even exercise their right to vote. An important strat-
egy in this mobilization was to create an informed woman citizen who understood
the policy debates and could influence the policy arena. A group of women called
the Fogelstad Group founded a Citizenship School to achieve this objective.
A feminist writing in their journal maintained, ‘The women’s movement needs its
work sites, a place where its members are trained for their duty as citizens’ (Spoof
1933: 4). Women’s journals provided extensive coverage of political debates, and
parliamentary proceedings were published verbatim in several women’s journals
to keep readers informed. Women were urged to exercise their vote, regardless of
party. That women would be seen as a constituency of voters increased the
brokering and bargaining power of women’s organizations to influence policy-
making in the Social Democratic Party.
Participation and representation were the mobilizing frames for gender politics
in the 1930s and would be the mainframe again for feminist mobilization in the
1990s, after a precipitous decline in women’s representation in parliament. This
framework was strategic in that it appealed to a broad spectrum of women voters,
but it also revealed the early precedents of institutional feminism, a perception
that women’s groups gained power resources through an integrationist strategy,
that policy-making influence came through involvement in policy-making insti-
tutions and party politics. As one author wrote in Hertha, the journal of the oldest
Swedish feminist organization, Fredrika Bremer, ‘We women do not wish to be
invited into the Folkhem once it is finished and ready. We will only be satisfied if
we are welcome to help lay the groundwork and build it’ (quoted in Karlsson 1990:
36; see also Hertha 1938: 98). Another Social Democratic feminist, writing in her
party’s journal, demanded that women be recognized as political actors: ‘We
believe it is politically unjust and poor tactics to treat women as political zeros’
(Västberg 1938: 5).
Women’s right to work

During the 1930s the wide-angle lens of participation as active citizenship
refocused and narrowed when women’s groups concentrated their energies on
the campaign against the threat to married women’s right to work. In Sweden, the
political debate over married women’s right to work arose in a context of high
rates of unemployment, as was true in other industrialized countries in the
156 Barbara Hobson
Depression era (Hobson 1993). The number of unemployed rose more than five-
fold between 1931 and 1932, and then reached its maximum in 1933 when over
186,000 Swedes were registered as unemployed (Tingsten 1973: 285). Married
women were seen as taking men’s jobs and every party from left to right
demanded restrictions. In the official statistics, only 10 per cent of working
women were married, but neither part-time workers nor married women working
alongside their husbands were considered as being in the labour force (Nyberg
1987). Moreover, it is important to keep in mind that half of the adult middle- and
working-class women lived alone with or without children and had to support
themselves up until the 1930s. Perhaps most importantly, there was the symbolic
import of denying married women the right to work. In the Social Democratic
worker state, in which paid work was the marker of citizenship, to exclude married
women was to create a gendered category of lesser citizens.
In response to the pressure for a marriage bar in employment, a branch of Open
Door International, an organization formed to protect women’s right to employ-
ment, was launched in Sweden. The whole spectrum of women’s organizations,
including the National Housewives Association, defended women’s right to work
on the basis that it was a citizenship right, though their preference was for
mothers to be at home (SOU 1938). An article that appeared in the Social
Democratic women’s journal Morgonbris was reproduced in two feminist journals
with a middle-class base. The link between participatory citizenship and the right
to employment was clearly made: ‘the right to walk through the door of the labour
market, not as a subservient, but as full citizens, equal to men. As long as there is
democracy, equality and justice, the bitter war between the sexes should not exist’
(Morgonbris 1937: 4; Tidevarvet 1937: 2; Hertha 1937: 3).
The Farmers’ Party, the Conservative Party and the right wing of the Liberal
Party all advocated laws restricting married women’s employment and asserting
the rights of male breadwinners. The majority of Social Democratic men (including
the leader of the party, Per Albin Hansson) hoped that women would return to the
home after they were married, but Albin Hansson was against restrictions.
However, he did not envision a law protecting married women’s rights to employ-
ment (Franguer 1998). The trade unions were against the marriage bar, not because
they supported women’s rights, but because they opposed any government inter-
ference in worker–employer contracts (Rothstein 1992). A small band of Social
Democratic men were strong supporters who joined forces with women’s groups
to turn the tide against the laws against married women’s work. They were
instrumental in getting a law passed that in fact prohibited the firing of married
women, pregnant women (married or unmarried) and women with children
(Hobson 1993).
Boundary-making: privileged speakers and discourses

in social democracy
The debates and public discourse around married women’s right to work signalled a
change in the framing of gender politics that reflected the prominence of feminists
on government commissions. They were feminist actors who championed paid
work as the path to women’s emancipation. Two key spokespersons in the debate
were the chair and secretary of the Government Committee on Married Women’s
Work. The chair, Kerstin Hesselgren, was the first woman to be elected to the first
chamber of the Swedish parliament in the Liberal Party. Her experience as the first
woman factory-inspector made her an obvious choice to speak for the women’s sit-
uation in the labour market. The secretary of the Committee, Alva Myrdal, was
chairperson of the Professional Women’s Organization, and a Social Democrat with
an inside track into party politics through her husband, Gunnar Myrdal.2
The committee not only steered the discourse in women’s journals and the
media, but also set the terms for the debate and the framing of gender equality in
the future. Myrdal represented a 180 degree turn from the earlier maternalist femi-
nism of Ellen Key, the most prominent writer and speaker on women’s issues in the
first decades of the twentieth century. Key opposed suffrage and advocated a femi-
nism of difference, in which woman’s power derived from her maternal role and
female consciousness. She maintained that feminists who championed the vote
and careers for women were misguided (Key 1912). In contrast, Myrdal imagined a
gender equality in which women and men would both be active participants in the
labour market, and put forward proposals for collective housing and cooperative
kitchens. These utopian schemes did not stick in Sweden, but what did and would
be revived in the 1960s and 1970s was her vision of women’s participatory citizen-
ship through paid work. By the end of the 1930s the maternalist position had lost
ground. Although issues involving motherhood were part of feminist claim –
making – women’s groups put forward demands for maternal health care, paid
vacations for mothers alongside childcare, maternal leave benefits for employment –
the basis for women’s activism was not bound by a maternalist world view that
women had unique political identities based upon their motherly roles.
In the 1930s women’s groups, through their pressure on Social Democratic
government, achieved rights and protections for women in paid work (Hobson
1993; Hobson and Lindholm 1997). These included:
● Protection for all women against the loss of their jobs after marriage or
pregnancy. It was against the law to fire a woman from her job because of
marriage or pregnancy (the law also protected unmarried mothers).
● Maternal leave for all women in waged work, but those in the public sector were
given a paid leave benefit.
● Income maintenance for unmarried and divorced mothers, in which the state
gave advance payments to divorced or unmarried mothers when fathers failed
to pay their support.
At the end of the 1930s, a decade after the Women’s Committee on Married
Women’s Work published its findings and recommendations, Alva Myrdal pro-
claimed the successes that women had achieved: ‘This is a year of social political
duties. We have a government under our rule. We have a Parliament elected on a
worthy social program. We have women’s increased political awareness and also
family politics in the forefront’ (Myrdal 1937: 21).
158 Barbara Hobson
This optimism was short-lived. During the war and postwar years, women’s
issues were placed at the margins of the ‘real politics’. Rationing produced a
reassertion of traditional female skills of canning and domestic activities.
Furthermore, as the postwar years revealed, the consolidation of the Social
Democratic power base shaded out women’s voices and weakened women’s
brokering power with party leaders. It should not be forgotten that the historic
Saltsjöbaden agreement was signed at the end of the 1930s, which laid the
foundation of corporatist bargaining structures between employers, unions and
the government. These postwar years from 1945 until the 1960s represented a
bracketing of feminist politics and it would take over two decades before a revived
feminist movement could emerge. Still, the contours of Swedish feminism that
emerged in the 1930s in this early phase of welfare-state formation shaped future
feminist politics and claims for gender equality.
1 Women’s collective agency was channelled through organizations and govern-

ment bureaucracies, what I have called institutionalized feminism. Hesselgren
and Myrdal were in fact the prototypes of the so-called ‘femocrats’ – a phrase
coined in our own day, applied to women with feminist orientations who have
chosen to enter public life and work within the system to promote women’s
interests (Cass 1994; Eisenstein 1996). Their activities in policy-making reveal
both the strengths and weaknesses of forms of institutional feminism. On the
one hand, women in key bureaucratic positions provide links between grassroots
feminist activism and civil society and government bureaucracy. They have
access to information (early drafts of legislation) and informal networks. On
the other hand, they are compelled to adapt to the rules of the game in bureau-
cratic organizations and this limits their ability to put forward radical agendas.
2 Claims for gender equality came to be more and more bound up with partici-
pation in paid work. This was to be the case for nearly all the Nordic welfare
states, but most true for Swedish women’s claim-making.3 Participatory citizen-
ship became coupled to paid work in this first phase of feminist mobilization
and legitimized through official government policy documents.
3 Feminist actors who were recognized as the spokespersons for women’s interests in
Swedish social democracy tended not to emphasize women’s difference or particu-
larized identities. It is interesting to note that even in the early 1900s the Swedish
delegation to the International Labour Organization (both men and women) was
the only group who refused to sign the protective legislation for women but sought
more universal protections for all workers. Those feminists who became privileged
actors in policy-making in the 1930s and later in the 1970s (by being put on gov-
ernment commissions and given policy-making roles) encased women’s claims in
a gender-neutral frame, one in which gender was not decoupled from class.
Welfare-state expansion and the dual breadwinner model
Unlike many welfare states in which dramatic increases in women’ labour-force

participation occurred in the late 1980s and 1990s, in Sweden the surge in
women’s employment took place during a period of welfare-state expansion and

political opportunity with new coalitions and discourses around equality (Hobson
2003).4 Called the ‘golden age’ of the Nordic welfare states, the 1960s and 1970s
brought to fruition the institutionalized welfare state with its comprehensive poli-
cies covering a range of protections and risks and reaching all citizens. According
to Huber and Stephens (1998) the main contours of these policies include: full
employment, generous welfare state, basic income security, tax system supportive
of production and redistribution and a wage determination system that supports
reduction of inequalities. These reforms in the Swedish welfare state were assumed
to be inclusive, that is among different classes of earners. High earners were
brought into the public system of earnings-related pensions (Korpi and Palme
1998). Everyone was entitled to a basic amount for part of his/her pension,
sickness insurance and parental benefit; however the major part of the benefit was
calculated on prior work experience and earnings (Lindqvist and Marklund 1995).
As the majority of women were out of the labour force, as housewives or workers
on the family farm in which the husband was the recognized breadwinner, they
received the lower flat-rate benefits in contrast to the majority of the ‘working’
population (mainly men) who received the more generous benefits, and the
difference was substantial (Sainsbury 1996).
Women’s interests and the contests over what kinds of redistributive systems
would best benefit women became central issues in political debates in this
changing policy constellation. Should there be a care allowance and housewife
pensions or publicly supported day care? Should there be individual taxation or a
tax for the male breadwinner and his family? Or should benefits and services be
linked to citizenship or paid work?
Looking from a long-distance lens of nearly 25 years, Gunnar Fredriksson, social
democrat, journalist, and one of the central opinion-makers in sex-role debates in
the 1960s, claimed that there were three main positions in the debate on women’s
role in society (Vestbro 1992). The conservative: women as housewives of male
breadwinners. This had been the ideal of the 1950s for both the middle and work-
ing classes. The moderate: in which women would have two roles as housewife/
mother and as participant in waged work, and men would have one role as workers.
She would be responsible for care of children and the home so therefore she could
not achieve an equal position in the labour-market work. The radical: in which
equality between men and women in work, family life and society were assumed.
This rejected biological determinism and saw equality as a question of justice so
that men and women could develop themselves (Vestbro 1992).
Fredriksson concluded that these differing positions represented clear-cut political
lines. The Church party espoused the conservative view; the right and centre par-
ties stood behind the moderate position and the social democrats and liberals took
the radical line. But the cleavages were not that clear and they reflected ideologi-
cal positions that sometimes cut across party lines in the 1960s. In fact,
Fredriksson was part of a network of intellectuals and radicals, called ‘Group 222’,
who orchestrated many of the policy debates and policy frameworks for equality
between the sexes. The participants now appear as prominent members of the
160 Barbara Hobson
Social Democratic Party, but many of the participants in the group were from dif-
ferent political parties (Social Democrats, Liberals and Communists), which is why
they met in secret.5 They were young radicals, men and women, often husband-
and-wife teams, and the most notable attending these meetings were Lisbeth and
Olof Palme.
Generational differences were a key divide in the contests over the male-
breadwinner norm and the state’s role in promoting women’s employment
(Bergqvist 1999). The male-breadwinner norm still had strong supporters in the
Social Democratic Party and even within the Women’s Federation. One of the stal-
wart defenders of the full-time housewife was a long-time activist in the women’s
section of the Social Democratic Party, Nancy Eriksson. She published her book
Just a Housewife to counter the images in the media of housewives as parasites or
prostitutes (Florín 1999). Eriksson launched a crusade against the individualiza-
tion of taxes, which as she rightly understood devalued the carework of mothers
in her generation, and which would ultimately lead to the demise of the male-
breadwinner wage that supported housewives. Realizing that she had sympathizers
among Social Democratic women, the party gave Eriksson a position on the com-
mission to investigate tax reform, but she was outflanked by the younger radical
men and women on the committee.
The winds of change were moving in another direction. In the 1950s, 77 per
cent of women without children were in the labour market, and 55 per cent with
one child were employed; about 28 per cent of those with two children were in
paid work (Florín and Nilsson 1999). A survey of young women in Swedish schools
in the 1950s revealed that young women did not want to follow in the footsteps
of their mothers as housewives ( Johansson 2000). By the time the tax reform was
implemented in the 1970s, nearly half of women were already in the labour market
(Axelsson 1992). Those women who championed the dual-earner model would sit
on the dozens of commissions on marriage, family and childcare throughout the
1970s. Within the Social Democratic Party, they became the privileged speakers for
women’s interests.
The dual-earner model and gender equality
As the research from the former Soviet countries has illustrated, activation of
women into the labour force in itself does not necessarily lead to gender equality
in outcomes, or even a discourse on equal participation of men and women in
paid and unpaid work. In Sweden, the dual-earner norm was coupled to a gender-
equality discourse in which men were considered to be part of the emancipatory
project. From the perspective of Eva Moberg, feminist spokesperson for the Liberal
Party, ‘the fulfillment of the goals of feminism requires a radical change of the
habits of living, attitudes and values of the average man …’ (Moberg 1962: 116).
Yet although radical men in Group 222 supported women’s participation in paid
work and even envisioned gender parity in the labour market, they could not
imagine the transformation in gender roles and the organization of paid and
unpaid work that this would entail (Hobson 2003).
Inclusion and participation parity

In the early debates at the end of the 1960s and 1970s, the assumption was that
women would take full-time employment and that men would take a larger share
of housework. The logic was that the greater employment of women would lead to
more equality in the family. The main policy study in the period, Women’s Life and
Work (Kvinnors liv och arbete), affirmed that there would be a convergence in men’s
and women’s roles (Dahlström 1962). The ideological core of the of the Swedish
gender equality model came to be participation parity: equal participation in the
spheres of work and family life for both men and women. Olof Palme interpreted
this model of gender equality in his keynote speech for the Social Democratic
Party congress in 1972. For him gender equality meant increasing employment for
both men and women as well as changing attitudes about male and female tasks
in the home and a sharing of the division of work at home. However, he did not
suggest policy incentives for men to increase their unpaid work in the home. He
saw women’s struggle as part of the labour movements’ struggle for a just society.
From the perspective of Fredriksson’s three positions (Baude 1992), we can say
that Social Democratic men sympathetic to gender equality, such as Palme,
affirmed a radical rhetoric on participation parity – equality between men and
women in work, family and society, but designed policies that reproduced the
moderate position. Women would have two roles as housewife/mother and par-
ticipant in waged work. Men would have two roles as well – as fathers, but not as
caretakers. Their main role would continue to be that of breadwinner. Both would
be parent-worker-citizens, though men and women would not participate equally
in both spheres of paid work and care.
Parent-worker-citizen: a gender-neutral frame of rights

The Swedish model of the worker-parent-citizen came into being during this
period. There was a purposeful approach to family and labour-market policy in the
1960s and 1970s, a set of initiatives that sought to alter the patterns of family life
so that more women would become labour-force participants. A set of coherent
policies emerged from political debate, government commissions on individual
taxation, family law, parental leave, and day care. Policies were aimed at shaping
the choices of families around paid and unpaid work. Individual taxation raised
the costs of women’s unpaid labour so that families with only a single earner paid
a penalty for maintaining a traditional housewife; every Swedish crown a woman
earned was extra income for the family (Gustafsson and Bruyn-Hundt 1991).
Generous parental leave and highly subsidized day care and elderly care were the
carrots to seduce women into the labour market. The expanded public care
facilities not only provided care for young and old, but also created thousands of
jobs for women.
Women were integrated into the parent-worker-citizen model on men’s terms.
All policies were gender neutral: leave for taking care of children was parental
leave; married and cohabitating partners were treated as gender-neutral individuals
in pension and tax systems; the last remnant of gender-distinctive policy, the
widow’s pension, was phased out in 1990. In December 1990, the month before
162 Barbara Hobson
the law ending widows’ pensions came into effect, there were 64,000 marriages
(the average had been about 40,000 a year (Hoem 1991)), a reflection of the
gap between gender-neutral ideologies and policies and gendered practices in
everyday life.
Sweden was the first society in the Western world to denaturalize child-leave
policy in 1975. No longer maternal leave, it became parental leave and the bene-
fits were generous and the conditions incredibly flexible: fathers and mothers
could divide the leave up in any way that they chose and use the time up to the
eighth birthday of their child. By 1989, parents could take up to one year and
were compensated with 90 per cent of their salaries up to a very high ceiling6 and
a further three months with a flat benefit of about 60 crowns a day (equivalent to
around 7 euros a day). The policy was gender-neutral; the practices were not. By
the early 1990s, less than a third of men took off any time for parental leave and
the proportion of days fathers took off was still no more than 10 per cent of the
couple’s permitted leave.
The idea that men should share the leave was an idea that had existed since
the 1970s in different policy circles (Bergqvist 1999). For example, when the
Commission on Family and Marriage published their report in 1972, they not only
recommended that the leave be gender neutral, but also suggested that the leave
be shared. This was meant to strengthen women’s labour-market position and
increase men’s responsibility for care (SOU 1972: 41).
Women’s organizations across the political spectrum sought legislation that
would designate a portion of the leave for men. They reasoned that otherwise the
reform would be meaningless and perpetuate lower wages and fewer career
prospects for women. Two parliamentary members of the Centre Party proposed a
12-month parental leave in which no parent would use more than eight months.
Perhaps succumbing to the realities of policy-making, Social Democratic women
insisted that one month be reserved exclusively for the father out of the eight
months allowable (Bergqvist 1998). It took two decades before the ‘daddy month’
(one month of parental leave reserved for the father) was enacted (1995), which in
1999 was extended to two months. Men took the lead in this reform; men in men’s
groups and men in key policy-making roles who had clout in political parties,
most notably the head of the Liberal Party and at that time, the Minister of Social
Affairs, Bengt Westerberg.
Though not equal participation by any stretch of the imagination,7 the ‘daddy
months’ embody an awareness that equal participation in the labour market
implies change in the division of unpaid work in the home. Moreover, this policy,
though couched in gender-neutral terms, which states that each parent shall have
at least two months of the leave, is understood as targeted and gender specific – for
fathers (Bergman and Hobson 2002).
Brokering: opportunities and constraints
There is a widely held assumption in the scholarly literature on the Swedish welfare
state that policy interventions around gender were a response to labour-market
shortages – that trade unions preferred women to immigrants to fill the expanding
service economy (Elman 1995; Jonung and Thordarson 1980). More recently
scholars have rejected this functionalist account of gender equality and high-
lighted the range of actors, men and women, who were involved in the debates on
sex roles and gender equality that began in the 1960s (Bergqvist 1999; Florín and
Nilsson 1999; Karlsson 1996). Rather than the main determinant of enabling poli-
cies for women to enter labour-market work, according to Florín and Nilsson
(1999), the shortage of labour offered a window of opportunity for brokering with
men in the Social Democratic Party who were lukewarm towards feminism and
with men in the blue-collar trade unions, the LO, who were even less sympathetic.
Unions became crucial allies of feminists in the 1970s in the debates around
publicly supported day care and individual tax reform. Their backing was essen-
tial. The large industrial unions (LO) were the main power bloc in Social
Democracy. Although they supported women’s issues that were workers’ issues –
they had opposed restrictions on married women’s right to work in the 1930s and
they backed equal pay for women workers in the 1950s – still they had little
sympathy for feminism. They tended to cast women’s movements for gender
equality as bourgeois, driven by professional women and often seen at odds with
the working man’s struggle for the male-breadwinner wage to maintain a full-time
housewife. The LO preference for women’s labour versus immigrant labour to
solve the labour shortage in Sweden is well known (Kyle 1979), and they put pres-
sure on the government to restrict immigration. Activist women in the unions and
Social Democratic Party seized upon this window of opportunity. Women in the
industrial and service-sector unions had been pushing the day-care issue since the
1940s, but in the 1960s they found allies among men in LO, and they were able to
push forward the day-care issue onto the LO agenda at a time when the shortage
of labour was on the minds of policy-makers, employers and union officials
(Hirdman 1998). Considering brokering in unions, we should keep in mind that
women had little voice in the shaping of policy within the trade unions. Although
LO was interested in women’s recruitment into unions, they did not recruit
women into the union hierarchies until the 1990s (Mahon 2002; Hirdman 1998).
Constraints
Alliances in policy-making necessitate forms of brokering that not only involve
strategic framing of issues, but also set the parameters for what kinds of claims will
have play in the policy-making forums. From this perspective, every alliance can
be an exclusion (Hobson 2003).
We can see specific examples of this in response to feminist claims around
working times and pay equity. Since the 1970s, women in the Social Democratic
Party continuously have sought a shorter working week as a women-friendly policy.
However, unions have always won out in the battles over shorter working days
versus longer vacations: the mandated vacation is now six weeks. Policies to alter
the discriminatory gendered patterns in pay, job recruitment and promotion have
engendered the most controversy within the unions. The industrial unions pro-
moted a wage-solidarity policy for low earners in their centralized bargaining with
164 Barbara Hobson
employers, which indirectly helped women who were over-represented in low-paid

unskilled jobs. Still, they were adamantly against any law against discrimination on
gender grounds in wages, hiring and promotion (Ruggie 1984).8 The LO maintained
that such an anti-discrimination law would interfere with collective bargaining.
It is no secret that Sweden has had one of the softest laws against gender
discrimination in employment among democratic welfare states (Eklund 1996; Gelb
1989; Ruggie 1984). Sweden did not have a law against gender discrimination until
the mid-1980s and it was extremely weak, since it did not cover gender discrimina-
tion under union/employer contracts. A revised law in 1991 eliminated this clause
but the main contours of the law remained. The parameters for claiming discrimi-
nation remain highly circumscribed. The only unlawful sex discrimination exists
when an employer passes over an objectively better-qualified candidate (Eklund
1996). Swedish law does not permit group or class actions based on systematic pat-
terns of discrimination (for example female employees in a firm who are systemati-
cally passed over in promotions). Not until Sweden joined the EU was the principle
of equal pay for work of equal value introduced in Swedish law (Eklund 1996).
The main employer of women in Sweden is the state (including the municipali-
ties) with nurses, teachers and health-system public-sector workers amongst their
employees. The social democratic welfare states offer women a family-friendly
workplace, but they also produce a female ghetto of jobs with low compensation
in a sex-segregated care sector (Huber and Stephens 2000; Mandel and Shalev
2003; Persson and Wadensjö 1997). Though Sweden has been one of the most obe-
dient of member states in implementing EU law in the area of gender equity in
pay, it has shown resistance toward implementing wage equity for women
employed in the highly segregated public-service sector. The equality ombudsman
has lost nearly every case brought before the Labour Court over the last several
years. One in particular stands out as it was appealed to the EU Court twice.9 It
concerns wage discrimination of midwives in comparison with male technical
assistants; the latter have less education and earn more than 4,000 crowns a
month more than the midwives. Though the case was sent back twice from the
European Union Court for review, the Swedish Labour Court first argued for
the pay differential based on market demand and then for supremacy of collective
agreements (that is, to increase the pay of the midwives would undermine the
union bargaining process).10 This case makes visible the contradictions in the notion
of a women-friendly state: can a women-friendly state be a gender-discriminatory
employer? Should union agreements be outside the purview of legal action when
the wages are systematically set lower for gender-segregated occupations? Can
women who are integrated in the Social Democratic state apparatus and party
machinery make challenges that prioritize a politics of redistribution that recognizes
gender inequalities over class inequalities?
The limits of participation parity
The belief that women’s and men’s roles would converge as more women entered
the labour market appears naïve with hindsight. And the data on practices
45
40
35
30
25
20
15
10
0
1990/91 1990/91 2001/02 2001/02
Men Women Men Women
paid work unpaid work
Figure 7.1 Women’s and men’s paid and unpaid work, weekly hours 1990–91, 2000–01
Source: Statistics Sweden: www.SCB-se/statistik7be0503/tidsavanv.asp
between couples underscores the distance from this idealized gender-equality

model that presumed equal participation in the workplace and the home (see
Figure 7.1). As Figure 7.1 illustrates, the much-heralded ‘daddy month’ has not
made much difference either; women still take the lion’s share of unpaid work. In
1992, women worked on average 27 hours of paid work and 33 hours of unpaid
work; for men the proportions were reversed, with 41 hours of paid work and
20 hours of unpaid work per week (Flood and Gräsjo 1997). More recent surveys in
2001 show only a slight change. Men have reduced their working hours by about
one hour a week, but they have not increased their share of unpaid work. In fact,
men and women have decreased their unpaid work.11
What has been achieved is participation parity in the labour market. There are
nearly equal numbers of women in the labour market, but participation parity
does not necessarily translate into equity. Employment statistics in 2002 show that
between the ages of 20 and 64, 80.5 per cent of men and 77 per cent of women are
in the labour force. Looking at the active labour market years (25–54), Swedish
women have one of the highest participation rates among OECD countries (nearly
90 per cent). But a significant portion of women – over one-third – are working
part time.
The Swedish model, built around the participation of men and women in
unpaid care and in paid work, in practice, is a one-and-three-quarters model. Men
work full time and invest in their careers and women work part time (though it is
a long part time, 20 to 30 hours) in the public sector, in workplaces where it is
easier to combine employment with having a family, in so-called women-friendly
jobs. Mothers choose women-friendly workplaces with low pay. This trade-off has
166 Barbara Hobson
70
60
50
40
30
20
10
0
SW95 SW00 UK99 FR94 GE94 GE00 NL94 SP90 IT95 IE96 US00 HU99 AS94
Figure 7.2 Poverty rates of solo mothers

Source: Derived from LIS (Luxembourg Income Study Data).
Sw-Sweden, UK-United Kingdom, Fr-France, Ge-Germany, It-Italy, IE-Ireland, US-United States,
HU-Hungary, As-Australia.
consequences for mothers after divorce. Women not attached to male earners are
poorer in terms of time and money. This has intensified since the 1990s, a period
of economic downturns and global restructuring and the fiscal conservatism that
EU membership requires.
Though Sweden continues to have the lowest poverty rates among solo mothers
compared to other industrialized welfare states (see Figure 7.2), since the 1990s,
there has been a decline in solo-mother earnings as a result of their deteriorating
labour-market position. Solo poverty rates have doubled (Fritzell 2000). The
employment rates for solo mothers are now lower than for partnered mothers,
which represents a reverse of the earlier pattern. Not only higher unemployment,
but also longer periods of unemployment, underemployment and withdrawal
from the labour market characterized the experience of solo mothers in the 1990s
(Björnberg 1997; Statistics Sweden 1998; Sainsbury 2000).
Non-Nordic immigrant groups are also bearing the brunt of a highly competitive
and discriminatory labour market. At the beginning of the 1990s, 70 per cent of
immigrants were employed, but by the end of the 1990s, the level had dropped to
55 per cent. Non-Nordic women immigrants fared the worst, with only 43 per cent
holding onto their jobs (Sainsbury 2000). And immigrant women have not had a
voice in feminist organizations or immigrant organizations (Hobson et al. 2006).
Participation and voice

The interplay between institutions and women’s collective agency has been
characterized as a virtuous circle in women-friendly states, in which social policies
and institutions create female constituencies who reinforce women-friendly
policies (Bryson 1999; see also Razavi 2002). This virtuous circle opens up oppor-
tunities but also closes off others, including types of claims that can be made and
discursive resources to open up new agendas. Institutionalized feminism (women’s

groups ensconced in political parties and unions) and femocrats scattered
throughout the Swedish bureaucracy gave women a voice in policy-making, and
in the making of the women-friendly welfare state. Staying within the circle allows
influence, but limits autonomy and inhibits public debate and criticism of
government actors and policies. Party discipline is strong.
Women’s claims-making has been lodged within political parties; each party has
its own women’s organization. This has inhibited coalitions across party lines and
the development of grassroots women’s movements and NGOs to push radical
agendas from outside institutional structures. It was not until the early 1990s that
feminist mobilization similar to that of the 1930s emerged. As was true of the
earlier era, representation in policy-making was the issue that led to cross-party
alliances just before the 1994 election, which took the form of loosely coordinated
organizations that were represented in the public arena as a network. Calling
themselves the Redstockings, feminist organizations flexed their muscle and
threatened to form a women’s political party during an election year. They were taken
seriously, as women made up more than half the Social Democratic voters. The
mobilization lead to a dramatic increase in women’s political representation,
which is now at 43 per cent;12 half the ministers are women.
Still, this representation near-parity does not necessarily translate into voice and
influence. The unfilled promises of the government concerning improved services
for care of the elderly and children alongside the frustration of women inside and
outside political parties to shape agendas has borne fruit with the emergence of
another feminist network. Unlike the earlier feminist network, which only threatened
to form a party, this coalition of feminists, called a feminist party by the press, has
been more open to the idea of actually forming a political party. However, recent
polls suggest that there has been a backlash and decline in support for the feminist
party. But its very existence shapes the processes of brokering and boundary-
making within parties The current network, with its grassroots and loose organi-
zational structure – lacking the strict party discipline of the traditional parties in
which women have been weaker brokers – has the potential to produce a wider set
of social-policy debates around work and care. It has already opened up new
debates and pressures for tougher laws against rape and domestic violence (two
issues that have galvanized younger feminists). Finally, the network has the poten-
tial to create a multi-layered feminism with new voices, including gay and lesbian
groups and immigrant women’s organizations, groups whose issues have been
shaded out in the institutionalized feminism of the past.
Whether we see a formal feminist party with representatives in parliament in
the next election or whether leaders decide to remain a cross-party network of
women’s groups may be less important than the fact that women have created a
political presence, a women’s public voice with a core of issues and a feminist iden-
tity. This sense of a virtual women’s constituency with a gender-equal profile has
made it difficult for political parties to gain advantage through appeals to tradi-
tional gender roles and family values. Women’s political presence in parties and
unions also can be seen as a factor in limiting the challenges to women’s social
168 Barbara Hobson
citizenship, the chipping-away of benefits and services, during an era of welfare-

state retrenchment. Retrenchment and restructuring in many countries have cut a
swathe into women’s social rights. In Sweden, many policies remain untouched,
including family allowance, day care and parental leave, that have become sacred
cows in Swedish society. The question that hovers in the minds of many feminists
in Sweden and in other countries, however, is whether a women-friendly state can
be sustained in globalized economies with the diminishing policy-making role of
the state.
Notes
1 The Social Democratic Women’s Organization increased its membership from 7,302 in
1930 to 26,882 in 1940. In 1937 the Fredrika Bremer society (a Swedish organization
dedicated to advancing feminism) had 37 local organizations with over 6,000 members.
The National Association for Housewives’ mailing list grew from 10,000 newsletters sent
in 1930 to 23,550. The Swedish Women’s Business Organization boasted 4,000 members
at the end of the 1930s (Åkerman 1983: 122).
2 The committee produced a 500-page report that included masses of statistics that under-
scored the extent of gender inequality in the labour market (SOU 1938: 2).
3 See Leira (1992) for a discussion of Norwegian exceptionalism in this regard.
4 One might compare the women-friendly policy initiatives to those developed in the
1990s, a period with dramatic increases in women’s labour-force participation in Europe
(see Bruning and Plantenga 1999).
5 The name of the group is the address of a well-known feminist, Annika Baude, who
became a key person in the debates on women’s work.
6 At that time it was the equivalent of 3,200 euros.
7 Though over three-quarters of fathers take some leave, the proportion of leave remains
at 17 per cent.
8 The first anti-gender discrimination law was passed during a term when the Social
Democrats were out of power in the mid-1980s.
9 Swedish Equality Ombudsman brought the case against the Örebro county council who
employed both workers at the hospital (case a 153/95).
10 The case was taken to the EU Court twice and rejected by the Swedish Labour Court: the
first time arguing that the two jobs were not comparable, and the second time claiming
that it would undermine collective bargaining (Case C-236/98). The case has not been
appealed again due to cost of EU cases (Interview with Jamställdhetsombudsmannen).
11 Women have decreased their hours by about two and a half hours per week; men by
about an hour per week.
12 The Social Democratic Party in the 1994 election made the decision to alternate women
and men on all of their party lists. This meant greater representation in real terms.
Bibliography
Acker, J. (1992) ‘Två Diskurser om Reformer och Kvinnor i den Framtida Välfärdsstaten’ (Two
Discourses on Reforms and Women in the Future Welfare State). In Acker (ed.) Kvinnors och
Mäns Liv och Arbete. Stockholm: SNS Förlag.
Åkerman, B. (1983) Vi Kan, Vi Behövs! – Kvinnorna går samman i egna föreningar ( We Can! We
are needed! – Women Organize themselves in Their Own Moverments). Stockholm: Förlaget
Akademilitteratur.
Åmark, K. (1992) ‘Social Democracy and the Trade Union Movement: Solidarity and the
Politics of Self-Interest’. In K. Misgeld, K. Molin and K. Åmark (eds) Creating Social
Democracy, A Century of the Social Democratic Party in Sweden. University Park PA:
Pennsylvania State University Press.
Axelsson, C. (1992) Hemmafru som forsvann. Övergången till lönearbete bland gifta kvinnor i
Sverige 1968–1991 (The Housewife that Disappeared: On the Shift to Wage Labour amongst
Married Women in Sweden 1968–1991). Stockholm: Swedish Institute for Social Research.
Baude, A. (1992). ‘Inledning’ (Introduction). In A. Baude (ed.) Visionen om jämstalldhet
(Vision of Equality). Stockholm: SNS.
Bergman, H. and B. Hobson (2002) ‘Compulsory Fatherhood: The Coding of Fatherhood in
Sweden’. In B. Hobson (ed.) Making Men into Fathers: Men, Masculinities and the Social
Politics of Fatherhood. Cambridge: Cambridge University Press.
Berqvist, C. (ed.) (1999) Equal Democracies? Gender in the Nordic Countries. Oslo: Scandinavian
University Press.
Björnberg, U. (1997) ‘Single Mothers in Sweden: Supported Workers Who Mother’. In
S. Duncan and R. Edwards (eds) Single Mothers in an International Context. London
University College: London Press.
Bruning, C. and J. Plantenga (1999) ‘Parental Leave and Equal Opportunities: Experiences in
Eight European Countries’. Journal of European Social Policy, Vol. 9, No. 3: 195–209.
Bryson, V. (1999) Feminist Debates: Issues of Theory and Political Practice. London: Routledge.
Cass, B. (1994) ‘Citizenship, Work and Welfare: The Dilemma for Australian Women’. Social
Politics, Vol. 1, No. 1: 106–124.
Dahlström, E. (ed.) (1962) Kvinnors liv och arbete (The Life and Work of Women). Stockholm: SNS.
Eduards, M. (1991) ‘Toward a Third Way: Women’s Politics and Welfare Policies in Sweden’.
Social Research, Vol. 58, No. 3: 677–705.
Eisenstein, H. (1996) Inside Agitators: Australian Femocrats and the State. Philadelphia: Temple
University Press.
Eklund, R. (1996) ‘The Swedish Case – the Promised Land of Sex Equality?’ In T. Hervey and
D. O’Keefe (eds) Sex Equality Law in the European Union. New York: John Wiley.
Elman, A. (1995) ‘The State’s Equality for Women; Sweden’s Equality Ombudsman’. In
D. McBride Stetson and A.G. Mazur (eds) Comparative State Feminism. Thousand Oaks CA: Age.
Esping-Andersen, G. (1990) The Three Worlds of Welfare Capitalism. Princeton NJ: Princeton
University Press.
Flood, L. and U. Gräsjo (1997) ‘Tid för barn, tid för arbete’ (Time for Children, Time for
Work). In G. Ahrne and I. Persson (eds) Familj, makt och jämställdhet (Family, Power and
Equality). Stockholm: Statens offentliga utredning (SOU), 138: 159–188.
Florín, C. (1999) ‘Skatten som befriar: Hemmafruar mot yrkeskvinnor in 1960-talets
skattningsdebatt’ (The Tax that Liberates: Housewives against working women in the
‘Separate taxation’ debate of the 1960s). In C. Florín, L. Sommerstad and U. Wikander (eds)
Kvinnor mot Kvinnor: Om systerskapests svårigheter (About the Problems of Womanhood).
Stockholm: Nordstedts.
Florín, C. and B. Nilsson. (1999) ‘Something in the Nature of a Bloodless Revolution … How
New Gender Relations became State Policy in Sweden in the 60s and 70s’. In Rolf
Torstendahl (ed.) State Policy and Gender System in the Two German States and Sweden
1945–1989. Stockholm: Opuscula Historica Upsaliensis, No. 22.
Frangeur, R. (1998) Yrkeskvinnor eller makens tjänarinna? Striden om yrkesrätten för gifta kvinnor i
mellankrigstidens Sverige ( Working Woman or the Servant of Power? The Fight about Labour
Rights for Married Women in the Mid-War Period ). Stockholm: Arkiv; PhD thesis, Department
of History, Stockholm University.
Fritzell, J. (2000) ‘Inkomstfördelningens trender under 1990-talet’ (Trends in Wage
Differences during the 1990s). In J. Palme and J. Fritzell (eds) Välfärdens finansering och
fördelning (The Financing and Distribution of Welfare). SOU: 57. Stockholm: Fritzes.
Gelb, J. (1989) Feminism and Politics: A Comparative Perspective. Berkeley CA.: University of
California Press.
Gustafsson, S. and M. Bruyn-Hundt (1991) ‘Incentives for Women to Work: A Comparison
between The Netherlands, Sweden and West Germany’. Journal of Economic Studies, 18.
170 Barbara Hobson
Hernes, H. (1987) Welfare State and Woman Power: Essays in State Feminism. Oslo: University
of Oslo Press.
Hirdman, Y. (1989) Att Lägga livet till rätta. Studier i sventsk folkhems politik (To Adapt Life the
Right Way. Studies of the Swedish ‘People’s Home’ Policy). Stockholm: Norstedts Förlag.
Hirdman, Y. (1998) Med kluven tunga: LO and Genusordningen (Divided Tongue: The Labour
Union and the Role of Gender). Stockholm: Atlas.
Hobson, B. (1990) ‘No exit, No Voice: Women’s Economic Dependency and the Welfare
State’. Acta Sociologica 33, 3: 235–50.
Hobson, B. (1993) ‘Married Women’s Right to Work in the U.S. and Sweden During the
1930s’. In S. Koven and S. Michel (eds) Mothers of a New World. Maternalist Politics and the
Origins of Welfare States. New York: Routledge, Kegan and Paul.
Hobson, B. (1994) ‘Solo Mothers, Policy Regimes, and the Logics of Gender’. In D. Sainsbury
(ed.) Gendering Welfare Regimes. London: Sage.
Hobson, B. (2003) ‘Recognition in Universal and Gender Distinctive Frames’. In B. Hobson
(ed.) Recognition Struggles and Social Movements: Contested Identities, Agency and Power.
Cambridge: Cambridge University Press.
Hobson, B. and M. Lindholm (1997) ‘Collective Identities, Power Resources, and the Making
of Welfare States’. Theory and Society (Fall): 1–34.
Hobson, B. and R. Lister (2002) ‘Citizenship’. In B. Hobson, J. Lewis and B. Siim (eds)
Contested Concepts in Gender and Social Politics. Cheltenham: Edward Elgar.
Hobson, B. and M. Takahashi (1997) ‘The Parent-Worker Model: Lone Mothers in Sweden’.
In Jane Lewis (ed.) Lone Mothers in European Regimes: Shifting Policy Logics. London: Jessica
Kingsley Publishers.
Hobson, B., M. Carson and R. Lawrence (2006) ‘Recognition Struggles in Trans-national
Arenas: Negotiating Identities and Framing Citizenship’, Critical International Reviews of
Social and Political Philosophy, 8(4).
Hoem, J. (1991) ‘To Marry Just in Case … ; The Swedish Widow’s Pension Reform and the
Peak in Marriages in 1989’. Acta Sociologica 34: 127–35.
Huber, E. and J. Stephens (1998) ‘The Internationalization and the Social Democratic Model:
Crisis and Future Prospects’. Comparative Political Studies, Vol. 31, No. 3: 353–97.
Huber, E. and J. Stephens (2000) ‘Partisan Governance, Women’s Employment and the Social
Democratic State’. American Journal of Sociology, Vol. 45: 323–43.
Johansson, U. (2000) Normalitet, klass och kön: Liv och lärande i svenska läroverk 1927–1960
(Normality, Class and Gender: Life and Learning in Swedish Upper Secondary Schools
1927–1960). Stockholm: Arkiv, PhD. dissertation, History Department, Stockholm
University.
Jonung, C. and B. Thordarson (1980) ‘Sweden’. In A. Yohalem (ed.) Women Returning to Work:
Policies and Progress in Five Countries. Montclair NJ: Landmark Studies.
Karlsson, G. (1990). Manssamhället till behag? (Conforming to the Patriarchy?) Tidens.
Karlsson, G. (1996). Från broderskap till systerskap, Det socialdemokratiska kvinnoförbundets
kamp för inflytande och makt i SAP (From brotherhood to sistership, The social democratic
womens unions’ struggle for influence and power in SAP). Lund: Arkiv.
Key, E. (1912) The Women’s Movement. New York: Putnam.
Korpi, W. (1980) The Democratic Class Struggle. London: Routledge.
Korpi, W. (2000) ‘The Faces of Inequality: Gender, Class, and Patterns of Inequalities in
Different Types of Welfare States’. Social Politics, Vol. 7, No. 2: 127–91.
Korpi, W. and J. Palme (1998) ‘The Paradox of Redistribution and Strategies of Equality and
Poverty in Western Countries’. American Sociological Review, Vol. 63: 661–87.
Kyle, G. (1979) Gästerbetskerka in mannsamhället (Female Guest Workers in the Patriarchy).
Stockholm: Liber.
Leira, A. (1992) Welfare States and Working Mothers. Cambridge: Cambridge University Press.
Lewis, J. (1992) ‘Gender and the development of welfare regimes’. Journal of European Social
Policy, Vol. 2, No. 3: 159–73.
Lindholm, M. (1991) ‘Swedish Feminism, 1835–1945: A Conservative Revolution’. Journal of

Historical Sociology, Vol. 4, No. 2: 121–42.
Lindqvist, R. and S. Marklund (1995) ‘Forced to Work and Liberated From Work: A Historical
Perspective on Work and Welfare in Sweden’. Social Welfare, Vol. 4: 224–37.
Mahon, R. (2002) ‘Sweden’s LO: Learning to Embrace the Differences Within?’ In F. Colgan
and S. Ledwith (eds) Gender, Diversity and Trade Unions: International Perspectives. London:
Routledge.
Mandel, H. and M. Shalev (2003) ‘Class Conditioning of Welfare State Effects on Gender
Inequality: A Preliminary Comparative Analysis’. Paper prepared for the Annual
Conference of Research Committee 19, Toronto, 21–24 August.
Mkandawire, T. (2001) ‘Social Policy in a Development Context’. Social Policy and
Development Programme Paper No. 7, Geneva: UNRISD.
Moberg, E. (1962) Kvinnor och människor (Female and Human). Stockholm: Bonniers.
Myrdal, A. (1937) ‘Familjefrägorna i riksdagspolitiken’ (Family issues in parliamentary
politics). Morgonbris, February: 1.
Nyberg, A. (1987) ‘Vad är försvarsarbete?’ (What is Gainful Employment?) Kvinnovetenskaplig
Tidskrift: 1.
Ohlander, A.S. (1992) ‘The Invisible Child? The Struggle over Social Democratic Family
Policy’. In K. Misgeld, K. Molin, and K. Åmark (eds) Creating Social Democracy: A Century of
the Social Democratic Labor Party in Sweden. University Park PA: Pennsylvania State
University Press.
Olsson, S. (1990) Social Policy and Welfare State in Sweden. Lund: Arkiv.
Orloff, A.S. (1993) ‘Gender and the social rights of citizenship: The comparative analysis of
gender relations and welfare states’, American Sociological Review, Vol. 58, No. 15: 303–328.
Persson, I. and E. Wadensjö (1997) Glastak och glasväggar? Den könssegregerade arbetsmark-
naden (Glass Ceilings and Glass Walls? The Gender-Segregated Labour Market), Report to the
Committee on the Distribution of Economic Power and Economic Resources between
Women and Men. Stockholm: SOU, Ministry of Labour: 137.
Razavi, S. (2002) ‘Project on Social Policy in a Development Context: Thematic Component
on Gender and Social Policy’. Research Proposal, UNRISD, Geneva.
Rothstein, B. (1992) Den korparativa staten. Intresseorganisationer och statsförvaltning i svensk
politik (The Corporative State. Special-Interest Organisations in Central Government
Administration in Swedish Politics). Lund: Arkiv.
Ruggie, M. (1984) The State and The Working Woman: A Comparative Study of Britain and
Sweden. Princeton NJ: Princeton University Press.
Sainsbury, D. (1996) Gender, Equality and Welfare States. Cambridge: Cambridge University
Press.
Sainsbury, D. (2000) ‘Välfärdutveckling för kvinnor och män på 1990-talet’ (Welfare
Developments for Women and Men in the 1990s). SOU, Stockholm: Fritzes: 40.
Sen, A. (1992) Inequality Re-examined. Oxford: Oxford University Press.
Siim, B. (1994) ‘Engendering Democracy: Social Citizenship and Political Participation for
Women in Scandinavia’. Social Politics: International Studies in Gender, State, and Society, Vol. 1,
No. 3: 286–305.
Statens Offentliga Utredningar (SOU) (1938) Kvinnoarbeteskommitténs betänkande 1938:47
(Report of the womens work committee 1938:47). Stockholm.
Statens Offentliga Utredningar (SOU) (1972) Familj och äktenskap 1972 (Family and Marriage
1972). Stockholm: SOU.
Spoof, C. (1933) ‘Kvinnorna behöver sina verkstäder’ (Women Need their Workshops),
Tidevarvet, July: 1.
Statistiska Centralbyrån (Statistics Sweden) (1998) at www.SCB-se.statistik7be0503/
tidsavanv.asp, accessed 10 February 2004.
Tingsten, H. (1973) The Swedish Social Democrats: Their Ideological Development, trans. Greta
Frankel and Patricia Howard-Rosen. Totowa NJ: Bedminster Press.
172 Barbara Hobson
Vestbro, D.U. (1992) ‘Från liberal rörelse till socialistisk kamp mot patriarkatet’ (From Liberal
Movement to Socialist Struggle against the Patriarchy). In Annika Baude (ed.) Visionen om
jämstalldhet. Stockholm: SNS.
Västberg, D. (1931) ‘Manlig konservatism och kvinnlig sterilitet’ (Male Conservatism and
Female Sterility). Morgonbris, March: 8–9.
Archives and women’s journals consulted

Hesslegren, K. (Archive), The Royal Library in Stockholm: Box L 55:6
Hesslegren, K. (Archive), Gothenburg University, Boxes A14, A 48.
Rydh, H. (Archive), Gothenburg University. Boxes A 12, A120, A30.
Arbetets Kvinnor (Federation of Women in Work) – associated with International Federation
of Business and Professional Women.
Hertha (Fredrika Bremmer Society journal) 1935–47.
Mödersförbundets Medlemsblad (Housewives Association journal) 1932–1947.
Morgonbris (Social Democratic Women’s Federation journal) 1933–1950.
Tidevarvet (Free thinking radical feminist group, Fogelstadskvinnorna, radical men and
women published articles on abortion, birth control, unemployment, women’s status in
labour market 1925–1940.
8
The Adult-worker-model Family
and Gender Equality: Principles
to Enable the Valuing and
Sharing of Care
Jane Lewis and Susanna Giullari
Introduction
There is now agreement in the literature that the male-breadwinner-model family,

in which men took primary responsibility for earning and women for the unpaid
work of care, has been substantially eroded (Crompton 1999; Lewis 2001a). There
is also substantial agreement that the actual patterns of family formation and dis-
solution and labour-market participation have changed, but there has been, if any-
thing, a more rapid shift still in normative ideas about the contributions that
women especially should make to families, and in the goals of social and labour-
market policies. There is of course substantial variation between countries in the
degree of change in social trends, policy assumptions and policy development. But
given the steady (and in some countries dramatic) rise in female labour-market
participation together with large-scale family change, which has resulted in a
rapid increase in lone-mother and single-person households, policy-makers may
feel justified in making assumptions regarding increasing individualization, in the
sense of increasing self-sufficiency. However, part-time employment for women is
extremely common in most Northern and Western European countries, which
alone is sufficient to undermine any such simple assumption. How care work is to
be accommodated in this new model is also a major issue, but one that has often
been subordinated to economic and labour-market concerns and hence has
received only partial acknowledgement in policy-making.
The balance of the contributions men and women make to households in the
form of cash and care is changing, but much more for women who tend to have
added paid work to their existing responsibilities for care, than for men who in
most Western countries have decreased the amount of paid work they do and
increased their carework only relatively slightly (Gershuny 2000). Indeed, in the
adult-worker-model family it is usually assumed by governments and policy experts
that the work of care will increasingly move to the formal, paid sector, which helps
to account for the secondary place accorded to care on the policy agenda.
173
174 Jane Lewis and Susanna Giullari
There have been a variety of reactions in the academic literature to the shift
towards an adult-worker-model family. Some have argued that women’s ‘moral
rationalities’ are at variance with the new model and that many want to, and do,
put the work of informal care first (Duncan and Edwards 1999; Barlow and
Duncan 2000). This position is in some respects bolstered by Hakim’s (2000) argu-
ment that only a minority of women want to be and choose to be ‘career women’.
Others have argued that because the shift towards an adult-worker-model family is
part of a much wider trend towards emphasizing the importance of responsibilities
rather than rights, and of so-called ‘active’ rather than passive ‘welfare’, it is
impossible to oppose (Orloff 2002). The issue then becomes the terms and condi-
tions on which the shift to the new model is undertaken. Many feminists have,
after all, long argued in favour of women’s financial independence. But the nature
of the choices women and men face and the pursuit of gender equality depend on
the extent to which policies address the issue of care work as well as on the posi-
tion of women in the labour market. Choice is socially ‘embedded’ and ‘genuine
choice’ or ‘real freedom to choose’ in respect of the balance of paid and unpaid
work at the level of the household (and hence gender equality) will involve not
only a rebalancing of paid work between men and women, but a complicated
rebalancing of unpaid work between the market, the state, and men and women.
A mother who wishes to take up a good job offer may well hesitate if she can find
only poor-quality (formal or informal) care for her child.1 Choice has become a
central value in the restructuring of welfare states ( Jenson and Sineau 2001), but
this example signals the central difficulty in ensuring real or genuine choice for
women, which in turn impacts on gender equality.
Something that is missing from the debate is a discussion of the principles on
which a move towards the adult-worker model should be based, especially in regard
to the issue of care work. Yet policy-makers usually operate on (often implicit)
assumptions as to what is ‘good’ and desirable, especially in the field of family policy
(Kaufman 2002). The first part of this chapter explores the very different ideas fram-
ing the shift to an adult-worker-model family at the EU level. These ideas tend to be
dominated by economic considerations, which means that the importance
attached to the choices of the individual are considered in relation to economic
goals. In addition, when gender equality is acknowledged it is often defined in a
particular, partial and instrumental way: in respect of the importance of labour-
market participation, but not care work. This also serves to limit genuine choice.
We review the nature of the policy discourse on employment and care at EU
level and go on to suggest that there are real limits to the pursuit of a full adult-
worker model based on the commodification of care. If this argument is accepted,
and given that care work is necessary for human flourishing (particularly, but not
exclusively, in respect of young, old and disabled people), it means that it must be
made possible to choose to care. This in turn raises issues about the recognition
and valuing of care, and about the way it is shared by men and women, and by the
household and the collectivity. In our search for firmer principles to inform
policies, we examine the possibilities offered by the capabilities approach (CA)
developed by Amartya Sen and Martha Nussbaum (Sen 1985, 1992, 1999;
The Adult-worker-model Family and Gender Equality 175
Nussbaum and Sen 1993; Nussbaum 1999, 2000, 2003), who have explicitly
addressed both the issues of gender equality and care. We suggest that the empha-
sis placed by the CA on the individual’s real freedom to choose something he or she
considers to be worthwhile to do or to be provides a substantial basis for the recog-
nition and valuing of care work, as well as providing a means of countering the
more rank instrumentalist concerns expressed at the level of EU policy-making.
However, gender equality also requires care to be more equally shared between men
and women, which is more difficult to encapsulate with a capabilities approach. In
societies where the position of men and women is unequal, it is possible that the
exercise of men’s freedom to choose between work and care serves to limit women’s
choices. This raises in an acute form the old issues of how to reconcile choice and
equality and what kind of state intervention can be justified.
Conceptualizing the adult-worker-model family

at the EU Level
The trend in Western welfare states has been towards a reworking of the relation-
ship between social provision and employment, with the emergence of what
Goodin (2001: 39) has called ‘a new constellation of work-and-welfare variables’.
Policy-makers have sought to shift the emphasis from rights to responsibilities,
and from so-called ‘passive’ to ‘active’ welfare, such that claimants on the welfare
system are ‘encouraged’ into work and the work is made ‘to pay’ (Lödemel and
Trickey 2001). The relationship between employment and social provision has
always been central to modern welfare states (Supiot 1999); what is striking about
this reworking of the relationship is the way in which it has been couched in
gender-neutral language. It is now assumed that women as well as men will be
‘citizen workers’.
Linking social to economic policy

At the EU level, employment growth has increasingly been seen as a key social
policy goal for the Commission, because it is viewed as the means of promoting
social inclusion, a position shared by European social democrats (e.g. Vandenbroucke
2001) and by many (mainly American) writers to the right of the political spec-
trum (most notably Mead 1986). As or more importantly, adult labour-market par-
ticipation has been seen as the way of increasing economic competitiveness: the
1993 White Paper of the Commission on growth and competitiveness was the first
major policy paper to make this case, and also identified the formal care sector as
a source of new jobs (CEC 1993). But implicit in these two positions was a loom-
ing tension between the idea of the wage as the best form of ‘welfare’ and the
inevitably more ‘flexible’ employment regime that is necessary for increasing
competitiveness.
Many commentators have argued that social policy goals are in fact subordinate
to the goals of economic integration policy, that is, to competitiveness and macro
economic stability (e.g. Scharpf 2002; Kleinman 2002). The European Commission’s
Social Policy Agenda, published in 2000, sought to strengthen ‘the role of social
policy as a productive factor’, pointing to the economic benefits of health and

education expenditure, of protection against a range of social risks, and of measures
to facilitate adaptability in the labour market (EC 2000: 5). In this approach, social
policies are justified in terms of the social investment that is necessary to sustain
competition and growth. Indeed, the notion of an agenda for ‘better jobs’ (EC
2003) seems to depend heavily on a ‘high level of social protection, good social
services available to all people in Europe, real opportunities for all and the guar-
antee of fundamental and social rights’ (EC 2000: 13), rather than good-quality
jobs per se, at least for new (often female) entrants at the low end of the labour
market in the short to medium term. The key adaptation strategy for achieving the
new work/welfare relationship was thus conceptualized as one of flexibility and
security.
Furthermore, the new expression of linkage between economic and social policy
has been parallelled by the adoption of a new form of governance – the Open
Method of Coordination – which abandons ‘harder’ instruments like ‘Directives’
in favour of committing member states to common objectives, while leaving
national governments free to determine how to make progress towards them.
Government by some combination of common standards and performance indi-
cators is one way, perhaps the only way, that states can hope to reconcile welfare
ends and market means. However, as O’Connor (2003) has observed, the imple-
mentation of policies to do with the care of children have had more conspicuous
success in terms of implementation in member states when governed by Directive
(as in the case of parental leave (EC 1996).
The implications for gender equality

In respect of female workers, the discourse of social inclusion on the basis of
employment has been folded into the need to promote gender equality, an idea
that has often taken a central position in key policy documents. In other words,
the social/economic policy divide is subsumed under the goal of gender equality.
However, the history of the treatment of gender equality in respect of paid and
unpaid work at the EU level has been complicated (e.g. Ross 2001; Hantrais 2000).
Over the last decade the Commission has increasingly stressed the importance
of the effective use of women’s skills in a competitive, knowledge-based economy
(EC and Council of the European Union 2002), and has received academic backing
for its position. Women have been seen as an untapped labour reserve. Reports to
both the Portuguese Presidency in 2000 (Ferrera et al. 2000) and to the Belgian
Presidency in 2001 (Esping Andersen et al. 2001) favoured higher female labour-
market participation both as a means of increasing both competitiveness and the
tax base of the continental European social-insurance welfare states. The European
Commission has set a target of 57 per cent for female labour-market participation
in member states by 2005, and 60 per cent by 2010 (EC 2002b; EC 2003).
However, generalizing the expectation of labour-market participation to women
further complicates the ‘better jobs’ goal of the Commission. First, there is the
issue of who has been most likely to fill the more ‘flexible’ jobs that are created. As
Rubery et al. (1998) have argued, flexible employment in the sense of part-time
work or short-term, more precarious employment can be seen as a way for women
to enter the labour market, or as a type of employment that women find very hard
to escape.
Furthermore, in the recent economic strategy documents of the Commission,
there has been relatively little reference to the problems raised by the need to
reconcile family and work responsibilities. Indeed, there is increasing evidence
from recent EU documents and from the work of academic policy advisers that the
extension of the ‘citizen worker’ strategy to women is part of the dominant
competitiveness and growth agenda. It is significant that recent EU-level employ-
ment policy documents have put more emphasis on the provision of formal child-
care services, which have been shown to be the most effective way of promoting
female labour-market participation (Bradshaw et al. 1996; Gornick et al. 1997).
In 2002 a benchmark for childcare was set, whereby by 2010 at least 33 per cent of
children under three years of age and at least 90 per cent of children between three
and the mandatory school age should have access to childcare (EC 2003).
The academic arguments put to EU level bodies since 2000 in favour of the
adult-worker-model family have tended to assume the desirability of a ‘de-
familialization strategy’ for care work (for children and adult dependants),
whereby it is commodified and put into the public, formal arena (Esping Andersen
et al. 2001, 2002). The overwhelming emphasis has been on gender equality
defined in terms of increased female employment. In practice, in those countries
where women’s employment has increased the most dramatically over the past
decade, there has been a tendency for women to enter more ‘flexible’ forms of
employment. Thus gender equality defined in terms of individualization and
labour-market participation has not managed to deliver ‘better jobs’ in the sense
of equally secure, equally well-paying jobs, to both men and women. Rather, in
the case of women, ‘better jobs’ have tended to be defined in terms of the extent
to which provision is made for reconciling work and family responsibilities, and
all too often the focus in respect of these policies has been on women alone, rather
than on men and women.
The economic productiveness and competition goal has been accompanied by a
‘work first’ strategy for women and an emphasis above all on the de-familialization
(and commodification) of care work via the provision of childcare services.
Differing amounts of emphasis have been placed on the provision of cash transfers
(e.g. for parental leave) as a means to achieve more female labour-market partici-
pation, but very little attention has been paid to the pursuit of gender equality in
respect of unpaid work at the household level. It is also significant that the princi-
ple of gender equality is deployed in EU-level documents in such a way as to
eclipse the equally difficult issue of ‘choice’. The economic imperative to increase
female employment has resulted in some extraordinarily instrumentalist arguments.
Thus high female labour-market-participation rates are required for the health
of the economies of EU member states and the EU social model, and are
justified in terms of achieving ‘gender equality’ (Esping Andersen et al. 2002). The
de-familialization of care work necessarily follows, and there is no further discussion
of gender inequalities in unpaid work. Nevertheless, very different ideas about the
desirability of ‘choice’ in regard to the performance of unpaid work exist, both at

the level of national governments and among people, and the willingness to
undertake informal care work is unlikely to become less important in the future.
Care and the limits of an adult-worker-model strategy

based on the commodification of care
The key issue underlying the goal of making further progress towards a full adult-
worker-family model within existing policy approaches is thus, above all, how far
it is possible to defamilialize and hence ‘commodify’ care. This in turn begs
questions about the nature of care work. It also raises issues about gender equality,
for policies to reconcile work and family have historically been focused in the
main on women only.
The commodification of care can take place in the formal and informal arena,
with very different implications for defamilialization. Ungerson (1997) argues that
the different kinds of cash payments for informal caregivers and care receivers
introduced across welfare states, indicate a general trend towards the commodifi-
cation of informal care, but obviously do not amount to a clear-cut defamilializa-
tion of care strategy. Still, payments for care can differ in relation to the distinction
between formal and informal care – some payments may not be used to pay for
relatives – and indeed this distinction is sometimes also differently drawn between
child and adult care, and between countries (McLaughlin and Glendinning 1993;
Land 2002; Millar and Warman 1996; Ungerson 1997). If care is commodified and
put in the formal arena either through the provision of cash subsidies to purchase
care on the market, or by directly providing public child and elderly care services,
the scope for defamilialization is clearly widened. Yet we shall argue that such a
strategy nevertheless results in incomplete defamilialization, because it is not pos-
sible fully to commodify care (Folbre and Nelson 2000). However, it should be
noted that we are not following those who argue that care should not be com-
modified, because the marketization of care jeopardizes the affective dimension of
care (e.g. Noddings 1984).
Care is more than a task; it involves emotional labour and relationship (Finch
and Groves 1983). Care is also active and passive, involving physical and non-
physical presence (Land 2002). Balbo’s (1987) concept of doppia presenza incorpo-
rates these elements of passive care and non-physical presence, which show up in
the uneasy accounts of mothers who report that they are thinking simultaneously
about their jobs and about whether their children are being looked after properly.
The interaction between the passive and active dimension of care suggests that
the shift from ‘passive’ to ‘active’ that has underpinned the reworking of the
work/welfare relationship is particularly disadvantageous for women who have
caring responsibilities, and makes the valuing and rewarding of care even
more difficult.
Folbre and Nelson (2000) show that there has been a rapid expansion of
commodified care services in the US, and Daly (2001) has also argued that care
services, public and private (market), have been an area of expansion in what has
otherwise been a period of retrenchment in European welfare states. Buying

services enables carers to focus on a smaller number of personal and emotional
caring activities, which cannot be commodified. Nevertheless, family life is
increasingly characterized by what Beck and Beck-Gernsheim (2002: 91) have
called ‘the divergence of tempos and abodes’, by which they mean the increasing
spatial and temporal fragmentation of family life. The demand for the organiza-
tional and personal dimension of care work identified by Balbo therefore shows no
sign of decreasing. In addition, welfare-state restructuring has tended to result in
the greater fragmentation of services and thus to demand more rather than less by
way of informal care.
Thus the commodification of care work in the formal arena neither fully substitutes
for women’s caring work, nor does away with their need to rely on the informal
care provided by relatives and friends. Even if cash payments for childcare were to
be used to pay relatives, adding such a partial commodification in respect of infor-
mal childcare to more formal strategies of commodification would not amount to
an adequate strategy for fully defamilializing care, and is additionally at odds with
the EU policy agenda on ‘active ageing’ (EC 2002a).
Policy-makers seem to have interpreted major social trends in respect of family
and labour-market change as providing evidence of increasing individualization,
which has in turn bolstered their views as to the feasibility as well as the desirabil-
ity of the adult worker model family. Academics have certainly emphasized the
importance of a shift from a ‘community of need’ to ‘elective affinities’ and ‘fam-
ilies of choice’ (Beck and Beck-Gernsheim 2002; Silva and Smart 1999), but they
have also stressed that this does not amount to a process of atomization. Evidence
indicates that obligations and practices of mutual support and care continue to
bind people together (Finch and Mason 1993; Lewis 2001b; Smart and Neale 1999).
Thus policies that assume a trend towards full individualization in the labour mar-
ket and the commodification of care, neglect the complexities of care relationships,
and may actually impede the goal of increasing female employment.
Care is embedded in personal relationships of love and obligation, and in the
process of identity formation. It is only when care is recognized as conflating
labour and love, as a key process in female identity formation embedded in a
gendered normative framework of obligations, and as relationship between
informal carers and between carer and person-cared-for, that it becomes possible
to understand why the commodification of care is too weak a strategy to address
gender inequality in employment and care.
It is also a strategy that poses problems for the issue of choice. Women
experience stronger pressures to care than do men. Finch and Mason (1993) found
that the negotiation of kin responsibilities, including those involving care, con-
tinues to be determined by gender, albeit indirectly. Because women are more
likely not to be in full-time work and to need childcare, they are also more likely
to develop cumulative commitments to care for kin, and in the process of so
doing, develop an identity and a reputation as a carer. This in turn serves to
weaken their negotiating power, because, having established such a reputation,
they are less likely to have ‘legitimate excuses’ not to care.
The adoption of the adult-worker-family model at the policy level has been
based on an instrumental approach to gender equality and care. But care cannot
be fully defamilialized or commodified. Thus we need to make a case that first,
enables the recognition of care work as something that is worthwhile and necessary,
which in turn necessarily involves valuing it. Human beings need care – it is a
human activity – which means that we must also consider the way in which care
work is shared in societies, between men and women at the household level as
well as between the individual and the collectivity. The current policy debates
tend to treat care work as something that has to be reconciled with employment
for women, and which can be addressed via commodification, whether through
cash payments of some kind (through the tax/benefit system or directly to carers)
so that carers may purchase care on the market, or through the provision of
public-sector care services. Such policies are poorly developed compared to the
other major social services (health and education) in most countries. But if it is in
any case not possible to commodify all care work, then the issue of how it is shared,
not just between individuals and the collectivity, but also between men and
women at the household level, must also be addressed.
The possibilities and problems of the capabilities approach
Amartya Sen and Martha Nussbaum’s work on the ‘capabilities approach’ (CA)
(Sen 1985, 1990, 1992, 1999; Nusbaum and Sen 1993; Nussbaum, 1999, 2000,
2003) is promising ground for establishing a basis on which to address the issue of
care, not least because gender equality is of central concern in the work of both
authors. There are important differences in their approaches, the most fundamental
being that Nussbaum argues for a universal list of core capabilities whilst Sen does
not (Robeyns 2003). In addition Nussbaum (2000, 2003) devotes a lot of attention
to care, which hardly appears in Sen’s writings. However, first, it is important to
recognize the way in which the work of both authors has served to shift attention
from inequalities in resources, outcomes and preferences, to inequalities in
capabilities, that is, to the ‘real freedom’ that a person has to be and to do what
she has reason to value, and the significance of this shift for addressing the issue
of care.
This notion of individual freedom is pivotal to the CA, which recognizes that
genuine freedom must be underpinned, while also making important distinctions
between: the resources available to a person (means); what she is and does
(functionings); the personal, social and environmental factors that affect her ability
to transform means into functionings (conversion factors); and the combination of
being and doings that she has the real freedom to achieve (her capability set).
A further crucial distinction is made in the CA between well-being freedom and
agency freedom (Sen 1985). Sen (1985) argues that the outcome measure of well-
being has major limitations because it fails to address the pivotal and distinctive
part that agency plays in a person’s life. He argues that it is important to take
account of agency above and beyond the role that it plays in securing well-being,
because of the importance that is attached to the individual’s ‘real freedom’ to
choose in accordance with her conception of ‘the good’, regardless of whether this
would actually be to her advantage (Sen 1985: 206), for instance, in respect of a
decision to forgo paid employment in order to undertake care work. Advocates of
the capabilities approach would argue that her capabilities set should enable her to
choose to care as well as to choose not to care.
There is another fundamental reason for making a distinction between well-
being freedom and agency freedom, which Sen (1999) stresses in Development as
Freedom, where he argues that the real freedom of individuals is the motor of
development and social change, rather than economic growth per se. The recogni-
tion that Sen accords women’s agency in achieving social change parallels the
increasing interest in the question of individual and collective agency in work on
gendered citizenship and on gender and the development of welfare states more
generally (Lister 1997; Misra and Akins 1998; Siim 2000; Williams et al. 1999).
Gender equality of capabilities: recognizing and valuing care

The key advantage of using the CA to address the problem of gender equality in
relation to paid work and care is that it provides a universal equality model rooted
in the recognition of human diversity. It thus offers a way of recognizing and
valuing care which does not get bogged down in the equality/difference debate.
This is a major advantage, given that ‘care-centred thinking’, as much as neoclas-
sical economic thinking (Folbre 1995), has tended to be biologically essentialist,
stressing the importance of gender difference as much or more than gender equality
(e.g. Noddings 1984).
Using the CA, equality is conceptualized in terms of the equal substantive or real
freedom to choose beings and doings and is applied to all universally. The notion
of capability enables us to account for the significance of human diversity in terms of
diverse preferences, as well as to recognize that they are formed in the context
of unequal conditions, which means that preferences are likely to be adaptive.
It follows that if women are to exercise real freedom to choose, there must be
equal freedom on the part of men and women to choose between alternatives
(e.g. between work and care). Thus agency freedom requires underpinning: real
freedom of choice has to be guaranteed independently of the utilitarian gains that
individuals with different preferences might make. The real freedom to choose
paid work and/or to choose carework requires that a comparable value be attached
to care, independent of different preference gains.
This is important not least because the preference theory commonly used by
neoclassical economists takes no notice of the way in which preferences are
socially embedded. Indeed utilitarian preference models are often invoked to jus-
tify why care work does not need monetary reward, arguing that the motivation
for care rests on the emotional and relational benefits that care work provides for
women, which make the low monetary value that is attached to it immaterial
(Folbre 1995). The tendency to define the work of care as passive as much or more
than active also helps to justify the low value attached to it. Thus the fact that care
work is devalued regardless of whether it is carried out in the formal or informal
sector (Ungerson 2000) is not of ethical concern within preference theory. Yet it
has long been argued by feminist analysts that the value attached to particular
activities is affected by who does them. The activities of the less powerful are
devalued (e.g. Bradley 1989).
The CA also accounts for human diversity in a more fundamental way, for in
order to guarantee equality of real freedom to choose and achieve functionings,
different individuals will require different kinds and amounts of resources (which
is why equality of resources is in and of itself an insufficient condition for the
achievement of chosen functionings). The notion of personal, social and environ-
mental conversion factors accounts for the fact that people have different capacities
to gain access to the same resource, and different potentials for converting
resources into chosen functionings. To choose employment and care an individual
requires either more money to pay for childcare or affordable childcare services, as
well as a right to reduced working time and paid time in order to care. But real
agency freedom also requires there to be a genuine choice to reject alternative
functions. A carer should be as free as a non-carer to choose to work. Clearly this
requires anti-discrimination legislation, the right to return to the same job, and
the offer of retraining. It also requires rights to income based on caring, at a suffi-
ciently high level to secure the effective freedom to choose care. Gender equality
in capabilities for paid work and for care highlights the need for a combination of
a variety of social entitlements, and brings to the fore the issue of time as well as
money. It thus makes manifest the biases and limitations of the current policy
trend that focuses on wages as the best form of welfare, active labour-market
policies, and care services aimed at the defamilialization of care.
Although Sen and Nussbaum both recognize that human beings differ in regard
to the values they attach to different beings and doings, they differ considerably in
respect of the process of defining and selecting functionings to be included in the
capabilities set. In regard to valuing care, Nussbaum’s approach appears at first
sight to be more useful than that of Sen, who leaves the construction of the capa-
bilities set open, and with it, the possibility of excluding care. But, despite
Nussbaum’s concern with gender justice, care is not explicitly included in her list
of the core capabilities necessary for human flourishing (unless it is taken to be
part of ‘capability for affiliation’, which includes ‘showing concern for other
human beings’), which is symptomatic of the problem of drawing up a core list of
capabilities that are universally applicable.
In contrast, in Sen’s approach the importance of human diversity is embraced
further, as it is argued that real agency freedom requires the inclusion and partici-
pation of diverse voices in the definition and selection of capabilities sets. As
Robeyns (2003) notes, Sen stresses that capabilities sets are context-dependent,
and that their legitimacy is dependent on democratic processes of inclusion and
on who is involved in their definition, valuation and selection (see also Bonvin
and Farvaque 2003). Thus, paradoxically perhaps, the fact that Sen leaves the def-
inition of capabilities open, insisting that in order to achieve equal agency free-
dom for men and women it is necessary to include diverse voices in the process of
defining and selecting the functionings that are to be part of the capabilities set,
is, if it can be achieved, a more gender-sensitive and gender-equal path to the
recognition and valuing of care. After all, policies based on universality, equality
and justice have all too often been modelled on the male norm, and have resulted
in the exclusion and subordination of women’s claims (Lister 1997; Okin 1989;
Young 1989).
In addition, by requiring the capability for voice and by arguing that policies
should be designed, implemented and evaluated on the basis of the real freedom
that a person has to be and to do what she has reason to value, the CA exposes the
democratic deficit of the instrumental approach that has informed the shift to the
adult-worker model.
Gender equality of capabilities: the problem of recognising

interdependency and sharing care
Many commentators have focused on the importance the CA attaches to the indi-
vidual’s agency freedom and the extent to which this differs from the neoclassical
economists’ emphasis on individual choice. Nelson (1996) has argued that within
economics Sen’s approach comes closest to appreciating that individual agency is
constituted through social relationships and thus requires material underpin-
nings, while Robeyns (2000, 2001, 2003) argues that the CA rejects ontological
individualism (which conceives of individuals as atomized) in favour of an ethical
individualism (which simply maintains that what matters for the purpose of
evaluation is the individual), and that this is most evident in its acknowledgement
of the importance of social conversion factors. We endorse these arguments and
agree that the freedom and well-being of the individual, rather than that of her
family or community, is particularly important for gender equality. Women’s
poverty, income levels, and control and consumption of resources have long been
hidden behind the categories of household and family (Glendinning and Millar
1987), and a focus on the individual becomes even more pressing in the context of
the shift towards the individualized adult-worker-model family.
However through a care lens we see that the CA approach remains based on a
notion of a separate and self-interested self. We would add that a capability
approach that takes gender equality seriously requires recognition of the self as
autonomous and interdependent, capable of making choices out of concern and
responsibility for others, as well as for one’s self.
In the first place a rejection of ontological individualism implies the recognition
that an individual’s capability to exercise agency freedom actually depends not
only on the freedom to choose to give or not to give care, but also on the receipt
of care. On this crucial point Nussbaum’s recognition of the need for care appears
more promising. However, Nussbaum (2003), like Kittay (1999), approaches the
issue of care from the perspective of the dependant – the person-cared-for – rather
than the carer. However, reconciling the demand for autonomy by both the carer
and the person-cared-for is described as a central dilemma in the literature on care
(e.g. Lewis and Meredith 1987; Morris 1993; Williams 2001).
A person’s real freedom to be and do what she has reason to value is not simply
dependent on social entitlements or procedural justice, or even simply on her nur-
turing by others. It also depends on the needs and actions of others. In other
words, the capabilities sets of individuals are interdependent (Basu and Lopez-
Calva 1999), and such interdependence is particularly acute at the household level
(Iversen 2003). A child’s capabilities are dependent in large measure on her
mother’s choice to care or not to care for her, while her mother’s and her father’s
capability to choose to engage in paid work or political activism are dependent on
each other’s choices to care or not to care for their child. Nussbaum’s conceptual-
ization of care is therefore limited because it does not enable us to recognize fully
that women’s agency freedom is unequally restricted by their caregiving.
In Sen’s (1990, 1999) cooperative–conflict bargaining model the individual
enters into cooperation solely to maximize his or her welfare, which makes it
difficult to acknowledge the kind of relational interdependency that care entails,
which in turn cannot be broken down into individual benefits and costs (Held
1993). As Peter (2003) argues, Sen’s account of women’s low perception of self-
interest ignores that women’s agency is situated in relationships of care, and there-
fore that concern for others needs to be taken seriously as an expression of
autonomy. Agency is unequally restricted along gender lines largely because men
can be concerned about their relationships without being constrained by under-
taking care work.
If we maintain, as both Sen and Nussbaum do, that agency freedom requires
that the real freedom to achieve the same capabilities set must be available to all,
then, given the interdependence that characterizes care relations, it would seem to
us that the key issue is not how to make women more self-interested – which Sen
(1999: 144) argues will come about via their increasing attachment to the labour
market – but rather how to promote conditions that foster responsibility for shar-
ing care between men and women and that enhance women’s agency freedom by
making men more accountable for their responsibility to care for others. But this
is particularly difficult within the CA.
The CA makes agency freedom of the individual pivotal, which means that care
can only be conceptualized as an opportunity, a real individual choice. But the
argument we have put forward about care suggests that it is inherently relational.
This means that the individual’s autonomy and interdependence must be recog-
nized. Ultimately, it is only when all persons are conceived from the start as
autonomous and interdependent – that is as persons who need, give and receive
care (Fraser 1997; Lister 1997; Tronto 1993) – that gender equality in respect of
agency freedom can be embraced. We all need others to care for us in order to
develop agency freedom, and gender justice therefore requires care to be valued
and conceptualized as both a ‘legitimate’ opportunity/choice and as a necessary
central human activity to be shared between men and women.
Choices are made in the context of gendered inequalities in power relations, in
all their economic, political and discursive manifestations, which skew the inter-
dependency of men and women’s individual capabilities sets at the household
level. Thus Phillips (2001) has criticized the CA’s departure from the more tradi-
tional concern with equality of outcomes and resources. It is impossible to ignore
group inequalities of this kind. Inequalities in ‘group functionings’ (for example,
the fact that women persistently do more care work than men) signify gender
inequalities in capabilities (unless of course we assume that the distribution of

preferences between men and women is rooted in sexual difference), and gender
justice requires that note be taken of them. In addition, the issue of gender
inequalities in control of resources needs to be addressed. But, unlike Phillips
(2001) we would argue that the idea of equality of capabilities remains important,
and that once the rationale for addressing men’s freedom not to choose to care is
accepted, the idea of personal and structural conversion factors, which is central to
the CA’s notion of positive freedom, can be stretched to address gender inequali-
ties in control over resources. In other words, a less powerful individual needs
more resources to be able to choose the functionings that she values out of her
capabilities set. While Sen’s CA enables us to make a strong case for the empower-
ing of women’s capabilities for voice, it is also important to acknowledge that
gender inequalities in political presence and power (Lister 1997; Phillips 1993;
Young 1990) undermine women’s capability for voice in the definition and
selection of capabilities, which means in turn that the issue of how to share care is
also less likely to be raised.
A more equal division of care work between men and women in the household
is more likely to increase women’s economic power than attaching a material
value to care work (necessary though this is), because it is unlikely that any
government will ever attach a high enough value to such work (in the informal or
formal sectors) (Bojer 2002; Lewis 1997). Sharing care work between the individual
and the collectivity, and between men and women, also increases the possibility of
women choosing to engage in some form of political participation. Thus in the
Scandinavian countries, where policies to reconcile work and family (mainly, but
not exclusively for women) have developed furthest, women’s participation in
local and central government is also highest (Siim 2000). A more equal gender
division of care work at the household level serves to increase women’s collective
economic, political and discursive power (which in turn makes it more likely that
material value will be attached to care).
Conclusion: policies to promote a gender

equality of capabilities
If there is to be ‘genuine choice’ between work and care for women and men, then
a wide range of policies – in respect of time to care, cash for care, care services and
the regulation of working hours – is necessary. In practice, existing measures have
usually been invoked to enable individuals (mainly women) to choose to add work
to care (the supported-adult-worker model of the Scandinavian countries and dis-
cernible in some, but not all, EU-level policy documents) and they rely primarily
on sharing care between (women in) households on the one hand, and some form
of paid provision on the other. However, we have argued that care must also be
shared between men and women at the household level, and that it is necessary to
recognize the importance of interdependence and unequal power, as well as the
individual’s agency freedom in order to achieve this. Is there a place for compul-
sion in this respect? When she considers the issue of sexual harassment in the
workplace, Nussbaum is prepared to advocate the curtailment of male freedoms,

something, as Bojer (2002) notes, she does not do in regard to the gender division
of labour. However, compulsion, sanctions and penalties run the risk of causing a
backlash and in any case threaten the moral qualities of attentiveness, respon-
sibility, competence and responsiveness that characterize care (Tronto 1993).
With this in mind, positive incentives for men as individuals to engage in care,
such as the ‘daddy leave’ offered by the Scandinavian countries, alongside the
reform of the male-employment model directed towards male workers as a group,
for example in the form of reduced working time (a shorter working week and the
regulation of overtime) appear more appropriate. Such interventions would
amount to a radical transformation of the male-employment model, which has
continued to dominate the shift toward the adult-worker model.
Thus, the logical end of the argument presented here is a universal
carer/worker–worker/carer model (see also Fraser 1997; Lister 1997, 2003), which
requires a commitment to provide time to care and affordable, accessible, high-
quality services, as well as cash for care. In brief, the policy grid to enable genuine
choice would require action on the following:
● Time: working time and time to care

● Money: cash to buy care, cash for carers
● Services for (child and elder) care
and in relation to both:
● The household level (in terms of the provision of cash and time for men and
women)
● The collective level (in terms of services and cash, and workplace-based changes
especially in respect of measuring the work accomplished, rather than the
‘face-hours’ devoted to it).
The task of devising social policies that promote real or genuine choice for both
men and women in respect of paid and unpaid work poses huge difficulties. Both
employment and care are necessary for human flourishing. Care is not just ‘a good
thing’, it is crucial to the welfare of the person-cared-for; employment is not just
‘a good thing’, it is crucial to the welfare of the adult (and likely to become more
so as governments assume the capacity for self-provision will increase). From the
point of view of human welfare, it is impossible to choose not to care or not to
work. In this sense, to choose between them has the hallmarks of tragedy as much
or more than opportunity, which makes it all the more important to address
seriously the policies that play a part in structuring people’s choices.
Acknowledgement
The authors gratefully acknowledge financial support from the EUROCAP project, funded by
the EC’s FP5 Programme and coordinated by Professor Robert Salais, and Susanna Giullari
also acknowledges the support provided by a British Academy Postdoctoral Fellowship.

They also acknowledge the stimulation provided by the invitation given by UNRISD to consider
these issues further. This chapter draws on two different versions of the article: ‘The Adult
Worker Model Family, Gender Equality and Care: the Search for New Policy Principles and
the possibilities and Problems of a Capabilities Approach’, published as an UNRISD Working
Paper (2005), and in Economy and Society 34 (1) 2005: 76–104.
Note
1 The reasons why the father of the child may not hesitate, at least to the same extent, are
much debated between those who assume biological essentialism, which underpins
much of the classical literature, e.g. Parsons and Bales (1995), Becker (1981), also Gilligan
(1982), and those who insist on the power of socialization which underpins most of the
feminist literature, e.g. Oakley (1981). New variants of the arguments have come to the
fore in the face of observations that some lone mothers put the obligation to care before
paid employment for reasons that include structural variables (Duncan and Edwards
1999), while Hakim (2000) argues that lifestyle choice alone is the main explanation.
Bibliography
Balbo, L. (1987) ‘Crazy Quilts: Rethinking the Welfare State Debate from a Woman’s Point of
View’. In A. Showstack Sasson (ed.) Women and the State. London: Hutchinson.
Barlow, A. and S. Duncan (2000) ‘New Labour’s Communitarianism: Supporting Families and
the “Rationality Mistake”, Part II’, Journal of Social Welfare and Family Law 22(2): 129–43.
Basu, K. and L. F. Lopez-Calva (1999) ‘Functionings and Capabilities’. In K. Arrow, A. Sen and
K. Suzumura (eds) Handbook of Social Choice and Welfare. Amsterdam: Elsevier
Science/North Holland.
Beck, U. and E. Beck-Gernsheim (2002) Individualization. London: Sage.
Becker, G. (1981) A Treatise on the Family. Cambridge MA: Harvard University Press.
Bojer, H. (2002) The Social Contract: Unpaid Child Care And Women’s Economic Capability. Paper
presented at the VHI conference ‘Promoting Women’s Capabilities: Examining
Nussabum’s Capabilities Approach’, Cambridge, 9–10 September.
Bonvin, J. and N. Farvaque (2003) Towards a Capability-Friendly Social Policy: The Role of Local
Implementation Agencies. Paper prepared for the VHI conference ‘Transforming Unjust
Structures: Capabilities and Justice’, Cambridge, 26–27 June.
Bradley, H. (1989) Men’s Work Women’s Work. Cambridge: Polity Press.
Bradshaw, J., S. Kennedy, M. Kilkey, S. Hutton, A. Corden, T. Eardly, H. Holmes and J. Neale
(1996) The Employment of Lone Parents. London: Family Policy Studies Centre.
Commission of the European Communities (CEC) (1993) Growth, Competitiveness and
Employment – The Challenges and Ways Forward into the 21st Century. Luxembourg: CEC.
Crompton, R. (ed.) (1999) Restructuring Gender Relationships and Employment: The Decline of
the Male Breadwinner. Oxford: Oxford University Press.
Daly, M. (2001) ‘Alternative Ways of Compensating for Care in Western Europe’. In M. Daly
(ed.), Care Work Security. Geneva: ILO.
Duncan, S. and R. Edwards (1999) Lone Mothers, Paid Work and Gendered Moral Rationalities.
Basingstoke: Macmillan Press.
Esping-Andersen, G., D. Gallie, A. Hemerijck and J. Myles (2001) A New Welfare Architecture for
Europe? Report submitted to the Belgian Presidency of the European Union, Final Version.
Esping-Andersen, G., D. Gallie, A. Hemerijck and J. Myles (2002) Why We Need a New Welfare
State. Oxford: Oxford University Press.
European Commission (EC) (1996) ‘Council Directive 96/34/EC of 3 June 1996 on the
framework agreement on parental leave concluded by UNICE, CEEP and the ETUC’.
Official Journal of the European Communities, L 145.
European Commission (EC) (2000) Social Policy Agenda. COM (2000) 379 final, Brussels.
European Commission (EC) (2002a) Increasing Labour Force Participation and Promoting Active
Ageing. COM (2002) 9 final, Brussels.
European Commission (EC) (2002b) Taking Stock of Five Years of the European Employment
Strategy. COM (2002) 416 final, Brussels.
European Commission (EC) (2003) Scoreboard on Implementing the Social Policy Agenda. COM
(2003) 57 final, Brussels.
European Commission and Council of the European Union (2002) Joint Report on Employment,
at http://www.europa.eu.int/comm/employment_social/employment_strategy/report_2002/
jer2002_final_en.pdf, accessed on 9 June 2005.
Ferrera, M., A. Hemerijck and M. Rhodes (2000) The Future of Social Europe: Recasting Work and
Welfare in the New Economy. Report for the Portuguese presidency of the European Union.
Finch, J. and D. Groves (eds) (1983) A Labour of Love: Women, Work and Caring. London:
Routledge & Kegan Paul.
Finch, J. and J. Mason (1993) Negotiating Family Responsibilities. London: Routledge.
Folbre, N. (1995) ‘ “Holding Hands at Midnight”: The Paradox of Caring Labour’. Feminist
Economics, 1(1): 73–92.
Folbre, N. and J. Nelson (2000) ‘For Love or Money – Or Both?’ Journal of Economic Perspectives
14(4): 123–40.
Fraser, N. (1997) Justice Interruptus: Critical Reflections on the ‘Post-Socialist’ Condition. London:
Routledge.
Gershuny, J. (2000) Changing Times: Work and Leisure in Post-industrial Society. Oxford: Oxford
University Press.
Gilligan, C. (1982) In a Different Voice. Cambridge MA: Harvard University Press.
Glendinning, C. and J. Millar (1987) Women and Poverty. Brighton: Wheatsheaf.
Goodin, R. E. (2001) ‘Work and Welfare: Towards a Post-productivist Welfare Regime’. British
Journal of Political Science, 31(1): 13–39.
Gornick, J. C., M. K. Meyers and K. E. Ross (1997) ‘Supporting the Employment of Mothers:
Policy Variation across Fourteen Welfare States’. Journal of European Social Policy, 7(1):
45–70.
Hakim, C. (2000) Work-Lifestyle Choices in the Twenty-first Century: Preference Theory. Oxford:
Hantrais, L. (ed.) (2000) Gendered Policies in Europe: Reconciling Employment and Family Life.
Held, V. (1993) Feminist Morality: Transforming Culture, Society and Politics. Chicago:
University of Chicago Press.
Iversen, V. (2003) ‘Intra-Household Inequality: A Challenge for the Capability Approach?’
Feminist Economics, 9(2–3): 93–115.
Jenson J. and M. Sineau (eds) (2001) Women’s Work, Childcare, and Welfare State Redesign.
Toronto: University of Toronto Press.
Kaufman, F-X. (2002) ‘Politics and Policies Towards the Family in Europe: A Framework and
an Inquiry into their Differences and Convergences’. In F-X Kaufman, A. Kuijsten,
H-J Schulze, K. P. Strohmeier (eds) Family Life and Family Policies in Europe, Vol. 2. Oxford:
Kittay, E. (1999) Love’s Labour: Essays on Women, Equality and Dependency. New York:
Routledge.
Kleinman, M. (2002) A European Welfare State? Basingstoke: Palgrave Macmillan.
Land, H. (2002) ‘Spheres of Care in the UK: Separate and Equal’. Critical Social Policy, 22(1):
13–32.
Lewis, J. (1997) ‘Gender and Welfare Regimes: Further Thoughts’. Social Politics, 4: 160–77.
Lewis, J. (2001a) ‘The Decline of the Male Breadwinner Model: The Implications for Work
and Care’. Social Politics, 8(2): 152–70.
Lewis, J. (2001b) The End of Marriage? Individualism and Intimate Relations. Cheltenham:
Edward Elgar.
Lewis, J. and B. Meredith (1987) Daughters who Care. London: Routledge.

Lister, R. (1997) Citizenship: Feminist Perspectives. Basingstoke: Macmillan Press.
Lister, R. (2003) Citizenship: Feminist Perspectives. 2nd edition, Basingstoke: Palgrave Macmillan.
Lödemel, I. and H. Trickey (eds) (2001) An Offer You Can’t Refuse: Workfare in International
Perspective. Bristol: Policy Press.
McLaughlin, E. and C. Glendinning (1993) Paying for Care – Lessons from Europe. London: HMSO.
Mead, L. (1986) Beyond Entitlement: The Social Obligations of Citizenship. New York: Free Press.
Millar, J. and A. Warman (1996) Family Obligations in Europe. London: Family Policy Studies
Centre.
Misra, J. and F. Akins (1998) ‘The Welfare State and Women: Structure, Agency, and
Diversity’. Social Politics 5: 259–85.
Morris, J. (1993) Independent Lives. Basingstoke: Macmillan Press.
Nelson, J. A. (1996) Feminism, Objectivity and Economics. London: Routledge.
Noddings, N. (1984) Caring: A Feminine Approach to Ethics and Moral Education. Berkeley:
University of California Press.
Nussbaum, M. (1999) Sex and Social Justice. Oxford: Oxford University Press.
Nussbaum, M. (2000) Women and Human Development: The Capabilities Approach. Cambridge:
Cambridge University Press.
Nussbaum, M. (2003) ‘Capabilities as Fundamental Entitlements: Sen and Social Justice’.
Feminist Economics 9(2–3): 33–59.
Nussbaum, M. and A. Sen (eds) (1993) The Quality of Life. Oxford: Oxford University Press.
Oakley, A. (1981) Subject Women. Oxford: Martin Robertson.
O’Connor, J. (2003) ‘Measuring Progress in the European Social Model: Policy Coordination,
Social Indicators and the Social Policy Agenda in the European Union’. Paper presented at
the RC19 Conference, Toronto, August.
Okin, S. M. (1989) Justice, Gender and the Family. New York: Basic Books.
Orloff, A. (2002) ‘Explaining US Welfare Reform: Power, Gender, Race and the US Policy
Legacy’. Critical Social Policy, 22(1): 96–118.
Parsons, T. and R. F. Bales (1995) Family Socialization and Interaction Process. Glencoe, IL: Free Press
(first published 1955).
Peter, F. (2003) ‘Gender and the Foundations of Social Choice: The Role of Situated Agency’.
Feminist Economics, 9(2–3) 13–32.
Phillips, A. (1993) Democracy and Difference. Cambridge: Polity Press.
Phillips, A. (2001) ‘Feminism and Liberalism Revisited: Has Martha Nussbaum Got It Right?’.
Constellations, 8(2): 249–66.
Robeyns, I. (2000) ‘An Unworkable Idea or a Promising Alternative? Sen’s Capability
Approach Re-examined’. Paper presented at the Workshop on Poverty, Social Capital and
Development, St Edmunds College, 23 May.
Robeyns, I. (2001) ‘Sen’s Capability Approach and Feminist Concerns’. Paper prepared for
the conference on Sen’s Capability Approach, St Edmunds College, 5–7 June.
Robeyns, I. (2003) ‘Sen’s Capability Approach and Gender Inequality: Selecting Relevant
Capabilities’. Feminist Economics, 9(2–3): 61–92.
Ross, G. (2001) ‘Europe: An Actor without a Role’. In J. Jenson and M. Sineau (eds) Women’s
Work, Childcare, and Welfare State Redesign. Toronto: University of Toronto Press.
Rubery, J., M. Smith and C. Fagan (1998) ‘National Working-Time Regimes and Equal
Opportunities’. Feminist Economics, 4(1): 71–101.
Scharpf, F. W. (2002) ‘The European Social Model: Coping with the Challenges of Diversity’.
Journal of Common Market Studies, 40(4): 645–70.
Sen, A. (1985) ‘Well-Being, Agency and Freedom: The Dewey Lectures 1984’, Journal of
Philosophy, 82(4): 169–224.
Sen, A. (1990) ‘Gender and Cooperative Conflicts’. In I. Tinker (ed.) Persistent Inequalities:
Women and World Development. New York: Oxford University Press.
Sen, A. (1992) Inequality Re-examined. Oxford: Oxford University Press.
Sen, A. (1999) Development as Freedom. Oxford: Oxford University Press.
Siim, B. (2000) Gender and Citizenship: Politics and Agency in France, Britain and Denmark.
Cambridge: Cambridge University Press.
Silva, E. B. and C. Smart (1999) The New Family? London: Sage.
Smart C. and B. Neale (1999) Family Fragments? Cambridge: Polity Press.
Supiot, A. (ed.) (1999) Au-dela de L’Emploi (Beyond Employment). Paris: Flammarion.
Tronto, J. (1993) Moral Boundaries: A Political Argument for and Ethic of Care. London:
Routledge.
Ungerson, C. (1997) ‘Social Politics and the Commodification of Care’. Social Politics, 4: 362–81.
Ungerson, C. (2000) ‘Thinking About the Production and Consumption of Long-Term Care
in Britain: Does Gender Still Matter?’ Journal of Social Policy, 29(4): 623–43.
Vandenbroucke, F. (2001) ‘European Social Democracy and the Third Way: Convergence,
Divisions, and Shared Questions’. In S. White (ed.) New Labour: The Progressive Future?
Williams, F. (2001) ‘In and Beyond New Labour: Towards a New Political Ethics of Care’.
Critical Social Policy, 21(4): 467–93.
Williams, F., J. Popay and A. Oakley (1999) Welfare Research: A Critical Review. London: UCL
Press.
Young, I. M. (1989) ‘Polity and Group Difference: A Critique of the Ideal of Universal
Citizenship’. Ethics, 99: 250–74.
Young, I. M. (1990) Justice and the Politics of Difference. Princeton NJ: Princeton University Press.
Part II
Labour Market Informality and
the Search for Social Security
9
Labour Market Informalization,
Gender and Social Protection:
Reflections on Poor Urban
Households in Bolivia and Ecuador
Lourdes Benería and Maria S. Floro
Introduction
This chapter is an attempt to understand the ways in which poor urban

households in developing countries cope with the nature and depth of current
labour market informalization, poverty and economic insecurity. It explores the
dynamics of informality and its distributional aspects, and it demonstrates how
job precariousness and vulnerability are interconnected as well as the importance
of their gender dimensions. The paper is based on a study of poor urban house-
holds in Bolivia and Ecuador, using sample surveys carried out in 2002.1 We focus
mainly on three areas of inquiry: the evolving informal economy and its implica-
tions in the way households organize their lives; gender roles and gender relations;
and the precariousness of work and lack of social protection.
Conceptual framework
The evolving informal economy

During the 1970s and early 1980s, studies on the informal sector mostly assumed
that, with economic growth, it would tend to be absorbed by the modern
economy. The tremendous growth of informality since then has reversed these
expectations, pointing to the effects of neoliberal policies, and to the strong ten-
dencies in the global economy to generate precarious jobs below subsistence
wages. According to ILO estimates, informal employment represents one half to
three-quarters of non-agricultural employment in developing countries (ILO
2002). Regional and country differences are significant but the growing impor-
tance of informality has been registered in high-income countries as well; self-
employment, part-time work and temporary work represent 30 per cent of overall
employment in 15 European countries and 25 per cent in the United States.
However, for developing countries, informality has clearly become the rule rather
than the exception.
193
194 Lourdes Benería and Maria S. Floro
Since its inception, the notion of ‘informal sector’ was subject to debates and
criticisms, particularly in terms of the heterogeneity and artificiality of the
formal/informal division. The initial 1970s formulations by Keith Hart and by
the ILO mission to Kenya in 1972 implied that modernization and economic
growth would facilitate the absorption of the ‘informal sector’ by the modern sec-
tor. By contrast, during the 1980s the literature emphasized the links rather than
the divisions between the two sectors, pointing to the dependency of the modern
on the informal sector as a way of lowering costs, thus implying that there was no
incentive for the former to absorb informal activities (Benería and Roldán 1987;
Portes and Castells 1989). The 1990s saw a deepening of these processes and an
expansion of informality, thus increasing the need to understand further the
factors that have reversed the prediction of the initial formulation. Hence the
continuing debates on the topic and their increasing relevance to understand
the persistence of poverty and the increasing vulnerability in the midst of an
expanding global economy (De Soto 2000; ILO 2002).
This raises questions regarding the differences between earlier and current stages
of informalization. Have conditions changed and, if so in what way? Several obser-
vations can be made in this respect.2 First, since the 1980s, the macroeconomic
context has changed significantly due to the implementation of neoliberal policies
in many countries, including Ecuador and Bolivia. Privatization, market deregula-
tion and economic restructuring have contributed to set up the conditions for an
expansion of informal activities and precarious working conditions. We provide
specific examples related to these countries in subsequent sections.
Second, the expansion and deepening of markets has extended the links, direct
or indirect, between formal and informal activities while the distinction between
the two has become increasingly blurred. Large firms have developed increasing
links with informalized production through outsourcing and subcontracting. This
trend has been examined by several studies in the case of Thailand as well as for
the Latin American region (Boonmathya et al. 1999; Balakrishnan and Huang
2000; Pérez-Sainz 1994, 2005; Buechler 2002). Nevertheless, as argued below,
many other types of informal work remain disconnected from more formal
production particularly those related to survival activities.
Third, informal activities are no longer viewed as the anomaly that eventually will
disappear with modern economic growth. On the contrary, the trend has been in
the opposite direction; the decline in formal employment, particularly in the public
sector, and the dynamism of the informal economy in many cases has enhanced the
relative attractiveness and the predominance of the latter. Likewise, social exclusion
has become more prevalent among unskilled labour and rural-to-urban migrants.
Fourth, although the traditional association of informal work with low skills and
low productivity still holds, the last two decades have introduced many changes in
this respect. With the increase in outsourcing and subcontracting, several
segments of informal activities have their centre of gravity in the core firms that
generate them. This implies that productivity levels in these segments tend to be
high and competitive, particularly because technology transfers and skills training
take place through these linkages.
Labour Market Informalization, Gender and Social Protection 195
Fifth, increasing informality and the deterioration of working conditions in

developing countries during the past two decades have taken place in a climate that
has emphasized citizenship, political rights, individual agency and democracy.
Hence the contradictions and social tensions between these discourses emphasizing
empowerment and equality on the one hand, and the realities of precariousness,
poverty and powerlessness associated with informality on the other. Globalization
has intensified these tensions by enlarging their scope and sphere of reference,
and by contributing to the factors that generate informality.
As a result, non-standard forms of employment have multiplied, thereby intro-
ducing a greater diversity in working-life patterns for both men and women. There
is wide variation in the range of activities, skill and capital requirements, and in
the working conditions under which informality operates. This paper makes a
contribution by illustrating the heterogeneity of the informal economy and its
significance for the lives of poor urban households.
It is a well-known fact that the increase in the size of the informal economy has
been parallel to that of women’s economic participation in paid work. For a variety
of reasons, women are disproportionately represented in informal employment,
which, according to ILO estimates, accounts for 60 per cent or more of female
employment in developing countries. In some labour categories such as home-
based work, their participation ranges between 30 and 80 per cent of all work-
ers, while 80 per cent of industrial homework is carried out by female labour
(ILO 2002).
Gender roles and gender relations

A close look at households involved in informal activities allows us to explore with
great empirical richness the linkages between paid and unpaid work, since many
informal activities take place within the space and temporal dimensions of care
and reproductive work. This is particularly true of home-based work. Its flexibility
allows the worker to move between paid and unpaid activities with as much
frequency as necessary, thus facilitating the increase in women’s participation in
the latter. It also facilitates the reduction of production costs for informal workers
by using home space, utilities and labour of other family members, including chil-
dren. Paid and unpaid work are often performed simultaneously (such as cooking
and child-caring while working on a garment). As will be shown below, our study
shows the existence of these overlapping activities which intensify especially
women’s work. The consequence tends to be increased work intensity, stress,
incentives for including children as unpaid family labour, and isolation from other
workers.
In a highly informalized economy, the fluidity of labour that shifts between
formal and informal activities and also between paid and unpaid work is captured
by our analytical model. It shows how individual workers (and their households)
are fully responsible for their own survival and for managing risks as economic
conditions change. Figure 9.1 depicts this fluidity and overlap; Figure 9.1b shows
how, during economic crises, the informal and household economy expand to
pick up the slack of the formal economy. Firms hire and lay off workers in order to
Figure 9.1a Areas of economic activity and labour use during economic growth
Combined informal Interlinked formal–informal activities

market and domestic
activities Informal
Economy
Formal
Economy
Household
Economy
Combined formal market
and household activities
Figure 9.1b Areas of economic activity and labour use during severe economic crisis
adjust labour input without much resistance from workers, thus providing
maximum flexibility for capital. At the same time, many governments, including
those of Bolivia and Ecuador, have adopted labour laws that tolerate and even
promote labour flexibility without much concern about safety nets and unem-
ployment compensation schemes during periods of high unemployment and
underemployment.
A different question is the extent to which women’s increasing participation in
paid work is transforming gender relations. At the general level, the transforma-
tion of gender roles during the past decades has been profound, affecting many
aspects of women’s and men’s lives. Broadly speaking, and for the purpose of this
paper, these transformations fall into two categories.
One has to do with the division of labour and the location of women and men
in the world of work, such as the increasing participation of women in the paid
labour force. To be sure, gender remains a major axis of differentiation in terms of
work and labour market adjustments. Women workers tend to earn lower wages
than their male counterparts as a result of discrimination, gender norms and other
factors that permeate economic and social institutions. These factors and norms
vary across countries and cultures, thus influencing the ways in which gender
inequality takes concrete forms (Kabeer 2003; Benería 2003; Salzinger 2003). In
Latin America, and despite the fluctuating economic performance of several
economies since the 1980s, there have been remarkable changes for women in the
labour market (Duryea et al. 2004). At the same time, although women tend to be
placed at the bottom of labour hierarchies, we have also observed increasing
differentiation among women in terms of work and income (McCrate 1995;
Lavinas 1996; Piras 2004). Thus, while educated women have made inroads into
the professional world that used to be predominantly male, the large majority of
women are confined within the more precarious jobs and informalized
working conditions; they continue to be constrained by their low mobility, few
alternatives and by social norms that reduce their bargaining power over their
earnings and working conditions.
The second category of gender transformations operates at a deeper level of
gender constructions and has to do with the way men and women, as well as society
in general, see themselves and each other with regard to gender norms and gender
identities. As women enter the labour force in increasing numbers, an understand-
ing of the relationships between income and capabilities becomes key to evaluate
these changes (Sen 1992; Nussbaum 2003). Whether or not a worker is able to trans-
late his/her earnings into capability-functioning depends upon a variety of factors,
such as the nature of household relations and decision-making processes in addition
to the terms of employment. As our study illustrates, we can observe both change
and continuity in the ways gender roles and gender differences manifest themselves.
Changes in gender constructions at the deeper level of the individual and society,
including the most intangible ways in which gender inequality operates, are
neither straightforward nor easy to measure even though they may be observable.
Yet, they are crucial for our understanding of the march towards gender equality.
Here an interdisciplinary analysis is important if we are to comprehend the
multiple levels through which they manifest themselves. To illustrate this with one
specific example from Latin America, Gutmann’s anthropological study (1996) of
gender transformations in a working-class community of Mexico City provides
very useful insights into how these changes can take place. His analysis of what he
calls ‘degendering’ – or processes through which gender stereotypes weaken –
shows how gender constructions are subject to change, even in the face of appear-
ances to the contrary. In this sense, and as Gutmann argues, ‘the gendered
character of social life is never transparent’ (p. 12). Thus, how the evolving
division of labour and the world of work translate themselves into changes at
the deeper level of identity construction may not be obvious or it may appear in
contradictory ways.
The precariousness of work and lack of social protection

According to recent ECLAC estimates (2004), over 44 per cent of the Latin
American population lived under conditions of poverty in 2003, over 19 per cent
of them in extreme poverty. Although there are differences between countries, the
extent of the problem is compounded by the weakness of social protection
programmes throughout the region. For the purpose of this chapter, we make a
basic distinction between poverty and vulnerability to emphasize the notion that
insecurity is not only about access to income, as defined by poverty lines.
Vulnerability has to do with the lack of protection against risk in which many
households find themselves, such as the lack of minimal access to health services
and health insurance, protection for old age, precariousness in housing condi-
tions, unsanitary neighbourhoods and a polluted environment affecting daily life.
Our study provides an illustration of how poor urban households are exposed to a
high level of vulnerability even in cases when their incomes are not below the
poverty line.
In Latin America, the deepening and extension of marketization during the past
two decades has diminished the role of the state in social protection (Molyneux,
chapter 2 in this volume). Protection has increasingly been left either to private
insurance schemes or to the individual, households and, to some extent, the com-
munity. Neoliberal thinking has de-emphasized not only socialized forms of
protection but also the distributive aspects of macroeconomic policies. As the
effects of structural adjustment programmes have accumulated throughout the
region, and as distribution, left to market forces, has led to increasing inequalities
in most countries, social protection has lagged behind, diminished or been absent
for a large proportion of the population. In Latin America, households and fami-
lies have had to pick up the slack, often playing heroic efforts of daily survival
(González de la Rocha 2000). ‘Abandoned by the state’, in the words of one
Bolivian woman, we have observed a process of privatization of family survival
and risks since the 1980s.
To be sure, poverty-alleviation programmes have spread widely, particularly
since the 1990s and as a result of the agenda set by international organizations and
development agencies. This agenda was fostered by the realization that the social
costs of the ‘lost decade’ of the 1980s were higher than initially expected and
unequally distributed. At the same time, the gains resulting from neoliberal ‘devel-
opment’ have not alleviated the magnitude of poverty and social inequality. As in
many developing countries, privatization schemes and a weak fiscal sector have
resulted, at most, in fragmented forms of social protection – the result of sporadic
efforts to cope with the most urgent problems rather than any systematic way of
dealing with universal provisioning.
For women workers, the fluidity between informal and formal activities and
between paid and unpaid work is linked to their concentration in domestic
responsibilities and tied to gender norms. Their precarious connection with the
formal labour market diminishes their access to social protection schemes. This is
typically illustrated by the case of homeworking. For this reason, the pioneering
work of SEWA in designing schemes of social protection for homeworkers sets an
example that has inspired further work (Lund, chapter 10 in this volume).
Ultimately, however, the discussion and design of social protection schemes closer
to the notion of universal provisioning remains a challenge that cannot go away.
We will return to this topic in the last section.
The growth of the informal economy in Ecuador and Bolivia
The socioeconomic developments in Ecuador and Bolivia since the 1980s display
both similarities and differences, and the expansion of the informal economy
during the period has been a common feature. In Ecuador, this expansion was
facilitated by the pattern of economic growth that resulted in the formal market
economy’s incapacity to generate adequate jobs. Industrialization during the
1970s was capital-intensive and generated relatively little employment growth. By
the early 1980s, serious balance-of-payments problems ushered in an IMF-led
stabilization and structural adjustment programme (Floro 1992).3
The deepening of trade and financial liberalization policies during the 1990s
brought about a pattern of uneven development characterized by greater concen-
tration of wealth and further deterioration of working conditions. The rapid inte-
gration of Ecuador’s economy into the global market led to the increased use of a
flexible labour force by the productive sectors. Additional measures were thus incor-
porated in the Ecuador labour code in the early 1990s such as the use of temporary,
part-time, seasonal and hourly contracts and the replacement of indefinite labour
contracts with fixed-term contracts. At the same time, the reforms increased
restrictions on the right to strike, collective bargaining and the organization of
workers (CELA-PUCE 2002; Weinberg 2002).
Ecuador’s informal economy further grew in the context of persistently high
income inequality and poverty during the 1990s and a severe macroeconomic and
financial crisis in the later part of the decade (CELA-PUCE 2002). The Gini index
of income inequality increased from 52 to 54 between 1995 and 1999 (Parendekar
et al. 2002). By 1999, the extreme poor accounted for 21 per cent of the total pop-
ulation (12.4 million) and the poor accounted for 52 per cent of the population
(UNDP 2003).4 The Open unemployment rate increased during the period
1989–99, with women’s and men’s reaching 14.4 per cent and 19.6 per cent (UNDP
2002; ILO 2002). At the same time, underemployment remained at about 60 per cent
of the economically active population (CELA-PUCE 2002).
The Ecuadorian government response to the economic crisis further sustained
the deterioration of macroeconomic conditions. Its decision to adopt the US
dollar as its currency (called dollarization) resulted in high inflation and further
reduced the purchasing power of workers’ incomes (Solimano 2002). By 2001,
after dollarization, the real minimum wage had been reduced to 20 per cent of its
1980 value (CELA-PUCE 2002). Moreover, Ecuador’s fiscal austerity measures and
regressive taxation system have led to inadequate, ill-funded and poorly designed
social protection programs. Social spending decreased by more than a third, as
public spending on social services and social programme fell from US$78 to US$51
per person between 1995 and 2000. Ecuador’s level of social health services as a
result were reduced. The cut in social-services expenditures adversely affected the
Bono Solidario, a cash-transfer programme targeted at poor households, the school
meals programme and integrated early childhood care, and prenatal and neonatal
programmes so that their scope, type and degree of protection became even more
limited (Parandekar et al. 2002). Only a small proportion of the population, mainly
regular workers in medium and large firms as well as the public sector had access to
the social security system. Even then, the beneficiaries suffered huge losses as the
purchasing power of their monthly benefits declined drastically.
In the case of Bolivia, the dramatic expansion of the informal sector has resulted
from deep economic restructuring and the neoliberal reforms which began in the
early 1980s and were intensified with the structural adjustment programme (SAP)
adopted in 1985. Throughout the 1970s, rising foreign debt and the fall of export
commodity prices generated an economic crisis that peaked during the first half of
the 1980s. The crisis manifested itself in negative growth rates, hyperinflation, and
the declining value of the Bolivian peso. By 1985, the inflation rate had reached
levels way above 1,000 per cent,5 with debt service representing 45–50 per cent of
GDP (Morales Anaya 1987). This situation led to the adoption of the IMF-
sponsored SAP in 1985, which resulted in the ‘new economic policy’ (NEP).
The 1985 policies focused on structural reforms that redefined the role of the
government, turned the economy in a free-market direction through liberalization
of both the goods and labour markets, and promoted opening of the economy to
the rest of the world. Typically, it imposed wage freezes, reduction of government
spending, and a shift towards an export-oriented model of development. At the
same time, the 1980s saw the closing of the tin mines that had been a key export
sector and had provided secure employment and a way of life for an important
proportion of the working population. This led not only to social upheaval but
also to a rapid relocation of the mining population in 1986 (McFarren 1989).
The marked tendency towards increasing labour market informalization in
Bolivia was intensified with the 1994 Law of Popular Participation which included
several components emphasizing administrative and fiscal decentralization,
privatization, educational reform, and the restructuring of the country’s social
insurance system, among others (Kohl 1999; Benería-Surkin 2003). The economy
went through a period of recovery during the 1990s and economic growth was pri-
marily based on the growth of capital-intensive sectors, such as electricity, trans-
portation and financial services rather than that of the more labour-intensive
sectors such as manufacturing and agriculture (Republic of Bolivia 2001).
This implied that growth was achieved with insufficient job creation, resulting
in the persistence of unemployment and underemployment. Other chronic
problems also reappeared: the economy has been stagnant since 1999, with an
annual growth rate of 1.6 while the urban population has increased at 3.7 per cent
annually (UDAPE 2003). The current conditions of high unemployment rates,
government deficits, and foreign debt pressures, as well as critical levels of poverty
and significant disparities in regional economic growth have resulted in political
and social tensions. Under these circumstances, the informal economy has
continued to absorb a large part of the growing workforce.
As in the case of Ecuador, the process of adjustment in Bolivia has resulted in a

higher level of integration with the global economy but under unfavourable
conditions. Exports depend on a few primary commodities and the economy has
not diversified with new products able to compete in global markets. Bolivia con-
tinues to be the poorest country in Latin America, with GDP per capita of $936 in
2001 ($2,300 on a PPP basis). Income-inequality figures are high, with a Gini coef-
ficient above 0.60. Throughout the 1990s, 20 per cent of the poorest population
accounted for 4 per cent of aggregate urban income, while the 20 per cent of the
highest income levels accounted for 54 per cent (Republic of Bolivia 2001). The
highest incidence of poverty seems to be related to four factors: indigenous popu-
lations, the unemployed, low educational levels and recent rural–urban migration.
During the 1990s, the absolute number of households below the poverty line
increased even though there was a relative decrease from 53.3 to 47.8 per cent
between 1990 and 1995 (Republic of Bolivia 2001; UDAPE 2003). At present, a
large majority of the population is engaged in informal activities, with estimates
reaching a level above 67 per cent.
One key issue in the case of Bolivia – very relevant for our study – is that the
impact of growth on poverty is low because growth has concentrated on high-
productivity sectors which employed less than 10 per cent of the population. The
persistent lack of detailed statistical information about the informal economy
remains a problem for understanding the key factors affecting the dynamics of
productivity and distribution, particularly at the level of individuals and house-
holds. Our study is an attempt to fill this gap by using survey data at that level.
In the following sections, we examine empirically the dynamics of informality
that characterize jobs in the urban poor neighbourhoods of Bolivia’s and Ecuador’s
largest cities. We use the information collected from a 2002 multi-country study of
528 urban, low-income households in these two countries to investigate the varied
dimensions of job informality and to highlight the interconnectedness of job pre-
cariousness, vulnerability and gender. Our analysis makes use of the urban poor
household sample (UPHWS) survey data and corresponding interview insights
collected in 2002. Several low-income communities in the cities of La Paz, El Alto
and Santa Cruz in Bolivia, and Quito and Guayaquil in Ecuador, were selected to
provide representation of the diversity of urban poor neighbourhoods in these
major urban areas.6 They were randomly selected from a purposive community list
of households with at least one member (household head or spouse) engaged in
home-based or other self-employed, informal sector work.
Understanding the nature of informality: evidence

from Bolivia and Ecuador
We first examine the nature of informality that characterizes jobs in the urban
poor neighbourhoods of Bolivia and Ecuador’s largest cities. To highlight the
gender dynamics, we focus on 506 husband-and-wife respondents in a purposive
samples of urban, low-income households in these countries. Using information
collected from these respondents, we attempt to capture the different levels of
economic security faced by women and men in urban poor communities by a)

measuring what is called ‘degrees of informality’; b) considering their access or
lack of access to work-related benefits; and c) by estimating their earnings
variability.
Degrees of job informality

Numerous studies have demonstrated the predominance of women workers in the
informal economy in many developing countries. In recent years, there is increasing
recognition of the importance of informality, but also of its heterogeneity. Jobs
held by women and men workers in urban, low-income communities, for
instance, are diverse not only with respect to the type of economic activity but
also to the regularity and stability of employment. To examine the latter further,
we develop a job classification according to the relative degree of informality by
combining pertinent, identifiable characteristics that demonstrate work security
or insecurity, whether these jobs take place in the formal and informal sectors.
That is, we drop the assumption that workers who are considered informal only
work in those activities that fall in the informal-sector category. This is because
formal-sector firms can also employ workers on a casual basis. These workers often
do not appear on the firm registers nor get any of the work terms and benefits that
apply to the other registered workers of the firm. Although the firm and its output
are considered as parts of the formal market economy, the casual workers remain
outside and their jobs are likely to be precarious compared to other firm workers. In
other cases, a worker may work in two or more activities in the same time period.
Some individuals may work as an own-account, unregistered worker on a part-time
basis while maintaining a regular salaried job in the public or private sector.
For purposes of simplicity, we classify jobs as having a low, medium or severe/
high degree of informality. Jobs with a low degree of informality refer to permanent,
regular jobs in private and public sectors as well as self-employed activities that a
person has steadily been engaged in for over 60 months with at least 19 steady
work days per month. Medium informal jobs refer to self-employed activities held
between 24 and 60 months with an average of 12–18 steady days of work per
month, and piece-rate work for private individual/contractor involving work con-
tracts with a time period of over one year. Highly informal or severely precarious
employment includes those formal and informal jobs classified as temporary or
casual, self-employed activities that have highly irregular days of operations or
below 12 work days per month, and subcontracted or piece-rate work involving
contracts with a time period of less than one year.
Although 6–24 per cent of women and men workers in the sample classify them-
selves as permanent or regular workers, the nature of the work arrangement for
some of these workers reveals some dimensions of informality. Using the above
job classification, Table 9.1 shows that among those employed, more than 85 per
cent of workers in the Bolivia sample and 70 per cent in the Ecuador sample have
primary jobs that are medium or highly informal. For many women, these are self-
employed activities in the form of street food-vending and home-based, variety
store-retailing. In the case of Bolivia, the majority are small-scale, home-based
Table 9.1 Employment pattern of husbands and wives in urban, poor households, 2002
Bolivia and Ecuador Sample Survey, by sex (percentage of total in parentheses)
Bolivia Ecuador
By degree of
informality of Women Men ALL Women Men ALL
main job1 N ⫽ 137 N ⫽ 137 N ⫽ 274 N ⫽ 116 N ⫽ 116 N ⫽ 232
A. No. of employed
1. Formal/low 0 (0) 10 (7) 10 (4) 11 (10) 34 (29) 44 (19)
2. Medium 97 (71) 96 (70) 193 (70) 60 (45) 52 (45) 113 (49)
3. High/severe 22 (16) 20 (15) 42 (15) 22 (19) 29 (25) 51 (22)
All workers 119 (87) 126 (92) 245 (89) 9 (80) 115 (99) 208 (90)
B. Not 18 (13) 11 (8) 29 (11) 23 (20) 1 (1) 24 (10)
employed2
All respondents 137 (100) 137 (100) 274 (100) 116 (100) 116 (100) 232 (100)
Notes:
1
The categories on degree of informality are defined in the text.
2
‘Not employed’ refers to all heads or spouses who did not look for work in the previous month. Rounding
of percentages may not add up to 100.
Source: Authors’ calculations.
producers in traditional crafts, furniture, shoes, leather and garments sectors who
supply most of their goods to middlepersons. Over all, women workers in our sam-
ple tend to have relatively more precarious jobs than men in both countries,
implying greater vulnerability.
Categorization of employment by degree of informality of a worker is made
complicated by the decision of some workers, with the decline in real wages and
job security, to take on multiple jobs. Nearly a sixth of total respondents in the
Bolivia sample hold more than one job. The proportion is less in the case of
women and men respondents in Ecuador – about 10 per cent and 8 per cent
respectively. The multi-job dimension for some workers in Bolivia shows a coping
mechanism that serves as a buttress against job and income insecurity. This how-
ever may have welfare consequences in terms of lengthening the working day or
intensifying the work activities performed that can lead to, among others, stress
and deterioration of the health of the worker.
There is also a general absence of labour or social protection among these
workers, as reflected in Table 9.2. Only a very small proportion of the respon-
dents receive social-security benefits, and they are predominantly male workers
in the case of Ecuador. The lack of job benefits indicates the extent to which
economic insecurity permeates the lives of these workers, particularly women.
This is reinforced by the high degree of household-income variability examined
below.
Table 9.2 Access to benefits by workers, 2002 Bolivia and Ecuador Household
Sample Survey, by type of benefit and by sex (percentage of total in parenthesis)
Bolivia Ecuador
By type of
benefit Women Men Women Men
None 118 (99) 121 (97) 88 (94.6) 91 (79.1)

Social security
only 1 (1) 0 4 (4.3) 22 (19.1)
Health/medical 0 4 (3) 0 0
benefit only
Pension only 0 0 1 1 (0.9)
Both social
security and 0 0 0 1 (0.9)
health benefit
All workers1 119 (100) 125 (100) 93 (100) 115(100)
1
These are workers who answered the question(s). Several respondents left the question
of benefits unanswered and are therefore missing in this table.
Volatility in earnings
While the average monthly earnings of workers in the sample are not the lowest in
Bolivia and Ecuador, we have found a high degree of fluctuation, particularly among
workers with moderate and highly informal jobs. For some workers, this variation is
somewhat predictable due to the seasonal character of their activities; thus, the prob-
abilities of expected earnings in ‘low season’ or in ‘high season’ are likely to be known.
But for others, these probabilities are unknown and income variation tends to be idio-
syncratic and unpredictable. Using information based on the estimated ‘high’, ‘aver-
age’ and ‘low’ by the workers and their predicted or ‘subjective estimation’ of length
of ‘high’, ‘average’ and ‘low’ profit periods, we estimate an earnings-variability index
that allows us to compare across workers with different earnings patterns.
Table 9.3 shows average monthly earnings among self-employed workers with
predictable income variability which can dip to as much as a tenth of what they earn
during ‘high’ profit periods. The earnings variability index, called Zvar, is calculated
as the absolute value of the ratio of the weighted difference of the incomes during
high and low earnings periods to the total sum of the weighted incomes between
high and low earnings periods or:
兩Zvar兩 ⫽ abs 兩 [(Ylow ⫻ Tlow)] ⫺ [(Yhigh ⫻ Thigh)]兩
兩 [(Ylow ⫻ Tlow)] ⫺ [(Yhigh ⫻ Thigh)]兩
where Zvar ⫽ income variability index which ranges from 0–1

Ylow ⫽ average monthly income during low earnings period
Yhigh ⫽ average monthly income during high earnings period
Tlow ⫽ number of days/months considered to be a ‘low earnings period’ and
Thigh ⫽ number of days/months considered to be a ‘high earnings period’
Table 9.3 Monthly earnings and earnings variability of self-employed workers, 2002
Ecuador and Bolivia Household Sample, by sex (earnings in US dollars)
Bolivia Ecuador
Women Men All Women Men All
Average monthly 46.1 160.4 109.2 125.8 257.5 177.6

earnings1
Average Days of
operation (per 13.3 17.7 15.8 21.9 25.0 23.1
month)
Estimated
earnings-
variability index2 0.609 0.551 0.573 0.452 0.439 0.447
Notes:
1
Refer to workers with predicted low- and high-level earnings period only. Reported monthly
earnings is for the previous month.
2
The earnings-variability index is discussed in the text.; a higher index (closer to one) means
higher variability; a lower one (closer to zero) means more stable earnings. Reference period is for
the past 12 months.
A higher value of Zvar (closer to 1) reflects higher variability of earnings while a

lower value (closer to 0) means less variability. Table 9.3 gives the estimated
earnings-variability index for self-employed workers in the 2002 survey sample.
Workers in Bolivia overall seem to experience greater fluctuation in their income
levels (Zvar ⫽ 0.573) compared to those in Ecuador (Zvar ⫽ 0.447). Women workers
in both countries, however, experience greater variability as reflected in the higher
index compared to that of men. Overall, these results reaffirm our distinction
between poverty and vulnerability. While a fair proportion of households in our
sample reach a level of income above poverty lines, their daily lives involve a high
degree of economic insecurity and vulnerability, particularly in the absence of
social protection. It will be shown in the next section that in the absence of any
socialized means of absorbing the risks associated with variability in earnings
and informality of jobs, workers and their families adopt a multitude of strategies
that have both short-term and long-term welfare and productivity implications.
The fluidity of labour, gender and the household economy
It is evident that low money incomes and job precariousness represent only one
facet of economic insecurity experienced by poor and vulnerable groups; intensi-
fication of work is another important facet that tends to shift the burden of
economic insecurity to certain members within the household. The interconnect-
edness of paid and unpaid work, vulnerability and gender roles is illustrated in the
case of home-based workers in two studies based on the 2002 UPHWS Thailand
survey data.
Home-based work is carried out by two types of workers who do remunerative
work within their homes, namely: a) independent own-account producers and
b) dependent subcontract workers (ILO 2002). The study of Antonopoulos and

Floro (2005) of 268 husbands and wives in three Bangkok squatter communities
shows that the majority (87 per cent) of the women work in the informal sector,
predominantly as home-based workers (72 per cent). Among these home-based
workers, about 40 per cent are piece-rate homeworkers. In contrast, 41 per cent of
the husbands are waged or salaried employees in the formal sector.
These informal shopkeepers, artisans and other types of own-account, home-
based workers provide goods and services to others in their communities at low
prices. They typify subsistence production activities without much possibility for
accumulation. Piece-rate homeworkers, on the other hand, provide their labour to
domestic and foreign firms through outsourcing and subcontracting – thus, they
are tied to global production although at the bottom of the labour hierarchy.
Several studies have shown that the associated arrangements are inherently less
protective; subcontracted workers receive lower payments for their labour and
little or no benefits (Boonmathaya et al. 1999; ILO and HomeNet Thailand 2002).
The location of these forms of economic activity enable women to resolve the
contradictions between women’s socially defined roles as wives and mothers
and the necessity of earning income. They also enable women, in particular, to
combine paid and unpaid work.
Another study explores in greater depth the intensification of work among
urban, home-based workers in Bangkok’s low-income communities that seriously
affects their well-being (Pichetpongsa 2004). Using the 2002 UPHWS Thailand
survey data, Pichetpongsa (2004) demonstrated the extent to which home-based
workers perform both paid market work (primary activity) and overlapped
childcare or household work (secondary activity) using time-allocation analysis.
His study shows that when overlapped or secondary work activities are taken into
account, the time spent by home-based workers on both paid market work and
unpaid, household work increases. For instance, the time spent in unpaid domes-
tic and childcare activities is shown to be nearly double what conventional time-
allocation methods indicate. This work intensity is typically experienced by
women workers who care for young children while minding a store, sewing shoes
or selling food in a stall.
Income variability associated with much of informalized jobs can complicate a
worker’s control over his/her labour time and influence his/her level of work bur-
den. For women home-based workers in particular, the resulting burden is
reflected in their multiple roles both in the informal market economy and in the
household economy, which they try to accomplish simultaneously within severe
time constraints. This intensification of work is likely to have adverse conse-
quences both for the well-being and for productive efficiency of the workers.
Certain members end up devoting more labour resources to stabilize the incoming
stream of income in order to protect the household from the dire consequences of
substantial income fluctuations. The result may likely lead to tensions among
household members. The distribution of work burden within the household
depends on the interplay of gender norms and intra-household negotiations.
A different question is whether a worker’s increased contribution to household
income translates into greater bargaining power and gender equality as some
studies have examined (Benería and Roldan 1987; Kabeer 2000). If this is the case
and a renegotiation of household division of tasks occurs, women’s increased
participation in paid work may result in having other members take on her
domestic chores. But if gender norms are resilient to any change in money income
contributions, women may be compelled to increase the length of the working
day or intensify their work effort by overlapping activities, say combining
childcare and market work. As pointed out earlier in this chapter, gender roles in
Latin America have been changing rapidly in terms of the more observable and
measurable indicators. However, at the deeper level of gender norms and distribu-
tion of tasks between paid and unpaid work, we observe change as well as conti-
nuity. For example, a subsample of 29 households in Santa Cruz, Bolivia,
interviewed in January 2005, typifies these trends. When asked whether they
thought that gender roles were changing, 82 per cent replied affirmatively.7
However, the answers as to how they saw this change were more ambiguous;
57 per cent pointed out that women had more financial responsibility now
because of their greater involvement in paid work; about 43 per cent agreed with
the statement that it was easier for women to find jobs than for men; and 82 per
cent responded affirmatively to the question of whether they thought that gender
equality had increased.
Table 9.4 lists several proxies or indicators of the relative position of workers in
household decision-making. Based on the 2002 Bolivia and Ecuador household
workers sub-sample, the results show that there are relatively more women respon-
dents than men respondents who say that they require permission to work from
their partners/spouses. Roughly the same proportion of women and men reported
that they decide solely on the use of their own earnings. Another indicator of
Table 9.4 Patterns of household decision-making in couple households, 2002 Bolivia and
Ecuador Household Sample, by sex
Bolivia Ecuador
Proportion of respondents Proportion of respondents

(in %) (in %)
Women Men Women Men
1. Requires 79.2 49.45 42.6 17.5

permission to work
2. Decides on use 22.5 10.2 61.7 60.4
of own earnings
3. Decides jointly 69.4 75.0 20.0 22.5
on use of earnings
4. Knowledge of 87.4 79.6 66.9 43.2
spouse’s earnings
5. Decides to 19.8 16.7 23.5 50.0
borrow loan
6. Responsible for
loan payment 40.5 28.7 63.5 58.5
bargaining power is with respect to loan decisions. About 20 per cent of workers in
the Bolivian sample reported having made sole decision of when to borrow; the
majority reported that this is a joint decision. The proportion of individual deci-
sions to obtain credit is much higher in the case of Ecuador, especially among
men. But a different pattern emerges when it comes to the question of who pays
the loan or shoulders the burden of loan payment. A higher proportion of women
respondents in both Bolivia and Ecuador reported that they are solely responsible
for loan repayment compared to the men respondents. These have repercussions
in terms of a worker’s vulnerability and her future earning capacity as our case
study of Ecuador shows.
Coping strategies and risk-sharing
The mere existence of employment does not define the economic status and well-
being of a worker or a household. It needs to be seen as acting together with a
number of other factors, the most crucial being access to and control over
resources – mutual support networks, credit, savings and physical assets – in the
face of income shortfalls and consumption expenditure shocks.8 Income variabil-
ity, alongside the persistence of job insecurity, heightens the household’s need to
find ways of smoothing its consumption and designing survival strategies. Coping
strategies among the poor have received much attention in the literature on eco-
nomic and financial crises since the 1980s. As this section will illustrate, there is a
strong propensity for households with irregular and variable sources of income to
become increasingly differentiated. There is also a tendency for the risk burden to
be shouldered unequally among household members.
Gender, credit and debt-servicing

Our study among Ecuadorian urban poor households shows the critical role of
credit in consumption-smoothing and in managing losses from shocks such as
family emergencies. The data enables us to identify the designated borrower in the
household by also taking into account intrahousehold decision-making in terms
of ‘who decides to borrow and how much’ and ‘who repays’. Our data and inter-
views reveal that there are clear demarcations on whose earnings and responsibility
it is to pay. In most cases, incomes are not pooled in these households, at least to
meet debt obligations.
Based on 194 households that include both households with couples and single-
adult-headed households in the 2002 Ecuador UPHWS (Urban Poor Home-based
Worker Survey), Table 9.5 shows that the majority (62 per cent) of the ‘designated’
borrowers in the urban, poor households are women. Given their relatively lower
earnings compared to men, more loans suggests a higher debt burden among
women in low-income households compared to men. Floro and Messier (2004)
estimated the debt-service burden of borrowers in the Ecuador sample by calculat-
ing the mean ratio of montly debt-servicing (the past month’s loan principal and
interest amortization payment) to the past month’s earnings of the borrower. The
results are shown in Table 9.5: a much higher debt-service ratio among women
Table 9.5 Mean debt service ratio of borrowers, 2002 Ecuador Household
Sample, by sex and type of household headship
Men Women Single adult- Couples

N ⫽ 133 N ⫽ 176 headed households households
Debt-service Ratio .0879 .3784 .1789 .2661
Source: Floro and Messier (2004).
borrowers (0.378) than that among men (0.088). Given their dual roles as income
earners and primary care providers, Ecuadorian women workers in our sample seem
to have a more compelling need for credit both to meet the needs of their livelihood
activities, and to meet their household consumption expenses. The economic inse-
curity associated with informalized jobs therefore may have increased the demand
for credit to help smoothen consumption but with lessened inability to repay.
There can be important, far-reaching effects on future income from job insecu-
rity if some of the loans that are intended for productive purposes are themselves
used to buffer current consumption from income fluctuations. Home-based workers
who may need their own resources to pay for raw materials, equipment and other
inputs in their job, can find themselves severely capital-constrained and hence
unable to increase their earnings. Using the 2002 Ecuador survey data, the study
by Messier (2005) on poverty traps and credit provides empirical evidence on how
the allocation of credit, especially by self-employed women workers, for con-
sumption use (e.g. to meet food and educational expenses, household repair,
health care, and so forth) can adversely affect the level of investment in their own
enterprises. As a result, their income in the next period is likely to be lower or more
precarious, thus threatening social reproduction and, again, reflecting risk and
vulnerability.
Informal social networks and mutual support mechanisms

The degree of social cohesion and trust prevailing in a given community can help,
to some extent, in providing insurance to workers and their households. Kinship-
based or other social networks allow households to share each other’s risk through
some community mechanism or institutional arrangements in a given setting.
These include kin-based or non-commercial loans as in Ecuador and community
services provisioning in both Ecuador and Bolivia. Our observation indicates that
in both countries extended family networks are most important in designing
strategies to cope with risk and vulnerability but also for planning for the future
(such as in decisions to migrate). We next examine these different mutual support
mechanisms as they affect the lives of workers in urban low-income communities.
Systems of ‘generalized reciprocity’ often take the form of kin credit. These loans
are provided by those whose income is temporarily high compared to those in
need. Relatives, friends and neighbours are approached for loans during times of
cash shortage or family emergencies. This is reflected in the credit patterns of
urban poor households in Ecuador. Table 9.6 shows that kin credit or loans
provided by friends, relatives and/or neighbours comprise nearly 32 per cent of the
total number of loans, indicating the presence of socialized income-maintenance
schemes based on either extended kinship ties or locational affinity in urban poor
communities. In the case of one urban poor community in Quito called Itchimbia,
the strong cohesion and active community organizations have facilitated a high
degree of social cooperation and mutual support among its residents.
While annualized interest rates are roughly 24 per cent on formal/semi-formal
loans (predominantly from banks and microcredit organizations), and 73 per cent
and 44 per cent on informal loans (from private moneylenders) to men and
women borrowers respectively, kin loans are interest-free and require no collateral
(Table 9.6). Overall, women more often approach their relatives, friends and
neighbours for loans. Loan provisioning, as our Ecuador case study demonstrates,
is part of the income-maintenance schemes and mutual support systems that have
evolved to cope with the basic requirements and contingencies of life. Women
have played an important role as participants in such networks.
To some extent, community-based formal and informal networks have played
some role in meeting the transport, health care and social-service needs of the
community. This is the case among the urban poor communities in Bolivia and
Ecuador that we have encountered in our field work. Community-based trash,
sewer and medical services as well as some form of recreation facilities are
Table 9.6 Loan characteristics by gender and credit source, 2002 Ecuador Household Sample
(percentages in parentheses)
Men borrowers Women borrowers
Formal/Semi- Formal/Semi-
formal Informal Kin formal Informal Kin
N ⫽ 29 N ⫽ 10 N ⫽ 19 N ⫽ 41 N ⫽ 25 N ⫽ 30
Mean loan size1 1045.86 598 77.053 865.93 786.60 270.27

Annual interest rate2 24.92 72.89 0 23.75 44.26 0
No. of instalments3 14.90 8.90 2.26 16.56 26.44 1.7
Payment per
instalment4 96.86 53.5 21.05 158.55 35.31 88.47
No. of loans with
collateral5 12 (41.38) 4 (40) 0 (0) 27 (65.85) 11 (44) 0 (0)
No. of loans with
cosigners6 16 (55.17) 2 (20) 1 (5.26) 15 (36.59) 7 (28) 2
(6.67)
Notes:
1 Loan amount in dollars.
2 Nominal interest rates were annualized as follows: Interest rate ⫽ [(Repaid amount⫺Loan)/(Number of
monthly payments]*12. the inflation rate in 2002 was 12.5 per cent, using consumer price index.
3 Instalments refer to single payments made by the borrower to cover loan interest and principal.
4 Average dollar value of the monthly instalment.
5 Number of loans that required some form of collateral to guarantee the loan. Collateral included land,
house, durable goods, etc.
6 Cosigners include spouse, parents, other family members and neighbours.
Source: Floro and Messier (2004).
provided by the government in nearly two-thirds of the community sample; but

they are often precarious and represent a bare minimum. In two of the Ecuador
communities, we found some form of primary-care medical services being
provided by non-governmental organizations. Although these services are initi-
ated or sponsored by the government and non-governmental organizations, they
also make use of volunteer labour provided by community members.
Asset ownership and asset pawning

There are however limits to risk-pooling mechanisms within communities and social
groups especially when there are adverse economic and social conditions such as eco-
nomic downturns, crises, labour market restructuring and disintegration of social
networks. For the most part, informal workers and their households often have to
deal with economic insecurity and cash shortages themselves. The study on credit
access by urban poor households in Ecuador shows the critical role of other types of
loans in consumption-smoothing and in managing loss from shocks such as family
emergencies. In addition to relatives and friends, there are moneylenders and traders
as well as microcredit organizations and even banks who have provided loans.
In Thailand, both men and women workers in urban squatter communties have
made use of their real assets, especially jewellery, livestock and appliances as buffer
assets to smooth their consumption in the face of irregular income streams.
The study by Floro and Antonopoulos (2005) on asset pawning among poor
households shows that nearly a fifth (17 per cent) of women workers in the 2002
Thailand UPHWS data have pawned their individually owned assets such as
jewellery and even shop assets in the previous six months compared to only 5 per
cent of men. Their findings suggest that women’s responsibility of reconciling
income flows and cash deficits in order to keep the children in school, meet social
obligations (e.g. weddings and funerals) and buy basic necessities are social con-
structs that affect the manner by which coping mechanisms are shared within the
household. Women, in performing their gender roles, are more likely to pawn
their assets for consumption-smoothing and to meet special household needs.
In the final analysis, our fieldwork interviews in Ecuador and Bolivia reveal that
households provide the last resort or means of protection whenever chronic or
unexpected shocks increase their vulnerability or threaten their survival. The shar-
ing of living and even working spaces between two or more families occurs in
some of the households we interviewed. This meant sharing the costs of food, elec-
tricity, and often overcrowded living space. This is true for workspace as well, as
with the variety of stores situated at the front part of some homes.
Concluding remarks
Returning to the three areas of inquiry referred to in our conceptual framework,

our empirical work brings further information to our understanding of how urban
low-income households are absorbed into an increasingly informalized economy
and of how they cope with the economic insecurity generated by the lack of what
the ILO has called ‘decent jobs’, and by poverty and vulnerability.
First, in terms of the evolving informal economy, we document its heterogene-

ity and classify jobs as having low, medium or high ‘degrees of informality’ in
order to analyse the wide range of combinations and fluidity between the
formal/informal insertions. The large majority of jobs in our sample (over 85 per
cent and 70 per cent in Bolivia and Ecuador respectively) were characterized by
medium or high degrees of informality. For the majority of workers, these were
self-employed activities such as street vending and home-based work, with women
predictably having relatively more precarious jobs than men. However, we point
out that the relationship between job informality and worker vulnerability is not
straightforward, due in part to the multiple-job dimension of some workers and to
the intrahousehold dynamics which enable the shifting of the risk burden among
household members. As a result, some workers, by virtue of their gender roles
and/or bargaining position within the household, become major bearers of risk.
The study reinforces the importance of distinguishing between poverty and
vulnerability, particularly because we find a high degree of income variability and
other factors placing households at risk. We have found that, even in non-poor
households, income variability leads them to borrow when earnings are low and
exposes them to the pressures created by debt. We emphasize that there is a strong
propensity for households with irregular and variable income to become increas-
ingly differentiated and to face increased economic insecurity.
Informality and job precariousness are particularly illustrated in the case of self-
employment and home-based work which involve two different types of workers:
a) those representing subsistence production and providing goods and services for
their communities at low prices, and b) waged or self-employed workers linked
through subcontracting to other productive processes, including global production.
In both cases, we describe the fluidity of labour between paid/unpaid and formal/
informal activities. This is particularly so in the case of women; the location
and informality of their activities allows them to negotiate the tensions between
their socially defined roles around domestic work and the need to earn an income.
This fluidity of labour, however, has led to an increased intensification of work
that engenders high levels of stress and a lower quality of life.
Second, in terms of the gender dimensions of these processes, we emphasize the
important contribution of women to household earnings and we ask whether this
translates into greater bargaining power for them. Our data points to both change
and continuity in this respect. On the one hand, there is a growing awareness of
women’s contribution to household finances and of changing gender roles with
implications for a higher degree of gender equality. On the other hand, women’s
greater participation in paid work is not compensated by a decrease in their
responsibilities in domestic work, thus resulting in work intensification and over-
lapping activities. Women are also burdened with their greater responsibility in
debt repayment and household maintenance as well as in domestic work; indica-
tors of vulnerability (e.g. debt-service ratio, the likelihood of pawning one’s assets
and intensification of work) reflect a higher incidence among women. Thus, we
observe both some degree of ‘de-gendering’ as well as continuity in the division of
gender roles.
Third, one of the most striking findings from our study is the extremely low
degree of social protection among the households in the sample. Only a very small
proportion of the respondents received any type of social security benefits – the
exception being 19.1 per cent of male workers in Ecuador. Neoliberal policies have
had an implicit bias against any type of social protection programmes; we have
observed a privatization of security in the sense that, for the most part, it has been
left to kin and family networks to provide some means of protection. Mutual
support mechanisms and systems of reciprocity exist for example to provide loans
and as a way of smoothing consumption. Families also design living strategies
around housing and living spaces, such as relatives moving into a family home at
a time of crisis or loss of employment, or in cases of home-switching among family
members according to the affordability of the highest rent.
Ultimately, our study leads us to emphasize the notion that poverty elimination
programmes are not likely to succeed unless they are accompanied with policies to
create decent jobs for those living under the conditions of informality described in
this paper. Second, the study joins the voices that call for new forms of universal
social provisioning. Both point to a more interventionist state.
Notes
1 The sample survey is referred to here as 2002 Urban Poor Home-based Worker Survey
(UPHWS). The authors would like to specially thank John Messier for the fieldwork in
Ecuador and data processing, UNIFEM-Andean Region for the financial support received
and FLACSO (Facultad Latinoamericana de Ciencias Sociales), REMTE (Mujeres
Transformando la Economia) and FEDESO (Guayaquil) for their institutional support.
Many thanks also to Lourdes Montero and CEDLA (Centro de Estudios sobre el Desarrollo
en America Latina) in La Paz for their invaluable help with the Bolivia survey. We also
want to acknowledge the assistance provided by Lucia Salamea Palacios (UNIFEM-Andean
Region), Gioconda Herrera (FLACSO), Magdalena de Leon (REPEM), Maria Rosa
Anchundia (REMTE and University of Guayaquil), and Elizabeth Estrella (UNIFEM-
Andean Region) for their help with the training workshop in Ecuador and conference in
Quito in March 2004. Our thanks also to Sonali Duggal and Carmen Triana for their
generous assistance.
2 An elaboration of some of these points can be found in Benería 2001.
3 One of the main elements of this package was a set of labour-policy measures. These took
the form of: a) reduction of mandatory employment stability to all activities, b) wage-
setting at market-determined levels, and c) abolition of a firm obligation to pay workers
on strike (World Bank 1988).
4 A household is classified extremely poor if its total consumption expenditure is below the
food poverty line.
5 In fact, prices increased by 20,000 per cent from August 1984 to August 1985, and by
60,000 per cent between May and August 1985 (Sachs 1987). These represent the highest
inflation rates in Latin American history and one of the highest in the world.
6 The diversity of low-income neighbourhoods in the metropolitan cities can take the form
of old versus newly settled areas, proximity to or distance from main city centres, density of
population, and the degree of social cohesion represented by the presence or paucity
of grassroots organizations. It is important to note that the community selection for the
survey took into account the presence of contact such as a local community organization
or a community-based researcher which helped establish trust and initial rapport.
This facilitated the ‘entry’ of the survey team in the area.
7 Interviews were conducted either with men or women household members.

8 Findings from a number of studies show that although household incomes may
fluctuate widely as result of weather, prices and employment-related shocks, consumption
doesn’t necessarily follow such a pattern of variability. Household members make use of
different mechanisms to smooth consumption (Bardhan and Udry 1999; Sebstad and
Cohen 2001).
References
Antonopoulos, R. and M. Floro (forthcoming) ‘Asset Ownership Along Gender Lines: Evidence
from Thailand’. Journal of Income Distribution.
Balakrishnan, R. and M. Huang (eds) (2000) The Hidden Assembly Line: Gender Dynamics of
Subcontracted Work in a Global Economy. Bloomfield CT: Kumarian Press.
Bardhan, P. and C. Udry (1999) Development Microeconomics. Oxford: Oxford University Press.
Benería, L. (2001). ‘Shifting the Risk: New Employment Patterns, Informalization, and
Women’s Work’. International Journal of Politics, Culture and Society, Vol. 15, No. 1,
September: 27–53.
Benería, L. (2003) Gender, Development and Globalization: Economics As If All People Mattered.
London: Routledge.
Benería, L. and M. Roldan (1987) The Crossroads of Class and Gender: Homework, Subcontracting
and Household dynamics in Mexico City. Chicago: University of Chicago Press.
Benería-Surkin, J. (2003) ‘Power, Conservation, and Indigeneous Livelihood: Guarani
Strategies for Conquering Political Space in Izozog, Bolivia’. PhD dissertation, University of
California, Los Angeles.
Boonmathya, R., Y. Praparpun and R. Leechanavanichpan (1999) ‘The Situation of Women
Subcontracted Workers in the Garment Industry in Bangkok, Thailand’. Working Paper,
the Asia Foundation, Bangkok, October.
Buechler, S. (2002). ‘Enacting the Global Economy in Sao Paulo, Brazil: The Impact of Labor
Market Restructuring in Low Income Women’. PhD dissertation, Columbia University.
CELA-PUCE (2002) ‘Evaluacion de Los Impactos Economicos y Sociales de las Politicas de
Ajuste Estructural en el Ecuador 1982–1999’, at http://www.saprin.org/ecuador/research/
ecu_res_ej_fin.pdf, accessed on 8 July 2003.
De Soto, H. (2000) The Mystery of Capital. Bantam Press/Random House.
Duryea, S., A. C. Edwards and M. Ureta. (2004) ‘Women in the LAC Labor Market: The
Remarkable 1990s’, in Claudia Piras (ed.), Women at Work: Challenges for Latin America.
Inter-American Development Bank, Washington DC: 27–56.
Economic Commission for Latin America and the Caribbean (ECLAC) (2004). Panorama
Social para América Latina y el Caribe 2002–2003. Santiago de Chile: CEPAL.
Floro, Maria (1992) ‘Macroeconomic Analysis Background Paper: The Case of Ecuador’.
Urban Poverty and Social Policy in the Context of Adjustment Project, Working Paper,
Floro, M. and J. Messier (2004) ‘Credit, Gender and Debt Servicing: The Case of Urban Poor
Households in Ecuador’. Paper Presented at the Southern Eastern Economics Association
Meetings, 19–20 November 2004, New Orleans.
Floro, M. and R. Antonopoulos (2005). ‘Asset Pawning among Urban, Low Income
Households in Thailand’, Levy Economics Institute Working Paper, forthcoming.
González de la Rocha, M. (2000) Private Adjustments: Household Responses to the Erosion of
Work. UNDP, Social Development and Poverty Elimination Division, Conference Paper
Series No. 6, New York: UNDP.
Gutmann, M. (1996) The Meanings of Macho: Being a Man in Mexico City. Berkeley/Los
Angeles/London: University of California Press.
International Labour Organization (ILO) (2001–02). Key Indicators of the Labour Market
(KILM). Geneva: ILO.
ILO (2002). Women and Men in the Informal Economy. Geneva: International Labour
Organization.
ILO and HomeNet Thailand (2002). Impact of the Economic Crisis on Homeworkers in Thailand.
Bangkok, International Labour Office.
Kabeer, Naila (2000) The Power to Choose: Bangladesh Women and Labour Market Decisions in
London and Dhaka. London and New York: Verso.
Kabeer, Naila (2003) ‘Labor Standards, Women’s Rights, Basic Needs: Challenges to Collective
Action in a Globalizing World’. In L. Benería and S. Bisnath (eds) Global Tensions:
Challenges and Opportunities in the World Economy. London: Routledge.
Kohl, B. (1999) ‘Economic and Political Restructuring in Bolivia: Tools for a Neoliberal
Agenda’. PhD dissertation, Cornell University.
Lavinas, L. (1996) ‘As mulheres no universo da pobreza: o caso Brasileiro’. Estudos Feministas,
4(2): 454–79.
McCrate, E. (1995) ‘The Growing Divide Among American Women’. Unpublished paper,
University of Vermont, Department of Economics.
McFarren, W. (1989) ‘Economic Crisis and the Politics of Adaptation: Survival Strategies of
Displaced Bolivian Mining Households’. MA thesis, Cornell University.
Messier, John (2005) ‘Dynamics of Poverty Trap and the Role of Credit’. Unpublished
doctoral thesis, Economics Department, American University, Washington DC, August.
Morales Anaya, J.A. (1987) ‘Estabilización y Nueva Politica Económica’. EL Trimestre
Economico, LIV: 179–211.
Nussbaum, M. (2003) ‘Women’s Capabilities and Social Justice’. In L. Benería and S. Bisnath
(eds) Global Tensions: Opportunities and Challenges in the World Economy. New York:
Routledge.
Parandekar, S., R. Vos, and D. Winkler (2002) ‘Ecuador: Crisis, Poverty and Social Protection’.
In P. Beckerman and A. Solimano (eds) Crisis and Dollarization in Ecuador: Stability, Growth
and Social Equity. Washington DC: World Bank: 127–76.
Pérez-Sáinz, J. P. (1994) El Dilema del Nahual. Globalización, exclusión y trabajo en Centramérica.
Costa Rica: FLACSO.
Pérez-Sáinz, J. P. (2005) ‘Labor Exclusion in Latin America: Old and New Tendencies’. In
L. Benería and N. Kudva (eds) Rethinking Labor Market Informalization: Poverty, Precarious
Jobs and Social Protection. Ithaca, NY: Cornell e-Publishing.
Pichetpongsa, Anant (2004) ‘Work Intensity and Well-being: The Case of Women
Homebased Workers in Thailand’. Unpublished doctoral dissertation, American
University, March.
Piras, C. (ed.) (2004) Women at Work: Challenges for Latin America. Washington, DC: Inter-
American Development Bank.
Portes, A. and M. Castells (1989) The Informal Economy. Baltimore: Johns Hopkins University
Press.
Republic of Bolivia (2001) Poverty Reductions Strategy Paper. La Paz.
Sachs, J. (1987) ‘The Bolivian Hyperinflation and Stabilization’. American Economic Review,
May: 279–83.
Salzinger, L. (2003) Genders in Production: Making Workers in Mexico’s Global Factories.
Berkeley: University of California Press.
Sebstad, J. and M. Cohen (2001) Microfinance, Risk Management and Poverty. Washington DC:
Consultative Group to Assist the Poorest.
Sen, A. (1992) Inequality Reexamined. Cambridge MA: Russell Sage Foundation and Harvard
University Press.
Solimano, A. (2002) ‘Crisis and Dollarization: An Overview’. In P. Beckerman and
A. Solimano (eds) Crisis and Dollarization in Ecuador: Stability, Growth and Social Equity.
Washington DC: World Bank: 1–16.
Unidad de Análisis de Politicas Sociales y Económicas (UDAPE) (2003) Crisis, Crecimiento y
Pobreza. La Paz, January.
United Nations Development Programme (UNDP) (2002). Human Development Report.

United Nations Development Programme (UNDP) (2003). Human Development Report.
Weinberg, S. (2002). ‘An Analysis of Informal Sector Expansion in Ecuador’s Urban Areas’.
Unpublished Paper, Washington DC: American University.
World Bank (1988) Ecuador Country Economic Memorandum. Report No. 7321-EC, August.
10
Working People and Access to
Social Protection
Francie Lund
Introduction
The informal economy is growing worldwide in the proportion of the workforce

who work in it, in the number of informal enterprises, and in the economic con-
tribution of the informal sector to the GDP of many countries (ILO 2002). There is
an increase in the numbers of self-employed people (own-account operators, and
those who employ others); growing numbers of people work on a contractualized
or outsourced basis; fewer people work with secure and lasting contracts; and
many formal workers are losing work-related social benefits such as access to
health insurance and pension provision. This applies to countries worldwide,
though ‘non-standard employment’ is the term usually used to refer to countries
in the North and ‘informal employment’ to those in the South.
There are variations in the position of men and women in the informal economy,
in terms of their share of employment, status of employment, and incomes earned
(ILO 2002; Chen et al. 2004; Heintz 2005). With few exceptions, however, infor-
mal employment in developing countries tends to be a larger source of work for
women than for men. Moreover, men tend to be found in the higher segments of
informal employment, notably as employers who run informal enterprises, while
women are more likely to be either unpaid workers in family enterprises or indus-
trial outworkers (or homeworkers). As a result, women working informally earn
less than men working informally. Women working informally also earn less than
women working in the formal economy. In developed countries, the emerging
trend towards part-time and short-term employment is strongly gendered, with far
more women than men in such jobs (ILO 2002; UNRISD 2005). This ‘non-standard’
work tends to come with fewer social benefits such as medical aid, pension
schemes, maternity benefits or disability insurance (Vosko 2005).
Parallel with and related to these changes in the structure of employment, the
global neoliberal policies of the last thirty years, as espoused in the ‘Washington
consensus’, have seen, in many developed and developing countries, a shrinking
of the size and functions of the state in favour of market-led growth, and a reduc-
tion of many governments’ commitments to social spending. Withdrawal of state
commitments to health and welfare means that more of the responsibility for this
217
218 Francie Lund
work falls to private individuals (in the family) and to ‘community care’ or
‘volunteer work’. It is overwhelmingly women who take the larger share of this
unpaid caring work. In the period post-Washington consensus, there is more scope
for the role of the state, and poverty relief is on the social policy agenda (see
Molyneux, chapter 2 in this volume) – but women’s low paid and unpaid work is
not central to the analysis.
Especially for poorer women who work informally, the labour market is itself a
source of risk. Formal social-protection provision is not accessible; private (self-
funded) provision is not affordable, and small project-based provision is often not
sustainable (or affordable by the very poor). At the same time, poorer workers are
excluded from possibilities for skills generation or regeneration. While women’s
responsibility for health and welfare – in their own homes and in their communities –
continues, there seems to be little gendered exchange of productive and reproduc-
tive roles, even where there are high rates of male unemployment. Where chronic
poverty coincides with the AIDS pandemic, the caring obligations of women, and
to a lesser extent of men, increase and this influences access to and exit from the
labour market.
In 1952 the International Labour Organization (ILO), listed its core of contin-
gencies of social security: healthcare, incapacity for work due to illness, disability
through work, unemployment, maternity, child maintenance, invalidity, old age,
and death of the breadwinner. In advanced industrial economies, there was the
expectation that being in formal employment meant that there were contractual
obligations between employers and employees, which included some employer
responsibility for protection against risk, whether ex ante, anticipating risks that
would happen, or ex post, in response to various contingencies after they occurred.
This was frequently done with contributions from the state as well.
While the ILO in the 1950s expressed its hope that the contingencies for social
security could be expanded to all workers worldwide, there has in fact been a
reversal in social protection since then, and this has happened not only in those
developing countries which had some social security, but also in developed coun-
tries. What, actually, are poorer people supposed to do to cover their own risks
when they cannot afford the services provided by the private sector and when
their attempts at creating their own long-term formal and informal savings
founder because of the pressing needs of the present? Sustainable schemes of co-
insurance among poor people are difficult to achieve, as the core insurance principle
of cross-subsidization is absent. Further, expecting poor people to build their own
‘institutions for the poor’ is contradictory, in terms of investments of both time
and money required.
This chapter explores an approach to social protection which places women’s
low paid work as an important element for consideration in social policy. It holds
that patterns of production generate inequalities which disproportionately affect
women, and the growing size of the informal economy means that this will be a
key policy issue for the foreseeable future. Social policies which seek to address
both poverty and gender inequality will need comprehensive interventions from
multiple interest groups – informal workers, organized labour, the private sector,
Working People and Access to Social Protection 219
the state, and organizations in civil society. It is too ambitious to expect that all
workers, everywhere, will be able to get access to comprehensive protection mech-
anisms in the near future. The search for appropriate economic and social policy
responses will be assisted by reframing the standard risks and contingencies with
a focus on informally working women. This should take into account vulnerabil-
ity over the life cycle, and the specific contingencies and risks faced by informal
workers in their different places of work.
Gendered vulnerability and the life cycle
Many systems of social protection were based on a ‘family model’ or ‘male bread-
winner model’, with the idea of a male ‘primary breadwinner’ in full-time employ-
ment, and with access to insurance benefits secured through the relationship of
family members to the working person. Accommodating the fact that increasing
numbers of women are in the labour force, some detect a shift in welfare states to
an ‘adult-breadwinner model’ – with the assumption still that being in employ-
ment will bring with it access to measures of protection against risk. Men’s and
women’s risks are dependent on their relationship to other people in the
household; and households themselves, and members within them, go through
life-cycle stages, which affect men, women and children in different ways.
Different types of risks and vulnerabilities are associated with different stages in
the individual’s life cycle as well, and there are deeply gendered dynamics at play.
There are clearly overlaps between the different life stages, and some people may
not go through some of the stages. There are gender similarities and gender differ-
ences. Boy children may be at least as likely as girl children, in some countries, to
be withdrawn from school for household or other labour. Childbearing may begin
in adolescence and continue in later life. Not all women and men marry; single or
lone parenting has become more common, as has divorce. A woman may become
a grandmother at 50, and be responsible for the care of the children of her work-
ing daughter at the same time as being responsible for the care of her own 75-year-
old mother. Older men who are widowed are far more likely to remarry, and not be
alone, than older women.
Women, the life cycle and precarious work
For people in precarious employment, there is a close link between economic secu-
rity and social security. Loss of income may have different causes, for example sea-
sonal failure of produce, through loss of a contract, through inability to secure a
contract, or through unreliable and late payment for work done. Loss of assets may
arise through theft, fire and natural disaster; poorer informal workers such as street
vendors may lose assets through the destruction or confiscation by local authori-
ties of equipment such as selling-tables, carts, sewing machines, shoe cleaning out-
fits, and other tools. Another common reason for loss of income is ill-health, of
the worker herself, or of someone in her family for whom she has to care. Even
short interruptions of work can be a major disaster for the worker who has no paid
220
Table 10.1 Gendered risks and vulnerabilities associated with different stages of the life
cycle
Life-cycle stage Types of associated risks and vulnerabilities
Very young children Nutritional risks faced by children of working and

0 to 5 non-working mothers, with lifelong and irremediable
developmental deficits
Delays in early cognitive development if left with an
inactive carer
Acute vulnerability to disease and infection
Children 5 to 12 The risk of not attending school because of domestic
or income-earning responsibilities.
The invisibility of young children’s work, whether as
‘normal’ household responsibilities, or in non-family
enterprises when it is tied to the employment of
parents
The double burden of occupation and schooling and
therefore long-term vulnerabilities of low
productivity and fewer opportunities
Adolescents 13 to 19 Vulnerability of (especially girl) children to early
withdrawal from school
Double burden of occupation and schooling and
therefore long-term vulnerabilities of low
productivity and fewer opportunities
Entry into high-risk employment categories,
hazardous industries, prostitution, etc.
Young adults and Heightened lack of access to financial institutions/
mothers in their twenties asset-building opportunities
and thirties Loss of employment, or employment insecurity,
through pregnancy
Loss of employment, or employment insecurity,
through time taken out for childcare
Middle-aged adults Loss of employment, employment insecurity,
decreased productivity, through care for younger and
older family members at the same time
Bearing the costs of disease and death
Social expenses especially the costs of marriages of
children, funerals, other rituals
Older people Loss of income when work is lost, in the absence of
work-related provision for retirement, and absence of
state support
Widows’ loss of assets to late husbands’ family
claims
Onerous childcare responsibilities in countries
where AIDS or military destabilization has produced
high numbers of absent middle age adults, and high
numbers of children in stress
Source: Adopted from Lund and Srinivas (2000: 37–8).

leave, no paid sick leave, and no healthcare scheme. There is a vicious circle where
poor health lowers income, which in turns contributes to poor health.
Women worldwide take greater responsibility for childcare than men do. As
increasing numbers of women enter the labour market, this has an effect on the
need for childcare. Even poorly paid women hold that an advantage of working
from home is that multiple tasks can be seen to simultaneously, and in particular
that an eye can be kept on children. However, the presence of children, and the
need for children to be attended, requires inputs of both time and energy from the
worker/carer, and this will affect her productivity and ability to earn. Most welfare
states have a number of benefits surrounding the bearing and rearing of children.
A comprehensive package would include free healthcare, the right to extended
paid leave, the right not to be retarded in progressing along the career ladder by
having taken such leave, and in some countries, the rights of fathers to some
paternity leave (‘daddy leave’) as well. Some countries allow time off for breast-
feeding while at work, and work-based childcare facilities are provided. Workers
with no maternity benefits are at risk of loss of income when they take time off
work to have children, and are likely to return to work sooner than is good for
their own health and for the health of the newborn child, in order to earn again.
Women in export processing zones may be required to sign an agreement declaring
that they are not pregnant at the time of hiring, and that they will lose their
employment if they do get pregnant.
Though there are marked country and regional differences, girl children may be
at particular risk of having their education cut short so that they can work in the
family enterprise, or look after siblings while mothers work elsewhere. This may
reduce their chances of acquiring the skills necessary for earning a decent income.
However other factors are at work besides the demand for labour. Caste, custom
and religion can mire girl children into trajectories of little education, early mar-
riage and early pregnancy.
The vulnerability of unprotected elderly people has become a significant social
problem. The elderly may have been economically active all their lives, and
women in addition may have played a significant health and welfare caring role
all their lives, yet there is no certainty of security in their older years. There are crises
in pension systems worldwide, with demographic shifts (more people living far
longer), and cutbacks in social spending. Even those with formal coverage find
that, especially where the levels are not inflation-linked, the pension does not
stretch far enough. Many women’s access to financial security in old age is
through their working husbands. In most societies, women live longer than men.
Widowed men are more likely to remarry and have someone to care for them. A
woman in an advanced welfare state may get widow’s benefits, benefits for chil-
dren, and a lifetime pension for herself. In some societies, on the contrary, women
who are widowed are especially vulnerable to asset-raiding by the late husband’s
family, and to a lowering of social status (see Chen 1998 and Chen 2000 for an
account of the status of widows in parts of India).
Specific risks are presented by the place of work itself. There are hazards
associated with all forms of work, and formal workplaces are typically regulated by
222 Francie Lund
the country’s legislation on occupational health and safety, as well as insured by

the owners. It has now become common for people’s private houses to have
become places of informal work – for example for garment workers, typists, back-
yard mechanics, caterers, incense-stick rollers, paid childcarers. Having a secure
shelter becomes an important component of a secure working life; by the same
token, damage to or loss of the shelter poses threats to the productivity of home-
based workers. Many homeworkers (industrial outworkers) are in fact employed
on a piece-rate basis, but the employer bears no responsibility for the safety of the
place of work. Industrial outworkers living and working in poor and overcrowded
housing may have poor lighting and ventilation. Street vendors are vulnerable to
the weather, to criminals, and to passing traffic.
The vulnerability of specific groups
Men and women face specific risks over the life cycle, and additional specific risks
are attached to working in precarious work, as noted above. Specific groups of
women workers are excluded from access to social protection in specific ways. Four
such groups are considered here: migrants, sex workers, employed domestic workers
and unpaid full-time carers.
Migrants
A key feature of globalization is the mobility of labour. Increasing numbers of
people migrate in search of work, within their own countries, and across national
borders. While historically most migrants were men, sometimes accompanied by
women as family members, the numbers of women who migrate autonomously
are increasing (Kofman 2004). The highly mobile cadre of professional self-
employed migrant workers, who face uncertainty of employment and high risks,
are paid well enough to make their own provision for savings and insurance
against work-related contingencies. For the majority, though, the act of migrating
is a step into vulnerability. Women and men leave behind relationships with fam-
ily and friends, as well as neighbourhood and religious networks. They may also
leave behind or lose formal entitlements which depend on their status as citizens
in their local areas, such as access to health, welfare and other services. Many
people migrate through kinship or village-based networks, and go to places where
there is a known receiving community. Yet migrants typically assume minority
status, may be unfamiliar with the dominant language and have little knowledge
of ‘how things work’ locally. Women who migrate also leave behind a care deficit
(Chen et al. 2005: 28–30).
The access of foreign workers to social protection through work has become a
complex contemporary policy issue. For more formal workers, the access to or
exclusion from entitlements is at least transparent: a Turkish migrant who works
in a formal firm in Germany for a number of years will get access to social benefits
while there, and may be able to transport some benefits back to his or her family.
However, for millions of migrants, work comes with no social protection, and may
be unregistered. Workers are unorganized in labour terms (though they may have
vibrant informal networks and associations), and are too dependent on the job to
be able to express demands. They have migrated precisely to work for an income,
and that income may be higher than they could ever earn in their home country.
Commercial sex workers

Statistics on the extent of sex work and the incomes earned by sex workers are dif-
ficult to obtain. In some countries it is not allowed as ‘occupational category’ in
labour-force surveys; in others, data is collected but is hidden in a category which
includes other occupations (such as beggars), and cannot be disaggregated. Some
sex workers are highly paid and work autonomously as self-employed persons.
However, most female and male sex workers work for someone else, and are
dependent on them in complex ways.
Sex workers should not be seen only as ‘victims’, and the occupation needs to be
understood from the workers’ perspective (e.g. Bandyopadhyay et al. 2004). Some
men and some women choose to enter sex work, albeit because there are limited
options, or because incomes are high comparative to other occupations. It remains
the case that men and women sex workers are highly vulnerable to specific risks:
to assaults on themselves; to illness and disease; and to involvement with the
drug trade. They are also likely to stop working earlier than people in other occu-
pational groups, and their low social status may mean that they have no secure
place in a family or community as they get older.
Domestic workers
Worldwide, most domestic workers are women. They may live away from their
own homes, and in the homes of their employers (especially likely if they are
cross-border migrants). Working conditions are characterized by long hours, low
pay, and lack of autonomy. Working in someone else’s private domain renders
them both invisible and dependent on the employer. There is usually a large class
difference between employer and employee, and this may be accompanied by a
language difference as well, making communication difficult.
The relationship between (woman) employer and (woman) domestic worker is
complex. As women, they have shared interests in, for example, child development
and education. The career prospects of the working woman employer are contin-
gent on the (typically poorly paid) work of the domestic; the employer’s children’s
day-to-day security depends on the employee, who has to be away from her own
children to do this work. While the employer–employee relationship offers the
worker little autonomy or voice, there may be informal measures of assistance
such as with health costs, or school fees of children. However these are not con-
tractual obligations, and as such depend on the whim of the employer, and this is
not a sound basis for building long-term security.
Unpaid homebased carers

The health and welfare care work done in hospitals, clinics, old age homes and
welfare institutions, which is subsidized by the state or paid for in the private
sector, is one small fraction of all the caring work that has to be undertaken in a
224 Francie Lund
society. Health and welfare policies are premised on some model of the family, typ-
ically one in which there is an infinite supply of women’s caring time. Decisions
to allocate more, or less, to the development of health and welfare services will
have direct effects on the time that has to be spent in families, and in communi-
ties, in unpaid care work. This work is overwhelmingly done by women, and is not
classified as work or employment in the systems of national accounts.
Concepts and practices such as ‘community care’ and ‘home-based care’ conceal
much of this work; it is people living in families and neighbourhoods who do the
caring. There can of course be a positive value-base to community care – people in
need of care may live a qualitatively better life when cared for at home and in the
community, as opposed to in an institution such as a hospital or home for elderly
people. Many people get deep emotional satisfaction from the caring role.
However, community care is too often introduced in a cost-cutting environment
as part of a neoliberal regime which also carries conservative messages about the
primacy of the family, and women’s place in the home.
AIDS is having a catastrophic effect in many countries of the world, and where
resources for formal health and welfare services are scarce, there will be increasing
calls for home care and community care. Though AIDS is no respecter of social
class, in developing countries the burden of AIDS falls most heavily on poorer peo-
ple, and in sub-Saharan African countries, on poorer women. It is contradictory for
international agencies and national governments to say that they are pursuing
gender-sensitive policies, and pro-poor policies, at the same time as they advocate
cutbacks in social spending, privatization of health and welfare services, and
community-based care.
Innovations in social-protection provision
In some parts of the world, labour-protection and social-protection mechanisms

have been extended to formerly excluded categories of workers. In Argentina,
domestic workers have been included in unemployment insurance coverage. In
Chile, statutory health insurance which had been available only to full-time workers
was extended to include the (predominantly women) temporeras – people working
in the horticulture industry (Barrientos and Ware Barrientos 2003). In South
Africa, agricultural workers have been integrated into the occupational health and
safety regime, through the Compensation for Industrial Injuries and Diseases Act
of 1994. Also in South Africa, domestic workers (nearly all of whom are women)
and agricultural workers (about four-fifths of whom are men) are now covered by
the Unemployment Insurance Fund. Together these two occupational categories
comprised 17 per cent of all measured employment, making this a significant
extension of coverage.
If social policies are to be devised which address both poverty and gender
inequality, the extension of social protection will need to reach poorer working
women, in large numbers. This is a difficult task, the challenges are daunting, and
political will on its own is not enough. Many of the barriers are administrative,
with the challenge of reaching people who are often working in concealed places,
with low and erratic incomes. In many countries, the implementation of labour
laws on basic conditions of employment and social protection is not properly
monitored even for those in formal employment. Bearing these difficulties in
mind, however, there are examples of innovation which show that progress can be
made. The first example below comes from India, and describes an integrated
insurance scheme, built by poor women, for poor women and their families, in
partnership with government and with external support. The second comes from
Costa Rica, which designed a voluntary scheme for independent and unremuner-
ated workers, which over time will become compulsory for independent workers.
The final example is the South African elderly pension, which provides basic
security for millions of older and poorer men and women, and assists also the
members of the households in which they live.
Comprehensive insurance built by women informal workers: SEWA’s

integrated social insurance1
There has been a great deal of experimentation with social insurance in general,
and health insurance in particular. Many attempts have been made to reach
women, and especially poorer women. These schemes have provided access to
many people to insurance for the first time. Many experience problems of sustain-
ability, and of expanding membership numbers. Few have focused specifically on
reaching women in their status primarily as informal workers (but see von
Ginnekin 1999 for documentation of schemes for informal workers in developing
countries).
One good example is the insurance scheme, Vimo SEWA, that has been built
over twenty years by the Self Employed Women’s Association (SEWA) in India.
SEWA is a trade union and set of cooperatives for women who work informally. It
was started some thirty years ago in the city of Ahmedabad, and by 2003, there
were more than 700,000 paid up members. SEWA as a whole focuses on women,
and on their earnings. SEWA defines social security as ‘at least health care, child care,
insurance, and shelter’ (Chatterjee and Ranson 2003: 281). One of its central ser-
vices is the integrated social insurance, Vimo SEWA.
SEWA had set up its own bank for savings and credit, and learned that a major
reason for people not repaying loans was sickness. It responded by starting a pri-
mary healthcare programme, and then in 1992 Vimo SEWA was set up to comple-
ment the healthcare programme. It includes life insurance, asset insurance and
health insurance. Chatterjee and Ranson (2003) state that by 2003 the insurance
scheme had over 102,000 members, in urban and rural areas. There was a marked
increase in membership in 2001 following the Gujarat earthquake.
Joining requirements are being a member of SEWA, having a SEWA bank
account, and subscribing to all three parts of the package. In 2003 the annual
contribution was 85 rupees (of which 22.5 rupees was for health insurance). An
interesting design feature is that members choose to pay either on this annual
basis, or with a once-off larger deposit of 1,000 rupees, invested in SEWA Bank. The
interest on this deposit then pays their annual contribution up to the age of 60
(the upper age limit of the scheme), at which point the amount of the deposit is
226 Francie Lund
repaid to the member. Those paying the fixed deposit receive a wider package of
benefits, in particular access to a maternity grant.
The life insurance covers natural and accidental deaths of the member (and
more recently her husband). Asset insurance covers losses experienced by the
woman due to fire, floods and riots, and reflects SEWA’s focus on responding to
disasters in such a way that members are able to restore their income-earning abil-
ity as fast as possible. The health component includes hospitalization up to a max-
imum amount per year; those paying premiums through the fixed deposit get
additional coverage for cataract operations, hearing aids and dentures. The scheme
allows members to choose their own healthcare provider, private or public, though
SEWA gives advice about reliable providers.
Husbands can get access to limited benefits; they receive the insurance against
wives’ natural and accidental death, and women get coverage for their husbands’
death. Children have recently been included in the scheme as well. The SEWA
member, though, retains sole control of the SEWA Bank account through which
these benefits operate.
Vimo SEWA is a most interesting partnership between government, the insur-
ance industry (which until recently was government-run) and SEWA. Essentially,
the Indian government subsidized two large insurance corporations to offer some
of their services to poor people, and SEWA was one of a few large NGOs or trade
unions to take advantage of this opportunity. Vimo SEWA started in a partnership
with government insurance; SEWA then took the health scheme over themselves
for a while; they have recently again entered a partnership in which they purchase
insurance from a government general insurance company, but SEWA keeps control
of enrolment, approval and processing.
SEWA attributes some of the success of the insurance to the interaction between
its different programmes: the SEWA Bank, into which annual premiums are paid;
the health education by SEWA Health, which heightens members’ awareness of
health problems; and the literacy training provided by SEWA Academy, which
enables members to interact with the literate world of formal finance. It also ben-
efited from a large grant from Germany’s GTZ which allowed it to secure a finan-
cial base, cover administrative expenses, and also to innovate and experiment.
SEWA’s documentation of the scheme shows a real alertness to emerging problems,
and a process of continual innovation and experimentation, in response both to
objectively changing risks, and to needs expressed by members. As an ILO study
put it, it is a story of ‘a process of continuous adjustment sustained by great
creativity and tenacity’ (ILO 2001: 79).
A government-initiated voluntary scheme for health and pension

provision for informal workers: an innovation in Costa Rica2
In Central America, Costa Rica provides an unusual example of a government ini-
tiating a new insurance programme which aims to draw into the social-security
umbrella informal workers who do not have access to a formal occupationally
related provision. This is at present a voluntary scheme, covering access to health-
care, and to a pensions savings scheme.
Costa Rica’s population of some four million is about 50 per cent urbanized, and
it scores high in the region in terms of its per capita income, which was $3,960 in
2000. It has a long history of extensive social-security coverage for its relatively
small and homogeneous population. Yet demographic changes and changes in the
labour-market are presenting new challenges to social protection. The last two
decades have seen a rapid increase in women’s labour market participation, but on
poor terms relative to men. Most women work in the services sector where the
gender earnings gap is pronounced; for those with little education, the 1999 aver-
age wage of urban women was 59 per cent of the average wage of urban men. The
unemployment rate in 1999 was 6 per cent, much lower for men (4.9 per cent)
than for women (8.2 per cent) – but it should be noted how low these unemploy-
ment rates are compared to many low-and middle-income countries.
An increasing proportion of the Costa Rican labour force is not covered by occu-
pationally related social insurance, partly because of the increase in numbers
working in the informal economy. Some transnational firms and factories have
imposed labour practices which are illegal in the country: it is common knowledge
that people work for less than the minimum wage and for illegally long hours, and
that the right of women to breastfeed during working hours is not respected.
The World Bank tried to introduce pension reforms in 1994. The unpopularity
of these reforms led to widespread unrest, and a ‘Forum of National Concertation’
was then set up to work towards consensus about restructuring the social benefit
system. One of the components of the new agreement was the introduction of a
voluntary insurance scheme. This is available for independent workers – own-
account workers and non-remunerated workers, such as family workers and house-
wives, as well as students. This income-tested scheme is aimed at those who have
never contributed to a health or pension plan, or who did so for too short a period
or at too low rates to accumulate adequate benefits. Introduced as a voluntary
scheme, it is to be made statutory by 2005, when independent workers will be
required to join. It is funded by contributions of the state and those who join, with
the state contributing 0.25 per cent and the independent worker contributing
7.25 per cent of reference income.
By 2003 it was estimated that nearly three-quarters (74 per cent) of indepen-
dent workers contributed to the health insurance, while nearly one-quarter
(24 per cent) contributed to the pension insurance. It is remarkable that so many
independent workers are reached by the health insurance component, and it
makes the comparison with the much lower numbers opting for an old-age bene-
fit very striking. Probably, people opt not to join because old age seems far away,
there are conflicting immediate demands on available resources, and the returns
to present contributions seem uncertain. In addition, there is a disincentive to
joining because there is a non-contributory pension scheme for poor adult Costa
Ricans who are not covered by another regime.
This extension of social protection is an interesting example where a country
with a good history of social provision is attempting to adjust to changes in the
labour market, as well as to demographic changes, in flexible ways. It is trying to
set up seamless links between contributory and non-contributory schemes. It does
228 Francie Lund
not yet, however, address social protection for those who are waged workers but
who are not properly covered.
The South African pension for elderly people

South Africa’s system of state social assistance was originally set up to protect the
interests of poor white people, and then was expanded under the former apartheid
government to include the whole population (who were classified into four ‘racial’
groups – African, coloured, Indian and white – by that government). One compo-
nent is for elderly people, and women at age 60 and men at age 65 become eligi-
ble to receive a monthly Old Age Pension (OAP) from the state, if they qualify in
terms of an income-based means test. It is a non-contributory scheme, payable
from general revenue, and it has become recognized as making a distinctive con-
tribution to poverty alleviation – both for pensioners themselves, and for people
in the households in which they live.
The majority of poor older people in South Africa, and especially in rural areas,
live in three-generational extended families. The pension is applied for by an indi-
vidual elderly person; when the pension is received there is extensive income
pooling of the OAP, where the whole or a large part of the pension amount con-
tributes to the common household purse. A growing body of research documents
the positive impact of this system (Ardington and Lund 1995; Case and Deaton
1998; Case 2001; Lund 2002):
● it is well targeted to poorer elderly people;

● it is well-targeted in racial terms (for example it reaches 80% of the African
population, most of whom are poor, and an insignificant number of the wealthier
white population);
● it is well targeted to rural areas;
● it is well targeted to women, because they live longer, draw the pension earlier,
and are poorer;
● it is valued for its reliability;
● it contributes to the security of the households in which elderly people live;
● it contributes to the production of livelihoods of elderly people themselves, and
of other younger family members.
Informal men and women workers thus effectively have a guarantee of partial
economic security in their elderly years, affording them an earned place in the
household.
There are of course problems with any system of this nature. Some are con-
cerned that a benefit targeted to one group – the elderly – is consumed by the
household at large. There is some evidence that it may reduce the amount of
remittances sent from migrant household members. Pensioners themselves com-
plain about administrative problems, particularly with delays in receiving pay-
ments, and with the review system (eligibility has to be reassessed from time to
time, to check whether people are still poor enough to receive the pension, though
escaping poverty in the elderly years is highly unlikely). There are relatively minor
concerns about leakage in the system to the better-off. Generally though the sys-
tem delivers pensions monthly to between 70 and 80 per cent of all elderly South
Africans, at an amount of approximately US$3 a day at present levels. Pension-
receiving households (which are larger than non-receiving households) are still
poor, and households in the very poorest income deciles are there because they
have no elderly person, and no disabled person (there is a similar social assistance
system for people with disabilities).
The numbers of people in the older years is a small fraction of the overall popu-
lation, and the OAP has been judged to be sustainable and affordable. The AIDS
epidemic has now reduced longevity drastically, and in future proportionally
fewer of the population will reach eligible age. Older people have already taken on
more responsibilities for caring for children whose parents have died of AIDS. The
OAP will continue to make a vital contribution to household security.
Learning from innovation
As noted, building large-scale systems of social protection that reliably reach

informal workers with low incomes is a daunting challenge. The SEWA integrated
insurance scheme provides convincing evidence that investing in the poor,
through social insurance, can be successful and sustainable. The grassroots trade
union of informal working women has ‘gone to scale’, to the point where it pro-
vides a comprehensive package of social insurance benefits to over 100,000 members.
It has built partnerships with government and the insurance industry on terms
favourable to itself, and left those partnerships when they became too problematic
or inflexible. It has been inspiring in its ability to continuously respond to
members’ needs.
The development of Vimo SEWA shows the importance of flexibility and
responsiveness to members’ needs. The attempt to be flexible is evident also in the
actions of the Costa Rican government. As the voluntary insurance scheme was
introduced relatively recently, it is too early to assess its performance. It has to be
accepted that there will be administrative costs involved in extending social pro-
tection to informal or more dispersed workers. Costa Rica, with its small popula-
tion, grafted the voluntary insurance onto an existing administrative system for
delivering social security. SEWA’s solidarity and unity have meant that a large part
of the administrative work is done by the members themselves. Perhaps the fact
must be faced that, where there is no SEWA or similar organization, ‘sustainability’
may need to include an element of long-term subsidy. Chatterjee and Ranson ten-
tatively suggest a direct transfer of resources to community-based health insurance
schemes, ‘perhaps earmarked for administrative costs or marketing and promoting
the scheme among the poor’ (Chatterjee and Ranson 2003: 296).
The Costa Rican case suggests that informal workers find it easier or more
preferable to insure against ill health than to commit resources to savings for their
elderly years, though in this particular case there was a clear disincentive to join
the old-age-pension scheme because of the existence of a non-contributory
pension scheme. ‘Adverse selection’ is one of the common problems of insurance
230 Francie Lund
systems. SEWA found that an obstacle in its health insurance is that ‘it has
attracted a membership that is older and more likely to experience illness’ than the
general population of SEWA members (Chatterjee and Ranson 2003: 294).
Chatterjee and Ranson (2003: 294) identify another lesson of Vimo SEWA:
that improving access to health care may be ‘a mixed blessing where the qual-
ity of care is variable and, at worst, dangerous’. Where poorer people can only
get access to poor quality health services, and facilities in which for example
there may be little in the way of reproductive health services for women and for
men, then it is not surprising that people may avoid going to those health
services.
None of the case studies focused on employer contributions, and there is a need
for further exploration as to how to secure employer participation in cases where
there is an employer–employee relationship. India has a system of welfare funds
for workers in certain economic sectors, such as tobacco, building and construc-
tion, the film industry, certain mining sectors including limestone and dolomite,
and iron ore, manganese ore and chrome ore. Levies or cesses are placed on the
production or sale of goods in these sectors, and the welfare funds are used for col-
lective benefits in housing and education for workers and their children
(Subrahmanya 2000: 70). This overcomes the problem of identifying and securing
contributions from individual employers. Subrahmanya draws attention, though,
to significant problems, among which are administrative inefficiency, under
spending of the funds, and the fact that in general they do not see to the social
security needs of individual workers.
A number of examples show the importance of the role of the state in being
able to deliver to large numbers of people, through existing and new institu-
tions. This is clear in the South African OAP and the Costa Rican scheme and
points to the need for greater acknowledgement of this role, and for seeking
ways of making state interventions even more effective. Thailand has recently
introduced the ‘30 baht’ health insurance scheme which is available to informal
workers (as it is to all Thai citizens). India has agreed to pilot a new scheme of
social security specifically for ‘the unorganized sector’ or informal workers. Both
of these new programmes should be monitored closely; both carry the potential
to reach previously uncovered categories of worker, but possibly not the very
poorest.
None of the case study examples involving informal workers addressed the issue
of AIDS directly. In countries with a high AIDS prevalence, the epidemic is eroding
social protection for all people. Formal insurance companies are raising premiums
for policy-holders. The actuarial basis of smaller and informal savings and insur-
ance mechanisms is affected, as members themselves become sick, or have to use
their savings and insurance on health and funeral costs for family members.
Healthcare schemes are capping the value or number of services. Even in countries
with relatively sophisticated health services, people in the advanced stages of AIDS
are sent home from hospital to die, as hospital wards are full and priority is given
to assisting those who are not yet in the terminal stage. This increases the need for
home-based care.
Conclusion
The world is facing a growing crisis with the problem of the increasing insecurity
of poorer workers. Millions of people work all their lives yet remain poor, and
whole age cohorts are entering their elderly years with no savings for retirement.
This crisis will not be solved by patchwork responses. Some of the policy pres-
criptions that are put forward to deal with this problem are destined to fail,
because they do not take into account the reality that poorer people face greater
risks, through their poverty and through the kinds of work that they do. Nor do
they take into account that women face specific risks of their own. In line with the
conclusion of Benería and Floro (Chapter 9 of this volume), this chapter points to
the need for substantial and sustained state intervention, and a role for significant
donor funding as well.
Examples have been given of innovations in social protection that have reached
large numbers of people, including large numbers of women. They show that the
state has a constructive role to play in extending existing measures of social pro-
tection to more categories of worker, and also in working in partnership with orga-
nizations of informal workers. Through the SEWA example, we see an organization
that both promotes self-reliance and advocates an active role for the state. Employ-
ers and owners of capital are, however, uniformly hard to identify and to reach.
It has been argued that the informal sector and workers within it are in fact cov-
ered by some aspects of labour standards (Trebilcock 2004). In addition to the call
for extension of labour legislation, though, a multi-pronged strategy is needed
which considers different conditions governing different industrial sectors, the
institutional capacity to implement labour codes, and the potential effectiveness
of different kinds of alliances between different interest groups, at different times.
In addition, there is a need to understand how both international labour codes
and national labour legislation can be mutually reinforcing, bearing in mind that
some of the fair-trade initiatives have voluntary codes of conduct which have in
fact fewer protections for workers than existing national labour legislation. A core
strategic question is: what combinations of local, national, regional and supra-
national alliances and interventions would give access to people in different
employment relationships?
The access of informal workers to better (or any) measures of social protection is
contingent on their building strong and effective organizations. Local-level orga-
nizations are the vehicle through which women informal workers especially can
practise empowerment in day-to-day ways. Organizations are also the vehicle
through which to express their voice and interests. Building organizations of
informal workers is a long struggle, and social-protection benefits can be a focus
around which to organize. Informal workers need to find institutional spaces
where they can directly represent their own interests, and influence the way poli-
cies are formulated and implemented.3 Homenet Thailand, an organization pro-
moting the interests of industrial outworkers in the garment industry, worked
actively in the Thai Parliamentary Commission on Universal Health Care Coverage,
which resulted in the introduction in 2002 of the ‘30 baht’ universal healthcare
232 Francie Lund
scheme. India’s SEWA has influenced a number of significant policy platforms in

India, on child labour, insurance, disaster mitigation and local government among
others. In the Philippines, the organization of homeworkers, PATAMABA, partici-
pates in the government’s Anti-Poverty Council, where it speaks specifically for
the interests of poorer working women. In South Africa, the Self Employed
Women’s Union engaged in a number of national policy platforms in the transi-
tion to democracy, including commissions on the labour market, rural financial
services, and trade and industry policy.
Women’s work in gendered labour-market institutions needs to be placed more
centrally in social policy. Scholarship on reproductive health and on caring should
better incorporate a working-woman lens; those concerned with women in their
role as workers likewise should continue to analyse the specific difficulties that
women have in reconciling their caring and their income-earning roles. And, as
Molyneux (Chapter 2 of this volume) shows, with respect to the Oportunidades
scheme in Mexico, the costs borne by women through their participation in con-
ditional entitlement schemes need to be made more visible. Gendered policy
analysis also needs to incorporate the differentiation between women in different
classes. The focus on women in precarious employment offers a fertile ground for
learning about how class, space, race, ethnicity, and gender interact to reproduce
inequality in society, and provides pointers to new forms of mobilization, new
organizational forms and alliances in the struggle for better conditions for work-
ing people.
Notes
1 This case study draws primarily from Chatterjee and Ranson (2003) and ILO (2001).
2 This case study relies exclusively on Martinez Franzoni and Mesa-Lago (2003).
3 There have been good examples in the last few years; for a comprehensive account of
examples of good practice, see Chen et al. (2005): 67–77.
References
Ardington, Elisabeth and Frances Lund (1995) ‘Pensions and Development: Social Security as
Complementary to Programmes of Reconstruction and Development’. Development
Southern Africa, Vol. 12, No. 4: 557–77.
Bandyopadhyay, Nandinee, with Swapna Gayen, Rama Debnath, Kajol Bose, Sikha Das,
Geeta Das, M. Das, Manju Biswas, Pushpa Sarkar, Putul Singh, Rashoba Bibi, Rekha Mitra
and Sudipta Biswas (2004) ‘ “Street Walkers Show the Way”: Reframing the Debate on
Trafficking from the Sex Workers’ Perspective’. IDS Bulletin, Vol. 35, No. 4: 104–11.
Barrientos, Armando and Stephanie Ware Barrientos (2003) ‘Social Protection for Workers in
the Horticulture Industry’. In Frances Lund and Jillian Nicholson (eds) Chains of
Production, Ladders of Protection: Social Protection for Workers in the Informal Economy.
Durban: School of Development Studies, University of Natal.
Case, Anne (2001) ‘Health, Income and Economic Development’. Paper presented at the
ABCDE Conference, World Bank, Washington DC, May.
Case, Anne and Angus Deaton (1998) ‘Large Cash Transfers to the Elderly in South Africa’.
Economic Journal, Vol. 108: 1330–61.
Chatterjee, Mirai and M. Kent Ranson (2003) ‘Livelihood Security through Community-
based Health Insurance in India’. In Lincoln Chen, Jennifer Leaning and Vasant
Narasimhan (eds) Global Health Challenges to Human Security. Cambridge MA: Global
Equity Initiative, Asia Center, Harvard University.
Chen, Martha (1998) Widows in India: Social Neglect and Public Action. New Delhi: Sage.
Chen, Martha (2000) Perpetual Mourning: Widowhood in Rural India. New Delhi: Oxford
University Press.
Chen, Martha, Joann Vanek and Marilyn Carr (2004) Mainstreaming Informal Employment and
Gender in Poverty Reduction: A Handbook for Policy-makers and Other Stakeholders. London:
Commonwealth Secretariat.
Chen, Martha, Joann Vanek, Frances Lund and James Heintz, with Chris Bonner and Renana
Jhabvala (2005) Progress of the World’s Women 2005: Women, Work and Poverty. New York:
UNIFEM.
Doane, Donna, Rosalinda Ofreneo and Paonoi Srikajon (2003) ‘Social Protection for Workers
in the Garment Industry’. In Frances Lund and Jillian Nicholson (eds), Chains of Production,
Ladders of Protection: Social Protection for Workers in the Informal Economy. Durban: School of
Development Studies, University of Natal.
Heintz, James (2005) ‘The Growing Informality of Labour Markets’. Paper presented at the
UNRISD/CGGS Conference on Gender and Social Policy, Marstrand, Sweden, 28–29 May.
ILO (2001) Women Organizing for Social Protection: The Self-employed Women’s Association’s
Integrated Insurance Scheme, India. Geneva: ILO
ILO (2002) Women and Men in the Informal Economy: A Statistical Picture. Geneva: ILO.
Kofman, Eleonore (2004) ‘Gendered Global Migrations’. International Feminist Journal of
Politics, Vol. 6, No. 4.
Lund, Frances (2002) ‘ “Crowding in” Care, Security and Micro-enterprise Formation:
Revisiting the Role of the State in Poverty Reduction, and in Development’. Journal of
International Development, Vol. 14, No. 6: 681–94.
Lund, Frances and Jillian Nicholson (eds) (2003) Chains of Production, Ladders of Protection:
Social Protection for Workers in the Informal Economy. Durban: School of Development
Studies, University of Natal.
Lund, Frances and Smita Srinivas (2000) Learning from Experience: A Gendered Approach to
Social Protection for Workers in the Informal Economy. Geneva: ILO.
Martínez Franzoni, Juliana y Carmelo Mesa-Lago (2003) La reforma de la seguridad social en
costa rica en pensiones y salud: Avances, problemas pendientes y recomendaciones. San José,
Costa Rica: Fundación Friedrich Ebert.
Orloff, S (2000) ‘The Significance of Changing Gender Relations and Family Forms for
Systems of Social Protection.’ Background paper prepared for the World Labour Report
1999–2000. Geneva: ILO.
Subrahmanya, R. K. A (2000) ‘Welfare Funds: An Indian Model for Workers in the
Unorganised Sector.’ In Renana Jhabvala and R. K. A Subrahmanya (eds) The Unorganised
Sector: Work Security and Social Protection. New Delhi, Thousand Oaks and London: Sage.
Trebilcock, Anne (2004) International Labour Standards and the Informal Economy. Geneva: ILO.
UNRISD (2005) Gender Equality: Striving for Justice in an Unequal World. Geneva: UNRISD.
Von Ginnekin, Wouter (ed.) (1999) Social Security for the Excluded Majority. Geneva: ILO.
Vosko, Leah (2005) ‘Precarious Employment: Towards an Improved Understanding of Labour
Market Insecurity’. In Leah Vosko (ed.) Precarious Employment: Understanding Labour Market
Insecurity in Canada. Montreal: McGill-Queens University Press.
Part III
Gender Dimensions of
Social-sector Restructuring
11
Gender and Health Sector Reform:
Analytical Perspectives on African
Experience
Maureen Mackintosh and Paula Tibandebage
Introduction: gender, the ‘silent term’ in health sector reform
Vignette 1
A small private dispensary in the Tanzanian Southern Highlands was run in 1998/9 on a
day-to-day basis by a trained nurse, a woman, and overseen by her husband, a medical
doctor, who was not often there since he was employed in a mission-run hospital. The
nurse saw patients, managed other staff, and did the books; unusually in a private dis-
pensary she provided some preventative care such as vaccinations and ante-natal care.
Most of the clients we met were women, many bringing young children or other sick
relatives.
Health systems are gendered institutions. That is, their organization reflects and
responds to gender inequalities in the wider society. The hierarchy among health care
staff places doctors, policy-makers and administrators – predominantly male –
above nurses, paramedical staff and orderlies who are more likely to be female.
Furthermore, in Africa as in many parts of the world, day-to-day working relations
in health care are predominantly between women: nurses and lower-level staff
dealing with women accompanying sick children and the elderly, women caring
for relatives who are in-patients, or coming for care themselves (Vignette 1).
Furthermore, mother and child health care (MCH) is known to be important to
society’s health outcomes (WHO 2003a, LaFond 1995).
We would therefore expect, when we turn to the prescriptive and analytical lit-
erature on the health-sector reforms that have been proposed and implemented
across sub-Saharan Africa, that gender would be a key tool of analysis, and that the
literature would explain how the reforms address women’s needs.
Yet this is not so. The influential multilateral policy documents of the 1990s are
largely silent on gender. The World Development Report 1993: Investing in Health
(World Bank 1993) mentions ante-natal, delivery and post-natal care and family
237
238 Maureen Mackintosh and Paula Tibandebage
planning as ‘clinical interventions’ that are ‘highly cost effective’; and the
importance of schooling for girls, of ‘educated mothers’, and of women’s access to
income. It discusses the gendered pattern of disease burden and the problem of
insensitivity of health care to women’s broader health needs. Yet the moment we
turn to proposed reforms gender largely disappears. Discussion of the core policy
recommendation to introduce user fees, and the sections on insurance, market
liberalization and structural reform do not refer to gender differences in need or
ability to pay.
Similarly, the World Development Report 1997: The State in a Changing World
(World Bank 1997) outlines ‘new public management’ reforms without reference
to differential effects on men and women. The World Development Report 2004:
Making Services Work for Poor People (World Bank 2004) discusses clients and
providers, citizenship and health services in a largely ungendered manner, exam-
ining ‘co-production’ of health between communities and health sector staff with-
out recognition of its reliance on low-income women’s unpaid work. The silence
also runs through the Millennium Goals agreements, which identify gender equity
as an objective but are mute on the institutional changes that might be needed to
eradicate gendered inequity from key public-service institutions.
The silence has now been broken however by the World Health Organization, in
the World Health Report 2003: Shaping the Future (WHO 2003a) which identifies
gender discrimination in the health professions as a serious problem for the deliv-
ery of services to poor and disadvantaged populations. The report also promotes
women’s participation in health care governance and management.
This chapter argues that despite the silences, health sector reform models are
implicitly gendered: that is, they have gender built inaudibly into their assump-
tions. Section 2 explores analytical approaches to the gendered nature and impacts
of health sector reform: gender equity, women’s health needs and gendered health
systems frameworks. Our objective is conceptual: to examine health-sector reform
through a gender ‘lens’, considering how the reform framework is gendered and
the extent to which that gendered process may operate to the detriment of
women, especially poor women (Sections 3–6). We illustrate our arguments with
empirical evidence drawn largely but not exclusively from Africa. Section 6
considers implications of the analysis particularly for the African context.
Gender perspectives on health, health systems and reform
Health sector reform in the African context

Health sector reform (HSR) is not easy to analyse because the phrase is used to refer
to two distinct things. First, it refers to the processes of institutional change that
have swept through African, as many other, health systems since the early 1980s.
And second, it refers to the prescriptive framework for institutional redesign
of health care provision and public health repeatedly proposed in multilateral
reports, donor policies and consultancy reports. The two do not, of course,
coincide. Health systems are complex institutions, deeply influenced by cultural
ideas about health and illness, by historical experience and by social structure.
Gender and Health Sector Reform 239
Health sector reform models are themselves not without internal contradictions,
and contain gendered assumptions that may be problematic.
Health sector reform as promoted in sub-Saharan Africa has particularly encom-
passed the following (WHO 2000a, Mills et al. 2001):
● liberalization of private clinical provision and pharmaceutical sales, and

promotion of a ‘mix’ of public, private and voluntary providers;
● retreat of government towards a mainly regulatory and priority setting role,
with responsibility for direct provision of some services in public health and for
ensuring access to ‘basic’ care by the poorest;
● user charges for government health services, for government-provided drugs
and supplies, and for community-based health services;
● increased contracting-out of government-funded services to independent
providers;
● decentralization of health systems to local government control;
● increase in autonomy of hospital management and finance; some hospital
privatization;
● a shift towards insurance rather than tax-based financing mechanisms.
We could summarize the model as a shift to greater ‘commodification’ of health

care – that is, its provision as a set of discrete services for market payment or gov-
ernment ‘purchase’ on behalf of citizens – plus reduction of government activity
and more systematic priority-setting in government spending based on cost-effec-
tiveness of interventions.
In sub-Saharan Africa, this HSR model has been promoted in a context of gener-
alized and severe poverty and has often been implemented in the wake of econ-
omic crisis. In Tanzania in the early 1990s user fees were introduced at a time
when cuts in government spending and removal of subsidies on basic goods were
disproportionately affecting the poor. Health care fees are frequently unaffordable
and impoverishing in a context where a significant proportion of the population
cannot afford basic needs including food; in 2002 an estimated 15–18 million
Tanzanians out of a population of 34.4 million were living below a poverty line of
US$0.65 a day (World Bank 2002, 2004).
Gender perspectives on health and HSR

Current literature offers two broad perspectives for assessing the differential
impact of HSRs on women (Standing 1997, 2002a). One is a ‘gender equity’ per-
spective. This applies the general concept of equity in provision of health services
to men and women, asking: Do health systems respond equally to men and
women in equal need? And does HSR (seek to) improve gender equity in this
sense?
The second perspective is the ‘women’s health needs’ perspective. It focuses on
the specific health needs of women, notably (but not only) reproductive health
care needs. Do health systems respond to these needs? Does HSR (seek to) improve
responsiveness to women’s health needs?
We build on these two perspectives in order to evaluate HSR as a gendered

process. As Standing (2002a, 2002b) points out, many aspects of health care and
health sector reform are marked by sharp gender differences, including the role of
women as workers in health care. In this chapter we develop this understanding of
HSR as gendered institutional change by drawing on the broader literature on gen-
der and health and also on gendered analyses of high-income countries’ health
and welfare systems.
We take from the gender and health literature the concept of a ‘gender lens’.
‘Gender’ here refers to structural inequalities and differences between men and
women, and associated sets of behaviours, expectations and roles for men and
women. Concepts of health and illness and the ideas that shape health-seeking
behaviour are shown to be gendered. Gender ‘permeates social institutions’ and is
an ‘organising principle of social life’ (G. Sen et al. 2002a: 6).
This gender and health literature raises, but does not explore in depth, the way
in which the institutions of health systems are permeated by gender. However,
feminist writing on rich-country health and welfare systems centrally addresses
this issue, and illustrates the type of conceptual question that can be addressed in
other contexts ( J. Lewis 2002; Williams 1989; G. Lewis 1998; Lister 2000):
● What are the assumptions about household and family structure that underlie
health sector reform and are they justified? How do health financing strategies
respond to changing social understandings and experience of women’s finan-
cial dependence or autonomy within the household?
● What are the assumptions about women’s labour market participation – including
participation in the health-sector labour force? How, for example, do wider
work opportunities for women influence the nursing labour market?
● What are the assumptions about women’s capacity to undertake unpaid care? Does
health care reform make unjustified assumptions about women’s availability to
care for sick children and the elderly?
● How do changes in health care funding and expenditure differentially affect
women and men in different social groups? What are the gender-differentiated
effects of changes in the pattern of public expenditure on health care, the struc-
ture of the tax system, and the balance between public and private health
finance?
● To what extent do health and social services influence women’s circumstances as
well as responding to them? How do these services (re)construct the nature of
women’s citizenship and women’s rights, and how these are differentiated
according to income, social class, ethnicity or location?
● What are the assumptions about health sector governance in reformed frame-
works: who will participate and how? How can users from different social
groups exert leverage over resources through governance structures?
This literature analyses health systems as themselves ‘gendered’ in the sense

outlined above. A health system as it has built up historically is expressive of the
gender (and other) hierarchies of the wider society; and the health system feeds
back into social patterns of disadvantage. The standards and principles of evalua-
tion of health systems and the reform models themselves reflect patterns of estab-
lished disadvantage including gender hierarchies (Twigg 2000; Mackintosh 1998).
Conversely, health systems’ immense political and social importance, and the
association of health care reform with political and economic crisis, implies that
health systems can be sites of challenge to gender disadvantage. We seek in this
chapter to contribute to a relevant set of analytical questions about gender and
health sector reform for the African context.
Health finance and gender disadvantage
Key questions:
How do changes in health care funding affect differentially women and men in
different social groups?
What are the underlying assumptions in HSR models about women’s access to
cash and are they correct?
Health sector reform in Africa brought four linked health-financing changes:

the introduction of user fees for access to government health services; liberal-
ization and promotion of non-government fee-for-service provision; experi-
mentation with health insurance schemes; and pressure to target public
expenditure on primary care for the poorest. We argue that the key change, to
more extensive and explicit reliance on private payment, is likely to have dis-
proportionately disadvantaged poor women. Yet the gendered impact of these
payment systems on access to affordable, good-quality health care remains
largely unresearched.
User fees and out-of-pocket payment

HSR in Africa has been an important driver of ‘informal commercialization’ of
health systems, in which small-scale unregulated provision coexists with formal
government user-fee systems and informal payments (Mackintosh and Koivusalo
2005a; Tibandebage et al. 2001). User fees in low-income government health care
are generally unaccompanied by an effective exemption system and cause wide-
spread exclusion of those unable to pay (Karanja et al. 1995; Booth et al. 1995;
Gilson 1997; Tibandebage and Mackintosh 2002).
There is some evidence that women have suffered disproportionately from fee-
based care. In a number of countries including Nigeria, Zimbabwe and Tanzania,
the introduction of user fees was associated with a decline in utilization of repro-
ductive health care: falling admissions of pregnant women, increased maternal
deaths and morbidity rates among delivering mothers and their babies, and
decline in the use of Maternal and Child Health (MCH) services were recorded
(Standing 2002a; Ekwempu et al. 1990; Kutzin 1995; Walraven 1996). Standing
(2002c: 17) notes that ‘anecdotally expenditure on health care for women and
girls in some contexts is even more sensitive to price changes’ than men’s expen-
diture; in Nairobi the introduction of user fees at an STD clinic caused a 56 per cent
drop in attendance by women (over a nine-month period) and a 40 per cent

decrease for men (Results 2004).
Health insurance schemes

In Uganda, fees in primary care have now been abolished, and in Ghana a fully
national health insurance scheme is under construction. However a more com-
mon policy response to the exclusionary effects of fees has been the promotion of
social and mutual health insurance.
In the mid-1990s, eight sub-Saharan African countries including Kenya had
some form of social health insurance (SHI), and five were considering it (Ethiopia,
Ghana, Nigeria, Mozambique and Zimbabwe) (Gilson and Mills 1995). Tanzania
introduced a social insurance scheme, the National Health Insurance Fund (NHIF)
in 2001. Social health insurance (SHI) covers only workers in formal employment,
generally a small and relatively advantaged part of the population in low-income
countries.
Some women will be covered in SHI as dependants of employed men, but as
income earners women are relatively disadvantaged: they are less likely than men
to be in formal-sector employment, and if formally employed tend to be concen-
trated in low status poorly paid occupations or lower-level positions.
Unfortunately data on wage employment in Africa rarely distinguish formal from
informal wage employment by gender. Women are highly active economically,
but are much less likely than men to receive a wage or salary for their work
(Table 11.1).
Community-based Mutual Health Insurance (MHI), by contrast, aims to pro-
mote inclusion of the poor and other vulnerable groups. Schemes have been
designed with features to promote inclusion such as paying premiums in instal-
ments, provision for exemption of those unable to pay premiums, and payment
during harvest time for those with unsteady seasonal income.
Table 11.1 Selected African countries: percentage of the labour force that is female, and
percentage of the male and female labour force receiving wages or salaries (most recent
year available)
% labour % male labour % female labour

Country Year force female force waged force waged
Ghana 1990 51 33 9
Ethiopia 2000 41 10 7
Kenya 1995 46 32 12
Nigeria 1995 36 49 33
Tanzania 2000 42 n/a n/a
Uganda 1995 48 21 7
Zimbabwe 2000 45 51 22
Zambia 1995 45 31 9
Source: World Bank GenderStats http://devdata.worldbank.org/genderstats, accessed on 27 April 2004 and

24 April 2005.
However most community-based MHI schemes face a problem of very low par-
ticipation rates. In Tanzania, many rural Community Health Fund (CHF) schemes
were found to have participation rates below 10 per cent of eligible households
(Tibandebage 2004).1 Such low enrolment is not unusual in developing countries;
reasons cited by non-members include inability to pay, lack of transparency and a
decision-making process that does not effectively involve members (Jutting and
Tine 2000; Tibandebage 2004). Implementation of exemption for poor households
has been weak.
The low enrolment may disproportionately affect women since they are generally
less able than men to afford payment at the time of illness, and are likely to have
more health needs. In one rural scheme in Tanzania, members also lacked knowl-
edge of some of the benefits they were entitled to, such as referral to a hospital,
which could be disastrous for women with, for example, complicated pregnancies in
need of hospital-level care but with no out-of-pocket cash (Tibandebage 2004).
Who pays? Assumptions about women’s access to income

HSR changes to payment mechanisms therefore appear to have relied upon
assumptions about control of income within the household that were never, as far
as we can find, made explicit in reform documents. Since women are dispropor-
tionately greater users of services than men, the key assumption seems to be that
they have access to household income and funds borrowed from others to support
these visits, or that they can pay from their own incomes.
Societies vary widely in the extent to which funds are ‘pooled’ in a household.
We can (tentatively) generalize that in most African societies, women need to earn
money for themselves and to contribute to the support of their children. African
societies are rich in mutual savings schemes also frequently run by women. There
exist well documented forms of mutual resource support for health care costs:
sharing of information and medicines, networks of personalized contacts with
health workers; sharing domestic work to allow others to work for cash to pay fees;
and mutual loans and gifts of cash for fees (Mackintosh 2000; Mackintosh and
Gilson 2002; Booth et al. 1995; MacCormack and Lwihula 1983; Gilson et al. 1994;
Moser 1998; Sauerborn et al. 1994).
HSR therefore, by generalizing pay-as-you-go, has increased women’s need for
cash to cope with their health care responsibilities towards family members, and
the strain on low-income women is severe. Furthermore falling incomes and eco-
nomic crisis undermine women’s participation in networks of mutual financial
support (Moser 1998).
In mutual health insurance, there is some evidence that building on existing
cooperative savings schemes works better than stand-alone schemes since they are
rooted in existing patterns of economic solidarity, yet World Bank and donor sup-
port has generally gone to stand-alone schemes (Atim 1999; Mackintosh and
Gilson 2002). The UMASIDA insurance system in Tanzania (Kiwara 2000) builds
on existing cooperative groups in the informal sector, including some women’s
groups; on a larger scale, the health insurance scheme built in India by the SEWA
organization has become well established (Ranson 2002).
A strong assumption behind the financing reforms in HSR was thus that
women’s incomes and mutual support networks were robust enough to withstand
the payments demanded. The sharp decline in usage and low levels of participa-
tion in community insurance schemes documented above show how unrealistic
that assumption was in contexts of generalized poverty.
The gendered impact of public health expenditure

Health sector reforms promoting inequitable private funding mechanisms have
sometimes been justified on the grounds that they shift the better-off to the
private sector, and allowing the targeting of public finance to the poor. Others
counter that privatization creates political pressures for subsidy and can leach the
public sector of assets through sale, theft and ‘informal’ use (K. Sen 2003). The
debate has generated poverty-focused assessments of public budgeting that unfor-
tunately do not generally disaggregate the poor by gender (Naschold and Fozzard
2002; Booth and Lucas 2002; van der Westhuizen 2004).
Most gender disaggregation of public expenditure and taxation is instead the
result of ‘women’s budget’ initiatives, including those in South Africa, Tanzania,
Uganda, Botswana, Malawi and Mozambique (Budlender 2000; TGNP 1999). The
techniques of gender budgeting have been disseminated by, among others, the
Commonwealth Secretariat (Mukhopadhyay et al. 2002) and UNIFEM.
The African initiatives have explored the accessibility of the budget-making
process to women, the extent of programmes specifically addressing women’s
needs (such as reproductive health) and, with difficulty, the gendered impact of
the main sectoral programmes including health (Budlender 1997). Financial data
can appear to be, as interviewees for a South African Women’s Budget study put it,
a ‘state secret’ (Klugman and McIntyre 2000). In South Africa, Tanzania and
Uganda women politicians have been important in pressing for and publicizing
information (Fleshman 2002). Tanzanian NGOs had trouble at first in getting
access to the national public expenditure review (PER) process but the Tanzania
Gender Networking Programme (TGNP) later became an invited member of PER
meetings. TGNP were also commissioned to contribute to the Poverty Reduction
Strategy Paper (PRSP) process, and have lobbied successfully for gender-relevant
indicators in monitoring health spending (Bridge 2004).
These are key contributions to our understanding of the impact of health-
finance changes on women. To assess the gendered impact of HSR-financing
changes requires an integrated assessment of fees and public finance in terms of
their differential impact on women; this assessment has barely begun.
Women as workers, users and citizens in the health sector
Key questions:
How has HSR affected the conditions of women as health care workers?
How have those effects in turn influenced women’s experience as health care users?
What are the assumptions underlying HSR about women’s interactions as workers,
users and citizens and are they correct?
As the ILO noted in 2004 in its social dialogue on gender and health services, ‘in
most countries the workforce in the health sector is predominantly female’, yet
‘they fall at the bottom of the hierarchy in terms of authority, remuneration and
qualification’ (www.ilo.org/dialogue, accessed on 27 April 2004). Case studies
repeatedly confirm this, and it is likely to hold across Africa despite a tradition in
some countries, including Mozambique, of men going into nursing (WHO 2003b;
Ngufor 1999; Corkery 2000). Therefore downward pressure on wages, working
conditions, employment and professionalism fall particularly strongly on women
health workers in all sectors.
Gendered private provision and its consequences

A key aspect of health sector reforms in sub-Saharan Africa has been liberalization
of private practice. Across the sub-continent from the 1980s onwards informal and
formal small-scale private practice expanded as health workers sought to supple-
ment plummeting public-sector incomes with fees from private patients. Dual
practice in both public and private sectors became common, as did informal
charging in public facilities. Private provision, largely unregulated, much of it of
dubious quality, became particularly widespread in urban areas (Tibandebage et al.
2001; Mackintosh and Koivusalo (2005b); Bennett et al. 1997; Mills et al. 2001;
Alubo 2001).
Yet we know little about women’s work conditions in the private health sector.
Expensive private clinics may provide nurses with better working conditions than
the public sector,2 but wages and working conditions can be worse in the small-
scale private sector than in the urban public sector. Private employers try to keep
down costs by reducing wages using poorly trained staff; those who employ
trained staff on decent wages find themselves undercut by those who do not
(Tibandebage and Mackintosh 2002). Users of the small-scale unregulated private
sector face problems including expired drugs, very poor diagnosis, over-prescribing
and outright cheating, failure to identify and refer serious cases, and low levels of
participation in immunization and public health education. As a result, low-
income women users in much of urban Africa find themselves paying money they
cannot afford to low-paid women working in dangerous and insecure jobs: a
vicious circle of gender disadvantage in which those with the lowest income are
the most vulnerable to abuse.
There are private-sector exceptions to these generalizations: some small private
dispensaries visited in Tanzania in 1998 were doing preventative care (Vignette 1),
and some private providers catering largely for the better-off displayed higher lev-
els of probity. Patients interviewed, in the majority female, indicated that they val-
ued the pleasant manner of the staff, and the short waiting times and lack of
hassle, but that finding the money was a serious problem.
Gendered hierarchy in the public sector and

its consequences post-reform
The liberalization of private practice alongside severe deterioration in the funding
of the government sector has created strain and demoralization for many nurses
working in African public sectors. Researchers document public dissatisfaction

with rude and abusive behaviour by some government health care staff, in partic-
ular on maternity wards (Bassett et al. 1997; Gilson et al. 1994; Tibandebage and
Mackintosh 2002). Jewkes et al. (1998) found in South Africa that low-income
patients, and those regarded as reprehensible such as teenagers, were most likely to
face such abuse.
There are however two sides of this story of mistreatment of women by (pre-
dominantly) women, and during some research the authors did in Tanzania in
1998, a matron in a maternity hospital put the nurses’ side of the story (Vignette 2).
Vignette 2
This kind of thing [bad behaviour] happens, and it is because of poor morale, low com-
mitment, severe overwork and low salaries. Imagine, you are a nurse on duty … You may
have a ward of 40–60 seriously ill patients … You are two trained people at best. How will
you divide yourself? You are constantly over-working and under pressure. You are wor-
ried about family problems and commitments. For twelve hours you do not know what
is happening to your children. And you may not have as much as a cup of tea. Then there
is the problem of the commitment of other staff. You are a nurse by profession. The doc-
tor, who is supposed to be responsible, works his official hours and goes away, he waits to
be called. You are there, someone is bleeding and you cannot help them. She needs to be
operated on, no one is there, there are no facilities … . Relatives rush to send the sick per-
son to hospital, then we are not in a position to help the patient.
Public-sector decline and health sector liberalization tend to widen the gap in
wages and working conditions between doctors and nurses. More doctors than
nurses do additional private practice, and the ‘going rate’ for informal payments
to doctors was substantially higher than for nurses.3 Nurses have far more contact
with patients than doctors, and when the service falls apart they take most of the
strain. The matron quoted above described an acute shortage of trained nurses;
very low pay associated with the loss of allowances (such as for transport and uni-
forms, which they now had to pay for) as a result of reform; and lack of access to
training opportunities since specialist nurse training courses now had to be paid
for by the nurses themselves ‘and if people have a family they simply don’t see this
as a possibility’.
The result, as this matron confirmed, was that nurses had started to go abroad,
initially from Tanzania to Zimbabwe. In the five years since then, this flow
has accelerated from many African countries, as high-income countries have
expanded international recruitment (Mensah 2005). Doctors migrate too and in
Ghana, for example, most incentives aimed at retaining health care staff have
been focused on doctors, increasing nurses’ resentment. Given that, as Mensah
puts it, nurses are the ‘backbone’ of African health services, the rising out-migra-
tion is having extremely serious effects on services, and very specifically on pri-
mary provision on which low income women predominantly rely.
Gendered assumptions and scope for change

Writing on the human-resource aspects of health care reform in the UK and their
‘lessons’ for other countries (for example, Buchan 2000) generally does not men-
tion gender. Yet UK health-service reforms in the 1990s had strongly gendered
effects, such as the shifting of work previously done by nurses to poorly trained
social carers and auxiliaries, to the disadvantage of both low-paid women staff and
low-income clients. Managerially led reforms tend to restructure in favour of the
more powerful: gender, status and social class can interact, as institutions change,
to the detriment of low-income women as both staff and patients (Towers et al.
2000; Mackintosh 1998).
In much of the world, furthermore, trade unions as a voice for low-paid staff are
weak. Trade unions have been ignored or denigrated by HSR policy-makers, who
characterized them as a voice for provider interests, defined by assumption as in
opposition to the interests of users. This lack of representation exacerbates deteri-
oration of morale and professional ethics, and can generate a loss of staff rights to
decent working conditions and wages, as research on Eastern European health sec-
tor reforms demonstrates (Afford 2003).
A gender ‘lens’ on reform suggests that better outcomes can be built on an
explicitly gendered framing of provider–user relationships. The ‘Health Workers
for Change’ (HWFC) projects, in Africa and elsewhere, have built efforts to
improve health care quality post-reform on the observation that interpersonal
aspects of health care – such as respect and ability to listen – are important to
quality of care and are also gendered. Female health workers have somewhat
different styles of work than men; female health care users have specific needs,
and may encounter gender-specific forms of discrimination and abuse (Haaland
and Vlassoff 2001; Hartigan 2001; Fonn and Xaba 2001). Projects such as the
Women’s Health Project in South Africa have built collaborative efforts with
health care staff to explore frustrations and shift behaviour. There is evidence of
positive impact on quality of care in four African countries, resulting in part
from greater recognition among health workers of the constraints on health-
seeking behaviour faced by poor women (Onyango-Ouma et al. 2001). The HWFC
literature argues that understanding health systems as gendered was crucial to
these outcomes.
While HSR models ignore trade unions, they do repeatedly propose commu-
nity participation in priority-setting, notably as part of decentralization of the
management of health care and Sector-Wide Approaches (SWAps) to funding
that pool some donor resources. In Bangladesh and India, local women’s NGO
groups have managed to bring women’s health needs to the attention of gov-
ernment officials and health care providers (Evers and Juarez, n.d). In Africa,
that kind of organizational initiative to ensure women’s representation, and to
support their capacity to take part effectively in decision-making and be com-
mitted advocates of the interests of those they represent, remains weak
(Foster 1999).
Reproductive health and HSR
Key question:
Has HSR improved health systems’ capacity to help redress women’s vulnerability
to reproductive ill health? If not, why not?
Reproductive health as a key aspect of women’s health
Vignette 3
A 16-year-old girl from a poor rural family was about to complete primary school, but had
often been sent home for lack of school supplies. One day a man offered her gifts and
bought her exercise books. In return he wanted a relationship. Because she felt she
needed his help, the girl accepted and knowing nothing about contraception she soon
got pregnant. She hid her pregnancy until an advanced stage and then had to quit school.
She went through almost the whole pregnancy without attending an ante-natal clinic. At
delivery she was taken to a nearby health centre and there was a delay as the mother
rushed back home to borrow the money. The girl had complications during delivery and
because of her young age and the delay she lost the baby. Luckily the girl survived.
Globally, 20 per cent of the burden of disease among women of reproductive age
is connected with sex and reproduction; in sub-Saharan Africa it is 40 per cent
(UNFPA 2002). Women and adolescent girls continue to face uncontrolled fertility,
maternal morbidity and mortality, unwanted pregnancy, early childbearing,
sexually transmitted diseases including HIV/AIDS, sexual violence, and female
genital mutilation (FGM) which often results in reproductive-health-related
complications. The average maternal mortality rate in sub-Saharan Africa (SSA) is
1,000 deaths per 100,000 live births as compared to a world average of 400 (UNFPA
2002) Africa has one of the highest teenage pregnancy rates in the world, often
resulting in unsafe abortion (WHO 2000b).
The patriarchal ideology which treats women as inferior to men and allows
them few decision-making powers also continues to bear heavily on women’s
sexual and reproductive health. Women lack freedom over their sexual lives,
putting them at risk of contracting STDs as unwilling partners, and lack resources
for reproductive and other health needs. Adolescent girls are immensely vul-
nerable, especially girls struggling to cope with poverty in ways that puts their
health at risk in a health system that cannot assure their protection nor meet their
health needs.
Contrast this situation with the essential elements of reproductive health agreed
at the International Conference on Population and Development (ICPD) held in
Cairo 1994 including: having a satisfying, safe sexual life; access to appropriate,
safe, effective, affordable and acceptable methods of family planning based on
informed choice and dignity; prevention, diagnosis and treatment of STIs/
STDs/HIV; services for safe pregnancy and childbirth; elimination of harmful prac-
tices (FGM, domestic violence, sexual trafficking); and emphasis on poverty
alleviation, girls’ education, women’s empowerment and reproductive rights. Also
emphasized was the role of civil society and community in promoting sexual and
reproductive health.
HSR should be judged on whether it has focused or diverted health systems from
tackling this huge burden of reproductive ill health in the context of the HIV/AIDS
pandemic. In 2001, SSA had 3.5 million people newly infected with HIV/AIDS
(UNAIDS 2002). There, women have generally higher infection rates than men.
Women accounted for 59 per cent of the 2.23 million adults estimated to be living
with HIV in Tanzania in 2001, and young women aged 15–24 had twice the preva-
lence rate of young men (8.06 per cent compared to 3.55 per cent) (World Bank
2002). In Uganda, HIV infection rates were rising fastest among Ugandan women
aged 15–25 years (Ochieng 2003).
Service disintegration and the implications for reproductive health

Given this crisis, one would expect that health sector reorganization would ensure
that sexual and reproductive health was given top priority. However, while fund-
ing for HIV/AIDS has risen, HSR has been associated with a much-contested retreat
from the policy of integrated primary health care for all established by the Alma
Ata declaration in 1977 (WHO 1978; LaFond 1995). The shift to targeted vertical
‘interventions’ associated with HSR pulled apart the principle of integrated
primary care for women’s and children’s needs that had previously been seen as
key to effective provision.
The disintegration of primary care was reinforced over time since private
providers are reluctant to undertake unprofitable aspects of preventative care and
donors prefer to fund ‘vertical’ programmes operating independently of other
provision. Women and men may therefore need to visit different providers for
different needs, a source of strain, time loss, confusion and expense. Oliff et al. (2003)
argue that new funds set up to address the crisis of HIV/AIDS have also favoured
top-down, selective and donor-driven activity, noting that in Tanzania the gov-
ernment has since 1995 (just a year after Cairo) actively been trying to integrate
reproductive health programmes with each other nationally and at district level
with very limited success.
It is widely accepted that effective and universal reproductive health pro-
grammes produce major benefits to women’s and children’s health, and disinte-
gration wastes resources and undermines access. Yet pleas for dialogue between
reproductive health policy-makers and campaigners and health sector reform pro-
ponents have been made for many years yet appear still to be going unheard
(Lubben et al. 2002; Tangcharoensathien 2002). Standing (2002b: 25) states that
‘Health sector reforms have largely marginalised the Cairo agenda’. Berer (2002)
argues that HSR has undermined reproductive health through disease-based prior-
itizing of public funding, insufficient attention to evidence on the effectiveness
and social importance of reproductive health care, and the loss of public-sector
capacity to provide care. These have combined with explicit rejection of sexual
and reproductive health rights by ultra-conservative religious movements and
institutions to undermine provision for women’s needs and the needs of the next
generation.
Gendered health sector reform meets the labour market:

paid work, unpaid work, poverty and a gendered crisis of care
Key questions:
To what extent is the context of HSR one of unmanageable burdens of paid and
unpaid work for some categories of impoverished women?
To what extent is HSR exacerbating or relieving those burdens?
Vignette 4
A middle-aged woman runs a fruit-and-vegetable stall near her home. Her husband is a
clerk in a government office, and she helps in buying food items and paying for her
health expenses and sometimes the children’s. She also does all the caring for her three
young children, cooking, cleaning and fetching water from a nearby well. She has to run
her petty business close to home because there is no one else to care for the children,
since all adults are now trying to earn some money. She is usually exhausted by the end
of the day, while her husband hardly does any housework. She says however that she
must struggle on with the business because everything now requires money, including
health care and education. Overwork is taking its toll on her health.
African women bear disproportionately the burden of unpaid work, and have long
needed to earn cash for their own and their children’s needs. The economic crisis
and subsequent economic and social-sector reform programmes have escalated
costs of basic needs such as food, health care and children’s education, making the
combined pressures of paid and unpaid work more severe. Women survive by
working longer hours and years in low-return income-generating activities in the
informal sector (Whitehead 2001). This survival strategy in turn generates overwork,
affecting women’s health and increasing their health care needs – yet ironically it
may not pay enough to cater for increased health care, creating another vicious
circle of ill-health. Because of multiple working and caring roles at home and in
the workplace, women find it difficult to allocate time to seeking health care when
they are sick; long distance, long waits and staff absent from government facilities
make the vicious circle worse.
The crisis of care in the midst of HIV/AIDS and the plight

of elderly women
The scale of the AIDS epidemic has cruelly exposed the inadequacy of the assump-
tions about women’s ‘coping’ capacity that implicitly underpinned health-sector
reform: notably that the fees were manageable and that care in the home to reduce
the need for paid care could be supported by women. The convergence of general-
ized poverty, fees, high work burdens, ill-health of both care givers and care receivers,
and the loss of substantial numbers of the economically active population has put a
completely unmanageable burden on the shoulders of, particularly, elderly women,
with very serious implications for the current generation of children.
It is in the plight of elderly women that these pressures are exposed at their very
worst. AIDS death rates and the numbers of AIDS orphans in Africa are staggering.
By 2001, 83 per cent of world AIDS deaths had occurred in sub-Saharan Africa,
including 90 per cent of those dying before 15 years. In 2000 92 per cent of the
global total of 13.2 million AIDS orphans were in SSA (UNAIDS 2002; Africa Action
2000). As women and men aged 15–49 succumb to the epidemic, the elderly are
forced to take care of their terminally ill adult children and then to care for their
orphaned grandchildren. In Tanzania an assessment of the economic and social
impact of HIV/AIDS found 34 per cent of carers of orphans interviewed in schools
were grandparents, some taking care of as many as eight grandchildren (ESRF
2004). In South Africa and Uganda 40 per cent of orphaned children were found
to be living with their grandparents; in Zimbabwe the proportion was more than
half (HelpAge International and the International Alliance on HIV/AIDS 2003,
citing Ainsworth and Filmer 2002; Subbarao and Coury 2003)
The gendered division of labour continues to ascribe domestic work and caring
for children and the sick to women, in this case grandmothers, and women also live
longer than men. In Zimbabwe, two-thirds of 685 older care givers interviewed were
women. Often in ill-health and unable to take care of themselves adequately,
elderly women are a vulnerable group needing care and financial support, yet the
epidemic robs them of the support of their adult children. Doubly disadvantaged –
as female and elderly – they are among the poorest groups in society, now forced to
take on the additional burden of caring and raising their grandchildren within the
context of decreasing economic means and disintegrating family networks. Taking on
the additional responsibility of caring for their orphaned grandchildren means strug-
gling to meet food, clothing, health and education expenses for the grandchildren.
Elderly women’s poor health is exacerbated by physical and emotional stress
when taking care of their terminally ill adult children, and they risk infection
because they lack information on ways to protect themselves (HelpAge International
and The International Alliance on HIV/AIDS 2003). After their children die, the
worsening strain and hardship is illustrated by an interview with an elderly
woman in a South Western town of Tanzania (Vignette 5).
Vignette 5
This elderly lady is a widow who has seen her two sons die and is caring for the children
of both sons. Her only way of making an income is to buy vegetables, split them up into
small parcels and resell them retail. The money is not enough to feed the children
properly, nor herself. The children now go to school (since primary school fees were
abolished) but cannot come home for lunch because there is no food for them. She
herself is weak and sick from poor diet. There will be cumulative effects of poor diet for
the children.
HSR has in no way alleviated this cumulative burden of caring for the sick and
dying, raising children in old age and struggling to make ends meet to pay for
food, school books and clothes. Rather, it has added to needs for cash and unpaid
work in the context of an eroding family support system and inadequate or total
lack of a formal social protection system for the elderly.
Conclusions
The social context of unequal gender relations adversely affects women’s health
and overall well-being, both directly and through its health-service effects
(Cuttingham and Mynttri 2002). The health sector reflects established gender
hierarchies within its staffing and priority-setting; when it responds poorly to
women’s needs it reinforces women’s relative social and economic disadvantage,
undermining women’s capacity to earn and care for themselves and others.
A discriminatory and inequitable health sector reinforces other aspects of women’s
relative lack of rights and dignity in societies that are highly unequal in terms of
both gender and social class.
Health sector reform, far from challenging these mutually reinforcing forms
of gender disadvantage, appears to have made some of them worse. This chap-
ter has argued that these failures appear rooted in invalid assumptions about
women’s responsibilities and capabilities to cope with rising work burdens. A con-
ceptualization of HSR as gendered institutional change would have highlighted
the dangers – for women and for population health – of exacerbating women’s
expenditure and work burdens to unmanageable proportions and of reinforcing
women’s relative disadvantage at the lower end of health systems’ work and care
hierarchies. A gender ‘lens’ on the proposed reforms would have questioned
the desirability, for example, of user fees and refocused attention on improving the
relationships and integration of services within primary care.
So what would it take to build instead a more gender-equitable set of health
sector institutions? Here are some propositions.
1. Design reform of health care finance in such a way that it does not increase the bur-
den on the incomes of poor women, nor the relative economic disadvantage of women
in general. This requires more information on women’s control over resources,
integrated into the design of health-financing mechanisms.
2. Design health sector reform to promote equality between male and female health
workers and to improve the working conditions of lower-level staff. This requires
close attention to the gender-differentiated and interacting needs of both staff
and patients.
3. Involve women equally with men, both as staff and citizens, in health sector reform
from the most local to the highest level. Women’s knowledge and perspectives are
essential to achieve policy design that is not discriminatory on gender lines,
basing itself rather on concepts of equal citizenship: draw on and support
women’s organizations for this purpose.
4. Do not undertake reforms that increase the work burden of women, especially poor
women; instead, design approaches that reduce those burdens. This may also require
men to carry a greater share of the burden of unpaid care.
5. Build links between reproductive health care and provision for the other health needs
of women in ways that reduce the strain upon women of accessing care. This requires
investigating women’s points of view on how to integrate services.
6. Introduce reforms outside the health sector that strengthen women’s rights and status,
hence their ability to deal with the health sector on more equal terms with men. This
requires strengthening women’s rights and dignity through supportive laws,
policies and organizations.
Acknowledgement
We thank Shahra Razavi, Debbie Budlender, Susan Himmelweit, Dzodzi Tsikata and all
participants in the UNRISD workshop on Gender and Social Policy in Maarstrand, Sweden,
May 2005, for helpful comments on an earlier draft. The chapter is the sole responsibility of
the authors.
Notes
1 One scheme has subsequently achieved higher rates: P. Mujinja, personal communication.
2 Anecdotally, this seems to be the case in Accra, Ghana, where highly skilled nurses have
moved from the government hospitals to private specialist clinics.
3 Authors’ fieldwork.
References
Afford, C. (2003) Corrosive Reform: Failing Health Systems in Eastern Europe. Geneva: ILO.
Africa Action (2000) ‘Africa HIV/AIDS update’ at http://www.africaction.org/docs00/
hiv0007//htm, accessed on 20 April 2004.
Ainsworth, M. and D. Filmer (2002) Poverty, Children, Schooling and HIV/AIDS: A Targeting
Dilemma. Policy Working Paper No. 2885. Washington, DC: World Bank.
Alubo, O. (2001) ‘The Promise and Limits of Private Medicine: Health Policy Dilemmas in
Nigeria’. Health Policy and Planning 16(3): 313–21.
Atim, C. (1999) ‘Social Movements and Health Insurance: A Critical Evaluation of Voluntary,
Non-profit Insurance Schemes with Case Studies from Ghana and the Cameroons’. Social
Science and Medicine 48: 881–96.
Bassett, M, L. Bijlmakers and D. Sanders (1997) ‘Professionalism, Patient Satisfaction and
Quality of Health Care: Experience During Zimbabwe’s Structural Adjustment
Programme’. Social Science and Medicine 45(12): 1845–52.
Bennett, S., B. McPake and A. Mills (eds) (1997) Private Health Providers in Developing
Countries: Serving the Public Interest? London: Zed Books.
Berer, M. (2002) ‘Health Sector Reforms: Implications for Sexual and Reproductive Health
Services’. Reproductive Health Matters 10 (20): 6–15.
Booth, D, J. Milimo, G. Bond and S. Chimula (1995) ‘Coping with Cost Recovery’. Report to
SIDA, Development Studies Unit, Department of Social Anthropology, Stockholm
University.
Booth, D. and H. Lucas (2002) ‘Good Practice in the Development of PRSP Indicators and
Monitoring Systems’. Working Paper No. 172, London: Overseas Development Institute, July.
Bridge (2004) ‘Gender and Budgets.’ Gender and Development In Brief, Issue 12, www.ids.ac.uk/
bridge, accessed on 4 May 2004.
Buchan, J. (2000) ‘Health Sector Reform and Human Resources: Lessons from the United
Kingdom’. Health Policy and Planning 15 (30): 319–25.
Budlender, D. (1997) The Second Women’s Budget. Cape Town: IDASA.

Budlender, D. (2000) ‘The Political Economy of Women’s Budgets in the South’. World
Development 28(7): 1365–78.
Corkery, J. (2000) ‘Public Service Reforms and their Impact on Health Sector Personnel in
Uganda’. Geneva, WHO/ILO at www.ilo.org, accessed on November 2004.
Cuttingham, J. and C. Mynttri (2002) ‘Reproductive Health: Conceptual Mapping and
Evidence’. In Gita Sen, Asha George and Piroska Ostrin (eds) Engendering International
Health: The Challenge of Equity. Cambridge MA and London: MIT Press.
Economic and Social Research Foundation (ESRF) (2004) The Economic and Social Impacts of
HIV/AIDS in Tanzania: A Final Report. Dar es Salaam, ESRF.
Ekwempu, C.C.D., M.B. Maine, B.E. Olonkoba and M.N. Kisseka (1990) ‘Structural
Adjustment and Health in Africa’. Letter, Lancet (July 7): 56–67.
Evers, B. and M. Juarez (n.d) ‘Understanding the Links: Globalization, Health Sector Reform,
Gender and Reproductive Health’. In Ford Foundation Globalization Committee,
Reproduction Health Affinity Group, Globalization, Health Sector Reform, Gender and
Reproductive Health at www. fordfound.org/publications, accessed July 2005.
Fleshman, M. (2002) ‘ “Gender budgets” seek more equity’. Africa Recovery 16 (1) April at
www.un.org/ecosocdev/geninfo/afrec, accessed on 4 May 2004.
Fonn, S. and M. Xaba (2001) ‘Health Workers for Change: Developing the Initiative’. Health
Policy and Planning 16 (Supp.): 1–6.
Foster, M. (1999) Lessons of Experience from Health Sector SWAPs. London: CAPE/ODI.
Gilson, L. (1997). ‘The Lessons of User Fee Experience in Africa’. Health Policy and Planning
12(4): 273–85.
Gilson, L. and A. Mills (1995) ‘Health Sector Reforms in Sub-Saharan Africa: Lessons of the
Past 10 Years’. In P. Berman (ed.) Health Sector Reform in Developing Countries: Making Health
Development Sustainable. Cambridge MA: Harvard University Press.
Gilson, L., K. Heggenhougen and A. Allilo (1994) ‘A Tanzanian Evaluation of Community
Satisfaction with Primary Care’. Social Science and Medicine 39 (6): 767–80.
Haaland, A and C. Vlassoff (2001) ‘Introducing Health Workers for Change: From
Transformation Theory to Health Systems in Developing Countries’. Health Policy and
Planning 16 (Supp.): 1–6.
Hartigan P. (2001) ‘The Importance of Gender in Defining and Improving Quality of Care:
Some Conceptual Issues’. Health Policy and Planning 16 (Supp. 1): 7–12.
HelpAge International and the International Alliance on HIV/AIDS (2003) Forgotten Families:
Older People as Carers of Orphans and Vulnerable Children at http://www.helpage.org/publi-
cations/PapersEtc/PapersEtcmiddle.html, accessed in May 2004.
Jewkes, R, N. Abranams and Z. Mvo (1998) ‘Why Do Nurses Abuse Patients? Reflections on
South African Obstetric Services’. Social Science and Medicine 47(11): 1781–95.
Jutting, J. and J. Tine (2000) ‘Micro Insurance Schemes and Health Care Provision in
Developing Countries: An Empirical Analysis of Mutual Health Insurance Schemes in Rural
Senegal.’ Center for Development Research (ZEF) Bonn, ILO/ZEF- Project No. 7359, Project
Report: 5.
Karanja, M., G. H. Bloom and M. M. Segall (1995) ‘Impact of User Charges on Vulnerable
Groups: The Case of Kibwezi in Rural Kenya’. Social Science and Medicine 41: 829–35.
Kiwara, A. D. (2000) ‘Health Sector Reforms and Alternative Health Care Financing in
Tanzania: Windows of Hope’. Paper for workshop on Managing and Regulating Mixed
Health Care Systems, Dar es Salaam, Tanzania.
Klugman, B. and D. McIntyre (2000) ‘From Policy, through Budgets to Implementation:
Delivering Quality Health Care Services.’ In D. Budlender (ed.) The Fifth Women’s Budget.
Cape Town: IDASA.
Kutzin, J. (1995) ‘Experience with Organizational and Financing Reform of the Health
Sector’. Current Concerns, SHS Papers 8 (SHS/CC/94.3). Geneva: WHO.
LaFond, A. (1995) Sustaining Primary Health Care. London: Earthscan.
Lewis, G. (1998) ‘Citizenship’. In G. Hughes (ed.) Imagining Welfare Futures. London:

Routledge.
Lewis, G., S. Gewirtz and J. Clarke (eds) (2000) Rethinking Social Policy. London: Sage.
Lewis, J. (2002) ‘Gender and Welfare State Change’. European Societies 4 (4): 331–57.
Lister, R. (2000) ‘Gender and the Analysis of Social Policy’. In J. Lewis, S. Gewirtz and J.
Clarke (eds) Rethinking Social Policy. London: Sage.
Lubben, M., S. Mayhew, C. Collins and Andrew Green (2002) ‘Reproductive Health and
Health Sector Reform in Developing Countries: Establishing a Framework for Dialogue’.
Bulletin of the World Health Organization 80: 667–74.
MacCormack, C. and G. Lwihula (1983) ‘Failure to Participate in Malaria Chemo-Suppression
Programme: North Mara Tanzania’. Journal of Tropical Medicine and Hygiene 86(3): 99–107.
Mackintosh, M. (1998) ‘Public Management for Social Inclusion.’ In M. Minogue,
C. Polidano and D. Hulme (eds) Beyond the New Public Management. Cheltenham: Edward
Elgar.
Mackintosh, M. (2000) ‘The Contingent Household: Gender Relations and the Economics of
Unpaid Labour’. In S. Himmelweit (ed.) Inside the Household. Basingstoke: Palgrave.
Mackintosh, M. and L. Gilson (2002) ‘Non-market Behaviour in Health Care.’ In J. Heyer, R.
Thorp and F. Stewart (eds) Group Behaviour and Development. Oxford University Press.
Mackintosh, M. and M. Koivusalo (2005a) ‘Health Systems and Commercialisation: in
Search of Good Sense.’ In M. Mackintosh and M. Koivusalo (eds) Commercialization of
Health Care: Global and Local Dynamics and Policy Responses. Basingstoke: Palgrave
Macmillan.
Mackintosh, M. and M. Koivusalo (2005b) Commercialization of Health Care: Global and Local
Dynamics and Policy Responses. Basingstoke: Palgrave Macmillan.
Mensah, K. (2005) ‘International Migration of Health Care Staff: Extent and Policy Responses
with Illustrations from Ghana.’ In M. Mackintosh and M. Koivusalo (eds), Commercialization
of Health Care: Global and Local Dynamics and Policy Responses. Basingstoke: Palgrave
Macmillan.
Mills, A., S. Bennett and S. Russell (2001) The Challenge of Health Sector Reform: What Must
Governments Do? Basingstoke: Palgrave.
Moser, C. (1998) ‘The Asset Vulnerability Framework: Reassessing Urban Poverty Reduction’.
World Development 26 (1): 1–19.
Mukhopadhyay, T., D. Budlender, D. Elson and G. Hewitt (eds) (2002) Gender Budgets Make
Cents: Understanding Gender-responsive Budgets. London: Commonwealth Secretariat.
Nanda, P. (2002) ‘Gender Dimensions of User Fees: Implications for Women’s Utilization of
Health Care’. Reproductive Health Matters 10 (20): 127–34.
Naschold, F. and A. Fozzard (2002) ‘How, When and Why Does Poverty Get Budget Priority?
Poverty Reduction Strategy and Public Expenditure’. Working Paper No. 165, London:
Overseas Development Institute, April.
Ngufor, G.F. (1999) ‘Public Service Reforms and their Impact on Health Sector Personnel in
Cameroon’. Geneva: WHO/ILO, at www.ilo.org.
Ochieng, R.O. (2003) ‘Supporting Women and Girls’ Sexual and Reproductive Health and
Rights: The Ugandan Experience’. Development, Vol. 46, No. 2: 38–44.
Oliff, M., P. Mayaud, R. Brugha and A.M. Semakafu (2003) ‘Integrating Reproductive Health
Services in a Reforming Health Sector: The Case of Tanzania’. Reproductive Health Matters 11
(21): 37–48.
Onyango-Ouma, W., R. Laisser, M. Mbilima et al. (2001) ‘An Evaluation of Health Workers
for Change in Seven Settings: A Useful Management and Health System Development
Tool’. Health Policy and Planning 16 (Supp. 1): 24–32.
Ranson, K. (2002) ‘Reduction of Catastrophic Health Expenditures by a Community-based
Health Insurance Scheme in Gujarat, India: Current Experiences and Challenges’. Bulletin
of the World Health Organization 80: 613–21.
Results (2004) ‘User Fees on Primary Health and Education’, at www.resultsusa.org, accessed
on 23 January 2004.
Rifkin, S. and G. Walt (1986) ‘Why Health Improves: Defining the Issues Concerning
“Comprehensive Primary Health Care” and “Selective Primary Health Care” ’ Social Science
and Medicine 23: 559–66.
Sauerborn, R., A. Nougtara and E. Latimer (1994) ‘The Elasticity of Demand for Health Care
in Burkina Faso: Differences across Age and Income Groups’. Health Policy and Planning
9(2): 185–92.
Sen, G., A. George and P. Ostlin (2002a) ‘Engendering Health Equity: A Review of Research
and Policy’. In G. Sen, A. George and P. Ostlin (eds) Engendering International Health.
Cambridge MA: MIT Press.
Sen, G., A. George and P. Ostlin (eds) (2002b) Engendering International Health. Cambridge
MA: MIT Press.
Sen, K. (2003) Restructuring Health Services: Changing Contexts and Comparative Perspectives.
London: Zed Books.
Standing, H. (1997) ‘Gender and Equity in Health Sector Reform Programmes: A Review’.
Health Policy and Planning 12 (1): 1–18.
Standing, H. (2002a) ‘Frameworks for Understanding Health Sector Reform’. In G. Sen,
A. George and P. Ostlin (eds) Engendering International Health. Cambridge MA: MIT Press.
Standing, H. (2002b) ‘An Overview of Changing Agendas in Health Sector Reform’.
Reproductive Health Matters 10 (20): 19–28.
Standing, H. (2002c) ‘Towards Equitable Financing Strategies for Reproductive Health’.
Sussex UK: Institute of Development Studies.
Subbarao, K. and D. Coury (2003) Orphans in Sub-Saharan Countries: A Framework for Public
Action. Washington DC: Africa Region (Human Development) and Human Development
Network (Social Protection), World Bank.
Tangcharoensathien, V. (2002) ‘ICPD Goals and Targets Worth Revisiting in the Context of
Health Sector Reforms’. Reproductive Health Matters 10 (20): 138–40.
Tanzania Gender Networking Programme (TGNP) (1999) Budgeting with a Gender Focus.
Dar es Salaam: TGNP.
Tibandebage, P. (2004) ‘Health Sector Reforms and Gender Implications: A Case Study of
Mutual Health Insurance Schemes in Tanzania’. Research paper for UNRISD, Geneva at
www.unrisd.org, accessed on November 2004.
Tibandebage, P. and M. Mackintosh (2002) ‘Institutional Reform and Regulatory
Relationships in a Liberalising Health Care System: a Tanzanian Case Study’. In J. Heyer,
R. Thorp and F. Stewart (eds) Group Behaviour and Development. Oxford: Oxford University
Press.
Tibandebage, P., H.H. Semboja, P. Mujinja and H. Ngonyani (2001) ‘Private Sector
Development: The Case of Private Health Facilities’. ESRF Discussion Paper, Dar es Salaam.
Towers, B., P. Smith and M. Mackintosh (2000) ‘Dimensions of Class in the Integration of
Health and Social Care’. Journal of Interprofessional Care 13/3: 219–28.
Twigg, J. (2000) Bathing, the Body and Community Care. London: Routledge.
UNAIDS (2002) Report on the Global HIV/AIDS Epidemic, UNAIDS, Geneva. July.
United Nations Population Fund (UNFPA) (2002) State of the World Population 2002: People,
Poverty and Possibilities. New York: United Nations Population Fund.
Walraven, G. (1996) ‘Willingness to Pay for District Hospital Services in Rural Tanzania’.
Health Policy and Planning 11(4): 428–37.
Van der Westhuizen, C. (2004) ‘Monitoring Provincial Budgets from a Pro-poor Perspective’.
Budget Information Service, Cape Town IDASA: February, at http://www.idasa.org.za/
accessed March 2005.
Whitehead, A. (2001) ‘Trade, Trade Liberalisation and Rural Poverty in Low Income Africa:
A Gendered Account’. Background Paper for the UNCTAD 2001 Least Developed Countries
Report.
Williams, F. (1989) Social Policy: A Critical Introduction. Cambridge: Polity.
World Bank (1993) World Development Report 1993: Investing in Health. New York: Oxford
University Press.
World Bank (1997) World Development Report 1997: The State in a Changing World. New York:
World Bank (2000) Advancing Gender Equity: World Bank Action since Beijing. Washington DC:
World Bank.
World Bank (2002) Tanzania at the Turn of the Century: Background Papers and Statistics.
A World Bank country study, Washington, DC: World Bank.
World Bank (2004). World Development Report 2004: Making Services Work for the Poor.
World Health Organization (WHO) (1978) Primary Health Care: Report of the International
Conference on Primary Health Care, Alma Ata 1987. Geneva: WHO.
World Health Organization (WHO) (2000a) World Health Report 2000. Health Systems:
Improving Performance. Geneva: WHO.
World Health Organization (WHO) (2000b) Abortion: A Listing of Available Data on the
Frequency and Mortality of Unsafe Abortion. WHO/RHT/MSM/97.4, Geneva: WHO.
World Health Organization (WHO) (2002) Impact of AIDS on Older People in Africa: Zimbabwe
Case Study. WHO/NMH//NPH/ALC/02.12, Geneva: WHO.
World Health Organization (WHO) (2003a) The World Health Report 2003: Shaping the Future.
Geneva: WHO.
World Health Organization (WHO) (2003b) Strengthening Nursing and Midwifery. Report by
the Secretariat, 56th World Health Assembly, Geneva.
12
Health Sector Reform in China:
Gender Equality and Social Justice
Jufen Wang1
Introduction
In any country the most important aspects of social policy are the social security
and social welfare systems. These systems are not only the results of social and eco-
nomic development but also, through the redistribution of national income, can
actively enable the harmonious, stable and healthy development of society.
Moreover, these systems can strengthen people’s capacity to identify and avoid
risks and accommodate unexpected events. The Constitution of China (PRC)
states that the ‘citizens of PRC are entitled to obtain physical assistance from state
and society in old age and ill health’.
There is insufficient theoretical research on social security in China. For a long
time, Chinese social security theory lacked a clear definition. Social security was
understood in terms of social insurance, which in turn was equated with labour
insurance. For example, social security rights were defined as
the rights enabling workers to access necessary help from the nation and
society when they lose their ability or opportunity to work temporarily or per-
manently through [factors such as] old age, disease, disability, unemployment
or other difficulties
Guo and Li 2001: 173
Another government publication stated that ‘social security in China is labour

insurance during the planned economy era’ (DoP 1998: 401). Only in 1999 did Ge
provide a clear definition of social security in his book Social Security Economics, in
which he stated that
Social security is a series of organized measures, systems and services in which

the state or nation offers concrete assistance for those in difficulty relying on
special funding forcefully raised through distribution and redistribution of
national income according to state legislation to secure social stability. Social
security is a basic right for members of society. The nation has a legal responsi-
bility to secure livelihood rights for the members of its society
Ge 1999: 2
258
Health Sector Reform in China 259
Ge was an early proponent of social security theory and an acknowledged expert

in the field in China. This new definition was based on the government’s new
vision of social security in a more liberalized context, which imposed a greater
degree of responsibility for welfare on the individual rather than the government.
Healthcare insurance is one of the fundamental components of social security.
China’s healthcare insurance system was originally instigated at the beginning of
1950s, soon after the establishment of the People’s Republic of China. The health
care sector began reforming some forty years later, in the late 1990s. The aim of
this chapter is to examine the nature of social justice within the reformed health-
care system in urban areas from a gender perspective. I integrate issues of poverty,
social differentiation, migration and income inequality with gender to examine
the impacts of social policy on both the pre-existing urban populations and the
huge new populations of migrants who are rapidly settling into Chinese cities and
towns. Migrants are excluded from the social security systems through a bureau-
cratic inflexibility predicated on the government’s binary conception of rural and
urban environments and people.
Local authority management is one of the main principles of the basic healthcare
insurance system for China’s urban employees. Different schemes reflect differen-
tiation in local economic development and differences in the length of operation
of the scheme itself. It is difficult to get comprehensive details about schemes
across all regions of China and, as a result, it is not possible to comment compre-
hensively on insurance reform in this chapter. Consequently, I have restricted my
focus to the Shanghai region alone. The level of economic development in
Shanghai is so far in advance of other regions, especially those in the west, that
Shanghai cannot, of itself, represent all of China. However, since the model of
insurance reform eventually adopted in Shanghai was based on the ‘Liangjiang’
model discussed in greater detail below, the research reported here outlines one
city’s experience with the implementation of a systematic reform option from
which useful lessons may be drawn.
Background to health sector reform in urban China
The main characteristic of the healthcare insurance system established in the

1950s was its dual operating regime: the ‘Laobao’ and ‘Gongfei’ healthcare
schemes operated in the cities benefiting urban workers. The former covered work-
ers in state-owned and cooperative businesses. The latter covered those employed
in the bureaucracy, service and administrative sectors of the state, which included
civil servants, teachers, doctors and so on (Zheng 2002). At the same time, the
‘Hezuo Yiliao’ or cooperative medicare scheme operated for the population in
rural areas. This dual system corresponded to the differing economic development
of urban and rural areas of the time and was also consistent with the linked but
separate production patterns in which state ownership dominated in cities and
collective ownership dominated in rural areas.
The Laobao scheme came into being with the 1951 ‘Labour Insurance
Regulations in China’. The beneficiaries of this system included workers and their
260 Jufen Wang
direct relatives based in state-owned enterprises. Insurance covered all the workers’
own medical expenses and 50 per cent of their relatives’ medical expenses. The
following year, in August 1952, the government released a set of detailed regula-
tions for a second healthcare insurance system that became known as the Gongfei
system. The latter applied to workers in government institutions, the medical and
education sectors and other public-sector-oriented services. In the post-Mao
economic reform era serious drawbacks in the Gongfei and Laobao systems have
appeared and the main problems are discussed here with particular reference to
the structural inequalities. The issue of geography is also pertinent as is the process
of devolvement of responsibility by the central government.
One of the first casualties of economic transition has been the two main
providers of healthcare within the social security system of Maoist, socialist urban
China: the Gongfei and Laobao systems. Neither system is now supported by the
realities of the market-oriented economy. Previously, state ownership had domi-
nated the Chinese economic structure and the dominant state-owned economic
structure ensured that all of the urban population had access to health insurance.
At the end of the 1970s, China began its economic reforms by restructuring the
patterns of ownership. The new policy accepted state ownership as the main type
of ownership, while all other patterns of ownership were also to be allowed.
Private enterprises, joint ventures and other forms of companies and institutions
were henceforth permitted, ushering in a new era of economic flexibility. With the
introduction of this major structural reform and the accompanying social change,
the urban population found that they had other options to add to the once abbre-
viated list of employment with either stated-owned or collective companies.
However, while non-state-owned enterprises and companies (SOEs) proliferated
they were also excluded from the established system of healthcare, something the
population rapidly became aware of. The continued benefits of the Gongfei and
Laobao systems not only affected the way in which the newly created labour
market operated but also the nature of uptake of care. A survey conducted in 1998
by the State Medical Service determined that the highest hospitalization rate in
the urban population was amongst those covered by the Gongfei (89.97 per cent)
and Laobao (60.32 per cent) insurance systems. Those who have to pay at their
own expense exhibited the lowest rate of hospitalization, at only 29.49 per cent.
The main explanation for this differential was the nature of entitlement to healthcare
insurance (Zhang et al. 2003: 148).
Furthermore, the original medical insurance system as it operated through the
Gongfei and Laobao systems led to conflicts concerning the schemes’ financial
solvency in the post-1978 reform period. Neither system had established its own
funding source: the Gongfei scheme was funded by governmental appropriations
while the Laobao scheme was funded from the business welfare funds within each
individual enterprise. In the post-1978 market-oriented economic system, the
Laobao insurance system had to rely heavily on the new concept of business
profits. In contrast the Gongfei system was affected by the capacity of local gov-
ernments to raise their own revenues rather than rely on centrally distributed
funds. A further problem lay in the way in which the compensation mechanism
for hospitals operated. Government budgets were proving insufficient to fund

hospitals, leading to an imbalance between the supply of medical services and
compensation for costs arising through the normal provision of those services.
Hospitals now increasingly depend on profits generated from the difference
between service price and the cost of the medicine, aided by discounts offered by
the pharmaceutical industry and the expansion of commercially oriented business
practices to fund hospital costs. State-owned hospitals are thus facing problematic
choices about how and whether to maximize profits.
One of the ways in which hospitals have pursued profits has been to increase the
charges for people treated under the Laobao and Gongfei systems. Unfortunately, the
new costs often exceed the financial capabilities of the government and enterprises
and result in difficulties in implementing either of these insurance systems. Drug
and treatment innovations, while of great benefit to patients, also increase the
cost of healthcare. For instance, from 1979 to 1999 the average rate of medical
costs for employees in Shanghai increased by 29 per cent per annum, significantly
exceeding the average annual rate of increase in GNP of 9.5 per cent (Chen and
Chou 2002). The regulations of the urban medical-care insurance system
constructed in the 1950s undertook to cover all medical expenses for all types of
treatment needed by the patient. This aspect of the insurance system further
exacerbated the growth of medical expenditure. Unsurprisingly, by the 1990s, the
traditional urban healthcare insurance system faced serious fiscal problems. In
particular, since Gongfei medical care was funded by local governments, the by-
now huge expenditure on healthcare placed large financial burdens on the entire
society, threatening and restricting the development of the other sectors.
Successful enterprises faced development problems as a result of bearing the heavy
medical expenses for those employees covered by the Laobao scheme. In the most
extreme cases, as a direct result of medical outlays, enterprise finances lay in
arrears. Such enterprises could not afford to continue paying the medical expenses
for Laobao patients, especially those with chronic illnesses. By the end of 1999 the
total medical expense arrears for Laobao insurance in Shanghai alone totalled
almost 20 billion Yuan (Chen and Chou 2002).
The Gongfei and Laobao urban healthcare schemes were based on the ‘Hukou’
or household registration system, according to which the Chinese population is
subject to a residence-based system which controls access to goods, services and
products. Those in urban areas were in receipt of considerable subsidies which cov-
ered a wide range of goods and services and which were denied to the rural popu-
lation. The Hukou system also prevented migration from rural to urban areas.
Since the establishment of the reform policy and mass urban-ward migration the
registered urban population under the Hukou system no longer corresponds to the
actual urban population. It is estimated that about 120 million rural workers are
living in, or migrating between, cities all over the country. Very few are granted a
formal registration permit allowing them access to the medical and education sys-
tems and the other benefits accruing to the recognized urban population. By 2004,
Shanghai’s migrant population is estimated to have reached 4.99 million (CPN
2005). Most did not have medical insurance.
262 Jufen Wang
Reform of the Chinese urban health care insurance system
Reform of the Chinese urban healthcare system started with the delivery of the
document on ‘Decisions on the establishment of basic healthcare insurance for
urban employees’ by the State Council in December 1996. There had been earlier
attempts to reform the health insurance system. In the early 1980s, reforms aimed
to reduce expenditure on medical services. These State Council-led experimental
strategies considered the establishment of special insurance funds for particular
diseases, cost-sharing strategies and careful accounting of the population covered
by insurance schemes (Liang et al. 2003: 40). A variety of healthcare models were
piloted at the local-area level, notably the Liangjiang Model named after the cities
of Zhengjiang, in Jiangsu province, and Qiujiang, in Jiangxi province. The three
central features of the Liangjiang model were: a public fund, personal accounts,
and a three-stage payment system. By 1996, the experiment had been expanded to
nearly forty cities.
At the end of 1998, the State Council promulgated a further document entitled
‘Urban employees’ basic healthcare insurance scheme’ which stated that basic
healthcare insurance schemes should be managed at the local authority level and
reflect current levels of economic development. The implication of this was that
areas with higher rates of economic growth would be expected to establish more
comprehensive forms of healthcare insurance for their population, while govern-
ments in low-growth areas would not be expected to achieve the same standards of
healthcare insurance, thereby resulting in a patchwork of differing health insurance
regimes across the country and with restrictions on the transfer of policies
between the cities. Despite this flexibility with respect to economic context, all
urban residents are required to participate in the scheme. The document also
requires that under this new national insurance system costs be jointly borne by
employers and employees (http://www.em800.com).
Health care reform in Shanghai

Under the articles of the 1998 State Council document, Shanghai was directed to
put together a multi-level medical insurance system, building on the reforms to
the Gongfei and Laobao systems. The aim was to enable different types of medical
insurance systems (including commercial insurance) to operate and respond to the
needs of different groups of people. Shanghai responded not just by implementing
a basic healthcare insurance scheme but also by broadening the remit of schemes
to include serious disease and hospitalization insurance plans under the aegis of
the Shanghai Trade Union, childbearing or maternity insurance, comprehensive
insurance for migrant workers and complementary healthcare insurance for civil
servants. These are discussed later. The trade union does not provide any services;
rather, the function of the trade union’s insurance programme is to only remit a
certain percentage of the costs to the individual.
The new basic healthcare insurance system for urban employees was implemented
in 2001 and currently represents the major part of the healthcare insurance
system. It covers workers recorded in Shanghai’s household system or who have
been granted ‘green cards’ for residence in Shanghai.2 Basic insurance is also
granted to retired workers, mainly from the city of Shanghai. It excludes those
retired workers who worked in other cities and returned to Shanghai upon retire-
ment. This exclusion results from the way in which healthcare insurance schemes
are tied to individual cities and reflects the local and non-transferable nature of
accumulation and social responsibility of devolved governments. Moreover, as
employees’ healthcare insurance is funded by both employers and employees
themselves, healthcare provision is situated within the workplace and is not trans-
ferable if or when an employee moves to another job. Participants in the basic
healthcare insurance programme can choose to access medical services anywhere
within a city, except private hospitals. Should workers move to another city, their
right to a health insurance scheme remains with their original workplace, and
they must start a new scheme with their new employers. Such health insurance
schemes exclude the unemployed who only have rights to unemployment
insurance. However, the self-employed can be included if they choose to pay the
insurance premiums themselves.
Five characteristics of the post-reform urban basic healthcare insurance, as it
operates in Shanghai, can be differentiated. Firstly, the new scheme not only
combines aspects of both the Gongfei and Laobao systems but has been extended
to employees of all sectors. Secondly, the establishment of public funds has been
an innovation. Public funds improve risk-management capacities and allow funds
to be used across broader sections of society, acting as a mechanism for the reallo-
cation of national income. Additionally, public funds support socioeconomic
development in a long-term and sustainable manner. Thirdly, the post-reform
scheme has emphasized the shared responsibilities of the state, employers and
employees. The public fund is financed both by employers and employees
through contributions from employers equivalent to 12 per cent of an employee’s
salary. Employees meanwhile contribute 2 per cent of their wages. The first 50 per
cent of the payment made by the employer goes into the public funds, henceforth
referred to as social funds, while the second 50 per cent goes to an employee’s
individual account along with the employee’s contribution. Fourthly, the new
basic healthcare scheme aims to be more financially efficient. It emphasizes indi-
vidual responsibility as well as the benefits of healthcare insurance. The amounts
transferred into a personal account from the social funds vary according to the age
and the level of individual contributions. Individuals do have a certain degree of
individual control over the funds. For example, since there is no time limitation
on the use of the funds in each personal account, should the fund not be used it
can be accessed during older age. It is hoped that this aspect of the policy will
motivate people to save into the funds as the monies are never lost to the individual
and can be transferred to relatives or children.
Fifthly, healthcare insurance is utilized via sequential stages of entitlement: per-
sonal account, individual payment and shared responsibility. With diagnosis of a
disease or injury, patients commence using the funds in their personal accounts to
pay for their healthcare expenses. After using up the money in this account, they
reach the second stage of individual payments whereby the costs must be borne by
264 Jufen Wang
themselves or their family. Once their payments for health care reach a sum
equivalent to 10 per cent of the average wage in Shanghai for the previous year,
the system moves into the third stage. At this stage, the costs are recouped from
individuals and the social fund. The level of individual financial responsibility is
correlated with both the grade of the hospital used by the patient and the age of
the patient. Individuals bear the largest portion of medical expenses when in
receipt of treatment from the most prestigious hospitals; individuals bear an inter-
mediate cost of medical expenses if they seek treatment in second-grade hospitals
while they pay the least amount if they use third-grade or community hospitals.
These sums are also calculated according to age, with older people paying a smaller
proportion of the incurred expenses compared to younger age groups. When
expenditure exceeds 400 per cent of the average wage of the preceding year public
insurance funds are closed to a particular patient.
Gender implications of urban health care reforms

The fundamental aim of the reforms was to secure the basic medical needs of
urban workers but the new health care scheme has also generated some urgent
problems. These problems have important implications for social equity, the avail-
ability of urban medical services and the coordinated development of society as a
whole. The two important indices used in examining and evaluating the degree of
social equity inherent in health care insurance systems are extent of coverage and
benefit levels. The following discussion explores whether the new health care
insurance scheme can bring equal benefits for male and female workers, and
whether or not it accords with the national policy of gender equity.
The conditions of admission to, or coverage guaranteed by, the insurance
schemes, shows that having employee status is essential. But the extent and types
of employment opportunities for women, especially formal opportunities, are
more restricted than for men. This is unquestionably a barrier for women wishing
to join the basic health insurance scheme; the number of female workers eligible
for health care insurance is certainly smaller than the number of male workers. As
insurance statistics published by the state and Shanghai city are not classified by
gender, it is only possible to infer from the principle of this scheme that it is likely
to cover male workers more effectively than women workers. Data from Zhenjiang
province supports this hypothesis. The numbers of participants in the Zhenjiang
basic health care insurance scheme in 1997, 1998 and 2001 remained almost
unchanged at nearly 280,000. Men made up more than 57 per cent of the scheme’s
population (Jiang et al. 2004). Different employment rates between genders will
unavoidably mean that women will not benefit as much as men from the urban
basic health care insurance system.
The current basic medical insurance for urban workers does not take into
account gender differences in income. Like other countries in the world, the
average wage for males in China is higher than that for females. Currently,
women’s wage levels are only 70.1 per cent of those of men. The implementation
of economic reforms in China has done little to reduce the gender wage gap;
instead the gap has been widening (Li et al. 2003). The present basic health care
scheme for urban workers fails to consider this gendered wage differential. As
contributions to the health insurance scheme are based on wage levels and as
men’s average wages are higher than women’s, men’s contribution to the social
funds will be greater than those of women. This means that women are unable to
build up the same levels of insurance premiums as men, which in turn means that
they will not be able to access the same levels of professional care as men.
In addition, the average age of retirement for women is five years earlier than
that of men. Thus, men will contribute more to insurance funds over their
working lifetime compared to women. Furthermore, as the regulations governing
medical insurance mean that the retired population can utilize the funds on retire-
ment age, women will access their funds sooner than men and over a longer
period as female life expectancy is higher. According to the Shanghai Statistical
Yearbook, the average life expectancy of women in Shanghai reached 81.8 years in
2003, while that of men reached 77.8 years (SSB 2004). If it is assumed that women
will retire five years earlier and die four years later than men, they will need to use,
potentially, some nine years more of funds compared to their male counterparts.
The insurance system therefore effectively discriminates against both men, who
make greater contributions, and women, who do not have access to the same level
of funds.
Age is also an important factor with regard to the basic health care insurance
program for urban workers in Shanghai. Age is important for a number of reasons.
Older people are more susceptible to ill-health. Different age cohorts experience
differing exposure to environments, diseases and events which may compromise
health and well-being. In general, the older population survive on their pensions
and rarely have access to surplus income which they can invest in insurance
schemes. Additionally, since the reforms’ principles mean that funds in individual
accounts accrue over a lifetime the older population have been excluded from the
long-term accumulation process. However, provision has been made to credit
the insurance premiums of the older population: retired workers are not required
to make contributions to health care insurance, rather their accounts are credited
using the social funds, and the payments they are required to make, for outpatient
care for example, have been reduced in comparison to the working population.
Considerable research underpins the identification of differences in life
expectancy for women and men. Women experience longer durations of ill-health
at older ages compared with men. In Spain, in the over-65 age group, women expe-
rience, on average, some 12 years of health problems at a level sufficient to require
bed-based rest; the average for men is nine years (United Nations 2000). Analyses
on the healthy-life expectancy of Shanghai citizens conducted by the Shanghai
Disease Prevention and Control Centre indicate that over 50 per cent of women’s
excess average life expectancy is lived in a non-healthy state (Zhou et al. 2001).
Chronic diseases result in medical expenses and high nursing fees, which can
easily become a heavy burden for older women. The current basic health care
insurance scheme stipulates that once the stage of shared responsibility has been
reached, pensioners must pay between 10 per cent and 20 per cent of their total
medical expenses, regardless of gender. Owing to women’s lower wages and
266 Jufen Wang
reduced pensions, the same share of total medical expenses imposes a heavier
burden on female pensioners.
The threshold set for individual payments is relatively higher for the female
population than the male. As outlined above, insurance scheme regulations sepa-
rate medical payments for health care into three sequential phases: a personal
account stage; an individual payment stage; and a stage in which costs are shared.
The third stage is only encountered once individual payments for health care
have exceeded a sum equivalent to 10 per cent of the average wage in the preced-
ing year. Since the amount of money in the individual payment stage is not
directly related to individual income, actual income will affect and restrict access
to medical services when needed. This is because 10 per cent of the average wage
of the whole city in the previous year is a lower sum for high-income earners
compared to those on low incomes. As women’s wages are generally lower than
men’s, this aspect of the insurance system has important implications for women
through restricting their access to medical services. Given that the average male
wage equals the average wage of the city as a whole, an individual payment rep-
resents one month’s salary for a man. For the average woman, whose wages tend
to be less than the city average, an individual payment represents a month and
one quarter’s salary. Consequently, with respect to proportional income, female
workers are required to pay more for the same medical services. The 10 per cent
threshold was originally intended to ensure that insurance participants did not
abuse the shared-responsibility aspect of the system by requiring patients to first
pay for a certain share of their own medical expenses. The unequal gender
impacts are an unfortunate outcome of this attempt to counteract abuse of the
system.
Finally, because of the exclusion of workers’ spouses and children the current
basic health care scheme for urban workers augments the burden on single moth-
ers. Though no document has ever announced the exclusion of spouses and chil-
dren to medical benefits (there were different provisions in the Gongfei and
Laobao systems), the majority of units have abandoned this original broader
aspect to medical health care. Some units still follow the previous Gongfei and
Laobao regulations on the inclusion of spouses. Since children of divorced families
tend to live with their mothers, any medical expenses will add to their mothers’
financial burdens. This is not the only kind of bias that adversely affects women in
a strictly individualistic system such as this; the other exclusion is of women who
are full-time unpaid carers (taking care of young children, the elderly and so on)
who will not be able to receive any coverage as ‘dependants’ (if their husbands are
covered).
Gender differences in the accumulation of individual insurance accounts

The accumulation of individual insurance accounts is a new aspect of medical
insurance in China. This component of the system was drawn from the
Singaporean experience which emphasized individual responsibility in order to
try and reduce overuse of medical services. But, as discussed above, a scheme based
Table 12.1 Shanghai social healthcare scheme, contributions by percentage and age
Age Percentage of wages contributed to scheme
Below 35 years 2% of the individual average wage ⫹

0.5% of the average wage
35–44 years 2% of the individual average wage ⫹
1% of the average wage
45 years to retirement 2% of the individual average wage ⫹
1.5% of the average wage
Retirement to 74 years 4% of the average wage
75 years and older 4.5% of the average wage
Source: Shanghai Health Care Insurance Bureau (2000).
on wage contributions and employment status results in gender differences in the

accumulation of funds in individual accounts. This is shown in the above table,
which illustrates the range of health care insurance contributions by age and wage.
Table 12.1 shows that, under the present scheme, the funds in an individual’s
account funds can be divided into two parts dependent on the sources of monies.
The first part consists of the health care insurance fees paid by individuals and cred-
ited into individual accounts; the second part consists of an allocation from the
social fund (calculated by average wage). Though the latter payment is the same for
both genders of the same age, the former payment is based on individual incomes.
Thus, women’s lower wages directly result in smaller average sums in women’s indi-
vidual accounts compared with those of men even if men and women had started
work and retired at the same age. Additionally, since wages increase with age and
experience, an earlier retirement age for women means that they lose out on the
opportunity to benefit from the most senior posts and most advantageous salaries.
This aggravates the gender inequity in individual account accumulation.
Furthermore, as part of the individual account comes from the social funds, and
is based on the social average wage, the earlier retirement policy for women deprives
them of a further opportunity to benefit from higher-level salaries. All these ele-
ments point to the lower capacity of female individual accounts to accumulate
funds compared to their male counterparts. This lower level of fund accumulation
combined with women’s longevity and higher probability of ill-health in extreme
old age will mean that women will encounter illness and poverty simultaneously.
Differential impacts of the shared responsibility aspects of the

insurance system
In the shared-responsibility stage, under the principles of joint individual and
social responsibility, individual payments are made according to age, and grade of
hospital, but regardless of income, and can range between 10 per cent and 55 per
cent of the total medical expenses. As a result of the gender wage differential, and
assuming an equal probability of disease and access to the same grade of hospital
care, women pay more, relative to income, than men of the same age.
268 Jufen Wang
In order to explore this further it is assumed that consumption and needs can be
divided into three categories in a model based on Maslow’s hierarchy of needs
(Maslow 1970). The first of these categories of demand is basic subsistence (food,
shelter and transportation), the second, health (medical care), and the last
category is leisure (culture, recreation and travel). These three categories differ in
income elasticity, the first category having the least income elasticity, the third the
greatest. Given gender income inequity (in both wages and pensions) women tend
to sacrifice leisure for the sake of health, and health is sacrificed for basic subsistence.
Research conducted by the author has revealed that, having used up the funds
in their individual accounts, some elderly women did not access necessary health
care until monies were credited to their accounts the following financial year;
some women postponed necessary treatments and operations through poverty.
Although this situation may also apply to men, it occurs relatively more
frequently in the female population especially in the lowest-income groups,
clearly revealing the problematic aspects of this medical insurance scheme.
Table 12.2 shows the relationship between age, hospital capacity and the percent-
age of costs paid by individuals and from social funds. Generally, the younger an
individual is and the further into the future their predicted retirement the lower
the percentage of medical costs paid for through social funds and the greater the
individual contribution.
Current practices embedded within the provision of basic medical insurance for
urban workers overlook physical differences, different morbidity patterns and con-
sequently different medical demands between genders. These patterns are particu-
larly evident in the older population. In 2003, elderly citizens aged over 65 years
accounted for 14.8 per cent of the total population of Shanghai (CPN 2005).
A recent survey conducted by Ministry of Public Health indicates that 71.3 per
Table 12.2 Standard scale for the shared-responsibility stage of urban insurance, Shanghai
Shared responsibility section
Hospital grade Individual Supplementary funds

payment % (paid by social
Age group contributions) %
Born before 31 Dec. Grade 1 15 85

1955, retire after 1 Jan. Grade 2 20 80
2001 Grade 3 25 75
Born 1 Jan. 1956–31 Grade 1 30 70
Dec. 1965, retire after 1 Grade 2 35 65
Jan. 2001 Grade 3 40 60
Born after 1 Jan. 1966, Grade 1 45 55
retire after 1 Jan. 2001 Grade 2 50 50
Grade 3 55 45
Start work after 1 Jan. Grade 1 45 55
2001, retire at future Grade 2 50 50
date Grade 3 55 45
Source: Shanghai Health Care Insurance Bureau (2000: 39).

cent of elderly citizens suffer from chronic diseases, the percentage being more
than twice that of the general adult population. Additionally, 42 per cent of
elderly citizens suffer from more than two chronic diseases (Ou 2002). In an
ageing society, older female citizens are subject to more health problems than
older male citizens. ‘Despite longer average life spans, the health circumstances
and quality of life of female elderly citizens are worse than those of the male’ (Zeng
and Gu 2002: 59–66).
The basic health care insurance system currently offers the greatest coverage and
best policies for urban employees in Shanghai. It also covers most of the demands
on the medical services. Compared with other cities, Shanghai’s citizens enjoy
one of the best insurance schemes in the country. In comparison, elsewhere
policyholders only receive help with medical expenses once hospitalized.
Complementary health care insurance
The complementary insurance systems operated by the Shanghai Trade Union

are supplementary to the basic medical insurance for workers and the retirees.
These types of insurance complement and extend the remit of the basic health
care insurance system. The most important forms of complementary insurance
are: serious disease insurance (e.g. renal failure, cancer etc.), hospitalization
insurance, and an insurance plan for pensioners. All these schemes include both
current and retired workers, but they must be members of the trade union and must
join the insurance in employment-based units (at least 75 per cent of employees of
a particular company or business must agree to participate). Therefore, if there is
no trade union in the work unit or insufficient people want to join, an individual
worker cannot participate. Membership of the trade union is a basic condition for
accessing this insurance scheme. In some areas there is an attempt to extend mem-
bership of the trade union to self-employed workers. Such schemes do not increase
the number of people able to receive health care insurance but they do improve the
nature of the insurance as those who become ill will receive greater compensatory
amounts. Further insurance packages are available to sub-groups within the trade
union, such as women who can choose to participate in insurance schemes which
cover for cancers of the reproductive system such as breast and uterine cancer.
Maternity insurance
Under the Gongfei and Laobao programmes, a woman’s entire medical expenses
during pregnancy and while on maternity leave were covered, with salaries
provided by companies and businesses. With health care insurance reform, the
restructuring of ownership of businesses and industry and especially the introduc-
tion of profit-focused, market-oriented managerial methods, companies started to
consider expenses related to child bearing as an unwelcome and additional
burden.
To provide salaries and medical expenses during pregnancy and the post-natal
period is to both acknowledge women’s particular social contributions and
promote women’s equal employment rights. With the support of the All China
270 Jufen Wang
Women’s Federation (ACWF) at national and local level, the State Council issued
documents establishing maternity insurance. This was part of a promise made at
the Fourth World Conference on Women in Beijing in 1995. As a part of the
reform of the whole of the Chinese urban medical insurance programme,
maternity insurance will play an important role in enhancing women’s health.
Maternity insurance guarantees the following two basic rights. Firstly, ante-natal
care and expenses related to childbirth is covered. Secondly, it guarantees the con-
tinuation of salaries during maternity leave. It could be argued that this insurance
represents progress in the protection of urban women’s rights to social security.
However, although restrictions based on household registration do not exist
theoretically – meaning that all working women of childbearing age can participate –
some migrant women are excluded because they are not covered by the basic
urban health care insurance system.
The Shanghai municipal government initiated maternity insurance in October
2001. This scheme aims to improve women’s employment circumstances and
secure their entitlement to maternity benefits. Prior to the implementation of this
insurance, employers paid medical expenses during pregnancy and the salaries of
women on maternity leave. Now, these expenses are funded through the new
maternity insurance system. However, monies are not specifically collected for this
system; rather they are transferred from the basic urban health insurance schemes
and from pension funds. The advantage is that there is no need for society to raise
money separately. However, coverage is limited to those female workers already
enrolled under the basic health care insurance system. Therefore, a large number
of migrant women, who work outside the formal employment sector, are excluded
from this insurance scheme although they too are working in Shanghai. Under the
terms of this new plan, eligible women are entitled to four and a half months’
maternity leave while receiving 100 per cent of their wages and near-free medical
care during pregnancy and birth if their child is delivered in mid-grade hospitals.
The levels of insurance guarantees for participants are higher in Shanghai than
elsewhere in terms of maternity wage levels, length of maternity leave, as well as
coverage of care.
In China as a whole, however, maternity insurance has developed comparatively
slowly. By the end of 2003, some 36,550,000 women had participated in maternity
insurance schemes throughout the country (Xinhua News Net 2005). Some 16
provincial capitals have formulated regulations on maternity insurance.3
Moreover, maternity insurance in many places only targets registered permanent
residents, a residual from the former planned economy, and thus excludes migrant
women.
Comprehensive insurance for migrant workers

The medical-insurance reform process and the resulting new insurance system are
applicable to registered permanent urban residents whilst labourers working in
cities without urban residence permits are excluded. Concerned government
departments identified this omission in the health care insurance system and
established a ‘comprehensive insurance programme for migrant workers’. This
insurance covers workers migrating into Shanghai, but it is based on formal

employment and consequently excludes certain populations such as those work-
ing in the informal sector. There are two competing theories for this omission.
Firstly, it is thought that this may be a consequence of the original aim of the
insurance programme which was to provide a fund for industrial injuries, assumed
to be irrelevant for domestic servants and agricultural workers. Secondly, the
omission may be due to the uncertainty experienced by agricultural workers and
domestic servants with regard to wages, which are often paid on an hourly basis,
as well as their high levels of mobility. Workers eligible for this programme are
predominately male and thus male migrant workers constitute the major beneficiary
group with respect to migrant insurance.
Comprehensive insurance for migrant workers was enacted in Shanghai on
1 September 2003. This insurance plan includes policies protecting against work
injuries, provisions for in-patient expenses, and old-age subsidies. Thus, strictly
speaking, this type of insurance is not purely health care based, but a broader
type of insurance devised for the population of migrant workers. Under the reg-
ulations governing coverage, all enterprises irrespective of ownership type are
required to participate in this scheme if they employ migrant workers, and
cover the costs. Some self-employed migrant workers who have community-
based employment cards are also required to participate in this scheme.
However, this scheme does not include migrant workers engaged in domestic
service and farming. Up to 98 per cent of the former are women; the latter are
often men.
Insurance for migrants includes industrial injury insurance (or accidental injury
insurance), insurance for hospitalization and subsidies for members of the older
population. While this is not medical insurance in the strictest sense, it provides a
certain amount of protection for migrant labourers. But if it was intended to pro-
vide social security for all migrant workers, it has badly neglected the interests of
female migrants. If this insurance is regarded as an initial attempt to protect the
rights and interests of the population lacking urban residence permits then more
importance has been continually placed on male health than on female health
(especially where industrial injuries are concerned). In particular, women, as the
population who bear the task of reproduction with its particular health care needs,
have been neglected.
Complementary health care insurance for government officers

A final new complementary health care insurance for government officers, which
will not be discussed here, replaced the Gongfei system for this employment sector
and guarantees a certain level of health care as per the principles of urban health
care reform laid down by the State Council.
Supplementary health care scheme policy and improvements in

gender equality
A supplementary policy to reduce the financial burden on poorer people in
chronic ill-health was set up on 15 October 2003 and enacted on 1 April 2004 in
272 Jufen Wang
Shanghai. The new policy favours those who are low-income earners and who
have been diagnosed with chronic diseases because the proportion of the medical
expenses borne by the individual is closely related to his or her wage or pension.
The supplementary policy was subsequently revised in April 2004 and its main
items are outlined in Table 12.3 below.
Table 12.3 demonstrates the way in which women, who tend to be low-income
earners, are more likely to be able to reduce the costs of their personal medical-
care expenses. People who for example earn a sum equivalent to, or less than, the
average minimum wage in Shanghai in the year preceding the onset of medical
expenses receive a 90 per cent reduction in their medical costs once those costs
have surpassed a figure equivalent to 30 per cent of the average minimum wage in
Shanghai in the preceding year.
The supplementary aspects of the basic health care insurance system have had
marked benefits for women as they challenge the gender inequalities of the previous
system. Thus, to some extent these subsidiary aspects have meant that gender
equality has been reached in Shanghai’s urban health care scheme.
Examination of the current basic medical insurance for urban workers shows that
it fails to consider gender throughout its structure including in the scope of insur-
ance and in the share of payment. Based on age alone, the scheme appears to be
gender neutral but, in effect, brings out inequities in rights and interests by ignor-
ing the differences between genders in social status, employment, income, age at
retirement and health status. Fortunately, the supplementary policy demonstrates
the gains which can be made in the implementation of national gender equality.
Table 12.3 Main aspects of the 2004 supplementary policy on reducing medical care
expenses, Shanghai
Personal wage level Individual payments for Reduction rate once

medical care expense identified proportion
has been exceeded
Wage level equal to, or less Accumulative expenses up to 90%

than, the minimum wage in 30% of minimum wage of the
the city in the preceding year city in the preceding year
Wage level higher than Accumulative expenses up to 90%
minimum wage but less than 40% of personal wage
1.5 times the average wage in
city in the preceding year
Wage level higher than 1.5 Accumulative expenses up to 90%
times minimum wage but less 50% of personal wage
than 3 times the average wage
in city in the preceding year
Pension is worth less than Accumulative expenses up to 90%
minimum wage level 30% of personal pension
Pension is worth more than Accumulative expenses up to 90%
minimum wage level 40% personal pension
Source: Shanghai Social Security Bureau 2004, No.126.

Further improvements to the urban health care insurance scheme

and policy implications
Expanding participation
The criteria for judging a social insurance scheme should highlight coverage and
benefit levels, although national and local levels of economic development, and
the redistribution of socioeconomic resources must also be taken into considera-
tion. Currently, basic health care insurance for urban workers, the principal part of
Chinese urban health care security, is based on employment and permanent resi-
dence permits and thus results in the exclusion of the lowest stratum of society
which encompasses the unemployed, informal workers and migrant labourers.
These aspects of the social security entitlement system not only hamper coverage,
but also compromise the size of available funds. For example, by the end of June
2003, the basic health care scheme covered 42 per cent of the total number of
urban workers in China (Chou 2005). Currently, this means that some 109.02 million
members of the urban population are included within the scheme, of which 79.75
million are workers and 29.27 million are pensioners (PRC Year Book 2005: 414).
Pensioners account for some 26.8 per cent of the population entitled to the
health care insurance system. Pensioners pay no contributions into the scheme
but are heavy users, which places a drain on insurance funds. Statistics from
the Medical Insurance Clearing Centre of Zhenjiang City reveal that in 2001 the
average sum paid out for medical expenses per worker was 673.05 RMB, while
that per pensioner was 1651.66 RMB, some 2.45 times higher (Li 2003: 14).
‘In 2001, retirees accounted for 27.40 per cent of the total participants [of the
health insurance scheme] but consumed 44.84 per cent of the total medical
expenses and 48.42 per cent of the total medical funds’ (Jiang et al. 2004: 18). Yet,
health care insurance policies that have favoured retirees still fail to meet the large
medical demands of the older population. If current practices of allowing pen-
sioners to consume the majority of social funds continue, not only will younger
participants’ interests be violated, but also the operation of the whole health care
insurance system will be compromised.
To solve this problem effectively, two measures have been identified. First, the
authorities should formulate a special programme offering compensatory benefits
to those pensioners who have made due contributions during the planned-
economy era, or establish a subsidiary health care insurance scheme for pensioners,
to alleviate the pressure on the basic health care insurance (which will require sig-
nificant subsidies from the state, both local and central). Secondly, manipulation
of the age structure of participants should take place by allowing more young
migrant labourers working in cities to participate in basic health care insurance.
Recruitment of part-time and informal workers

The rules of entitlement to the current basic health care insurance also create
gender inequity. Under the Gongfei and Laobao systems women in China enjoyed
an employment rate of up to 98–99 per cent and, as the spouses of workers, had
the right to have 50 per cent of their medical expenses reimbursed. Thus women’s
274 Jufen Wang
health care insurance rights were largely secure. However, the rate of women’s
participation in the formal workforce has decreased sharply. A 2003 survey in
Shanghai indicated that the employment rate of women in the 16–54 age group had
declined to 52.6 per cent, and their unemployment rate had reached 22.5 per cent
(Xu and Yuan 2004). Urban basic health care insurance schemes based on employ-
ment therefore presented an institutional barrier to women’s health care insurance
rights. At present, differences between genders in medical insurance rights have
not became obvious because contracts between laid-off women, early retirees and
their former work units require those units to continue to pay contributions
towards women’s health care insurance and pension schemes, thus fundamentally
reserving their rights to basic health care. In the near future when the levels of
unemployment among young women, the increase in the proportion of female
informal workers and the trend towards short-term employment become significant,
the gender differences in urban health care insurance rights will become a
more visible form of social injustice in China. Apart from taking measures to
secure women’s equal employment rights, the government should also further
improve the regulation of basic health insurance, through the inclusion of part-time
workers and informal workers in the urban basic health care insurance scheme.
For a significant period, the health care security system in China was binary in
nature, being divided between urban and rural locations with city residents
enjoying higher standards of living compared with rural residents. Before health
care insurance reform, the high employment rate in urban areas guaranteed that
nearly every registered permanent urban resident could access health care under
the Gongfei and Laobao systems while the rural population relied on a less
comprehensive cooperative medical service. Entitlement to this latter system
was based on access to agricultural land. In some areas the cooperative medical
services did not survive the agricultural reforms, which led to the financial collapse
of rural health care institutions. This has severely reduced access to any form of
health care in rural areas.
According to incomplete estimations, there are some 120 million migrants from
rural areas working in China’s large cities. Part of this population has been living
and working in the cities for up to ten years and can be considered at the least semi-
permanent residents. This shift of rural labourers to cities, as we have seen, has
brought with it medical security problems. Despite reforms, entitlement to the
basic health care insurance in urban areas still rests on registered permanent
residence or ‘green cards’ which grants access to urban privileges for a select
population. Chinese health care insurance reform is still bound by traditional
thinking, which formulates social or public policies based on the inflexible
residence-registration system. As a result the differences in urban and rural social-
security systems intensify and social equity problems crystallize. The form that any
social security system for rural migrants may take has aroused heated discussion. In
Shanghai, while a comprehensive insurance for rural migrants has been estab-
lished, only a minority of rural migrant workers can participate. Due to the high
mobility of rural migrant labourers and the desire to reduce costs, employers have
not made contributions. Additionally, a large number of rural migrant labourers
work within the informal employment environment and are excluded from this
insurance system.
More attention should be paid to the increasing number of women migrating
into cities who are excluded from medical insurance systems. In 1988, the gender
ratio of the migrant population in Shanghai was 241 men to 100 women. By 2003
it had fallen to 134 men for every 100 women. More female migrants are excluded
from the medical insurance system because more undertake independent, informal
work for which there are no insurance provisions. The most significant health
problems experienced by female migrant workers are those related to the repro-
ductive health system and are also an important barrier to the further improve-
ment of maternal health standards in Shanghai. Migrant women are less likely to
participate in ante-natal care (Zhan et al. 1999) and experience problematic
outcomes. In Beijing, the perinatal death rate for migrant pregnancies is 5.7 times
higher than that of the local population (Huang 1999). Different levels and
patterns of health care for women with different residences have affected the
health of mothers and foetuses. Some pregnant migrant women, excluded from
urban health care, return home to deliver their babies and to search for under-
qualified doctors whose fees are cheaper. Thus, they are further marginalized with
respect to their maternal health compared with women holding Shanghai
residence permits. The main solution to the problem of the provision of health
care insurance for migrant women would be to break the link between registered
residence and access to health care and admit migrant women workers into the
basic health insurance system as well as maternity insurance. An alternative would
be to establish migrant childbearing insurance via an individual payment system
(but without government subsidies this will not be viable given migrants’ irregular
incomes and the unwillingness of employers especially in the informal economy
to make contributions). Relevant government departments should also consider
subsidizing the migrant childbearing insurance system. The migrant women’s
maternity insurance could be run by insurance companies under government
supervision. This would also aid the government’s aim to establish a multi-level
health insurance programme for the population of China.
Conclusion
China is a developing country with a huge population and had experienced nearly
twenty years of intense reform and change by the end of 1990s. By this stage there
were serious fiscal concerns with regard to the provision of social security. This
chapter has outlined the development and progress of health care reform from
1998 to date. Bearing in mind the nature of the needed change and the size of
the population, the new policy has been a marked success.
Yet, detailed examination reveals that the current basic medical insurance for
urban workers failed to consider gender at the design stage, whether in terms of
the scope of basic insurance or in the share of payment. Based on age, the scheme
appeared to be neutral and equal but in practice exacerbated inequalities. Shanghai’s
health insurance scheme overlooked initial differences between genders in social
276 Jufen Wang
status, employment, income, retirement age, physical features and health status at
the inception of the scheme, resulting in gender inequity in rights and interests.
Fortunately, the supplementary policy has brought about reductions in medical
expenses based on individual wage and pension levels and presents a national
example in the implementation of China’s national policy of gender equality.
Results have shown that there is still room for improvement as the current scheme
excludes informal workers, migrant workers and worker’s spouses. However, it repre-
sents a significant move forward especially when considering that the supplementary
policy was not designed specifically to resolve gender inequalities. It is also true that
the solution to certain social issues also requires economic development and time.
The lessons we can learn from China’s urban health care reform are: firstly, the
realization that mainstreaming gender issues is still a challenging process for all of
society; and secondly that the design of any social policy should include gender
considerations from the start and any potential negative impacts on vulnerable
populations should be avoided.
Notes
1 This chapter has benefited from Caroline Hoy’s excellent editing.
2 ‘Green cards’ are normally granted to individuals who originate from outside the city of
Shanghai but who have obtained a post in the formal employment sector in the city, or
contributed to the development of Shanghai’s economy in some way.
3 Namely, Chengdu, Nanjing, Changchun, Ji’nan, Hefei, Yinchuan, Guangzhou, Hangzhou,
Shijiazhuang, Taiyuan, Xi’ning, Kunming, Nanning, and Wulumuqi.
References
Chen, Yani and Yi Chou (2002) ‘Record on Shanghai Health Care Reform’, at
www.people.com.cn, accessed 10 August 2002.
China Statistics Bureau (2004) People’s Republic of China Year Book. Beijing: People’s Republic
of China Year Book Press: 414.
Chou, Yulin (2005) ‘The Circumstances and Improvement of the Health Care Insurance
System in China’. Monthly Journal of Social Security 2, 39. Information Center for Social
Science, People’s University, Beijing.
China Population Net (CPN) (2005) at www.chinapop.gov.cn, accessed 13 January 2005.
Department of Propagation of Shanghai Municipality and Association of Social Science
(DoP) (1998) The Development Report of Shanghai Philosophy and Social Science Research.
Shanghai: Shanghai People’s Press: 401.
Ge, Shouchang (1999) Social Security Economics. Shanghai: Shanghai University of Economics
and Finance Press.
Guo, Weihua and Fucheng Li (2001) The Legal Protection of Workers’ Rights. Beijing Xiyuan
Publishing House.
Hu, Shanlian (2002) ‘Experiences and Issues in Shanghai Health Reform’. Journal of China
Health Resources 5, 5: 202.
Huang, Xiaoling (1999) ‘Study on Perinatal Infant Death in the Migrant Population’. Journal
of Maternal and Child Hygiene 14, 4: 242.
Jiang, Licheng, Linping Xiong, Yiping Li and Xiuqiang Ma (2004) ‘An Analysis of the Medical
Expenses of Insured Government Employees and Workers in Zhenjiang City’. Journal of
Chinese Health Economics 23, 6: 18.
Li, Qiufang, Simei Wang and Xiuhua Jiang (2003) ‘The Circumstances of and Measures for
Chinese Women’s Employment’. China Women’s Express Vol. No. 6.
Li, Yiping (2003) ‘A Study of the Aging of Health Care Insurance Participants’. Journal of
Chinese Health Economics 22, 8: 14.
Liang, Hong, Yuan Cheng, Ying Zhou and Ying Zhu (2003) ‘Inequality Problems with
Medical Services in Cities from the point of View of the Health Care System’. Journal of
Chinese Population Science 1: 40.
Maslow, Abraham (1970) Motivation and Personality. Second edition, New York and London:
Harper & Row.
Ou, Shuisheng (2002) ‘Concerns about the Health Care Security of the Urban Poor’. Journal
of Soft Science of Health 16, 4: 11.
Shanghai Health Care Insurance Bureau (2000) Hand Book of Shanghai Basic Health Care
Insurance for Urban Workers. Shanghai: Shanghai Health Care Insurance Bureau 19.
Shanghai Statistics Bureau (SSB) (2004) Shanghai Statistical Yearbook 3.3 at www.
stats-sh.gov.cn, accessed 5 August 2005.
State Council document on ‘Decisions on Setting up Urban Basic Health Care Insurance for
Urban Employees’, at http://www.em800.com, accessed 14 October 2001.
United Nations (2000) The World’s Women, Trend and Statistics. New York: United Nations: 75.
Xinhua News Net (2005) at www.xinhuanet.com/, accessed on 15 February 2005.
Xu, Ke and Jing Yuan (2004) ‘Investigation into Shanghai Women’s Employment’. Shanghai
Women 10, October: 8–13.
Ying, Xiaohua, Wen Chen, Lijun Huang, Qin Jiang and ShanLian Hu (2004) ‘Equity of
Health Care and Health Finance’. Journal of Chinese Health Economics 23,1: 52.
Zeng, Yi and Da’nan Gu (2002) ‘Dynamics in International Research on the Quality of Life of
the Elderly Population’. Journal of Chinese Population Science 5: 60–1.
Zhan, Shaokang, Lingyu Xie, Changqin Liu and Yangguang Wang (1999) Research on the
Maternal and Prenatal Care of Migrant Women in Xuhui District, Shanghai’. Journal of
HaiXia Health Protection 5: 1–2.
Zhang, Taihai, Maojin Cheng, Hong Wang, Yun Peng and Fang Wang (2003) ‘Thoughts
and Measures for Health Care Insurance of the Urban Poor’. Journal of Soft Science of
Health 17, 1: 7.
Zheng, Gongcheng (2002) Evaluation of China’s Social Security System. Beijing: China People’s
University Press: 120–2.
Zhou, Feng, Guixiang Song, Songbai Lin, Yaling Sun and ZhiYing Wang (2001) ‘Analysis on
Shanghai Population’s Health Expectancy’. Journal of China Health Statistics 18, 2: 86.
13
Secondary Education in the Indian
State of Uttar Pradesh: Gender
Dimensions of State Policy
and Practice
Jyotsna Jha and Ramya Subrahmanian
Introduction
Linking gender and education to the social policy agenda is not straightforward.
Many questions arise, most of which hinge on the kinds of impact education is
expected to have on human behaviour, capacities and skills, and on gender iden-
tities and relations. In that sense, education within the wider social policy debate
is linked to questions of society and citizenship, understandings about modernity
and national identity, and the recasting of notions of masculinity and femininity
at different historical moments of time. The content of education, the impact of
policy choices adopted in the process of meeting universal-education agendas,
and the ways in which equity and social-inclusion concerns are addressed through
public institutions, are among the several issues that arise in relation to the
reframing of social policy from a gender perspective.
A central proposition of this chapter is that the focus on minimum ‘thresholds’
for public investment, in turn derived from the analysis of rates of return to
education, has contributed to the neglect of female post-primary education.
Influenced by Human Capital theory (HCT), ‘gender’ and female education have
been central framing discourses of education policy, resulting in substantial policy
rhetoric and concern about women’s and girls’ education as a lever of develop-
ment and progress. In India, acceptance of this global rhetoric has been mediated
by particular policy choices, which have resulted in the neglect of the secondary
sector, the rise of for-profit schooling at all levels of education, and a fragmented
formal elementary-education system, with particular implications for achieving
gender parity and equality. This has resulted in a range of issues relating to female
well-being being erased from the policy map. Girls disappear off the formal educa-
tion policy agenda past the age of 14, at a crucial age when aspirations can be
channelled into opportunities. In this chapter, we focus on secondary schooling,
which we believe best serves the interests of girls, especially if supported by poli-
cies that expand its availability, address socio-cultural constraints that exclude
278
Secondary Education in the Indian State of Uttar Pradesh 279
girls (both within society and within the school), and keep its costs low. We argue
that the lack of policy focus on secondary schooling for girls is linked to the
curious contradictions between policy rhetoric, on the one hand, and policy
prescriptions on the other, where development visions are not matched by
policy decision-making processes that can realize these visions.
Our choice of Uttar Pradesh (UP) as a case study is guided by three factors. One,
UP is the most populous state in the country, and also considered one of the most
socially ‘backward’ in terms of development indicators. Two, there is recent and
high-quality research material on education in the state, on the political economy
of education (Kingdon and Muzammil 2003), on sociology, gender and education
(Jeffery et al. 2004), and more generally on education provision (Dreze and Gazdar
1996). Three, there has been a concerted attempt in the state to focus on
reproductive health and fertility decline.1 There are also innovative programmes
for empowering women through education that operate outside the formal
schooling system. Given the overall correlations drawn between fertility decline
and female education, exploring secondary schooling for girls against the backdrop
of this orientation towards fertility decline, on the one hand, and empowerment
on the other, provides an opportunity to delve deeper into the gender politics of
investment in females.
In this chapter, we report primarily on material garnered through secondary
research, as well as field work undertaken to explore the status of secondary
schooling in UP, particularly in relation to patterns of financial investment and
provisioning of single-sex schools for girls. Our empirical research uncovered a
vital consideration for gender-policy analysis in education – the ways in which
wider discourses get played out through particular policy processes at state level,
which in turn are dictated by the compulsions of democratic politics as they are
played out in India. In that sense, UP offers both an interesting case study but also
an impossibly complex one, given the multiple political actors who inform the
education policy agenda (see, for instance, Kingdon and Muzammil’s (2003)
fascinating account of teacher politics in Uttar Pradesh). Gender seems almost
irrelevant in this tableau, although its very irrelevance is in itself a revealing
insight into the ways in which rhetoric on gender equality that aims to please
diverse publics, from vote-banks (perhaps) to union government and donor
agencies, can mask actual practice. In that sense, the account that follows
illustrates the ‘reality’ of policy-making in a developing-country context, not least
the opaqueness of the concept of policy, and the difficulty of tracking what is or is
not evidence of ‘policy’ in an intensely political policy-making environment.
The political economy of market and commerce, on the one hand, and complex
socio-cultural norms, on the other, find their own ways to influence the realm of
policy and education system, often leading to inconsistencies between policy
intents and practices adopted at ground level. An analysis of policy remains
incomplete unless traced to its translation into implementation and practices at all
levels. Policy practices need to be traced through penetrations and informal
interactions at various levels, making it difficult to collect evidence following
‘scientific’ methods. Much of the empirical information presented in this chapter
280 Jyotsna Jha and Ramya Subrahmanian
is based on ‘leads’ gained through informal interactions. It is particularly difficult

to track policy practices when there is a conflict between vested interests of
socioeconomic-political considerations and stated policy priorities, as we found to
be the case with girls’ secondary education in UP.
State policy and practice on female secondary schooling in

Uttar Pradesh
In this chapter we focus primarily on the resourcing, management and provisioning

of secondary schooling in UP, with a view to assessing implications for gender
equity. Characterized by very low participation rates and a slow pace of change in
educational indicators over the years, secondary education has not received the
desired attention in terms of either policy initiatives or resource allocation. The
increase in real per student expenditure in nominal terms at secondary level has
not kept pace with that in elementary education. The major proportion of state
expenditure has been on maintaining the existing provisioning, mainly on teach-
ers’ salary in government and aided sectors with little emphasis on expansion of
services. Instead, private investment has been encouraged by relaxing the norms
and conditions required for recognition of schools and therefore compromising
the range and quality of facilities available. A policy initiative in the form of infra-
structure grants to direct private investment for expanding provision of single-sex
girls’ exclusive schools has been diluted by allowing boys also to be admitted. A
gender-differentiated scheme of studies offering ‘feminine’ subjects to girls further
reinforces gender stereotypes. In an environment where the majority of out-of-
school children are girls belonging to disadvantaged socioeconomic groups and
where girls’ education is not a social norm, increased commercialization proves
counter-productive. These issues are elaborated and discussed in the following
subsections.
Gender and secondary schooling in Uttar Pradesh

Secondary education in UP is characterized by overall low participation rates and
sharp gender differentials. While the participation rates are lower, gender dispari-
ties are higher than the national average in the state. The overall Gross Enrolment
Rate (GER) was close to 25 per cent during the late 1990s, this being only about
15 per cent for girls. The Gender Parity Index (GPI) for GER is only 0.45 in the state
as against 0.65 for the country as a whole (Table 13.1). Only about 27 per cent of
girls enrolling in grade I reach grade X and only about 60 per cent of these
complete the grade (Table 13.2). Despite substantial increase in enrolments,
the GERs are not improving in the state, reflecting the fact that the rate of increase
in enrolment is barely enough to keep pace with the rate of increase in the
population.
An interesting feature of girls’ schooling participation in UP is that though
notable gender differentials exist in favour of boys in transition rates from primary
to upper primary and from upper primary to secondary – especially the latter – the
Table 13.1 Gross Enrolment Ratios at secondary level (Grades IX to XII)
Uttar Pradesh All India
Year Total Boys Girls GPI Total Boys Girls GPI
1980/81 32.1 39.4 11.3 0.28 17.3 23.1 11.1 0.48

1990/91 23.82 34.79 11.15 0.32 19.28 33.89 10.27 0.30
1994/95 25.51 33.31 15.64 0.47 30.95 37.19 23.82 0.64
1995/96 25.45 33.36 15.48 0.46 31.00 37.10 23.80 0.64
1996/97 25.12 33.04 15.17 0.46 31.4 37.55 24.42 0.65
Notes: GPI (Gender Parity Index) ⫽ GER for girls divided by GER for Boys
These are actual data. Data available for later periods are provisional.
Source: A Handbook of School Education and Allied Statistics (1996), Ministry of Human Resource
Development (MHRD), Government of India (GOI), 1996 for 1980/81 and 1990/91 figures; Education in
India (1994/95, 1995/96 and 1996/97), MHRD, GOI, 2001, 2002, 2003 for 1994/95, 1995/96 and 1996/97
Table 13.2 Crude drop-out rates and proportion of repeaters
Total Boys Girls Total Boys Girls
Drop out rates (Grades 56.51 52.93 62.11 40.67 39.71 41.90
I to V) 2000/01
Drop out rates (Grades 61.02 56.26 68.54 53.67 50.33 57.95
I to VIII) 2000/01
Drop-out rates 62.11 56.22 73.17 68.58 66.41 71.51
(Grades I–X), 2000/01
Repeaters (%) in
Grades IX and X, 13.84 14.88 10.68 6.49 7.02 5.57
1996/97
Source: Annual Report, MHRD 2001/02 for drop-out rates and education in India (1996/97) MHRD;
Government of India 2003 for proportion of repeaters.
trend changes when it comes to the transition rate between secondary and senior
secondary (Table 13.3). A significantly smaller proportion of boys studying in
class X continue with their senior secondary schooling as compared to the proportion
of girls studying at the same level. Pass percentages are also higher for girls at both
grades X and XII, explaining to some extent the higher transition rate at that level
(Table 13.4). A high proportion of boys join the labour force at this age, which also
is partially responsible for their discontinuation of schooling after grade X.
Relatively low transition rates from primary to upper primary, and upper
primary to secondary, for girls indicate that the secondary-schooling participation
patterns cannot be understood in complete isolation. This is especially true for the
fact that the largest drop-out incidence takes place within primary level and only
38 per cent of girls enrolled in grade I reach grade V (Table 13.2). As a corollary to
Table 13.3 Crude transition rates for different levels of school education (1998/99 to
1999/2000)
Transition Rates Total Boys Girls Total Boys Girls
Primary to upper 92.6 96.1 85.8 89.0 91.6 85.5

primary
Upper primary to 83.2 91.7 69.5 83.0 83.1 82.7
secondary
Secondary to senior 40.2 34.7 56.8 72.2 72.9 71.1
secondary
Note: This has been calculated by taking enrolment in the entry year as a percentage of enrolment in the
terminal year of the previous level during the previous year. This does not take contributions of migration
from other states, repetition, etc. into consideration. Also, the fact that a large percentage of children
never reach Grade V, the terminal grade for primary education, is not reflected in this table.
Source: Selected Educational Statistics, MHRD, Government of India.
Table 13.4 Gender-wise pass percentages for Grades

X and XII, Uttar Pradesh
Class X Class XII
Year Boys Girls Boys Girls
1996 37.01 65.61 68.16 84.17

1997 40.32 69.57 62.41 82.39
1998 18.21 54.92 46.31 74.24
1999 25.62 66.91 53.06 78.49
2000 24.46 47.19 60.37 81.88
2001 26.78 53.27 63.49 81.42
2002 31.73 59.39 62.34 84.56
Source: Board Examination results at a glance (2000, 2001,

2002), Board for Secondary and Intermediate Education,
Uttar Pradesh.
this, the GERs at upper primary level are significantly lower than that at the
primary level (Table 13.5). A combination of high drop out within primary stage,
better academic performances at secondary level and a high transition rate from
secondary to senior secondary indicates that though a relatively small proportion
of girls continue with their post-primary schooling, those who do continue
perform better in examinations and a greater proportion is likely to complete the
senior secondary level. However, higher transition rates for post-secondary stage
might be indicative of gender differentiation taking a different shape where girls
do not have equal opportunities to join the labour market.
Table 13.5 Gross Enrolment Ratio (GER) at different stages of school education
Primary level Upper primary level Secondary level

(Grades I–V) (Grades VI–VIII) (Grades IX–XII)
Total Boys Girls GPI Total Boys Girls GPI Total Boys Girls GPI
1980/81 62.50 81.80 40.30 0.49 32.10 56.30 18.00 0.32 32.10 39.40 11.30 0.28
1990/91 71.06 89.13 50.98 0.57 27.33 63.13 25.62 0.40 27.33 34.79 19.69 0.56
1994/95 61.20 72.03 48.86 0.68 25.51 55.52 31.28 0.56 25.51 33.31 15.64 0.47
1995/96 61.19 71.95 48.92 0.68 25.45 53.24 29.96 0.56 25.45 33.36 15.48 0.46
1996/97 61.13 71.71 49.02 0.68 25.12 51.24 28.91 0.56 25.12 33.04 15.17 0.46
GPI (Gender Parity Index) ⫽ GER for girls divided by GER for Boys
Source: A Handbook of School Education and Allied Statistics (1996), MHRD, Government of India, 1996 for
1980/81 and 1990/91 figures; Education in India (1994/95, 1995/96 and 1996/97), MHRD, Government of
India, 2001, 2002, 2003 for 1994/95, 1995/96 and 1996/97.
The resource gap

A perusal of the trends in the intra-sectoral distribution in the education-financing
pattern makes it obvious that, starting from the 1960s, school education as a
whole received more emphasis in terms of financial allocation. This was a shift
from the relatively greater emphasis being laid on higher education in the past.
However, within school education, the relative stress in favour of elementary
education as against secondary education has been greater especially since 1980s.
The low priority accorded to secondary education, we argue, adversely affected the
expansion of state-sponsored schooling facilities for girls at post-primary level,
thereby affecting their participation. The absence of gender-segregated data for
finances stops us from taking the analysis further.
Table 13.6 shows that school education occupied about 70 per cent of total
expenditure in the sector during 1951/52, which went down to about 53 per cent
in 1960/61. It then increased to more than 72 per cent in 1970/71 and went on to
occupy nearly 88 per cent of the total education budget in 2001/02. However, the
increase was largely due to enlarged expenditure on elementary education and the
relative share of secondary education remained static around 30–4 per cent of total
education expenditure during the 1980s and 1990s. What is more revealing is that
nearly 95–8 per cent of the total expenditure on secondary education has been the
non-plan expenditure, i.e., the expenses incurred on maintaining the system and
only about 2–5 per cent is being spent annually on plan head or the new activities
such as expansion of coverage by opening new schools or improvement in the
quality of teaching by providing more facilities or organizing professional devel-
opment activities for teachers (Table 13.7).
The proportion of plan expenditure or the new investments, on the other hand,
has been higher for elementary education, especially since the early 1990s. This was
the period when the Uttar Pradesh Basic Education Project (UPBEP) followed by the
District Primary Education Project (DPEP) was launched in a large number of dis-
tricts of the state. While the former was an exclusive project for UP and covered the
entire elementary-education sector, the latter was part of the national initiative and
Table 13.6 Expenditure on school education by levels in Uttar

Pradesh, India (million rupees)
Elementary Secondary
Year education education Total educationa
1951/52 35.03 (47.16) 16.78 (22.58) 74.29

1960/61 60.32 (33.89) 35.03 (19.68) 178.02
1970/71 364.26 (48.33) 179.48 (23.81) 753.70
1980/81 1 714.47 (49.42) 1 097.20 (31.63) 3 469.20
1981/82 1 882.91(48.25) 1 379.97(35.36) 3 901.80
1990/91 12 109.42 (58.22) 6 149.82 (29.57) 20 798.40
1991/92 10 659.08 (53.70) 6 781.22 (34.16) 19 849.50
1992 /93 12 146.87 (53.57) 7 670.17 (33.82) 22 676.80
1993/94 13 335.88 (53.89) 8 034.28 (32.47) 24 745.90
1994/95 16 101.16 (55.10) 9 505.75 (32.53) 29 221.99
1995/96 18 621.84 (55.80) 10 881.83 (32.60) 33 374.74
1996/97 21 306.52 (55.73) 11 976.857 (31.33) 38 232.36
1997/98 22 521.59 (54.43) 13 498.13 (32.62) 41 375.12
1998/99 33 168.06 (56.76) 18 145.52 (31.05) 58 432.71
1999/2000 26 892.21 (51.88) 17 546.05 (33.85) 51 830.51
2000/2001 34 271.07 (55.91) 19 156.93 (31.25) 59 865.41
2001/2002b 34 446.74 (57.88) 18 041.08 (30.31) 59 517.16
a
including higher and other education
b
Budget estimates
Notes:
Figures in parenthesis show the percentage share of respective level in total
education expenditure.
These figures reveal expenditure on Revenue Account of the Education
Department. In addition to these, a small amount of expenditure is incurred
on Capital Account of the department, by other departments and ministries.
However, together they account for less than one to two per cent of the total
expenditure on revenue account of the department.
Source: Budgeted Resources for Education (1951/52 to 1993/94), Department of
Education, MHRD, Government of India, New Delhi, 1995 and Analysis of
Budgeted Expenditure on Education, Government of India (different years).
covered primary grades (I–V) only. Both these projects in UP were funded by the
World Bank, the former being the first basic education project being financed by
them in India. Although the contribution of these externally-aided projects to total
expenses of the sector has not been considerable, they formed a significant propor-
tion of the new investments. The total amount coming from external sources
amounted to nearly Rs.20,000.00 million during 1993–2002,2 which was about
82 per cent of total new investments on elementary education in UP during the
period. The project design for both UPBEP and DPEP demanded that the state
government provide about 15 per cent of the total project cost. Taking both exter-
nal and internal allocations together these two projects accounted for more than
96 per cent of total plan expenditure or the new investment during the period. This
means that the high allocation for new investment during this period was mainly
due to the presence of externally aided projects and only about 4 per cent of the
total allocations was spent on areas/items not covered by these two projects.
Table 13.7 Percentage distribution between plan and non-plan expen-

ditures on elementary and secondary education, Uttar Pradesh, India
Elementary education Secondary education
Year Plan Non-plan Plan Non-plan
1970/71 12.24 87.76 4.93 95.07

1980/81 3.33 96.67 2.17 97.83
1981/82 4.39 95.61 3.07 96.92
1990/91 8.34 91.66 2.91 97.09
1991/92 8.06 91.94 6.04 93.96
1992/93 8.26 91.74 4.78 95.22
1993/94 11.81 88.19 4.27 95.73
1994/95 12.73 87.27 4.32 95.68
1995/96 12.84 87.16 6.19 93.81
1996/97 13.52 86.48 4.61 95.39
1997/98 12.95 87.05 1.66 98.34
1998/99 11.84 88.16 3.05 96.95
1999/2000 4.18 95.82 1.42 98.58
2000/2001 9.33 90.67 4.64 95.36
2001/2002a 7.03 94.93 6.07 93.93
a
Budget Estimates
Source: Based on figures provided in Budgeted Resources for Education (1951/52 to
1993/94), Department of Education, MHRD, Government of India, New Delhi,
1995 and Analysis of Budgeted Expenditure on Education, Government of India (different
years).
The neglect of secondary education in UP is obvious from the slow increase in

per student expenditure in the sub-sector during the 1990s, especially in compari-
son to that at the elementary stage (Table 13.8). In real terms, the per-student
expenditure rose by 8.6 per cent per annum for elementary education between
1980/81 and 2000/01 at 1993/94 prices whereas it was only 3.79 per cent for per-
student expenditure in secondary education during the same period. The annual
rate of growth in real per student expenditure between 1993/94 and 2000/01 was
7.25 per cent for elementary education and 3.17 for secondary education. It could,
however, be pointed out that the per student expenditure itself has been signifi-
cantly higher at secondary level and hence the higher rate of growth for elemen-
tary education is to level these two. This needs to be understood in terms of the
different nature of education processes and educational management structures at
these two levels. The high per student expenditure at secondary stage in compari-
son to elementary stage is due to substantially fewer students, a significantly lower
teacher–pupil ratio and relatively higher salary levels in the former. The size of
physical infrastructure including laboratories and libraries at this stage is huge
compared with the elementary stage and it is natural for per student expenditure
to be higher. The low rate of growth in real per student expenditure at secondary
stage coupled with the fact that almost the entire amount goes to pay salaries sug-
gests a decline in expenditure on development and quality improvement.
Table 13.8 Per student expenditure at elementary and secondary education levels,
Uttar Pradesh, India
Elementary education Secondary education
Per student Per student

expenditure expenditure
Total Total Total Total
Year expenditure enrolment Monetary Reala expenditure enrolment Monetary Reala
1980/81 1 714.47 12 762 037 134.34 419.81 1 097.20 18 53 459 591.97 1 849.90
1990/91 12 109.42 16 466 813 735.24 1 007.17 6 179.82 29 55 073 2 081.11 2 850.83
1993/94 13 335.88 17 593 289 758.00 758.00 8 034.28 30 15 817 2 664.04 2 664.04
1994/95 16 101.16 17 835 564 902.75 828.21 9 505.75 30 45 167 3 071.11 2 817.53
1995/96 18 621.84 18 055 640 1 031.35 866.68 10 881.83 31 54 719 3 449.38 2 898.63
1996/97 21 306.51 16 602 422 1 283.33 1 002.60 11 976.85 32 05 232 3 736.65 2 919.25
1997/98 22 521.59 18 481 210 1 218.62 889.50 13 498.13 32 39 513 4 166.71 3 041.39
1998/99 33 168.06 18 680 569 1 775.53 1 175.84 18 145.52 32 74 449 5 541.54 3 669.89
1999/00 26 892.21 19 019 535 1 413.92 912.20 17 546.05 33 34 369 5 262.18 3 394.95
2000/01 34 446.74 18 049 991 1 908.40 1 142.75 18 041.06 33 18 222 5 436.97 3 255.67
a
At 1993/94 prices (taken from Economic Survey, Government of India)
Note: While total expenditure is given in terms of million rupees, per student expenditure is in terms of
rupees.
Source: Table VII for total expenditure, Shikhsha Ki Pragiti (different years), Government of Uttar Pradesh
for enrolment except for 1993/94 and 1994/95 and Education in India, MHRD, Government of India for
1993/94 and 1994/95.
Although it is difficult to surmise that the presence of externally funded projects

led to diversion of funds from other sectors to elementary education, it is obvious
that these projects forced the state government to maintain its own investment
expenditure for the sub-sector. Once the project period came to an end, the state
was also committed to maintain all the investments, raising the non-plan or the
maintenance expenditure in the sub-sector. Though there cannot be two opinions
about the need for increasing the level of investments in elementary education,
doing so at the cost of other sub-sectors is questionable. In the absence of a clear
policy focus and any direction from central government or an external agency, no
such pressure existed for maintaining or enhancing expenditure for secondary
education in UP. As discussed earlier, the external agencies, in this case the World
Bank, emphasized increased investment in primary education, though not always
explicitly so, even at the cost of other sectors, on the grounds of a higher rate of
social and private return.
The argument based on the rate-of-returns approach also ignores the point that
the different sub-sectors of education are interdependent and lopsided invest-
ments could prove counter-productive in the long run. A number of studies on
elementary school participation in India have indicated that the availability of
schools teaching higher grades also encourages enrolment and completion at
earlier stages.3 A comparison of statistics showing girls’ enrolment at primary stage
in schools where only the primary stage is available, and in schools which provide
education up to secondary level in UP, showed that the proportion of girls’ enrol-
ment at primary stage is higher in the latter.4 However, these complications rarely
receive attention in the policy-making process because of a short-sighted sub-

sectoral approach followed increasingly by Indian states especially since the early
1990s.
There also exist other important linkages between different levels of education.
For instance, the spread of higher education is associated with an increase in
income levels and its distribution, which in turn is associated positively with
increased demand for girls’ primary education. The quality of teachers available
for primary education is directly dependent on the quality and spread of sec-
ondary education in any country. The availability of women teachers, which is
considered critical for promoting girls’ participation and completion at primary
level, cannot be ensured in rural primary schools without encouraging secondary
education among rural girls. Women teachers constitute only about one-quarter of
the total number of teachers in UP, their proportion being lower than that in rural
areas.5 Therefore, it is difficult to achieve universalization of even the elementary
stage without greater investments in other stages of education.
The policy and management context

The absence of a clearly outlined policy and comprehensive management framework
in secondary education further highlights the neglect of secondary education in
the state. The lack of any comprehensive policy and focus to guide resources and
management of secondary education in the state makes it difficult to trace policy
discourse as understanding is gained through reviewing reports of the committees,
legal Acts, schemes and office orders.6 Although UP was one of the first Indian
states to establish an autonomous board of secondary education,7 and appointed
a Commission for Secondary Education immediately after independence,8 the
issues of girls’ participation in secondary schooling and making education gender-
sensitive has received scant attention so far. The Secondary Education Commission
(1952) dealt with a number of important aspects but did not include girls’ educa-
tion as a separate issue in spite of large gender disparities existing in the schooling
participation patterns. However, the report mentioned several measures including
the building of separate sanitary facilities and retiring rooms, and hiring of women
teachers in co-educational institutions to increase girls’ participation.9
One of the important features of secondary education in UP as in many other
Indian states is the vast presence of privately managed educational institutions.
Secondary schools in Uttar Pradesh can be put in three broad management
categories: Government, aided and unaided private schools. Government schools are
fully financed and managed by the state government. Aided schools are managed pri-
vately by individuals, trusts, societies or corporate bodies but funded almost entirely
by the government. The government bears almost the entire recurrent costs of these
schools by providing grants-in-aid for salaries. Unaided private schools are managed
and financed privately but recognized by the government. While the government
and aided schools charge no tuition fee and only nominal other charges, unaided
schools are for-profit organizations and charge substantial tuition and other fees.
The proportion of government schools remained almost static during the 1980s
and 1990s while that of aided schools increased till 1986 after which no new
school was brought under the grants-in-aid scheme. The growth of private
unaided secondary schools has been spectacular during the 1980s and 1990s. The
number increased from a modest 845 in 1994/95 to 6,541 in 2001/02, registering
a nearly eight-fold increase. The proportion of private schools, especially unaided
ones, has gone up to more than half of total secondary schools in UP after the
bifurcation of the state, reflecting the fact that the parts of UP that constituted the
newly carved state, Uttaranchal, had a higher number of government and aided
schools (Table 13.9). Near-stagnation in the number of government and aided
schools which charge only nominal fees and expansion only through fee-charging
private unaided schools reflects the state’s policy of promoting commercialization.
This is despite the fact that there exists apparently a greater demand for admis-
sions in government school as compared to private, including aided, schools
because of a common perception that these are better-resourced and higher-
quality.10 The results of the terminal stages at grade X and XII of all schools affiliated
to the UP board indicate that this perception is not ill-founded (Table 13.10).
The increased focus on commercializing secondary stage of education has negative
implications for girls’ participation. In a situation where girls’ education is not
valued, the demand for a fee-charging school system is bound to be low as parents
would not be willing to pay high charges for their daughters’ education. As mentioned
earlier, the increase in schooling participation at primary level witnessed during the
1990s is largely on account of demand from poorer and disadvantaged sections,
whose capacity to pay is also limited. The continuation of schooling for girls from
these sections depends crucially on state support and increased commercialization
Table 13.9 Percentage distribution of secondary

schools by management in Uttar Pradesh, India
Secondary schools (Grades 9–12)
Private
Government ⫹
Year local bodies Aided Unaided Total
1980/81 16.04 N/A N/A 83.95

1990/91 17.79 73.86 8.34 82.20
1994/95 23.00 64.85 12.25 77.0
1995/96 22.96 64.41 12.15 77.04
1996/97 22.87 64.61 12.45 77.13
2001/02a 4.70 38.50 56.80 95.20
a
2001/02 distribution denotes the position of schools that
remained in UP after bifurcation of the state into UP and
Uttaranchal.
N/A: Not available
Source: Education in India (1994–95, 1995–96 and 1996–97),
MHRD, Government of India, 2001, 2002, 2003 for 1994/95,
1995/96 and 1996/97, and Directorate of Secondary Education,
Uttar Pradesh for 2001/02.
Table 13.10 Number of examinees and pass percentages in different types of schools (Grade
X and Grade XII examinations – 2002), Uttar Pradesh, India
Class X examinations Class XII examinations
Govt. Aided Unaided Govt. Aided Unaided

schools schools schools Others schools schools schools Others
Number of 464 4467 3 917 273 237 3506 1301 156

schools (5.09) (48.97) (42.94) (2.99) (4.56) (67.42) (25.02) (3.0)
Number of 80 267 1 043 817 706 016 68 531 44 499 4 73 530 1 18 962 10 869
examinees (4.16) (55.02) (37.21) (3.61) (6.87) (73.09) (18.36) (1.68)
Students 78 920 421 868 310 199 26 367 32 922 3 41 978 86 695 8 239
passed (4.47) (53.13) (39.07) (3.32) (7.0) (72.79) (18.45) (1.75)
Pass 45.01 40.42 43.94 38.47 73.98 72.22 72.88 75.80
percentages
% schools 14.65 18.27 16.93 56.77 0.42 0.77 1.31 2.56
with pass
% 0–20%
Note: Figures in parenthesis depict percentages for respective types of schools.

Source: Based on data provided by Board for Secondary and Intermediate Education, Uttar Pradesh.
works against that. But ironically, the only two schemes started to support girls’
education during the late 1990s focused on promoting unaided, fee-charging
schools. These two schemes were meant to direct private investments towards the
opening of single-sex schools for girls, but even these were subverted with the
outcome of further discouraging girls’ access to secondary schooling. How political
considerations successfully subverted the policy goal is discussed in detail in the
next section.
The massive presence of privately managed, aided and unaided schools makes
the role of the regulatory framework crucial. A regulatory framework is important
for directing resources and provisioning, maintaining essential physical facilities
and an adequate number of teachers, and improving the quality of teaching–
learning processes, in order to establish a system of accountability. Perusal of the
existing norms, guidelines and directives indicates a near-absence of accountabil-
ity mechanisms not only for unaided schools but also for aided schools that
survive on substantial grants from the state exchequer. Gender focus even in terms
of ensuring facilities and creating an enabling environment for girls is entirely
lacking.
Successive committees and national policy statements have highlighted the
need for an increase in the number of women teachers in the light of parental inse-
curity and the resultant restricted mobility of girls as well as to create an enabling
environment by providing positive role models in an otherwise non-inspiring
environment. This is especially true for rural areas. However, the state does not
have any policy of reservation for hiring women teachers at secondary level either
in government or aided schools. Though the government funds the entire salary
bill for aided schools, making this item the single largest constituent of the
government’s expenditure on this sub-sector, and the teachers enjoy all the
advantages of government service, no regulatory frame for their recruitment and
conduct exists in the state. Teachers from government and aided schools enjoy
substantial political clout and resist any effort to have accountability measures
in place. Kingdon and Muzzamil’s (2003) research shows that UP is one among
only four states that has an upper chamber of the state legislature, and teachers
have guaranteed representation in this body. They note two kinds of political
influence:
Political influence has been of two types: one, from above which has been
instrumental in shaping the education system; and two, the lobbying and
pressure groups from within the system originating at the local levels (and uniting
at the state-level) in the form of organizations of teachers. Education-related
legislation in UP has often been framed under immense lobbying pressure
from teachers, particularly at the primary and secondary levels. Teachers
in school (as opposed to higher education) have been instrumental in
determining the local base of political parties in the state.
2003:6, our emphasis
Provisioning female secondary education: the politics of investment

Since pre-independence days UP has followed the practice of providing single-sex
schools at all stages of education. Considering the fact that the availability of a
school within reach was more important for the participation of both girls and
boys as compared to the school being single-sex for primary age-group children,
the provision was discontinued at primary level during the 1970s. This measure
was based, and perhaps rightly so, on the belief that girls in this age-group are
young and parents do not mind sending their daughters to co-educational insti-
tutions provided the school is not situated very far from their habitation.
However, the same logic does not apply to the secondary level as parents are
reluctant to send their adolescent daughters to co-educational schools. The avail-
ability of single-sex schools, therefore, becomes critical at this level. Tables 13.11
and 13.12 show that the number of single-sex secondary schools has been
increasing throughout the second half of the last century, the rate of growth
being the highest during the first thirty years of independence, i.e., during the
1950s, 1960s and 1970s. The annual rate of growth slowed down during the
1980s and early 1990s but again picked up during the late 1990s. Although disag-
gregated statistics by management are not available, it is obvious that the increase
in the number of single-sex schools for girls during the late 1990s is largely on
account of unaided schools. No new school was added in the grants-in-aid cate-
gory, and the number of new schools upgraded to secondary level in the govern-
ment sector has not been large. Increase in the number of unaided girls’ schools
during the late 1990s can be attributed directly to two schemes introduced during
this period, aimed at creating incentives for investment in secondary single-sex
schools for girls.
Table 13.11 Growth of schools in Uttar Pradesh, India
Number of educational institutions
Upper primary level Secondary Level

Primary education (Grades 6–8) (Grades 9–12)
% of % of % of
Girls girls Girls girls Girls girls
(single- to (single- to (single- to
Total sex) total Total sex) total Total sex) total
1950/51 31 979 2520 7.88 2 854 468 16.3 987 154 15.6
1960/61 40 083 4927 7.93 4 335 661 15.2 1 771 282 15.9
1970/71 62 127 11624 18.71 8 787 2 008 22.8 3 415 581 17.0
1980/81 70 606 13 555 3 200 23.6 5 178 758 14.6
1985/86 73 490 14 728 3 377 22.9 5 709 848 14.8
1990/91 77 111 15 072 3 319 22.0 5 999 886 14.7
1991/92 78 085 15 328 3 447 22.4 6 060 889 14.6
1992/93a 74 889 17 730 3 610 20.4 6 115 560 9.2
1993/94 79 522 15 546 3 509 22.5 6 637 962 14.4
1994/95 82 023 15 976 3 817 23.8 6 637 962 14.4
1995/96a 88 817 19 516 2 651 13.6 6 977 1 042 14.9
1996/97 91 093 19 917 2 746 13.7 7 003 1 057 15.0
1997/98 92 554 20 436 2 960 14.4 7 135 1 127 15.7
1998/99 94 476 20 675 3 051 14.7 8 339 1 364 16.3
1999/00 96 764 21 678 3 237 14.9 8 549 1 427 16.6
2000/1b 86 361 19 639 3 021 15.3 8 459 1 501 17.7
2001/2 88 927 20 429 3 102 15.1 9 063 1 608 17.7
Notes:
a
The particular year issues of the Government of Uttar Pradesh publication were not available to the
researchers.
b
The state was bifurcated in 2000 into Uttar Pradesh and Uttaranchal. Hence, the number of schools
shown in 2000/01 and 2001/02 does not include schools located in Uttaranchal. For this reason all trend
analysis is up to 1999/2000.
Sources: (i) Shiksha kee Pragati (Education’s Progress) (Government of Uttar Pradesh), Directorate of
Education, Allahabad, different years.
(ii) Growth of Schools 1950–1992 (State-Wise), MHRD, Department of Education, 1997 for 1992/93 data.
(iii) Education in India 1995–96, MHRD, Department of Education, 2002 for 1995/96 data.
The first of these two schemes was introduced by the state government of UP in
1997 making any new private single-sex school for girls opened in a hitherto
uncovered block headquarter eligible for a one-time infrastructure grant of
Rs.1 million.11 Once all the block headquarters were covered, a new scheme was
brought to provide Rs.2 million to any new girls’ school being opened in an
uncovered Nyaya Panchayat.12 The fact that a good number of unaided private
girls’ secondary schools were opened reflects a positive response to these schemes.
However, these are fee-charging schools with no aid from the government for pay-
ing salaries or incurring any other recurrent expenditure. This means these schools
have to survive on their own revenue, the fees collected from parents being the
Table 13.12 Annual growth rate for the number of institutions at different school levels in
Uttar Pradesh, India
Annual growth rate for the number of institutions
Primary level Upper primary level Secondary level

(Grades 1–5) (Grades 6–8) (Grades 9–12)
Girls Girls
Year Total Total (single-sex) Total (single-sex)
1950/51 to 1980/81 4.03 12.49 19.46 14.35 15.69

1980/81 to 1990/91 0.92 1.12 0.37 1.58 1.69
1990/91 to 1994/95 1.59 1.50 3.75 2.66 2.14
1994/95 to 1999/2000 3.59 7.14 ⫺3.04 5.76 9.67
1990/91 to 1999/2000 2.83 5.24 ⫺0.27 4.72 6.78
Source: calculated on the basis of Table XI.
main source. The demand for fee-charging private schools for girls in these interior
areas did not seem to take off. As mentioned earlier, girls’ education not being
highly valued and the paying capacity of poorer households being low together
act to keep down demand for fee-charging schools for girls. The owners of the
schools opened under these schemes demanded that they be allowed to admit
boys as well. Initially, in 1999, all schools opened under these schemes were appar-
ently allowed to admit boys by an executive office order, which was later revised
after a gap of two years. Under the revised order, only those schools that are
located in rural areas were allowed to admit boys along with girls.13 It implied that
schools opened under this scheme and located in urban areas could not admit
boys. Separate statistics for enrolment are not available in the public domain to
indicate the proportion of boys and girls in these schools.
This experience shows that girls’ schooling participation in rural areas cannot be
promoted by the expansion of provisioning through fee-charging schools. It also
reveals the role of political considerations in policy-making and their implemen-
tation. While the larger concern for girls’ education at the national and interna-
tional levels made it a good political decision to have directed schemes to improve
provisioning of girls’ schools, narrow political considerations at the local level
demanded measures which subverted the very basis of allowing subsidy for a
private enterprise. Most of these schools were opened by local politicians who first
used the scheme to access public money for their private initiative and later
exercised their political influence to get an order passed which allowed them to
make their enterprise initially viable and later profitable. These facts are not easily
traceable as the available aggregate statistics do not provide any pointers and only
a deeper analysis of executive office orders and other such directives help reveal
such subversive attempts.
The rate of growth in enrolment slowed down during the 1990s, especially
the later part of the decade, for all stages of education. The growth rate for girls’
enrolment, however, has been higher as compared to boys at all stages. This is natural
as girls constitute the vast majority of out-of-school children in all age-groups.

Nonetheless, despite the higher growth rate for enrolment, girls constitute only
about 36 to 39 per cent of total enrolment at primary and upper primary stages
and only 26 per cent at secondary stage in 2001/02. The fact that a major expan-
sion at secondary stages has taken place in fee-charging private unaided schools is
certainly one of the factors for the slowing down of the enrolment growth rate for
both boys and girls, especially the latter, in the 1990s. In an environment where
girls’ education is not a highly valued choice, increased commercialization is
bound to act against their schooling participation.
The above analysis makes it clear that secondary education in UP is character-
ized by a resource gap in public funding with its implications for the expansion of
provisioning at this stage. As a corollary to this there has been an increasing focus
on promoting commercialization through direct and indirect measures.
Commercialization has not been complemented by a suitable regulatory framework
outlining directions and accountability mechanisms, which has allowed market
forces and political considerations to play an unchecked role. The absence of con-
crete and fool proof interventions to enhance appropriate provisioning for girls’
education proves that there is more rhetoric than substance. The fact that the rate
of growth for boys’ enrolment has been even slower than that for girls shows that
increased privatization has hurt their participation as well.
The socio-cultural context, gender stereotyping and the secondary

curriculum
The fact that despite very low enrolment ratios for girls at secondary level in UP,
fee-charging single-sex schools for girls have failed to attract adequate numbers of
girls in the state indicates lack of sufficient demand for secondary schooling of
girls, at least in rural areas. It also shows that though provisioning plays an impor-
tant role, provisioning alone, particularly through private profit-making institu-
tions, is not the way to increase girls’ participation at the secondary stage of
schooling. However, no policy initiative aimed to influence demand for education
can be found at the secondary stage in UP. Some scholarship schemes exist for
Scheduled Caste and Scheduled Tribe students but they are not necessarily meant
for girls. The coverage is also extremely small and they constitute less than 1 per cent
of total expenditure on secondary education. However, it is important to recognize
that several schemes exist, focusing on encouraging girls’ enrolment at primary
and upper primary levels and these appear to have played their role in increasing
participation at those stages and to some extent at secondary stage as well. These
schemes include free textbooks and provision for scholarships for girls. Under
UPBEP and DPEP, efforts were also made to review the textbooks for gender bias
and train teachers for gender sensitivity.
The scheme of studies, curriculum design and content at secondary and senior
secondary stages could itself be important means for initiating a change in
existing gender positions and could influence the constructs of masculinity and
femininity in society. In this context, the 1968 national policy’s emphasis on
undifferentiated curriculum till grade X assumes some significance. The role of
pressures to join ‘womanly’ courses, the non-availability of non-traditional courses

for girls and women, gender stereotypes in both the official and the hidden
curriculum, and the negative attitude of teachers have been recognized as con-
straints by the Ramamurti Committee report (1991) which formed the basis of the
Programme of Action (POE), 1992. Although it is not possible to delve into all
these aspects of the curriculum, the analysis undertaken shows that some of the
existing practices in UP reinforce rather than question gender stereotyping. Apart
from the fact that the curriculum design has no space for some critical skills,
information and knowledge areas crucial for adolescent girls in their specific
contexts, the existing scheme of studies also practises gender differentiation to the
disadvantage of girls.
With the adoption of the 10 ⫹ 2 ⫹ 3 system and a unified curricular approach,
the state adopted a new scheme of studies in 1998. Under this scheme, all students
are supposed to complete seven subjects successfully. These include five subjects
where these is no or little choice: Science, Social Science, Hindi, another Indian
language, and Mathematics. While Mathematics is compulsory for boys, it is not
for girls. Girls have a choice of opting for Home Science in place of Mathematics.
The state offers two types of courses in Mathematics – Mathematics and
Elementary Mathematics. While boys have a choice between Mathematics and
Elementary Mathematics, girls have an additional option available in Home
Science. This practice not only goes against the very philosophy of an undifferen-
tiated curriculum, it also strengthens the existing stereotype that girls are not
capable of doing well in Mathematics. What makes the situation worse is that
the majority of single-sex girls’ schools in rural areas do not offer the choice of
Mathematics, the only available options being Elementary Mathematics and
Home Science. In the case of two additional subjects where students have wider
choices available, the single-sex girls’ schools usually offer limited options of
‘womanly’ subjects such as sewing, cooking, and so on. The scheme provides for a
number of options including ‘non-womanly’ courses such as commerce, agriculture
and accountancy, but most of these are not offered in the majority of girls’ schools
and students are forced to opt for among whatever is available.
The rationale for having the choice of Elementary Mathematics and Home
Science, and ‘womanly’ optional subjects, stems from the need to respond to ‘demand’.
The very presence of this option of Home Science in place of Mathematics for girls
strengthens the existing notions of masculinity and femininity. It is rarely realized
that the school is also responsible for creating ‘demand’, not only responding to it.
In any case these are reflections of societal expectations of the feminine role and
whether girls are also interested in these has rarely been investigated. Societal
expectations and family pressures to opt for ‘womanly’ subjects find an expression
in this choice. Moreover, schools strengthen the demand for stereotyped courses
since the choice of opting for non-womanly subjects is only notional in most of
the single-sex girls’ schools. The ‘practical’ problem of not finding enough teach-
ers for Mathematics and other such subjects to teach in girls’ schools is often cited
as a reason for not offering the course. It is not considered viable to have a teacher
if only a few students are opting for the subject. What is not realized is that unless
more students including girls are encouraged to opt for those subjects the shortage
of women teachers is going to continue. The fact that these practices have been
retained in the latest curriculum changes reflects the lack of a guiding vision based
on notions of gender equality to guide policy-planning. These practices perpetuate
gender disadvantage and the school system becomes an agent of reinforcement
rather than of change in the process.
Conclusion
In this chapter we have argued that education policy initiatives remain divorced
from broader visions of gender justice and social policy, partly on account of the
continuing dominance of discourses of investment in female education, which
rely on identifying thresholds for minimum public action. Although there are
multiple discourses influenced by diverse actors in the education policy arena,
these do not seem to translate into policies and practices that have an impact on
female schooling in a way that reflects the reality of girls’ education today.
We make three specific points. First, we argue that, at a symbolic level, the
continuing sway of the idea of the ‘educated person’ purely in terms of human
capital has resulted in mechanistic definitions of what it is people are assumed to
gain from education. This, arguably, emphasizes elementary education invest-
ment over subsequent levels of education, reduces the focus on developing poli-
cies appropriate to girls and boys at different stages of the life cycle, and places
issues of quality and equity at the fringes of policy action. Combined with insti-
tutional biases and the political challenges of promoting gender equality, claims
for investment in women that continually emphasize women’s role in helping
their families, communities and nations end up reinforcing the core messages of
human capital.
In our attempt to map the operation of discourses and practices in the secondary
schooling arena, we found that at all levels the neglect of secondary education for
girls was reinforced, thus appearing to be a direct translation of assumptions
related to human capital. However, we also argue that ‘gender’ does not operate
only in the symbolic arena and that symbolic discourses are not all-powerful.
Discourses can and do get renegotiated in the realm of practice, driven by varied
motivations and compulsions. State policies for equity, such as in the UP case, are
often implemented through market-mediated processes. While we do not wish to
suggest that this is the result of a linear process of translation of a particular (albeit
hegemonic) policy discourse at different levels of policy decision-making and exe-
cution, our case study has shown how political actors at the local level can subvert
policies enacted in favour of women, resulting in the dilution of state commit-
ments to gender equality, even if these were weak at the outset.
In the case of the policy enacted in UP to encourage female secondary school-
ing, we argue that this was weak to start with because it omitted to build up the
demand-side of girls’ secondary schooling, resulting in a low uptake of the
schools made available. The neglect of a focus on the demand for girls’ school-
ing also meant that when the policy was reversed, through the actions of
private providers, girls and parents affected by this move had barely any voice
to challenge the change in policy. This leads us to argue strongly for the active
role of states, even where private providers are in play, not just to subsidize the
costs of schooling, but also to build up demand to ensure that education provi-
sion is seen as a matter of right, and not a privilege to be bestowed and then
taken away.
Our third and final point refers to the issue of the outcomes of policy. Given the
emphasis on gender parity in secondary education as a Millennium Development
Goal (MDG) and Education For All (EFA) target, as well as the emphasis on gen-
der equality in the EFA goals, the issue of gender and education cannot rest at the
level of grand policy pronouncements on the importance of investment in
women. This point has global relevance, but here we speak to the Indian context.
While elementary education is finally getting the impetus it has been promised
for decades, care has to be taken that the policy directions followed for this sub-
sector do not have negative implications for the sub-sector that follows. If all
schools are not improved in terms of their quality, transition to secondary
schools in terms of access of the poorest, and of girls, will continue to be difficult
to achieve. This fact, combined with the neglect of the secondary sub-sector in
terms of providing girl-friendly single-sex schools that are relatively cost-free, will
mean that girls are pushed out of education at a time when they need resources
to develop the capacities, skills and aptitudes that will support their transition to
adulthood.
Notes
1 Notably, through a large USAID funded programme in the state, Innovations in Family
Planning Services (IFPS).
2 Based on estimates provided in the Implementation Completion Reports (World Bank)
of the two programmes.
3 See Vaidyanathan and Nair (2000) and Jha and Jhingran (2002)
4 These statistics have not been included in this chapter.
5 See Jha and Bhardwaj (2001) for detailed analysis on women teachers in rural India.
6 Office orders are executive measures often based on decisions taken by the bureaucracy,
which do not require to be approved by elected legislative bodies.
7 The Board of High School and Intermediate Examinations was established as an
autonomous body functioning under the Directorate of Secondary Education in 1922
in UP.
8 A special Secondary Education Commission (known as Acharya Narenda Dev Committee)
was appointed in UP which submitted its report in 1952.
9 Based on Sharadindu (2001)
10 Based on personal interactions with people and functionaries belonging to government
and private school systems.
11 Block is a sub-district administrative unit in India. It usually covers about 50 to 100 primary
schools.
12 Nyaya panchayat is a smaller administrative unit covering several villages and about
15–20 primary schools.
13 The revised order of the state government (order number 2501–8-2001–3009 | 5 | / 94) is
dated August 17, 2001. It revised the order issued on September 29, 1999 (order no.
15–8–99–3009 | 5 | / 94).
References
In addition to the references cited below, a number of Acts, unpublished reports, internal
analyses papers, handouts, office orders, etc. were referred to.
Board of High School and Intermediate Examinations, Uttar Pradesh (different years),
Analyses of Examination Results. Allahabad Board of High School and Intermediate
Examinations.
Council of Boards of School Education In India (COBSE) (2002). Sahodaya, Newsletter,
Special Number, January 2002, Vol. XXXXI. New Delhi: COBSE.
Dreze, J. and H. Gazdar (1996) ‘Uttar Pradesh: The Burden of Inertia’. In J. Dreze and A. Sen
(eds) Indian Development: Selected Regional Perspectives. Oxford and New Delhi: Oxford
University Press.
Government of India, Ministry of Human Resource Development (different years).
A Handbook of School Education and Allied Statistics (1996). New Delhi: Department of
Education.
Government of India, Ministry of Human Resource Development (different years). Annual
Report. New Delhi: Department of Education.
Government of India, Ministry of Human Resource Development (different years). Analysis
of Budgeted Expenditure in Education. New Delhi: Department of Education.
Government of India, Ministry of Human Resource Development (different years). Education
in India. New Delhi: Department of Education.
Government of India, Ministry of Human Resource Development (1996) Growth of Schools
1950–1992. New Delhi: State-Wise.
Government of India, Ministry of Human Resource Development (1997) Growth of Enrolment
in School Education 1950–51 to 1993–94. New Delhi: State-Wise.
Government of India, Ministry of Human Resource Development (different years) Selected
Educational Statistics. New Delhi: Department of Education.
Government of Uttar Pradesh (2002) Vittiya Akhya, Mahyamik Shiksha (Financial Statement
Secondary Education) (2002–03). Lucknow.
Government of Uttar Pradesh (different years) Shiksha Ki Pragati (Progress of Education).
Allahabad: Directorate of School Education.
Jeffery, P., R. Jeffery and C. Jeffrey (2004) ‘Islamisation, Gentrification and Domestication:
“A Girls’ Islamic Course” and Rural Muslims in Western Uttar Pradesh’. Modern Asian
Studies, Vol. 38, No. 1: 1–53.
Jha, Jyotsna and Dhir Jhingran (2005) Elementary Education for the Poorest and Most Deprived
Groups: The Real Challenge of Universalisation. New Delhi: Manohar.
Jha, Jyotsna and Meenakshi Bhardwaj (2001) Women Teachers in Rural India. New Delhi:
UNESCO and Ministry of Human Resource Development (MHRD).
Kingdon, G. M. Muzammil (2003) The Political Economy of Education in India: Teacher Politics
in Uttar Pradesh. Oxford: Oxford University Press.
Ministry of Human Resource Development (MHRD) (1997) Remodeling of School Education
Board. Report of the Task Force on the Role and Status of Boards of Secondary Education.
New Delhi: Department of Education, Government of India.
Ministry of Human Resource Development (MHRD), Programme of Action 1992 (National
Policy on Education 1986). New Delhi: Department of Education, Government of India.
Ministry of Human Resource Development (MHRD), Report of the CABE Committee on Policy
1992. New Delhi: Department of Education, Government of India.
Ministry of Human Resource Development (MHRD), Towards an Enlightened and Humane
Society, NPE- A Review. Ramamurti Committee Report. New Delhi: Department of Education,
Government of India.
National Council of Educational Research and Training (NCERT) (1998). Sixth All India
Education Survey Report, New Delhi: Government of India.
Planning Commission (2002) Tenth Five Year Plan 2002–2007. Volume II, Sectoral Profiles and
Programmes. New Delhi: Government of India.
Sharadindu. (2001) School Education in U.P., Status, Issues and Future Perspectives. State Council
of Educational Research, Lucknow; mimeo of draft submitted to the National Council of
Educational Research and Training, New Delhi.
Stromquist, N. (1999) ‘The Impact of Structural Adjustment Programmes in Africa and Latin
America’. In C. Heward and S. Bunwaree (eds) Gender, Education and Development: Beyond
Access to Empowerment. London: Zed Books.
Subrahmanian, R. (2002a) ‘Engendering Education: Prospects for a Rights-Based Approach to
Female Education Deprivation in India’. In M. Molyneux and S. Razavi (eds) Gender, Justice
and Development. Oxford: Oxford University Press.
Subrahmanian, R. (2002b) ‘Gender and Education: Review of Issues for Social Policy’. Social
Policy and Development, Paper No. 9. Geneva: UNRISD.
Subrahmanian, R. (2003) ‘Gender Equality in Education: Definitions and Measurements’.
Background paper for the Global Monitoring Report on Education for All 2003–04. Paris:
UNESCO.
Tyagi, R. S. and Sharadindu (1999) Educational Administration in Uttar Pradesh. National
Institute of Educational Planning and Administration. New Delhi: Vikas Publishing House.
United Nations Educational, Scientific and Cultural Organization (UNESCO) (2002)
Education for All: Is the World on Track? Global Monitoring Report 2002/03. Paris: UNESCO.
United Nations Educational, Scientific and Cultural Organization (UNESCO) (2003) The Leap
to Equality: Gender and Education for All. Global Monitoring Report 2003/04, Paris:
UNESCO.
United Nations Children’s Fund (UNICEF) (2003) The State of the World’s Children 2004: Girls,
Education and Development. New York: UNICEF.
United Nations Development Programme (UNDP) (2003) ‘Millennium Development Goals:
A Compact among Nations to End Human Poverty’. Human Development Report 2003.
New York: United Nations Development Programme.
Unterhalter, E. (2003) Education, Capabilities and Social Justice. Paper prepared for the EFA
Global Monitoring Report 2003/09. Paris: UNESCO.
Unterhalter, E. and S. Dutt (2001) ‘Gender, Education and Women’s Power: Indian State and
Civil Society Intersections in DPEP (District Primary Education Programme) and Mahila
Samakhya’. Compare, Vol. 31, No. 1.
Vaidyanathan, A. and P. R. Gopinathan Nair (eds) (2001) Elementary Education in Rural India.
New Delhi: Sage.
Verghese, N. V. (2003) ‘Universalisation of Higher Education and Equity Concerns’. Paper
presented at the Oxford Conference on Education and Development, Oxford, UK.
Wazir, Rekha (2000) ‘Profiling the Problem’. In Rekha Wazir (ed.) The Gender Gap in Basic
Education: NGOs as Change Agents. New Delhi: Sage Publications.
Woodhall, M. (1987) ‘Human Capital Concepts’. In G. Psacharapoulos (ed.) Human Capital
Concepts. Page references relate to the reprint in A. H. Halsey et al. (eds) (1997) Education:
Culture, Economy and Society. Oxford and New York: Oxford University Press.
World Bank (2003a) Secondary Education in India. South Asia Human Development Sector
Report No. 2, New Delhi.
World Bank (2003b) Opening Doors: Education and the World Bank. Washington DC: World Bank.
World Bank (2003c) A Policy Note on the Grant-in-Aid System in Indian Education. South Asia
Human Development Sector Report No. 3, New Delhi.
Part IV
Financing Social Provisioning and
Counting in Women
14
Gendered Implications of Tax
Reform in Latin America: Argentina,
Chile, Costa Rica and Jamaica
Evelyne Huber
In Latin American and Caribbean countries, poverty and inequality have been
longstanding problems, and the momentous economic and social policy changes
over the past two decades have done little to ameliorate them. The most
effective means for reducing class- and gender-based poverty and inequality would
be citizenship-based entitlements to basic (i.e. allowing basic subsistence) income
support, healthcare, and education. In advanced industrial societies, public spending
is an extremely important instrument for the alleviation of class- and gender-
based poverty and inequality (Moller et al. 2003; Bradley et al. 2003; Huber and
Stephens 2001), and it could potentially play a similar role in Latin America and
the Caribbean. However, responsible, that is non-inflationary, financing of such
programs requires a sound system of taxation, something that is scarce in
developing countries, including in Latin America and the Caribbean. Systems of
taxation on their part have important implications for class and gender equity.
This chapter explores changes in the systems of taxation in four Latin American
and Caribbean countries – Argentina, Chile, Costa Rica, and Jamaica – from the
point of view of their gendered impact.
The starting point of the argument is that effective tax collection is a necessary,
though not sufficient, condition for the amelioration of gender-based poverty and
inequality. Low-aggregate tax collection hurts women because it prevents the
establishment of programmes that counteract market distribution of income, in
which women are generally disadvantaged. They are disadvantaged in market
income because they provide the bulk of the non-paid care work, because their
paid work takes place in the informal sector to a greater extent than men’s paid
work, and because – if they work in the formal sector – they tend to be employed
in smaller enterprises and to earn less than men.
Latin American countries as a whole have been undertaxing their populations,
with an average tax burden of 14 per cent of GDP in the first half of the 1990s,
compared to 17 per cent of GDP in a group of East and Southeast Asian countries
(IADB 1996: 128). Direct taxes amount to about 25 per cent of tax revenue only,
and of this amount some 60–80 per cent typically comes from corporate tax
301
302 Evelyne Huber
payments, while only 10–15 per cent comes from private individuals (ECLAC
1998: 72). Interestingly, the situation in the English-speaking Caribbean has been
very different, with an average tax burden in the first half of the 1990s of
27–28 per cent of GDP, essentially double the rate of Latin America, and direct
taxation on individuals and corporations accounting for some 40 per cent of tax
revenue (ECLAC 1998: 66–72). This contrast suggests that the fundamental reasons
for the poor tax collection performance in Latin America are poor policy choices,
rather than low levels of economic development and technological capacity.
Changes in tax structures since the 1970s
The dominant pattern of tax reforms over the past three decades in Latin America
and the Caribbean, which – with some exceptions – holds true for our four cases,
has consisted of reductions of marginal tax rates on corporate and individual
incomes and attempts to broaden the tax base, but these have not succeeded in
significantly increasing income-tax collection. At the same time, revenue from
imports and exports and other international transactions declined significantly in
many cases due to the lowering of tariff rates and other forms of deregulation.
To make up for this decline and to facilitate tax collection, most countries have
shifted more weight to value-added taxes (VAT). There has been more variation in
the treatment of social-security taxes, corresponding to the different approaches
with regards to social policy reforms.1
Advocates of this pattern of tax reform, prominent among them representatives
of the World Bank, have argued that reduction of marginal tax rates reduces the
incentive for tax evasion (not declaring income or transactions) and tax avoidance
(manoeuvreing within the tax code to reduce tax liability) and that broadening of
the tax base reduces non-neutral tax treatment that encourages resource shifts to
relatively low tax activities and thus distorts resource allocation. Moreover, reducing
the complexity of the tax structure makes tax collection less costly and more effective
Table 14.1 Tax categories
Direct taxes Income, individual

Income, corporate
Property, wealth and inheritance
Indirect taxes on Value added (VAT)
consumption Sales
Excise on selective products;
‘sin’ taxes, e.g. alcohol, tobacco
Trade taxes Import duties
Export duties
Social security Employees/ self-employed in
taxes formal sector
Employer payroll in formal sector
Other taxes
Gendered Implications of Tax Reform in Latin America 303
(Thirsk 1997: 8–11). VAT in particular is supposed to be easy to collect and – if

broadly based – not to have distorting effects on resource allocation.
In all these countries, though, tax evasion and avoidance have remained
fundamental problems. Both individuals and corporations continue regularly to
engage in both. Despite administrative reforms to upgrade the status, equipment,
and quality of personnel of the tax administration agencies, these agencies have
fewer human resources and smaller budgets but higher collection costs than their
counterparts in the more developed countries (ECLAC 1998: 83).
Data and sources

Tables 14.2 and 14.3 show the overall level and the composition of taxes in 1980/81,
before the onset of the general Latin American debt crisis, and in 1999/2000, after a
decade or more of economic reforms. A note of caution is in order here. These data
Table 14.2 Taxes as a percentage of GDP
Argentina Chile Costa Rica Jamaica
Tax category 1980–81 1999–2000 1980–81 1999–2000 1980–81 1999–2000 1980–81 1999–2000
Income, profits, 0.0%a 2.7% 5.5% 3.8% 2.7% 2.9% 10.8%b 9.2%b
capital gains
& property
Goods and services 5.4% 5.9% 12.0% 10.0% 5.0% 8.0% 14.5% 9.3%
International trade 0.9% 0.7% 1.5% 1.4% 4.1% 1.1% 1.3% 2.2%
Social-security 2.7% 3.4% 4.8% 1.4% 4.8% 6.2% 1.2% 0.0%
contributions
Other 4.4%a 0.0% 1.8% 0.8% 0.2% 0.0% 1.0% 2.0%
Total 13.5% 12.7% 25.7% 17.4% 16.8% 18.2% 28.8% 22.8%
a
Carciofi and Cetrángolo (1995) show a different figure in ‘Income, profits, capital gains & property’, and
a lower in ‘Other’.
b
Includes some payroll taxes that are social policy schemes.
Source: IMF Government Finance Statistics, and IMF International Financial Statistics.
Table 14.3 Taxes as a percentage of total revenue
Argentina Chile Costa Rica Jamaica
Tax category 1980–81 1999–2000 1980–81 1999–2000 1980–81 1999–2000 1980–81 1999–2000
Income, profits, 0.0%a 20.9% 21.5% 21.8% 15.8% 15.8% 37.6%b 40.6%b
capital gains &
property
Goods and services 41.1% 46.7% 46.8% 57.3% 29.8% 44.0% 50.3% 40.9%
International trade 7.1% 5.8% 6.0% 8.8% 24.4% 5.9% 4.5% 9.7%
Social-security 19.6% 26.5% 18.7% 8.2% 28.7% 34.3% 4.1% 0.0%
contributions
Other 32.1%a 0.1% 7.1% 4.7% 1.4% 0.0% 3.5% 8.9%
Total 100.0% 100.0% 100.0% 100.7% 100.0% 100.0% 100.0% 100.0%
a
Carciofi and Cetrángolo (1995) show a different figure in ‘Income, Profits, Capital Gains & Property’, and
a lower in ‘Other’.
b
Includes some payroll taxes that are social policy schemes.
Source: IMF Government Finance Statistics, and IMF International Financial Statistics.
304 Evelyne Huber
come from the IMF International Financial Statistics (1991, 1997, 2003) and
Government Finance Statistics (1988, 1991, 1995, 2003). Other authors cite data that
may vary quite considerably from those presented here. For instance, Tanzi (2000)
cites ECLAC data that show significant discrepancies to our data for Chile and Costa
Rica in the earlier period and for Argentina and Costa Rica in the later period.
Carciofi and Centrángolo (1994), citing Tanzi (1987), present data for 1980–82
for Argentina that are much higher than ours, with total tax revenue accounting
for 19.9 per cent of GDP, and Gómez Sabaini et al. (2002) present figures for
Argentina for 1997 that show central government tax revenue as 19.7 per cent of
GDP, provincial tax revenue as 3.7 per cent of GDP and municipal revenue as
1.2 per cent. Thus, one would have to add about 25 per cent to our total central
government tax-revenue figures to arrive at the total tax burden. In addition, different
authors and institutions may classify certain taxes differently.
Having acknowledged these data problems, we can make three points to affirm
the usefulness of the IMF data for the comparisons to be made here. First, our data
all come from the same source and thus are more comparable than data pieced
together from different sources. Second, IMF figures are important because they are
used in setting targets for governments in the formulation of fiscal policy. Third, if
we look at the pattern of change in composition of tax revenue over time, and if we
are attentive to and correct for different classification decisions, the various sources
and authors paint a similar picture: They show a decrease in reliance on foreign-
trade taxes and an increase in reliance on indirect taxes, with the share of direct
taxes remaining roughly constant. The important exception is Jamaica.
The incidence of different categories of taxes

In general, the following assumptions are made regarding the incidence of
different kinds of taxes. Direct taxes on income, profits, and property tend to be
universally progressive and indirect taxes generally regressive, though the degree
to which they are regressive varies with the exemption of basic goods and services. If
such exemptions are targeted narrowly on a basket of basic goods and services
consumed by low-income earners, indirect taxes such as sales taxes or VAT become
less regressive (Hossain 2003).2
Social-security taxes are ambiguous; the share paid by employees is generally a
percentage of earnings and thus proportional, but since there are often caps on
contributions, these taxes become regressive at the upper end. The share of social-
security taxes paid by the employer may be passed on to the final price of the prod-
uct and thus at least partly shared by the consumer and work like an indirect tax.
However, in the new open economic environment, possibilities for passing on
these costs to the consumer are more limited, so we can assume that a larger share
is carried by the employer and – even more likely – the employees (in the form of
lower wages to absorb some of the employer’s share).
The situation in the early 1980s

In 1980/81, Chile and Jamaica were clearly at the upper end of Latin American
countries in terms of total tax collection as a percentage of GDP, with central
government tax revenue accounting for 25.7 per cent of GDP and 28.8 per cent
respectively, whereas Costa Rica with 16.8 per cent of GDP and Argentina with
13.5 per cent of GDP were closer to the mean. It is important to note that Chile
had already undergone radical economic reforms and a tax reform by this point.
Indeed, Chile already relied on indirect taxes for close to half of total tax revenue,
with 47 per cent. Direct taxes, that is, income and capital-gains taxes on individuals
and corporations, together with property taxes, amounted to between 16 per cent
of total tax revenue (in Costa Rica) and 38 per cent (in Jamaica).3 The figure for
Jamaica is inflated because some earmarked taxes for social policy are added to the
income tax, which in other countries would be classified as social-security taxes.
Chile and Argentina had already lowered taxes on foreign trade by the early
1980s, whereas they continued to play a very important role in Costa Rica, more
important than income taxes. For Jamaica, we are dealing with a classification
problem. Bahl (1989), for instance, shows 25.5 per cent of total government revenue
in the early 1980s coming from export duties. In addition, he shows 20 per cent
coming from import duties, so the total contribution of foreign-trade taxes to gov-
ernment revenue in his data is 45.5 per cent, as in Costa Rica clearly higher than
income taxes. In the IMF data, stamp duties on imported goods must be classified
under taxes on goods and services and thus greatly inflate that figure. If we count
them as taxes on foreign trade, then Jamaica is in an intermediate position with
regard to the relative share of foreign trade and domestic indirect taxes, compared
to Chile and Argentina on the one hand and Costa Rica on the other. The former
two countries have a low share of total taxes coming from foreign trade and a high
share from domestic taxes on goods and services, and the latter has a comparatively
higher share coming from foreign trade and a lower one from domestic taxes on
goods and services.
Tax reform in Argentina

In Argentina, one really cannot speak of a particular episode of tax reform, but
rather has to consider a protracted process of changes in the tax system, responding
to fluctuations and shocks in the macroeconomic environment.4 In the 1980s, the
crushing debt burden, runaway inflation and recession required efforts to increase
tax revenue with a series of different measures. The successful stabilization pro-
gramme of the early 1990s then opened the way for a more coherent tax-reform
programme focused on increasing overall tax receipts and relying heavily on VAT.
From the mid-1990s on, several minor tax changes were implemented in response
to recessions and growing budget deficits, but they had generally little impact
(Gaggero and Gómez Sabaini 2002: 38).
The VAT rate rose from 16 per cent in 1980 to 21 per cent in 1995. In 1998 the
base for VAT was expanded to include new services, but at the same time the
rate was lowered for agricultural products and some others. As Table 14.3 shows,
the most remarkable change in the tax structure between the early 1980s and 1999/
2000 was indeed the increase in revenue from indirect taxes. However, expressed
as a percentage of GDP, the increase in revenue from indirect taxes was only slight,
because total tax revenue of the central government as a percentage of GDP declined
306 Evelyne Huber
slightly. Nevertheless, this category of taxes is by far the biggest source of tax
revenue, bringing in more than twice the revenue from income taxes. As a
reminder, we should add tax revenue at the provincial level and at the municipal
level amounting to roughly 25 per cent of tax revenue of the central government,
so that by the late 1990s total tax revenue for all levels of government reached
roughly 16 per cent of GDP. This is still lower than tax revenue in Chile, Costa Rica
and Jamaica.
A study of the distributive impact of the tax system, based on the 1997 house-
hold survey in urban areas and a survey of the consumption structure, shows that
it is slightly regressive (Gómez Sabaini et al. 2002). Since 88 per cent of the popu-
lation reside in urban areas (World Bank 2003), these data characterize the tax
system as it affects close to nine-tenths of the population. The GINI index for
income distribution among deciles increases from 0.55 before taxes to 0.58 after
taxes. This is mainly due to the VAT burden that weighs particularly heavily on the
lowest decile and least on the highest. The structure of social-security contribu-
tions is equally regressive, though it has a somewhat lower weight on the lowest
decile than VAT. The income-tax burden is progressive. However, since VAT has
such a central role in the tax structure, the system as a whole is shaped by its
incidence. Thus, looking at the growing role of VAT and of social-security contri-
butions from the early 1980s to 1997, we can conclude that it has become more
regressive. Indeed, the authors analysed data from 1986 with the same methodology
as the 1997 study and found that the degree to which the tax system was regressive
was more marked in the later period (Gómez Sabaini et al. 2002: 67).
Tax reform in Chile

In Chile, major reforms of the tax system began comparatively early, as part of the
neoliberal restructuring of the Chilean economy under the Pinochet dictatorship.
In 1975, Chile introduced VAT at a rate of 20 per cent, the highest rate in Latin
America and the Caribbean at the time. As of the early 1980s, close to half of Chilean
tax revenue came from domestic taxes on goods and services (see Table 14.3).
VAT remained broadly based and there were very few exemptions; the rate fluctuated
between 20 and 16 per cent. In 1984, the government lowered income tax, reducing
the marginal tax rate and widening the tax brackets. Savings were exempt from the
tax base and deductions for some financial investments were introduced, clearly
reforms benefiting upper income earners.
Tax reform and increases in social spending became a major issue in the elec-
tions of the first democratic government. In 1990, the newly elected government
arrived at an agreement with business and the opposition that allowed it to raise
taxes temporarily, for a period of four years, to finance an increase in social expen-
ditures. The government raised the VAT rate from 16 to 18 per cent and reversed
some of the more regressive changes in the income tax that had been imple-
mented in 1984. It also closed some loopholes in corporate taxation and imposed
taxes on all business profits, raising the rate for 1991–93 from 10 to 15 per cent.
The increase in tax revenue resulting from this reform was estimated at 1.5 per cent of
GDP, which in turn allowed an increase in social expenditures of 17.4 per cent
over the original budget for 1990, and an increase in real expenditures on social
investment, mainly in healthcare, education and housing, of 12.2 per cent from
1990 to 1991 (Schkolnik 1992: 27–8). Tax revenue as a percentage of GDP
increased from 14.5 per cent in 1990 to 18 per cent in 1993.
Similar to Argentina, the tax structure in the second half of the 1990s
emphasized indirect taxes, particularly VAT, and de-emphasized direct taxation,
particularly individual income taxes (see Table 14.3). One of the main reasons for
the small contribution made by personal income taxes is the fact that declared
incomes are low, even in the top decile. Wealthy Chileans have the option of
forming a company, transforming part of their personal incomes into business
income, computing various expenses as costs, buying assets that are used in
personal consumption, and paying the flat 15 per cent business tax on profits. This
is a perfectly legal form of tax avoidance (Engel et al. 1998).
It is no surprise, then, that the tax system as a whole is somewhat regressive, just
like in Argentina. The Gini index before taxes in 1996 was 0.49, and after taxes
0.50 (Engel et al. 1998). Again, the income tax is clearly progressive and VAT clearly
regressive, but since VAT accounts for a much larger share of taxes, the system as a
whole is regressive. Gómez Sabaini et al. (2002) cite results from studies of tax
incidence in Chile in 1969 (Foxley et al. 1980) and 1997 (Engel et al. 1998), which
clearly show that the system was somewhat progressive in 1969 and became
significantly more regressive with the series of reforms since the 1970s. What is
important to note is that public expenditure in Chile in the 1990s did ameliorate
inequality; Engel et al. (1998) estimate that the Gini index fell from 0.5 after taxes
but before spending to 0.43 after taxes and spending.
Tax reform in Costa Rica

In Costa Rica, the same general pattern is visible as in Argentina and Chile, with a
decline of reliance on international trade taxes and an increase in reliance on
indirect taxes (see Table 14.2). It is interesting to note here that Costa Rica does not
have a generalized VAT; it only exists in some industries. Most of the indirect tax
revenue comes from sales and excise taxes. Tax reform was part of the three
structural adjustment agreements that the country entered into with the World
Bank covering the period 1985 to 1998 (Quesada et al. 1999: 109–12). The two
motives for tax reform were stimulation of non-traditional exports and increases
in tax revenue in order to reduce the budget deficit. Under the 1985 agreement
with the World Bank, taxes on profits from non-traditional exports were elimi-
nated, as were import taxes on inputs for export production, and customs duties
in general were lowered. These efforts were continued under the following two
agreements, with an emphasis on simplification of the tax system, broadening of
the tax base, and efforts to strengthen tax administration. The top income-tax rate
was lowered from 50 per cent to 25 per cent. In 1995, the rate of the sales tax was
increased from 10 per cent to 15 per cent. In 2003, a new tax-reform project
was accepted by a congressional committee, envisioning the introduction of a
VAT with a rate of 13 per cent, and a uniform company tax with a rate of 18 per
cent (www.nacion.com; accessed on 1 December 2003).
308 Evelyne Huber
In contrast to Chile and Jamaica, where total tax revenue as a percentage of GDP
fell rather markedly between 1980/81 and 1999/2000, Costa Rica managed to
increase revenue slightly in this period, from 16.8 per cent to 18.2 per cent of GDP,
which put it second only to Jamaica among our four countries. What makes Costa
Rica distinctive is the high share of total tax revenue accounted for by social-
security taxes. These increased their share from 29 per cent to 34 per cent of the
total in this period, whereas they increased also, but at a lower level, from 20 per cent
to 26 per cent in Argentina but declined from 19 per cent to 8 per cent in Chile.
The large decline in Chile is heavily due to the fact that employer contributions
were abolished when the pension system was privatized and employers contribute
nothing for health insurance for their employees either.5 As noted above, the data
for Jamaica are not comparable because taxes designated for social insurance and
services are included under income taxes. As in Chile, direct taxes on income,
profits and property retained a roughly constant share of the total in Costa Rica,
with 16 per cent.
I was not able to find any studies of the incidence of taxation in Costa Rica.
However, if we look at the composition of taxation in Costa Rica and Argentina in
1999/2000, we find a remarkable similarity (see Table 14.3). Thus, we might
assume that the distributive impact of the tax system is similar. However, this
assumption would ignore the differences in the structure of indirect taxes in the
two cases shaped by exemptions, as well as the differences in social-security taxes.
In Costa Rica, a basket of basic goods, veterinary products and agricultural
inputs are exempt from sales tax. In Argentina, a wide range of goods and services
is exempted from VAT, but these exemptions tend to benefit all income groups,
rather than being focused on low-income groups (CIPPEC 2002). Since the basket
of basic goods and agricultural inputs absorbs a large part of the income of low-
income groups but not of high-income groups, these exemptions make the Costa
Rican sales tax less regressive than the Argentine VAT. As noted, Gómez Sabaini
et al. (2002) found social-security taxes to be regressive in Argentina. These are
high and they have caps. Social-security taxes in Costa Rica are much lower
and there are no caps (Social Security Administration 1995). Thus, these taxes are
proportional to income.
Given the similarity in weight but differences in the structure of indirect taxes
and social-security taxes between Costa Rica and Argentina, with exemptions from
the sales tax more focused on goods consumed by low-income groups and social-
security taxes proportional to income in Costa Rica, we would expect the Costa
Rican tax system as a whole to be distributionally neutral or slightly progressive.
In addition, we need to keep in mind that redistribution happens mostly on the
expenditure side, and thus the fact that Costa Rica has the second highest level of
total taxation, behind Jamaica, puts the country in a comparatively favourable
position to effect redistribution. Indeed, in studies of household income, Costa
Rica along with Uruguay consistently shows the lowest Gini indices for income
distribution among 17 Latin American and Caribbean countries in the 1980s and
1990s (Morley 2001).
Tax reform in Jamaica

In Jamaica, tax reform was also begun and shaped in the context of structural
adjustment programmes negotiated with the IMF, the World Bank and USAID in
the 1980s. Indeed, the US government provided the funding for the comprehensive
tax-reform project (Bahl 1991: 9). Given the comparatively high level of taxation
in Jamaica in the early 1980s, the primary motivation was not to increase revenue
but rather to simplify the tax system and make it consistent with a more liberal-
ized economy. Among other things, this meant reducing taxes on high-income
earners and corporations who might be more likely to invest their money abroad
in the more liberalized economic environment.
The progressive rate structure of the personal income tax, with a top rate of
57.5 per cent, was replaced by a flat tax rate of 33 1/3 per cent in 1986, and in 1992
it was reduced to 25 per cent. A system of multiple tax credits was replaced by a
standard deduction; the size of this deduction was increased in 1992, when the
rate was lowered. In addition, before the reform Jamaican firms had offered their
employees a number of nontaxable ‘allowances’, such as cars, to reduce both the
firms’ and the employees’ payroll tax liability. Most of these ‘allowances’ were
made taxable. Also, the tax reform included interest as taxable income and
abolished the preferential treatment of overtime pay. The tax rate on corporations
was lowered also, from 45 per cent to 33 –13 per cent. On the other hand, there was
little change in various payroll tax schemes (for education, training, housing and
social security), which generated revenue roughly equivalent to half of individual
income-tax collection (Bahl 1991: 38). Thus, if we look at the total effect of the tax
reforms in the 1980s and 1990s, the share of direct taxes, including the payroll
taxes that in other countries would be considered social-security taxes, actually
increased from 38 per cent to 41 per cent of total tax revenue.
Jamaica’s system of indirect taxation before the reform was exceedingly
complex, and indirect taxes accounted for a majority of tax revenue in the early
1980s already. In 1991, Jamaica introduced a general consumption tax with a rate
of 10 per cent; by 1997 the rate had risen to 15 per cent. Before the reforms, 80 per
cent of indirect taxes came from excise taxes, and the largest share came from alco-
hol, tobacco, and gasoline. The emphasis on these tax sources was maintained in
the design of the tax reforms. The share of taxes on international trade in total tax
revenue more than doubled over the two decades of the 1980s and 1990s (Bahl
1997). Compared to the other three countries, it had started at the lowest level but
ended up at the highest, with close to 10 per cent, followed by Chile with 8 per cent.
According to figures presented by Bahl (1997: 219), based on Wasylenko (1991)
and Sjoquist and Green (1992), the tax system as a whole was already progressive
before the reforms and was made slightly more progressive by the reforms, at least
up to 1991 (when the data for these studies were collected). As to developments
after 1991, we can outline several factors that had countervailing effects: the
income tax rate was reduced to 25 per cent, which benefited higher-income earners,
but at the same time the standard deduction was increased, which benefited
lower-income earners; the general consumption tax rate was increased, but other
310 Evelyne Huber
indirect taxes were decreased; the exemption from the general consumption tax of
food stuffs, health supplies, and books and materials for education, among other
categories, cushioned the burden on lower-income earners.
Looking at the composition of tax revenue as a whole at the end of the 1990s
and comparing it with the composition in the early 1980s, we can venture the
assessment that the tax system developed in a slightly more progressive direction,
with direct taxes increasing their share and taxes on goods and services decreasing.
These developments stand in contrast to the other three countries, where the
share of taxes on goods and services increased, sometimes dramatically so,
whereas the share of direct taxes remained stagnant. If we look at the sum of direct
taxes and social-security taxes as a percentage of GDP, a pronounced difference
emerges between Jamaica and Costa Rica on the one hand and Argentina and
Chile on the other (see Table 14.1). In Jamaica they accounted for 9.2 per cent
of GDP at the end of the 1990s and in Costa Rica for 9.1 per cent, compared to
5.2 per cent in Chile and 6.1 per cent in Argentina. Thus, the Jamaican and Costa
Rican tax systems were clearly more effective in collecting direct and social
security taxes.
Politics of tax reform and enforcement
Despite the tax reforms, all four countries continue to suffer from widespread
tax evasion and avoidance. The tax administrations all lack sufficient qualified
personnel and equipment for effective monitoring of tax compliance and for
auditing tax returns. There are many potential taxpayers, both individuals and
corporations, who are not even registered. Of those who are registered, only a frac-
tion actually file returns, and most of them under-report their income. For
instance, Bahl (1997: 220) points out that in Jamaica in 1992 there were about
80,000 registered self-employed taxpayers, and of those only about 25 per cent
filed returns. In a random sample of self-employed people from nine occupations,
the per cent of income tax filers was 17 per cent in 1982 and 11 per cent in 1983
(Alm et al. 1991: 196). Among large enterprises, compliance with the general con-
sumption tax was around 85 per cent but among all other enterprises it was in the
low 60 per cent range (Bahl 1997). For Chile, Engel et al. (1998) estimate that
around 23 per cent of the total tax base is not reported. In Costa Rica, a report
issued by the Controller General’s Office found that sales tax evasion in 2002
amounted to 46 per cent of the total, or roughly 2 per cent of GDP (La Nación,
8 March 2004). The evasion rate for VAT on domestic transactions in Argentina in
1993 was estimated at 32 per cent, a huge improvement from the 62 per cent in
1989, but still a large gap (Rosenwurcel 2000: 45).
An important part of the reason for continued tax evasion is the reluctance of
governments to punish it, particularly among major taxpayers. The example
of Jamaica shows that low to intermediate levels of development are not insur-
mountable barriers to tax collection due to insufficient technical capacity. In 1980/81,
Jamaica collected 29 per cent of GDP in tax revenue, with GDP per capita at
US$3,452 in 1980. In 1999/2000, Chile collected 17 per cent and its GDP per
capita in 1999 stood at US$9,568, and Argentina collected 13 per cent with a 1999
GDP per capita of US$11,181.6 It is clearly not technically feasible to enforce total
compliance (this is true for any of the advanced industrial democracies also), but
it is feasible in all four countries to improve tax collection capacity, if governments
develop the political commitment and support to do it.
If we look at the politics behind the tax changes, we find that these changes
were for the most part treated as a necessary component of general economic
reforms. In Costa Rica and Jamaica, they were explicitly and openly tied to donor
conditionality. In Argentina and Chile, the influence of the IFIs was exerted more
through technocratic networks of the sort analysed by Teichman (2001). She traces
the establishment of networks among officials of IFIs, particularly the World Bank
and the Inter-American Development Bank, and high-level government officials
in Argentina, Chile and Mexico. Through regular interaction over long periods of
time these networks became the carriers of neoliberal ideas and thus shaped
reform designs in a neoliberal direction even in the absence of explicit conditionality.
The fact that many of the high-level government officials shared educational
backgrounds with the IFI officials, such as PhD degrees in economics from US
universities, made them more susceptible to the transmissions and reinforcement
of neoliberal ideas.
There are two different paths to tax reform, exemplified by Argentina and Jamaica.
In Argentina, tax changes were generally ad hoc reactions to economic changes,
whereas in Jamaica they were part of a major coherent reform of the entire system of
taxation. In all four countries, the debates about tax reform were generally confined
to a relatively small circle of technocrats in the Ministry of Finance or Economics,
with minimal public involvement. Parliamentary debates tended to be comparatively
short also. Compared to other economic reforms, such as reduction of subsidies for
basic goods, energy, and the like, or privatization of state-owned enterprises, tax
reforms stimulated little popular protest. These other reforms had much more direct,
sudden and visible impacts on the population than the tax reforms that were phased
in more gradually. In some cases, such as the forced-savings scheme in Argentina
introduced in 1985, there were protests, but mainly after the fact.
Of course, one should also ask why Jamaica started from such a different
pre-reform position and remained more concerned about protecting lower-income
earners than Chile and Argentina. The key is the strength of the People’s National
Party (PNP), a democratic socialist/social democratic party that alternated in office
with the Jamaica Labour Party (JLP). The PNP has had a clear commitment to
reduction of poverty and inequality and when in office oriented both tax and
expenditure policies towards those goals (e.g. Manley 1982; Stephens and
Stephens 1986).7 Moreover, both the PNP and the JLP had close links to labour
unions, so the JLP governments were constrained in rolling back programmes that
the PNP had established. Even the tax reform discussed here, under JLP Prime
Minister Edward Seaga, had as one of its goals to avoid making the lower classes
worse off (Bahl 1997: 186). The National Liberation Party (PLN) in Costa Rica used
to play a similar role in promoting social programmes and a mode of financing
these programmes that would protect lower-income earners.
312 Evelyne Huber
In Chile, in contrast, major reforms of social policy and the tax system were
carried out under the dictatorship of Pinochet and were oriented towards stimu-
lating economic growth in the private sector and reducing the responsibilities of
the state for social welfare. Since unions and parties of the left were ruthlessly
repressed, these reforms could be carried out without having to deal with real
opposition. In Argentina, the major neoliberal reforms of economic and social
policy were carried out by President Menem, ironically from the Peronist party, the
traditional pro-labour party in the country. The party had, however, always been
highly personalistic and had turned into a non-programmatic, clientelistic party.
Major changes in the tax system towards increasing reliance on indirect taxes
had already been effected by the military government, and the various steps going
further in this direction with complicated packages did not evoke much public
debate.
Given the general absence of public debate about tax reform, it is little surprise
that concerns about the gender-specific impacts of the changes were hardly raised at
all. Effects of tax changes were not near the top of the agenda of women’s
movements in these countries either. In Argentina and Chile, human rights, democ-
ratization and cooperative economic-survival strategies dominated the agenda. To
the extent that women protested against economic policy in all of the four coun-
tries, it was mostly against the general austerity measures, visible in higher unem-
ployment and higher prices. Deteriorating pensions and deteriorating healthcare
were additional issues that elicited protests from women (as well as men). In short,
there were so many policy decisions taken to transform these economies in a neolib-
eral direction, in addition to general austerity measures, that tax-reform decisions
did not stand out in any major way in the public mind in general and in the mind
of women activists in particular. This may be in part cause and in part effect of the
lack of studies of the gender-specific effects of tax reforms in Latin America.
Gender-specific effects of tax reforms
In the absence of disaggregated data, tracing the implications of changes in tax

legislation for different groups by necessity has to be based on logical reasoning
and educated guesses as to tax incidence. As ECLAC states, in its most authoritative
publication on the issue,
… the skepticism of the 1980s and early 1990s with respect to the redistributive
capacity of taxation has resulted in an almost total lack of up-to-date, detailed
empirical studies on the implications and distributional effects of various taxes
or combinations thereof in countries of the region. After almost two decades of
intensive and far-reaching reforms of the tax, tariff and social security systems,
there is not enough data to permit an informed assessment of such issues
ECLAC 1998: 96–7.
The main results of the few studies that exist for our countries were discussed
above. All of them looked at incidence on income groups, not a single one contained
any data on gender. Thus, the attempt to improve our understanding of the
gender-specific implications of these changes is more speculative, based on a
combination of theoretical and comparative knowledge and reasoning.
Many analysts would argue that the most important issue for married women is
whether the taxes of married couples have to be filed jointly or can be filed
separately. Joint taxation in the presence of a progressive income tax works as a
disincentive for married women to join the labour force.8 From that point of view,
all four cases are gender-egalitarian. In Chile, the wage tax and general income tax
are personal, levied on individuals and not on households (Engel et al. 1998).
In Argentina and Costa Rica married couples are taxed separately on their income,
but only one of them (in Argentina the husband) is taxed on property income and
property, and only one of them can take deductions for dependants (in Costa Rica
they can be claimed by either spouse). Where there is a flat-rate income tax, as in
Jamaica, individual or joint filing does not matter.
In order to assess the gendered implications of the changes in the tax structure, we
have to begin by situating women in the occupational and earnings structure of their
countries. In every one of our four countries, female labour-force participation is lower
than male. Though female labour-force participation increased significantly between
1990 and 2000 in Chile and Costa Rica, from 35 per cent and 39 per cent to
42 per cent and 43 per cent, respectively, and in Argentina slightly from 44 per cent
to 45 per cent, this rate remained clearly below the male rate, which varied from
73 per cent to 77 per cent (ECLAC 2002). The women who are in the labour force are
more likely to work in the informal economy, that is, in small unregistered enterprises,
own-account and non-remunerated labour, and as domestic workers (Table 14.4).
Women who work in the formal sector are over-represented in the public sector,
most probably as teachers and nurses. In the private sector, they have roughly
equal representation in professional and technical jobs to men but are greatly
under-represented in medium and large enterprises, that is, enterprises that employ
more than five workers (ECLAC 2002: 179–82). Finally, open unemployment
is higher among females than among males (ECLAC 2002: 207). Women earn
less than men, and the differences are greater if we look at the entire employed
population than if we consider just wage-earners. In 1999/2000, the ratio of average
female to average male incomes among the entire employed population in urban
areas was 61 in Chile, 65 in Argentina and 70 in Costa Rica (ECLAC 2002).
Given the lower labour-force participation rates and lower earnings of employed
women, the low share of revenue that comes from direct taxes clearly favours
males, because they are higher-income earners and more likely to be owners of
corporations or shares in corporations than women. Accordingly, men constitute
a stronger tax base and would have to pay more if the rate structure was more
progressive and exemptions lower, and – most of all – if the laws were enforced.
The same is true for the low share of total taxes coming from property taxes.
Social-security taxes are tied to earnings of employees and of those self-employed
who are covered, so their incidence on males and females should be similar to the
incidence of income taxes. However, to the extent that social-security taxes have a
cap and men tend to be higher-income earners, the effect of social-security taxes is
314 Evelyne Huber
Table 14.4 Population employed in the urban informal sector
Wage
earners in Own-account and Domestic
Year Country Total businesses non-remunerateda servants
Argentinab
1990 Men 37.3 12.4 23.1 1.8
Women 45.4 10.2 22.7 12.5
1994 Men 35.7 14.5 20.8 0.4
Women 42.5 11.5 18.7 12.3
1997 Men 34.1 17.5 16.2 0.4
Women 39.8 9.6 17.5 12.7
1999 Men 34.2 15.9 18.1 0.2
Women 38.3 10.3 15.3 12.7
2000 Men 35.1 16.4 18.6 0.1
Women 40.6 12 15.7 12.9
Chile
1990 Men 32.2 10 22 0.2
Women 45.5 8.2 18.2 19.1
1994 Men 27.5 9.1 18.3 0.1
Women 40.3 7.7 15.8 16.8
1996 Men 26.9 9.7 17 0.2
Women 39 8.2 14.5 16.3
1998 Men 26.2 9.7 16.4 0.1
Women 38.1 9.7 13.2 15.2
2000 Men 23.8 7.9 15.8 0.1
Women 36.7 7.4 13.3 16
Costa Rica
1990 Men 28.6 10.3 18.1 0.2
Women 37.2 8.6 16.6 12
1994 Men 28.6 11.6 16.7 0.3
Women 36.5 10.3 16.1 10.1
1997 Men 29.7 12.4 17.1 0.2
Women 37.1 9.2 18.7 9.2
1999 Men 30.4 13.3 16.7 0.4
Women 40.1 9.4 18.1 12.6
2000 Men 31.2 12.4 18.5 0.3
Women 38 10.9 15.7 11.4
a
Does not include professional and technical jobs.
b
Covers Greater Buenos Aires only.
Source: Tables 4.1 and 4.2 in Panorama Social de America Latina 2001/02 – ECLAC 2002.
likely to be more regressive on women than on men. Moreover, we need to keep

in mind that social-security taxes also entitle the payers to benefits and that the
relationship between social-security taxes and benefits is much closer than the
relationship between general taxes and benefits from all government activities.
This means that men will tend to benefit more from these taxes than women.
The implications of the changes in tax legislation regarding direct taxes are
partly contradictory. Clearly, the reduction of marginal tax rates and of payroll
taxes benefited mostly men, because they are the high-income earners and formal-
sector employers and employees. On the other hand, to the extent that exemp-
tions and loopholes were closed, such as car allowances, these changes affected
men more negatively than women, as men were more able to take advantage of
those means before the reforms to reduce their tax payments. If we look at the
overall picture, we can assume that the burdens of direct taxes on men and women
remained roughly unchanged. If we introduce social-security taxes into the picture,
we see an increase in the share of these taxes in Argentina and in Costa Rica, and
a decrease in Chile. Accordingly, we would conclude that the total tax burden on
men, from direct taxes and social security taxes, was increased in Argentina and in
Costa Rica to a greater extent than on women, and that men experienced a greater
decrease in their tax burden than women in Chile.
We can begin an assessment of the gendered effects of the shift to greater
reliance on indirect taxes with some general observations. To the extent that
indirect taxes are regressive, which they clearly are in Argentina and Chile (Engel et al.
1998; Gómez Sabaini et al. 2002), we can assume that women as lower-income
earners are more disadvantaged. Where the degree to which they are regressive is
tempered through exclusions focused on basic goods and services in health and
education, as in Costa Rica and Jamaica, the incidence is presumably more gender-
neutral. To the extent that a large share of indirect taxes comes from excise taxes
on alcohol and tobacco, women as lower consumers of these goods would enjoy
an advantage – assuming that expenses for alcohol and tobacco do not come out
of the shared household budget. This forces us to think more about women’s
position in the household and about the determination of household budgets, a
topic to which I shall return in a moment.
The final incidence of taxes on foreign trade depends on the share that is passed
on to the consumer. In the case of import taxes on consumer goods, they simply
increase the price of those goods in the domestic market, and the incidence depends
on whether these goods are essential to the poor. Import taxes on intermediate and
capital goods increase the price of goods manufactured with the imported ones,
and the extent to which those increases are passed on to consumers depends on the
competitiveness of markets. In more competitive markets, presumably a greater
share of the import taxes for intermediate and capital goods would be absorbed by
the manufacturers, and in less competitive markets a greater share would be passed
on to consumers. In the latter case, we would again need to know whether those
goods are part of the necessary consumption of poor households. Everything else
being equal, a lowering of taxes on imports should have lowered prices to the con-
sumer and thus partially offset an increase in VAT, sales or consumption taxes.
In order to advance further in assessing the gendered impact of changes in the
tax systems towards greater reliance on indirect taxes, we need to situate women
not only in the occupational and earnings structure, but also in the structure
of households. We can think of at least eight combinations of marital status
(married/cohabiting in stable relationship or single/widowed/divorced), labour-force
316 Evelyne Huber
participation (working or not working), and children (with or without children).

For all married/cohabiting women, working or not working, with or without
children, it seems that the crucial distinction is between household budgets that
are fixed by the husband/partner and those that are consensually decided upon.
If a married woman is working, we can assume a greater probability that the house-
hold budget will be consensually decided upon, and this probability increases with
the share of total income earned by the woman. In the case of consensual budget-
making, one could assume that income and expenditures are shared and thus the
final experienced incidence of taxation is the same for both spouses/partners,
whether the woman works or not, and whether children are present or not. In real
life, it is probable that the woman will be in charge of administering the commonly-
decided-upon household budget and thus will experience price increases caused
by rising indirect taxes more directly. Therefore she will spend more time and
energy to compensate for these rising prices.
In the case of unilateral budget-making, where the woman gets a certain amount
of money to take care of the needs of the household and the husband/partner
retains the rest for personal consumption and investments, we can distinguish
gendered impacts of indirect taxes even more clearly. To the extent that the fixed
household budget is small, women will be disproportionately affected by indirect
taxes that increase the price of basic food items and health and education, and
that effect will be magnified by the presence of children. Again, where these
items are exempt from the VAT or sales or consumption taxes, as they largely are
in Costa Rica and Jamaica, the negative impact of indirect taxes on women is
mitigated.
Excise taxes on gasoline affect bus fares and other transportation costs. Similar
to other indirect taxes, women on low fixed household budgets will be more
affected than their husbands/partners, given that they bear the responsibility for
getting the children to where they need to go and for procuring household goods
that may require considerable journeys. On the other hand, among middle- and
upper-income earners, men/husbands are more likely to own and use cars and
thus to carry the burden of taxes on gasoline.9 The gendered impact of excise taxes
on alcohol and tobacco depends on one’s assumptions about their effect on the
share of total household income designated by husbands/partners for household
needs versus their personal consumption. In general, women are less likely to
consume alcohol and tobacco than men, so they are less directly affected. For
low-income earners, it seems reasonable to assume that higher excise taxes will
shrink the share of the budget allocated by men for household needs and thus will
affect women with fixed budgets negatively. However, in a study of the impact of
indirect taxes on poor households in Jamaica, the authors found that excise taxes
on alcohol and tobacco, which were comparatively high, did not have a major
impact, as reported expenditures on alcohol and tobacco constituted only 2.2 per cent
of total expenditures for the sample of low income households in 1983/84 (Bird
and Miller 1989).
Single working women without children are affected primarily by the greater
probability of earning lower incomes than men and thus are disadvantaged by
regressive taxes of all kinds, social security and indirect. Single women with
children are generally in the economically weakest positions; not only are they
likely to be disadvantaged in the labour market but they also bear the responsibil-
ity of providing for their children. Thus, they tend to have low incomes and have
to spend more time and energy to provide for the needs of their families if prices
increase due to tax increases. Unless the exemptions from the indirect taxes are
heavily directed towards basic foods, transportation, health and education services,
single women with children will shoulder the greatest burden from increases in
indirect taxes, that is, they will pay the largest percentage of their income for these
tax increases.
Summary and policy lessons
We could detect two clear tendencies as a result of the tax reforms effected in
Argentina, Chile, Costa Rica and Jamaica, over the last two decades of the twentieth
century. First, in three of the countries overall tax receipts as a percentage of GDP
fell; the exception is Costa Rica. Second, three of them came to rely more heavily
on indirect taxes levied on domestic goods and services; the exception here is
Jamaica. Both of these developments are negative from the point of view of reduction
of poverty, inequality and gender-based disadvantages through state intervention.
Lower tax revenue means a reduced capacity of the state to correct market outcomes
that disadvantage the poor and women by financing health and education services
and anti-poverty programmes. Indirect taxes are generally regressive, though the
extent to which they are regressive depends on exemptions of basic goods and
services. Women as lower-income earners and as mothers responsible for the needs
of the family, particularly if on fixed budgets, are more negatively affected than
men by changes that increase regressive indirect taxes.
In this light, Chile and Argentina have clearly done worse by their low-income
earners and by women than Costa Rica and Jamaica. They both lowered their tax
burden, which now is lower than in both Costa Rica and Jamaica, even though
their GDP per capita is much higher. They both increased their reliance on indirect
taxes on domestic goods and services, and their share from these taxes is higher
than the shares in Costa Rica and Jamaica. Their structure of exemptions does not
focus on basic goods and services to the same extent as that in Costa Rica and
Jamaica. Accordingly, their tax systems are regressive, whereas the Jamaican tax
system is progressive. Studies cited above have shown that they are regressive
across income classes, and we can conclude that they are also regressive in gender
terms. The lesser reliance on indirect taxes on domestic goods and services and the
more focused exemptions from these taxes prevent the indirect taxes in Costa Rica
and Jamaica from making the whole tax systems regressive, from both a class and
gender point of view.
Two major policy lessons emerge from this analysis. First, it is possible to
construct overall progressive tax systems that are effective in raising at least moderate
levels of tax revenue in Latin American and Caribbean countries, even at com-
paratively low levels of development, as demonstrated by the example of Jamaica.
318 Evelyne Huber
The keys are reliance on direct taxes with an adequate standard deduction and few
special deductions for higher-income earners for an important share of the total,
a structure of indirect taxes that exempts only basic goods and services, and social-
security taxes without a cap. In Argentina and Chile, income taxes are progressive
but the tax systems on the whole are regressive, because of the heavy reliance on
and the regressive structure of VAT. However, the extent of redistribution that can
be achieved through the tax system in practice seems to be limited. In particular,
the very top incomes are difficult to tax for political reasons. The wealthiest
(potential) taxpayers are also investors who need to be cultivated by governments
in order to generate economic growth, and they may also be donors to political
campaigns who need to be cultivated by politicians to win elections. Moreover, in
this era of liberalized capital markets and internationalized production it has
become more difficult to tax corporations. Finally, powerful international actors
oppose high tax rates. A World Bank study argues that, ‘most economists and
policymakers consider that high tax rates not only significantly discourage
and distort economic activity but are largely ineffective in redistributing income and
wealth’ (World Bank 2004: 130).
Second, it may be as important to enforce the tax laws that are already on the
books as to spend energies and political capital on changing the tax systems in a
major way, such as by introducing a VAT or increasing its importance on the
grounds that VAT is easier to collect than other kinds of taxes. This by no means
implies abandoning efforts to improve the distributive profile of tax systems, for
instance by changing the structure of the VAT to focus exemptions on basic goods
and services or by eliminating caps on social-security taxes, but it does mean to
make enforcement a priority. Arguably, the low capacity to enforce tax legislation
remains the key problem for Latin American tax systems, including those in our
four countries. It is not a new problem, but the reforms since 1980 have not been
able to solve it. Poor enforcement of legislation and consequent low tax collection
clearly are detrimental to the poorer sectors and to women, because redistribution
is in practice accomplished mostly on the expenditure side, and solid tax revenue
is needed to finance expenditures on transfers and social services.
The degree to which social expenditures in Latin America are redistributive
depends on the category of expenditures. Expenditures on health and education
tend to benefit lower income groups more than upper income groups in relation
to their incomes, whereas social-security expenditures tend to favour middle and
upper income groups (World Bank 2004, ch.4). The same argument can be made
for women and men; if women on a fixed budget are predominantly responsible
for providing for the health and education of their children (and for their own
healthcare), better public health services and educational facilities will benefit
them. In contrast, women are less likely to be covered by social-security schemes
as they are constructed at present because they are more heavily represented in the
informal sector. Since benefits are generally earnings-related, even those women
who are covered are disadvantaged as lower-income earners. Thus, increased tax
revenues devoted to high-quality and highly accessible education and health
services and to anti-poverty policies such as non-contributory flat-rate pensions,
unemployment and disability schemes, and family allowances paid to mothers,

are the most promising avenue for mitigating the class and gender inequalities so
prevalent in Latin American and Caribbean societies.
Acknowledgement
I would like to thank Inés Valdez for excellent research assistance, and Debbie Budlender,
Shahra Razavi and John D. Stephens for comments on earlier drafts.
Notes
1 One should also note that most countries have introduced user fees for a variety of social
services, most prominently health services. Strictly speaking, these user fees constitute a
kind of tax, but they are not captured in any figures for tax burdens. To the extent that
these user fees are flat rate, they are regressive.
2 In liberalized economies without price controls, it is possible that the benefits of such
exemptions could partly be captured by producers or retailers rather than consumers,
which requires vigilance and potentially corrective regulatory action on the part of the
authorities.
3 In Table 14.1, property taxes are included in the category of direct taxes, along with
individual and corporate income taxes. They are low everywhere, between 4 per cent of
total tax revenue of the central government in Chile (Engel et al. 1998), 2 per cent in
Jamaica (Bahl 1997: 212), and less than 1 per cent in Argentina. In addition, in Argentina
provincial and municipal real-estate taxes amounted to 1 per cent of GDP (Gómez Sabaini
et al. 2002).
4 This section relies heavily on information provided by Carciofi and Centrángolo (1994),
Gaggero and Gómez Sabaini (1999), and Gómez Sabaini, Santieri, and Rossignolo (2002).
5 The Chilean pension system was completely privatized by the Pinochet dictatorship
and turned into a compulsory private savings scheme where issues of redistribution are
eliminated.
6 Data for GDP per capita are from the Penn World Tables, in constant dollars, adjusted for
purchasing power parity.
7 We know from comparative studies of advanced industrial societies that incumbency
of social democratic and Christian democratic parties is a strong predictor of generous
welfare states (Huber and Stephens 2001).
8 One might argue that this factor is less important in countries where a relatively large
proportion of lower-class women never marry even when they have children, which is
the case in Jamaica. However, since most of the direct income taxes are paid by middle- and
higher-income earners to begin with, who are also more likely to marry, this issue remains
relevant.
9 This is important to point out from the point of view of policy lessons; it may be better
to keep high excise taxes on gasoline and subsidize bus transport and transport of basic
foodstuffs, both from a class and gender point of view. Again, we have to consider the
expenditure side along with the tax side.
Bibliography
Alm, James, Roy Bahl and Matthew Murray (1991) ‘Income Tax Evasion’. In Roy Bahl (ed.)
The Jamaican Tax Reform. Cambridge MA: Lincoln Institute of Land Policy.
Bahl, Roy (1989) ‘The Political Economy of the Jamaican Tax Reform’. In Malcolm Gillis (ed.)
Tax Reform in Developing Countries. Durham NC: Duke University Press: 115–76.
320 Evelyne Huber
Bahl, Roy (1991) ‘The Economics and Politics of the Jamaican Tax Reform’. In Roy Bahl (ed.)
The Jamaican Tax Reform. Cambridge MA: Lincoln Institute of Land Policy: 1–62.
Bahl, Roy (1997) ‘The Jamaican Tax Reform: Its Design and Performance.’ In Wayne Thirsk
(ed.) Tax Reform in Developing Countries. Washington DC: World Bank: 167–232.
Bird, Richard M. and Barbara Diane Miller (1989) ‘The Incidence of Indirect Taxes on Low
Income Households in Jamaica’. Economic Development and Cultural Change 37(2) (January):
393–409.
Bradley, David, Evelyne Huber, Stephanie Moller, Francois Nielson and John D. Stephens
(2003) ‘Distribution and Redistribution in Post-Industrial Democracies’. World Politics,
Vol. 55, No. 2, January: 193–228.
Carciofi, Ricardo and Oscar Cetrángolo (1994) ‘Tax Reforms and Equity in Latin America:
A Review of the 1980s and Proposals for the 1990s’. Innocenti Occasional Papers; Economic
Policy Series, No. 39. Florence, Italy: UNICEF International Child Development Centre.
Centro de Implementación de Políticas Públicas para la Equidad y el Crecimiento (CIPPEC).
(2002) ‘Hacia un sistema tributario permanente para Argentina’. Buenos Aires: CIPPEC.
Economic Commission for Latin America and the Caribbean (ECLAC) (1998) The Fiscal
Covenant: Strengths, Weaknesses, Challenges. Santiago, Chile: ECLAC.
Economic Commission for Latin America and the Caribbean (ECLAC) (2002) Social Panorama
of Latin America 2001–2002. Santiago, Chile: ECLAC.
Engel, Eduardo M.R.A., Alexander Galetovic and Claudio E. Raddatz (1998) ‘Taxes and
Income Distribution in Chile: Some Unpleasant Redistributive Arithmetic’. Working Paper
6828. Cambridge MA: National Bureau of Economic Research.
Foxley, Alejandro, Eduardo Aninat and Jose Pablo Arellano (1980) ‘Las Desigualdades
Economicas y la Accion del Estado’. Mexico City: Fondo de Cultura Economica.
Gaggero, Jorge and Juan Carlos Gómez Sabaini (1999) ‘El Sistema Tributario Federal:
Diagnóstico y Propuestas’. Unpublished manuscript, Universidad Nacional de Buenos
Aires, Argentina.
Gaggero, Jorge and Juan Carlos Gómez Sabaini (2002) ‘Argentina: Cuestiones Macrofiscales y
Reforma Tributaria’. Buenos Aires: Fundación OSDE-CIEPP.
Gómez Sabaini, Juan Carlos, Juan José Santieri and Darío Alejandro Rossignolo (2002)
‘La equidad distributiva y el sistema tributario: un análisis para el caso argentino.’ Serie
Gestión Pública. Santiago de Chile: Comisión Económica para América Latina.
Hossain, Shahabuddin M. (2003) ‘Poverty and Social Impact Analysis: A Suggested
Framework’. IMF Working Paper WP/03/195, International Monetary Fund.
Huber, Evelyne and John D. Stephens (2001) Development and Crisis of the Welfare State:
Parties and Policies in Global Markets. Chicago: University of Chicago Press.
Huber, Evelyne, David Bradley, Stephanie Moller, Francois Nielsen and John D. Stephens
(2001) ‘The Welfare State and Gender Equality’. Luxembourg Income Study Working Paper
No. 279, September.
Inter-American Development Bank (IADB) (1996) Economic and Social Progress in Latin
America: 1996 Report. Washington DC: IADB.
International Monetary Fund (IMF) (1988, 1991, 1995, 2003) Government Finance Statistics.
Washington DC: IMF.
International Monetary Fund (IMF) (1991, 1995, 2003) International Financial Statistics.
Washington DC: IMF.
La Nacion, at www.nacion.com, accessed 8 March 2004.
La Nacion, at www.nacion.com, accessed 1 December 2003.
Manley, Michael (1982) Struggle in the Periphery. London: Third World Media.
Moller Stephanie, David Bradley, Evelyne Huber, Francois Nielsen and John D. Stephens
(2003) ‘Determinants of Relative Poverty in Advanced Industrial Societies’. American
Sociological Review, Vol. 68, No.1, February: 22–51.
Morley, Samuel (2001) The Income Distribution Problem in Latin America and the Caribbean.
Santiago, Chile: ECLAC.
Quesada Camacho, Juan Rafael, Danile Masís Iverson, Manuel Barahona Montero, Tobías
Meza Ocampo, Rafael Cuevas Molina and Jorge Rhenán Segura (1999) Costa Rica
Contemporánea: Raíces del Estado de la Nación. San José, Costa Rica: Editorial de la
Universidad de Costa Rica.
Rosenwurcel, Guillermo (2000) ‘Fiscal Reform and Macroeconomic Stabilization in
Argentina’. In Guillermo Perry, John Whalley and Gary McMahon (eds) Fiscal Reform and
Structural Change in Developing Countries, Vol. 1: 26–58. Basingstoke: Palgrave.
Santiere, Juan J., Juan C. Gómez Sabaini and Darío A. Rossignolo (2000) ‘Impacto de los
Impuestos sobre la Distribución del Ingreso en la Argentina en 1997’. Banco Mundial,
Secretaría de Programación Económica y Regional.
Schkolnik, Mariana (1992) ‘The Distributive Impact of Fiscal and Labour Market Policies:
Chile’s 1990–1 Reforms’. Innocenti Occasional Papers; Economic Policy Series, No. 33.
Florence, Italy: UNICEF International Child Development Centre.
Sjoquist, David L. and David Green (1992) ‘Distribution of Tax Burdens’. Jamaica Tax Review
Project, Report No. 3. Atlanta GA: Georgia State University, Policy Research Center, College
of Business Administration.
Social Security Administration (1995) Social Security Programs Throughout the World.
Washington DC: SSA.
Stephens, Evelyne Huber and John D. Stephens (1986) Democratic Socialism in Jamaica:
The Political Movement and Social Transformation in Dependent Capitalism. London:
Macmillan and Princeton: Princeton University Press.
Tanzi, Vito (1987) ‘Quantitative Characteristics of Tax Systems in Developing Countries’. In
Newbery and Stern (eds) The Theory of Taxation for Developing Countries. New York: Oxford
University Press: 205–41.
Tanzi, Vito (2000) ‘Taxation in Latin America in the Last Decade’. Paper prepared for
the Conference on Fiscal and Financial Reforms in Latin America, Stanford University,
9–10 November.
Teichman, Judith (2001) The Politics of Freeing Markets in Latin America: Chile, Argentina, and
Mexico. Chapel Hill NC: University of North Carolina Press.
Thirsk, Wayne (1997) ‘Overview: The Substance and Process of Tax Reform in Eight
Developing Countries’. In Wayne Thirsk (ed.) Tax Reform in Developing Countries.
Washington DC: World Bank: 1–32.
Wasylenko, Michael (1991) ‘Tax Burden Before and After Reform’. In Roy Bahl (ed.)
The Jamaican Tax Reform. Cambridge MA: Lincoln Institute of Land Policy: 751–92.
World Bank (2003) World Development Indicators. Washington DC: World Bank.
World Bank (2004) Inequality in Latin America and the Caribbean: Breaking with History?
15
Expectations versus Realities in
Gender-responsive Budget
Initiatives
Debbie Budlender
Introduction
Analyzing the impact of government expenditure and revenue on women and

girls, as compared to men and boys, is fast becoming a global movement to
build accountability for national policy commitments to women.
UNIFEM 2001 Annual Report: 17
The quote above and similar quotes from other reports give either implicit or explicit
indications of the ambitious expectations that the gender-responsive-budgeting
(GRB) idea evokes. The key question addressed in this chapter is: How does what
GRB initiatives have done in practice compare with the claims and expectations
about what they can achieve? In asking this question, the chapter does not aim to
detract from what has been achieved. Instead, it attempts to bring some realism
into the discussion, planning and assessment of these initiatives. The chapter also
stresses that different initiatives have different objectives and different outcomes
which depend on context, who is involved, and a host of other things. There is
therefore no single ‘correct’ approach.
The chapter does not explain in any detail what GRBs are. There is a range of
other sources which provide such a description and discussion (see, for example,
Reeves and Sever 2003; Budlender et al. 2002; Budlender and Hewitt 2003). The
working definition of a GRB for this chapter is that it involves an analysis of the
government budget in terms of its reach and impact on women and men, girls and
boys. A GRB is thus, in effect, a form of policy analysis from a gender perspective.
GRBs do not focus only on the numbers contained in the budget. They focus as
much – if not more – on the policy and programmes underlying those numbers.
Ideally, they also focus on what happens when the policies and programmes are
implemented. The ‘added value’ of GRBs in terms of policy analysis is that they
recognize that any other government policy or programme will not be effective
unless adequate resources are allocated to implement it.
The chapter is divided into three sections. The first section deals with issues
related to budgets and their relationship to conceptualizations of the economy
322
Expectations versus Realities in Gender-responsive Budget Initiatives 323
and economic and social policy. The second deals with issues related to gender as
a critical variable that structures the economy and society, alongside other axes of
difference such as race, class and age. The third deals with issues related to policy-
and budget-making as a process. The different sets of issues are often related to
each other, and there is thus some overlap between the sections.
Budgets and economic and social policy
Macroeconomic policy
GRB initiatives are often described as being about engagement with economic
policy. This is true because the budget constitutes a major part of the fiscal macro-
economic policy of the country. However, the assertion can be misleading if it
relies on too narrow a notion of the ‘economic’.
First, while the budget is part of economic policy, the way it is distributed affects
the effectiveness of the policies of all parts of government – economic, social and
protective. The budget is, in effect, the monetary reflection of all the policies of a
government. Second, talking about the budget as ‘economic’ policy can detract
from the aim of using gender budget work to integrate social and economic con-
siderations. Third, the understanding of budgets as part of economic policy high-
lights the very limited progress that has been made in addressing other parts of
economic policy. Broad statements about the gender impacts of macroeconomic
policy are common, and relatively easy to make. Translating these into concrete
analysis and proposals is a harder nut to crack.
If GRB initiatives are serious about influencing macroeconomic policy, they will
need to engage with the Ministry of Finance beyond the budget division which is
usually involved in GRB work. They might well discover at this stage that the
divisions responsible for macroeconomic issues have limited power and that it is,
instead, the international financial institutions (IFIs) who call the shots. The
World Bank has shown its support for GRB work to the extent of attempting to
establish initiatives in several countries. The International Monetary Fund has
published a working paper (Sarraf 2003) on the topic. However, both institutions
might balk at allowing GRB work to ‘interfere’ with macroeconomic strategies,
policies and models.
Mainstreaming of gender in social and other policies

GRB work is often touted as a useful tool to support mainstreaming of gender
in government policies. It does, indeed, have enormous potential to be used in
this way. However, as discussed in more detail below, often those involved in
GRBs end up focusing on targeted allocations for women rather than consider-
ing how to make all government policies and allocations gender-sensitive.
Targeted allocations are sometimes necessary as a form of affirmative action or to
cater for special needs. However, true mainstreaming requires changing of the
‘ordinary’ programmes and budgets which account for the bulk of government
activity.
324 Debbie Budlender
Enhancing rights
GRB is sometimes seen as a way of monitoring the Convention on the Elimination
of All Forms of Discrimination Against Women (CEDAW) and socioeconomic
rights more generally. However, the area of rights, like that of macroeconomic policy
beyond budgets, has proved one of the more difficult areas of GRB work. One chal-
lenge is that those who get involved in budget work usually have a different
approach to those who get involved in rights work. In particular, those who work
on budgets tend to have a more technicist approach, while the rights advocates
adopt a more idealist approach. Stated crudely, the technicists tend to argue in
terms of ‘prioritization’ and choices within a given budget ceiling, while the rights
advocates are more inclined to argue for ‘more for everyone’.
A related potential obstacle is that the rights discourse is sometimes not persua-
sive when talking to bureaucrats, especially those from technical ministries such as
Finance. In addition, reference to international instruments will often have little
sway with technocrats, especially where they already feel themselves under con-
siderable other pressures from IFIs and donors who wield power over their policies
and budgets.
Impact
GRBs are political interventions which aim to change the nature of budgets and
the policies and programmes that underlie them so as to change the situation of
(poor) women and men, girls and boys in a particular country. As such, the ques-
tion naturally arises as to their impact – whether they have effected any of the
intended changes. Unfortunately, impact is not easy to measure or to attribute
to particular interventions, whether for GRBs or any other political and social
interventions.
Some initiatives can claim some budget changes. In Mexico, for example, the
Commission on Gender Equality of the Chamber of Deputies succeeded, on the
basis of GRB analysis conducted by civil society, in its arguments for increased
allocations for reproductive health. Additional funding was also allocated for a
programme to address maternal mortality and for programmes for women in agri-
culture and immigrant women. In the United Kingdom, the work of the Women’s
Budget Group convinced the Treasury to redesign the Child Tax Credit so that it
was payable to the main caregiver (likely to be a woman) rather than to the main
breadwinner (likely to be a man). In South Africa, the GRB can claim some credit
for the introduction of the child support grant which is given to the primary
caregivers of young children from poor households.
Some GRBs can point to statements by finance ministers and others in which
they acknowledge the importance of gender. Some can point to budget documents
which now more clearly report on gender issues. All these achievements are impor-
tant, at least symbolically. Overall, however, the more than fifty GRB initiatives
around the world have probably produced relatively few budget changes.
One reason for the limited changes in budget figures is that policy does not
usually change because of a single initiative, but rather because of the combina-
tion of a range of forces. It is thus difficult for a single group to claim credit for a
change in policy. Another reason for the limited claims that GRB initiatives can
make in respect of budget changes is that policies seldom change simply because new
and better ‘facts’ are presented, much as we would all like to believe in evidence-based
policy-making.
A third reason for the limited claims that can be made about changes is that
those involved often do not have the power themselves to change budgets. This
lack of power is clearly the case for civil-society actors. I argue below that it is also
the case for most parliaments. And I argue further that it is often the case for civil
servants.
The limitations on the power of civil servants to change budgets is particularly
acute if the GRB initiatives involve only gender focal points or gender ministries.
Gender ministries should, in theory, have more power to influence others as their
mandate is usually stated as being to ensure mainstreaming through all other min-
istries. However, the weakness of these ministries in most countries is well-known.
Further, the staff involved often do not engage very energetically in GRBs. Often
the officials fear the technical nature of the venture. Many feel more comfortable
doing gender-awareness training within government and in civil society. Many
feel more comfortable focusing on women’s projects which, while outside the real
mandate of most gender ministries nowadays, continue to be funded by donors
who want concrete ‘deliverables’.
Gender as a critical variable in budget and policy work
Gender, women and special interests

The growth and nature of civil-society budget work
The worldwide interest in GRB work during the 1990s and since occurred alongside
a more general interest in budget work within civil society. The interest has been
encouraged by funders. However, the growth in interest in budget work is not
simply donor-driven. The growth also reflects real interest on the part of civil-
society actors. And the environment has been conducive in many respects. First,
the current focus of donors and the international financial institutions (IFIs) on
governance often encompasses the need for participation and engagement with
(at least parts of) civil society. Second, the increase in the use of performance-based
budgeting approaches in many countries has increased the amount of information
available, and made clearer the links between policy and budget. Third, ministries
of finance and financial considerations have become increasingly dominant in all
policy-making.
Special-interest budget work

Many of those involved in ‘general’ budget work see the GRB work as an example
of ‘special interest’ budget work. For many of those involved in GRB work, how-
ever, the ‘special interest’ tag is inaccurate and misleading. The first objection is
that, even if some GRBs focus on women rather than gender, it seems odd to see
women (and girls) as a ‘special interest’ group when they constitute more than
half of the population in most countries of the world. The second objection is that
if the GRB truly focuses on gender, it concerns every individual in that every
individual is part of the gendered make-up of society. Thus while ‘special interest’
budget initiatives tend to argue for ‘more’ for children, people with disabilities, or
whatever group is involved, theoretically GRBs should be advocating for ‘more
equitable’ programmes and allocations – ones which address the different needs
and interests of individual from different (overlapping) social groups.
‘More equitable’ is, of course, not a straightforward criterion to apply. Many
would see a 50–50 allocation between male and female as ‘equal’, and therefore
‘equitable’. A more sophisticated – and more valid – conception sees ‘equitable’ as
requiring recognition of the different situations of women and men, girls and
boys, and other social groups which give rise to their potentially different needs
and interests.
I wrote above that ‘theoretically’ GRBs should be advocating for more equitable
programmes and allocations. In practice, the situation is more complicated.
Special allocations for women and gender

On the one hand, many GRBs tend to focus primarily on allocations for women.
This tendency is found in some of the analytical work on GRBs where the analysts
are looking for categories with neat boundaries. It is found in some advocacy work
that call for budget lines for an area such as education to allocate separate amounts
for girls and boys, rather than asking for a single amount with monitoring of
whom it reaches. The characteristic is also found in many local-level initiatives
which record as a success the establishment of a special (and small) ‘women’s
fund’. Only too often, the women councillors and other participants in these GRBs
then focus all their energy on determining how this fund should be spent, leaving
the men and bureaucrats to get on with spending the rest of the budget.
The danger of focusing only on special-interest allocations was recognized in the
1980s, when Rhonda Sharp was assisting the government of South Australia in
setting up their women’s budget. To avoid focusing all attention on these allocations,
each agency was required to assign every single expenditure into one of three
categories, namely:
● Expenditures targeting women (the special-interest allocations);

● Expenditures promoting equal opportunity for women and men civil servants;
and
● All other expenditures, assessed for their impact on women citizens.
Several later initiatives in other countries used similar categories. Ideally, the
three-way categorization should highlight the necessity for focusing on the third
category given that this accounts for the overwhelming majority of the funds
allocated in the budget as a whole.
In Philippines and Indonesia (women) politicians have, perhaps unwittingly,
encouraged an approach of focusing on ‘special’ allocations for women or gender.
In Philippines in 1995 Senator Shahani proposed and pushed through a law which
stipulated that every national agency should allocate 5 per cent of its budget for
‘gender and development’ (GAD). This allocation soon became known as the
‘GAD budget’. The ruling was later extended to all local government units.
The position of the National Commission on the Role of Filipino Women
(NCRFW), which is the Philippines’ national gender machinery, is that the 5 per cent
GAD budget should be used to promote gender mainstreaming in the budget as a
whole. It asks that every government agency and local government unit draw up a
GAD plan, and use at least 5 per cent of the budget to implement this plan. Women
activists in the Philippines defend the approach strongly. They and the NCRFW
have put significant energy into training of officials and other activities to assist in
the drawing up of GAD plans and budgets. The energy has not always paid off. The
most notorious example of misallocation is the agency that included money for
ballroom dancing lessons for female civil servants in its GAD allocation.
One can understand the move to specify a percentage of the budget as a strategy
to have something which is measurable and thus more easily monitored. One
wonders, too, to what extent the percentage approach was influenced by the quota
debate in respect of representation of women.
Unpaid labour
A key aspect that distinguishes (or should distinguish) gender budget work from
other budget work is the focus on unpaid labour.
Both government and non-government budget work is often dominated by
economists. And in virtually all schools and all countries, economists are trained
to analyse markets and production that goes through the market. Feminist econo-
mists have taken the lead in pointing out the blindness of an approach which
ignores the non-market production which occurs in all economies and societies,
and which is necessary to ensure the ‘production’ and ‘maintenance’ of the people
who do the monetary production.
While unpaid labour is unseen in traditional economics, in my experience it is
very easily and quickly grasped by diverse audiences. Ordinary women tend to
understand the issue immediately, as they have daily experience of what it entails.
Middle-class women – and civil servants in particular – sometimes do not ‘feel’ the
issue in the same way because they have the resources to employ other women to
carry some of the burden of unpaid labour for them. However, they will have
usually experienced the difference in expectations of male and female when children,
as well as when adult in respect of childbearing and childrearing. In addition, most
men will – after initial puzzlement – start nodding their heads in agreement when
the issue is discussed.
The inclusion of unpaid labour in one’s approach to economics is in many
ways revolutionary. However, as one senior economist in a Ministry of Finance
remarked after a short presentation on the topic, it is ‘completely logical’ that
unpaid labour should be included in macroeconomic and other models. However,
despite its importance, and the fact that it is ‘obvious’ once pointed out, the
unpaid-labour aspect is not present in all GRB work. And when it is present, it is
often not strongly developed.
One obstacle is the lack of available data. Relatively few countries in the world
have undertaken the time-use studies which provide estimates which can be con-
verted into monetary equivalents. Detailed data are, however, not necessary for
unpaid labour to be considered in GRB initiatives. If – as is usually the case – everyone
acknowledges that women bear the main burden of most types of unpaid labour,
exact data strengthen the case and make it appear more scientific, but are not
required for appropriate policy-making.
Unpaid labour is completely absent from some GRB initiatives. In others, it is
mentioned as part of the situation analysis, but not necessarily taken further when
it comes to discussing policy and associated budgets. Given the importance of
unpaid labour as a basis for gender inequality, there are surprisingly few concrete
suggestions as to how it should be recognized in policy- and budget-making
beyond broad calls for increased provision of safe water, electricity and childcare.
The unpaid-labour focus probably emerges most strongly in the work of
the United Kingdom’s Women’s Budget Group around tax credits. The UK example
is, however, not immediately useful to developing countries because few have
revenue/benefits systems. Unpaid labour should, in theory, be even more of an
issue in developing countries than in developed because there is less public provision
of relevant services by the state or by the market. The HIV/Aids pandemic and
its attendant need for additional (unpaid) care should increase the importance of
the issue.
Pro-poorness
The World Bank has suggested that GRBs can ‘engage in a process of transformation
to take into account the needs of the poorest and the powerless’ (Blackden and
Bhanu 1999: 64–5). However, GRBs do not all necessarily have a pro-poor agenda.
Some groups get involved in GRB work out of a concern with disadvantage and, in
particular, the disadvantage that female people usually experience when compared
with male people. These groups will usually also be concerned about other aspects
of disadvantage. Other groups have a narrower focus on ‘gender equality’. This
often translates, in practice, into the promotion of women’s interests. These
groups would be as interested in promoting the interests of women entrepreneurs
as promoting the interests of poor women (and men).
There are several examples from gender analysis of revenue which illustrate the
tension between a pro-women and an equity approach. The first example comes
from the South African GRB’s second attempt to analyse taxation from a gender
perspective. During the apartheid years, the South African personal tax system
provided separate schedules for ‘married persons’ (actually married men), married
women and unmarried people. ‘Married persons’ were taxed at a lower rate on the
assumption that they would be supporting others. Married women were taxed at a
higher rate on the assumption that theirs was a second income in a household.
In subsequent years this was changed so that personal tax is now determined on an
individual basis, irrespective of the sex or marital status of the person concerned.
However, Smith (2002) points out that while the change has done away with
formal discrimination, it has probably left – and perhaps introduced – substantive
discrimination. This anomaly arises because while tax is now determined on

an individual basis, income is consumed by a household.
The new system discriminates, in particular, against households with only one
income earner – households in which only too often that income-earner will be a
woman. Smith illustrates the discrimination by comparing the situation of two
households comprising two adults and two children. In the first household, the
husband earns R2,000 per month and the wife earns R1,000 per month. Together
they provide for their two children. In the second household, the only paid worker
earns R3,000 per month, and supports her mother and her two children. Both
these households have the same total income and must support the same number
of adults and children. In 1994/95, before the change which allegedly did away
with discrimination, the first household would have paid R3,435 in tax annually
and the second would have paid R5,055. In 1999/2000, under the new system,
the first household would pay R850 while the second would pay R3,460 – over
four times as much as the first household.
The example has been described at some length because it illustrates one of the
more important contentions made in this chapter. The example suggests that a
simplistic approach which treats all ‘units’ in an identical way irrespective of the
social and other differences, results in inequitable outcomes. Such an approach
might produce formal equality, but it does not result in substantive equality
or equity. The example also, implicitly, raises questions about intra- and inter-
household equality. These need to be considered in GRBs, but are not discussed
further here.
Smith (2002) acknowledges that there is no easy solution to this problem. In
particular, it is not possible to draw on tax-benefit solutions used in developed
countries such as the UK because such a small proportion of the South African
population – and not the most needy – pay tax in the first place. Smith’s example
illustrates the point that simplistic elimination of sex discrimination does not
necessarily bring equity if we are interested in looking beyond a simple male–
female divide. However, calculation of tax on the basis of a household raises a wide
range of difficulties as to how one defines such a household, and what assumptions
one makes about marriage and other social arrangements. In a developed-economy
context, it is exactly these questions that have been the focus of the Women’s
Budget Group in the United Kingdom. The situation is further complicated by the
fact that women tend to have less power and control than men over household
income. The previous South African system exacerbated the likely power imbalances
by taking the extra tax off the women’s income.
TGNP’s experience on the revenue side provides a further example of the
difficulty of addressing equity, even in an initiative that is very serious about
addressing all aspects of inequality. In late 2002, TGNP convinced those in charge
of the annual public expenditure review (PER) process that they should commis-
sion a paper on gender and revenue to supplement the general review of revenue
that they were planning. The two gender researchers went into a lot of detail in
examining different aspects of revenue rather than simply focusing on direct
taxation (Meena and Mhamba 2003). However, many of their recommendations
focused on the interests of middle-class women rather than the poorest women
(and men). They raised, for example, the high taxes placed on imported khangas
and kitenge, baby food and sanitary pads. None of these are goods which are likely
to be of concern to poorer women. Similarly, their call for the increase of the
amount at which people start paying tax would benefit more men than women,
while simultaneously decreasing the amount of money available to government to
provide services to those who earn too little to pay direct tax with the current
cut-offs. In contrast, the UK Women’s Budget Group has consistently argued
against tax cuts or raising tax thresholds even at the very bottom of the income
scale on these grounds.
The politics of policy-making
Many commentators talk about the potential of GRBs in promoting women’s

political participation. Some also suggest that GRBs have a role to play in promoting
women’s leadership in the economic as well as political sphere.
Participation
There is often slippage in discussions of ‘progressive’ budget work in that
GRB work is assumed, by definition, to involve a more participatory approach. The
slippage can be partly explained by the fact that GRB work is often taken up
by organizations and actors with an interest in women’s political participation.
The slippage is, however, unfortunate. First, greater political participation by
women does not, in itself, always result in greater attention to gender issues or
women’s interests, although – overall – it is likely to do so. Second, there is a
chicken-and-egg question – does greater political participation by women result in
GRBs, or does GRB work tend to result in greater political participation by women?
Third, we need to be clear whether we are talking about participation in the
bureaucracy, in parliament, or within civil society. Finally, we need to avoid the
limited understanding of some governments who seem to feel that participation
should be by invitation and that any participation by non-governmental actors is
sufficient, no matter who the actors are. In these cases, participation is often
restricted to academics, think-tanks and/or government-aligned organizations
who are not necessarily those who might raise the issues which should be at the
heart of a GRB approach.
Increased political participation and GRB work

Uganda and South Africa are examples of countries where GRB initiatives emerged
subsequent to a significant increase in women’s political representation in the
national parliament. And in both countries the women parliamentarians played a
significant role in establishing and leading the GRBs.
Uganda’s GRB is housed in the Forum for Women in Democracy (FOWODE), an
NGO established by a group of women parliamentarians, and led by the powerful
and dynamic MP Winnie Byanyima. FOWODE draws on the support of the wider
women’s caucus within parliament. Uganda has a system of special representation
in parliament for women and other ‘marginalized groups’ (the military, workers,
youth and people with disabilities). At the time FOWODE established the GRB in
the late 1990s, the women’s caucus had more power and impact than similar bodies
in many other countries because of the absence of parties. Indeed, by uniting with
representatives of other ‘marginalized’ groups, the women’s caucus was, in effect,
the single largest caucus in the Ugandan parliament. The GRB thus had an immediate
audience and advocate.
In the first years, FOWODE’s GRB activities focused on the national-level budget.
FOWODE later extended activities to the local level as a direct result of its concerns
about participation, as well as out of concern as to what decentralization might
mean in terms of gender equality. The GRB initiative was established around the
time of the Constituent Assembly, which drew up a new constitution for the coun-
try. The women parliamentarians put a lot of effort into the Constituent Assembly
and managed to win a quota for women at the local level. FOWODE followed up
on this victory by providing support and training to women to stand for and win
local elections. They then went a step further by providing training to the success-
ful women to ensure that the victory went ‘beyond numbers’ to bring gender-sen-
sitive policy into local government. FOWODE’s local-level GRB was part of the
training and other activities intended to take the local-level quota a step further.
In South Africa, the GRB initiative was established one year after the first demo-
cratic elections which brought an end to apartheid. Over a quarter (27 per cent) of
the membership of the national parliament in the first post-apartheid parliament
were women – a significant increase from the 3 per cent of the final apartheid
years. Most of the new women representatives had been part of the struggle
against apartheid. Many had participated in the Women’s National Coalition
(WNC), a cross-party and cross-race grouping which, among others, ensured that
the interim constitution of the country included strong clauses on gender equity.
The WNC had also drawn up a Charter on Effective Equality for presentation to
the Constitutional Assembly. This Charter was drawn up through a fairly extensive
process of consultation around the country and represented a wish-list of what
women wanted to see in the ‘new South Africa’. Pregs Govender, previously the
national organizer of the WNC, was one of the new parliamentarians. She recog-
nized that women parliamentarians would need some way to ensure and monitor
that the demands of the Charter were being met, and that the budget could be a
powerful tool to do so. She therefore met with two policy-oriented research NGOs
to discuss how this could be done. This saw the birth of South Africa’s GRB.
In both the Ugandan and South African cases, then, the establishment of the
GRB resulted fairly directly from increased participation of women in parliament.
GRBs have, however, not been established in all countries with high levels of
participation. Some quota systems might, in fact, discourage the establishment of a
GRB. This could happen if the GRB is seen as a too effective means of oversight,
and if the women elected through a quota system feel beholden to government
and fear losing their seats if they are too critical.
The Ugandan and South African GRBs were established at a time when women
representatives were largely supportive of government policy. In both countries,
the executive, on its side, had a strong commitment to gender equality. Thus while
both GRBs styled themselves as exercising oversight over government, they were
largely advocating what the executive itself was committed to doing, but was
perhaps not doing forcefully enough, or with enough resources. In both countries
the leaders of government had no problems with the GRB initiatives, and
sometimes openly supported them. Indeed, Pregs Govender was rewarded by
being made chairperson of a new Parliamentary Committee on the Improvement
of the Status and Quality of Life of Women. In Uganda, FOWODE was invited to
participate in the sector committees which meet with the Ministry of Finance in
the early months of the budget cycle each year.
In later years the leaders of the GRB initiatives in both countries increasingly
criticized government policy. In Uganda, the situation was complicated by the fact
that Winnie Byanyima’s husband stood for the presidency (and lost) against Yoweri
Museveni. In South Africa, tensions developed when Pregs Govender started
criticizing the stance of the government on issues such as large-scale military
expenditure and HIV/Aids policies, and took many of the members of her
committee along with her. Both women leaders were prepared to stand up to gov-
ernment and pay the consequences. Winnie Byanyima suffered ongoing harassment.
For Pregs Govender, the consequences were leaving parliament. Many other
women parliamentarians might not be prepared to take this strong a stand.
GRBs in situations where women’s representation is lower

In Philippines, the non-governmental GRB at the local level had direct links with
action around increasing political participation for women although there were
not, as in South Africa and Uganda, significant numbers of elected women leaders
in the cities involved. The non-governmental GRB was initiated by the Asia
Foundation (TAF), a USAID-funded organization with offices in more than ten
Asian countries. In the Philippines, TAF had previously provided funding to the
Development through Active Women Networking (DAWN) Foundation, a local-level
NGO in Bacolod City which supported women to stand for local elections, and
which was led by Celia Flor, a city councillor. For the initial years of the GRB, TAF
provided funding and technical support to DAWN, as well as to a similar councillor-
led women’s organization in Angeles City, and a mixed people’s organization
working in Surallah in the Mindanao region.
DAWN’s work was the most successful and sustained of the three pilot cities.
Towards the end of the research period, Celia Flor’s party won control of the city
council. Andrea Si, one of the other leaders of DAWN, became the city adminis-
trator and was thus in a position, as they put it, to ‘implement what they had
learnt’ in the research and advocacy of the previous period. DAWN also dissemi-
nated the idea of GRBs to other cities in Negros Occidental where it was active in
encouraging and supporting women to run for elections.
While it is dangerous to generalize from a few examples, the cases of Uganda,
South Africa and the Philippines suggest that high levels of political representation
of women might provide a conducive environment for GRBs, but by no means
guarantee it. Rather, it is the presence of strong women leaders who recognize the
potential of GRB work, and are prepared to put in the effort needed to sustain it,
that are crucial. Further, all three countries are ones where there has been a recent
transition away from authoritarian government, and where women’s groups have
been very active and mobilized within this transition. One must therefore be wary
of generalizing from these examples.
Nevertheless, what the cases suggest is that GRB work can be an important
supplement to work around increasing women’s political participation. In particular,
GRB initiatives can help to ensure that women’s participation means more than a
change in the profile of political decision-makers. It can do so by increasing the
representatives’ understanding of what gender-sensitive policy means and helping
them focus on a key policy without which any other policy will not be effective,
namely the budget.
Engagement of (women) parliamentarians with GRBs

Despite the seeming potential, in practice there are relatively few countries where
(women) politicians have shown much interest in gender budget work. One
reason for this might be that in many countries parliamentarians have very little
effective power over the budget. In all countries the budget is a law which must
be passed by parliament. However, in many countries there are serious constraints
on the extent to which parliaments can change the budget proposals tabled by the
bureaucracy.
In countries like South Africa, parliament must either accept or reject the
budget as a whole – it cannot make any amendments (Krafchik and Wehner 1999).
Outright rejection is very unlikely, as it effectively constitutes a vote of no confi-
dence in the government. In some other countries, parliament may have greater
powers to increase or decrease particular votes, or particular sources of revenue.
However, in virtually all countries parliament cannot alter the overall level of
spending, or the balance between expenditure and revenue.
A further constraint on the power of many parliaments is that the budget is
often tabled only a short time before the end of the financial year. As a result,
much of the budget debate occurs after spending of the new financial year
has already begun. The debates provide an opportunity for parliamentarians,
including women, to raise issues. But their comments have no real impact on the
allocations.
Parliamentarians’ conception of their role in respect of budgets can also be
an obstacle. Many parliamentarians see their primary role as one of oversight.
In effect, they see their role as checking that the budget is spent as allocated, rather
than ensuring that the allocations themselves, and the policies that underlie
them, are correct. Performance budgeting, with its emphasis on the need for
budgets to reflect policy, should help to change this perception. However, in
most countries parliamentarians have not yet seized the potential of performance
budgeting.
Parliamentarians have also often not taken advantage of other opportunities
offered by budget reforms. In particular, many countries have introduced medium-
term expenditure frameworks (MTEFs) in recent years. These involve the provision
of definite figures for the following year, with indicative figures for the following
two or four years. This approach to budgeting should provide greater opportuni-
ties for parliamentarians and other players to have an impact on the actual budget
figures, as there is a longer time in which to advocate for or against government
plans. In practice, the opportunities do not seem to have been utilized, either by
GRB players or others.
Finally, parliaments in developing countries usually have little, if any, research
capacity. At first glance this provides an obstacle to GRBs. However, it can also
provide opportunities as parliamentarians who lack their own research resources
might be more prepared to receive input from outside and, in particular, from
NGOs and other gender advocates. Several non-governmental organizations
(NGOs) have taken advantage of this gap. Perhaps the most surprising of these is
Zimbabwe, where, despite the political turmoil, the Zimbabwe Women’s Resource
Centre Network (ZWRCN) has on several occasions been invited by cross-party
parliamentary committees to provide input based on its GRB research into key
ministries.
Participation in GRBs at the local level

GRBs have been initiated at national, provincial/state and local level in different
countries. The local-level GRB work has often had participation as one of the main
foci. The Philippines and Uganda examples cited above constitute two cases of a
GRB located at the local level. In both these cases, the GRB was seen as a way of
strengthening women’s political participation. Many other countries have also
done GRB work at the local level, often with an explicit focus on participation.
However, while many of the initiatives describe themselves as focusing on the
‘grassroots’, some who do so target their activities at local leaders rather than
‘ordinary’ citizens.
Some purposefully make little distinction between ‘grassroots’ and ‘leaders’ in
the local context. In Uganda, for example, FOWODE explicitly argued that the
local level provided greater opportunities for ‘ordinary’ women to engage in GRBs
and in political leadership. First, local-level politics often involves the basic, day-
to-day issues which concern women and their families, rather than ‘bigger’ issues
about which they feel less competent to speak. Second, because of domestic and
other ties, women often find it easier to engage at the local than at national level.
Decentralization has dangers as well as potential benefits if we are interested in
enhancing equality in gender and other terms. In particular, if devolution of
responsibility is not accompanied by the necessary funds, decentralization tends
to decrease the possibility of national redistribution as each area becomes res-
ponsible for providing for itself. In terms of gender, actors are often more conservative
and power relations cruder at the local than at the national level. Nevertheless, if
we are interested in GRB work out of a concern with all forms of inequality, local-
level activity has the advantage that local councillors are more likely than those at
national level to be aware of the poorest and most disadvantaged women.
Many national parliaments have regulations in terms of educational level, or
money required to stand, that result in national parliamentarians being an elite.
While some might be concerned about people who are poorer and more
disadvantaged than themselves, they are less likely than leaders from less advantaged
backgrounds to have an understanding of the real needs of those involved. Similar
regulations about who can stand for election are less common at the local level.
A focus on the local level is also important when – as is often the case – key
services such as basic health and education services are decentralized. In these
cases the national level’s responsibility is usually restricted primarily to policy-
making. It is at the local level that the impact of the implementation – or
non-implementation – of the policies will be felt. It is thus at the local level that
citizens – and particularly women – will often be most open to mobilization
around the needs of their children and other family members.
Women bureaucrats and GRBs

The above discussion in this section has focused on political representatives.
In several other cases, the presence of women in decision-making positions has led
to the establishment of GRBs within the bureaucracy. South Africa is again an
example. The post-apartheid government moved quickly to ensure that the
bureaucracy was more representative in terms of both race and gender. Within a
few years the public service had almost equal numbers of women and men. Men
still dominated the decision-making positions. However, there were significant
numbers of women among the Cabinet ministers and deputy ministers who
‘instruct’ the bureaucrats as well as among high-level officials within the ministries
(called departments in South Africa). The presence and support of these women
facilitated South Africa’s becoming the African pilot for the Commonwealth
Secretariat’s GRB within bureaucracies. While the initiative within the national
treasury lasted only for the two years of Commonwealth support, highly-placed
(mainly female) officials in some of the provinces and in some line ministries have
ensured that gender budgeting continues to live within the bureaucracy.
Demystification of budgets
The section on participation above focuses on participation at the official decision-
making level. Many donors are primarily interested in initiatives which focus on
official decision-makers, whether in the bureaucracy or parliament. However,
some GRB initiatives (for example, South Africa, Philippines, Indonesia, India)
have also been interested in extending participation among civil society, and in
demystifying budgets.
Who does the work

The Philippines initiative is interesting in illustrating the opportunities when
activists are prepared to take the plunge and confront the technicalities. As noted
above, the TAF GRB initiative involved three case studies. While all adopted a
common framework and covered similar topics, the way in which they conducted
the research differed. In Surallah, the work was led by a left-wing political organi-
zation that put much of its energy into bringing together citizens and local
decision-makers to discuss gender in planning and budgets. In Angeles City,
the lead organization contracted researchers from the university to do the bulk of
the research. In Bacolod City, in contrast, after initial hesitation on the basis that they
were not researchers, DAWN itself conducted the research.
DAWN’s approach was the most effective both in terms of producing a gender-
analysis of the city’s budget and in terms of providing a conducive environment
for implementation of recommendations. What was particularly interesting was
the extent to which the GRB research exposed Celia Flor, an energetic councillor
of several years’ standing, to new information and understanding of how the
council worked both in gender terms and more generally. Flor and her co-researchers
were also visibly empowered and proud of what they had achieved in doing the
research themselves. And the fact that they had done so facilitated the subsequent
dissemination of the approach to other cities in Negros Occidental.
In South Africa, the non-governmental Women’s Budget Initiative (WBI)
was conceived from the outset as a multi-stakeholder exercise. The core group was
made up of parliamentarians and two policy-research NGOs. From the start, this
core group involved people outside the core as central players. In the first three
years, the initiative produced research chapters on each of the departments
(equivalent to ministries in other countries) of the national government, as well as
on several other topics such as public-service employment, taxation, the budget
process, and intergovernmental fiscal relations (the allocation of money raised at
one level of government to other levels for spending). Members of the core group
coordinated the research and provided back-up to the researchers, but did very
little of the research and writing themselves. Instead, they drew on people from
other NGOs as well as academics to do the research in their areas of expertise.
In addition, they set up a reference group consisting of at least one government or
non-government person knowledgeable about each sector being researched. These
were generally people who were too busy to do the research themselves, but who
were considered important sources of knowledge or access to information, as well as
key people to whom the WBI wanted to spread its new approach to understanding
budgets.
There were several reasons why the inclusive approach was adopted by the WBI.
First, the initiators recognized that analysis of budgets from a gender perspective
requires a multidisciplinary approach and knowledge of several different fields.
They felt that it would be easier for people who already understood the gender
issues in a particular area to acquire the necessary skills to analyse budgets than for
people with budget analysis skills to acquire adequate knowledge of gender issues
in each sector. Second, and related, the initiators hoped to spread the relatively
uncommon skill of budget analysis so that NGO workers in different areas would
have this extra ‘tool’ to strengthen their advocacy and engagement with govern-
ment. Third, the approach had the by-product of generating relatively widespread
knowledge of, interest in, and buy-in to the project. The interest and buy-in was
increased by a large public launch of the first year’s research which was opened
by the then Deputy Minister of Finance. Fourth, the two core NGOs did not see
themselves primarily as advocacy organizations. Instead, they saw themselves as
assisting others to obtain budget-related skills and information which they could
use in advocacy. The parliamentary partners were the first target for such skills and
information. The researchers and their organizations were a second target which
could increase spin-offs.
The inclusive approach was successful in spreading awareness and buy-in. It also
resulted in more spin-off work than in most other countries. Many of those who
acted as researchers commented on how empowered they felt in having been able
to do the budget analysis. A few continued to do similar work outside of the WBI.
Several other NGOs which had not been involved themselves took up the idea of
GRBs. Some of these came back to the core WBI and asked them to do the research.
A few took the work forward themselves, perhaps with back-up from the core WBI.
Most successful among these was an NGO which had previously engaged in
extensive research and advocacy on gender violence, and saw the potential of the
GRB approach as an additional tool. Still others both inside and outside government
spoke about the importance of looking at budgets, although this often did not go
beyond talk. Several universities included GRB in their gender and development
studies courses.
In South Africa the composition of the core group, which included researchers
in the NGOs and advocates in the parliamentarians, appeared to provide an ideal
mix in that the researchers could provide the parliamentarians with information
to strengthen their advocacy. In practice, however, the parliamentarians did not
use the research to the extent hoped for. The Tanzanian initiative provides further
lessons about the dangers in separating the research and advocacy roles. TGNP is
the central GRB organization in Tanzania. When TGNP initiated their gender
budget initiative (GBI) they did so in collaboration with the broader Feminist
Activism grouping of which they form part. Nevertheless, they played the main
role from the very start.
TGNP’s approach mimicked that of South Africa in many respects in the first
year. In particular, one of the first activities was to commission research into the
budgets of key ministries, as well as into the budget process. For each of the papers
TGNP commissioned a team of at least two people. Typically, one was an academic,
while the second was someone from inside the relevant ministry. Instead of
coordinating and editing the work themselves, TGNP commissioned a feminist
political scientist at the university to do so.
TGNP’s choices reflected its own perception of itself – as an activist organization
rather than a research or academic one. The choice of academics reflected the
smaller number of other civil-society organizations and individuals who might
be persuaded to do this work. One weakness of the approach was that TGNP had
less knowledge of the detail of the research. This is an important weakness in
advocacy with ‘opponents’ such as government technocrats as the latter use tech-
nicalities as one of their key weapons. In subsequent research and related work,
TGNP has increasingly tried to participate more in the research process.
Economists and non-economists

Where GRBs contract researchers, there is often a tendency to think that the
researchers should be economists. In truth, budget analysis requires non-economics
as much as it requires economics skills – it requires a multidisciplinary approach.

It is also necessary that the economists who become involved are able to see the
pitfalls of some of the standard economic paradigms because traditional economics
training often militates against a researcher being able to see some of the important
gender issues – in particular, unpaid labour.
The tendency to think that economists must be the main – or only – players
occurs in GRBs both inside and outside government. In inside-government initiatives,
this is often understood as meaning that the financial or budget officers should be
the main players. The problem here is that financial officers often have limited
knowledge of policies, let alone of gender issues in the society and possible gendered
aspects of policies. At the least, GRB initiatives need to involve planners as well as
budget officers. Exactly who to involve depends on the way planning and budgeting
happen in a particular country.
In outside-government initiatives, non-economists might require more support
initially in understanding terms and concepts and in overcoming a fear of (large)
numbers. However, an interdisciplinary approach has many benefits both in terms
of the breadth of the resultant work, and in terms of the power it gives to the non-
economists to use budget arguments in their advocacy.
Involvement of less technically oriented and more organizationally based people
will also usually result in different, and potentially useful, outputs. For example,
the more technically oriented organizations tend to produce detailed documents
which require a reader with relatively high levels of numeracy, literacy and knowl-
edge of technical terms. If they move beyond written dissemination, they are likely
to do so through a lecture-style format. In contrast, more activist organizations will
usually attempt to disseminate knowledge in much simpler terms, even if they
thereby lose some nuance. They will also utilize more participatory and ‘democratic’
methods in workshops and other activities. These are more likely to enable
participants to use what they learn in their organizational activities.
Demystification and ‘popularization’ force those involved in GRB initiatives to
face the tension between proving to the economic planners in government that
they understand the intricacies and complications of budgets, and not confusing
the people who need to do the advocacy around the issues with unnecessary tech-
nicalities. It involves an understanding that different audiences require different
products. But the products aimed at the advocates also need to provide them with
sufficient understanding that they can confront the technicians.
One challenge in achieving greater participation is getting people to confront
their fear of numbers. Another challenge arises from the fact that budgets are, at
heart, about prioritization. Some civil-society groups may resist work that requires
them to prioritize and advocate less than the ideal on the grounds that it is
reformist.
Conclusion
The rapid growth in the number of GRB initiatives over the last few years testifies
to the potential which many see in these ventures. Yet the discussion above
suggests that often the initiatives do not fulfil all the roles which are attributed to
them. The chapter has tried to illustrate that one of the reasons for this apparent
failure is the great diversity in GRB initiatives. As a corollary, it implies that
non-fulfilment of each and every goal should not be seen as a failure. Whether an
initiative can play a particular role depends on the actors, their goals, their under-
standing, and the activities they undertake, as well as the political and social
context in a particular country. Those promoting GRB work would be more helpful
if they promoted it as a tool that can be used at many different stages in the policy-
making process, by many different players in many different ways to advance
many different causes in addition to the broad cause of gender equality. A single
initiative cannot, and should not, aim to be all things to all people. GRB work is
not a panacea for anything.
Bibliography
Blackden, C.M. and C. Bhanu (1999) ‘Gender, Growth and Poverty Reduction’. Technical
Paper No. 428, Washington DC: World Bank.
Budlender, D. and G. Hewitt (eds) (2002) Gender Budgets Make More Cents: Country Studies and
Good Practice. London: Commonwealth Secretariat.
Budlender, D. and G. Hewitt (2003) Engendering Budgets: A Practitioners’ Guide to Understanding
and Implementing Gender-responsive Budgets. London: Commonwealth Secretariat.
Budlender, D., D. Elson, G. Hewitt and T. Mukhopadhyay (2002) Gender Budgets Make Cents:
Understanding Gender Responsive Budgets. London: Commonwealth Secretariat.
Fontana M. (2002) ‘Modelling the Effects of Trade on Women: The Case of Zambia.’ Working
Paper No. 155, March 2002. Sussex: Institute of Development Studies.
Fontana, M. and A. Wood (2000) ‘Modelling the Effects of Trade on Women, at Work and at
Home’. World Development 28(7): 1173–90.
Krafchik, W. and J. Wehner (1999) The Role of Parliament in the Budget Process. Cape Town:
Institute for Democracy in South Africa.
Meena, R. and R. Mhamba (2003) A Gender Study on the Tanzania’s Government Revenue and
Taxation. Dar es Salaam: Tanzania Gender Networking Project on behalf of Gender Macro
Policy Working Group.
Pillay, K., R. Manjoo and E. Paulus (2002) Rights, Roles and Resources: An analysis of Women’s
Housing Rights – Implications of the Grootboom Case. Cape Town: Women’s Budget Initiative.
Reeves, H. and C. Sever (2003) Gender and Budgets: Key Texts, Case Studies, Tools, Guides and
Organisations. Sussex: BRIDGE Support Resources Collection, Institute of Development
Studies.
Sarraf, F. (2003) ‘Gender-Responsive Government Budgeting’. Working Paper No. 03/83,
Washington DC: International Monetary Fund.
Sharp, R. (2003) Budgeting for Equity: Gender Budget Initiatives within a Framework of
Performance Oriented Budgeting. New York: United Nations Development Fund for Women.
Smith T. (2002) Women and tax in South Africa. Cape Town: Women’s Budget Initiative.
United Nations Development Fund for Women (UNIFEM) (2001) Annual Report. New York:
UNIFEM.
Index
abortion, 77, 78, 79–80, 83n20 Basic Income Grant (BIG), 116, 120–6
Adato, M., 57, 63n28 impacts, 123–4
Adjustment with a Human Face (Cornia et al. Bates, R., 28
1987), 47 Baude, Annika, 168n5
adult literacy, 2, 4, 121 Beijing, China, 275
adult-worker family model, 173–87, 219 Beijing Platform for Action (PFA), 32, 53
in EU policy-making, 174, 175–8, 180 benefits see social benefits
Africa, 19, 28, 30, 33, 245, 247 Berer, M., 249
health sector reform, 237, 238–9 biological determinism, 159
see also sub-Saharan Africa; and individual see also maternalism
states Bojer, H., 186
AIDS pandemic, 26, 116, 121, 218, 229, 249 Bolivia, 2, 3, 4, 5, 6, 11, 45, 194, 196, 207
care implications, 218, 224, 230, 250–1, 328 1980s economic crisis, 200
orphans, in Africa, 251 kin networks, 209–10
and social protection, 230 labour informalization, 200, 201, 202–4,
in sub-Saharan Africa, 249 212
in women, 121, 249 poverty, 201
Albertyn, C., 127n6 structural adjustment, 200, 201
alimony benefit, 70, 73, 76, 80, 83n12 tin mining, 200
All China Women’s Federation (ACWF), 270 unemployment, 200
Alma Ata declaration, 1977, 249 urban, 201, 209–10, 211
Angeles City, Philippines, 332, 335–6 Bradshaw, S., 59
An-Naim, 101 Brazil, 8, 44, 45, 51
anti-poverty programmes, 58–62, 198, 213 1988 constitution, 51
and gender, 58–60 Ceará province, 53
see also poverty alleviation; and individual Bremer, Fredrika, 155
programmes Britain see United Kingdom
Antonopoulos, R., 206, 211 Brock, K., 48–9
Arat, Y., 106 brokering, 153, 162–4
Argentina, 2, 3, 4, 5, 6, 8, 43, 44, 62n8, 224, boundary-making, 153, 155, 156–8
312 budgeting, 29–30, 316, 323, 333–4
neoliberal reforms, 312 analysis, 337–8
taxation (reform), 304, 305–6, 307, 308, demystification, 335–6, 338
310, 311, 313, 315, 317 and economists, 337–8
Asia, 8, 28 gender-responsive see gender-responsive
1997 economic crisis, 14, 132, 144 budgeting
see also East Asia; and individual Asian medium-term expenditure frameworks
states (MTEFs), 333–4
Asia Foundation (TAF), 332, 335 performance-based, 325, 333
asset pawning, 211, 212 policy changes, 324
Australia, 11, 326 research, 334
Austria, 11 Byanyima, Winnie, 330, 332
Bacolod City, Philippines, 336 Canada, 11

Bahl, R., 310 capabilities approach, 152, 174–5, 180–4
Balbo, L., 179 and gender equality, 181–3
Bangkok, Thailand, 206 and individual freedom, 183–4
Bangladesh, 247 and interdependency, 183–5
340
Index 341
Carciofi, R., 304 Chile, 2, 3, 4, 5, 6, 43, 44, 62n4

care(work), 1, 2, 7, 22, 26, 35–6, democratization, 306, 312
37n7, 186 Pinochet regime, 312, 319n5
affective nature of, 178–9, 184, 186 taxation, 304–5, 306–7, 308, 309, 310–11,
as a choice, 174–5, 177, 179, 180–1, 312, 313, 315, 317
182, 183 China, People’s Republic of, 3, 5, 17, 275
commodification, 173, 177, 178–9, 180, All China Women’s Federation, 269–70
239: see also defamilialization; constitution, 258
privatization gender wage gap, 264–5, 266
community, 224 health care insurance, 259–76: funding,
devaluation, 181–2 263; Gongfei scheme, 259, 260, 261,
familial responsibility, 21, 23, 26, 35, 69, 262, 263, 266, 269, 271, 274; for
174, 224 government officers, 271; Hezuo
formal, 223 Yiliao, 259; Laobao scheme, 259–60,
full-time, 223–4 262, 263, 266, 269, 274; for migrant
informal, 179 workers, 270–1, 274
by men, 173, 184, 186: see also under men health sector reform, 258–75
shared, 184–5 Liangjiang health care model, 262
unpaid, by women 2, 7, 10, 11, 22, 26, life expectancy, 265
31, 35, 61, 70, 77, 78, 116, 165, 173, market economy, 260
218, 224, 250–2, 301 maternity insurance, 270
Caribbean countries, 23, 301, 302 medical service, 260
see also individual states migrants, 259, 269, 270–1, 273, 274–5
Carnegie Commission of Enquiry, 1929, retirement age, 265
111 retired workers (pensioners), 273
casual work, 202: see also labour rural areas, 259, 274
informalization; migrant workers savings, 263
Catholicism, 46, 77–8 social security, 258, 274–5
Central America, 8 state ownership, 260: restructuring, 260
see also individual states unemployment insurance, 263
Central Asia, 9 urban areas, 259–60, 26, 274: see also
see also individual states Shanghai region of China
Central Europe, 9, 68–84, 102 Zhenjiang province, 264
Communist parties, 69 see also Shanghai region of China
see also individual states civil servants, 97, 98, 325, 327
Centrángolo, O., 304 collectivist strategies, 27
Chatterjee, M., 225, 229 Colombia, 62n4
child allowances see family benefits colonial legacy, 8
childcare, 31, 35, 71, 83n21, 120, 126, 144, commercialization, 2, 12, 15–16, 18
146, 147n11, 177 of social services see under social services
formal, 179, 222: see also nurseries commodification, 16, 173, 177, 178–9, 180,
gendered, 161, 162, 165, 173, 221 239
subsidized, 161, 179, 180 Commonwealth, 335
see also care(work); parental leave community
child mortality, 97 importance of, 27
children, 20, 22 responsibilities of, 55
AIDS orphans, 251 Confucian principles, 22, 132
as care givers, 116, 250 contraception see family planning
childcare see childcare Convention on Economic, Social and
in poverty, 119 Cultural Rights (ICESCR), 90, 93
provision for, 44, 54–5, 58, 117, 119: see and Islamic family law, 93
also family benefits Convention on the Elimination of All
sick benefits, 70, 71, 83n11 Forms of Discrimination Against
see also under women Women (CEDAW), 92, 324
342 Index
cooperatives, 17, 225 demographic changes, 6, 131, 133–5, 141,

core values in public policy, 1 227
Cornia, G., 47 ageing, 134, 141
Costa Rica, 43, 62n5 divorce rates, 31, 134, 135, 136
demographic changes, 6, 227 fertility rates, 134, 141
informalization, 9, 12, 227 in Japan see under Japan
maternal mortality rates, 5 in Korea see under Korea
National Liberation Party (PLN), 311 Denmark, 11
pensions, 227 development
school enrolment data, 3, 4 models, 27–8, 44
social security ‘umbrella’ scheme, 225, policy, 48–9, 50
226–8, 229–30, 308 and social welfare, 114–15, 133
taxation, 23, 304, 305, 307–08, 310, 311, see also under social policy
313, 315, 316, 317 Development through Active Networking
urbanization, 227 (DAWN) Foundation, 332, 336
wages, 227 division of labour, 197
women MPs, 5 gendered, 251
crèches see nurseries household, 207
credit see also care(work); labour markets; work-
access to, 211 family balance
kin loans, 209–10 divorce, 90, 96, 98, 100, 101, 136, 141, 166,
micro-, 20, 48, 210 219
see also debt rates, 31, 134, 135, 136
Czechoslovakia, 70–1, 74, 83n7 see also alimony benefit
see also Czech Republic domestic workers, 223, 224, 271
Czech Republic, 6, 68, 69, 72–80, 83n6, dual-earner model, 7, 158–62
83n7
Department for Equality of Men and earnings variability, 205, 212
Women, 76, 80 index, 205
economic reform, 26, 73–4 East Asia, 21, 22, 29, 31, 87, 301
education enrolment data, 3, 4 welfare states, 131, 132–3: comparison of,
family benefits, 22, 72–5: compared with 132–3
Poland, 75–82 see also individual states
maternal mortality rates, 5 Eastern Europe, 9, 25
women MPs, 5 Communist parties, 69, 71
see also Czechoslovakia see also Central Europe; and individual
Cuba, 43, 62n4 states
Ebadi, Shirin, 103
Daly, M., 178–9 ECLA(C), 44, 62, 198, 304, 312
data problems, 11–12 economic activity, 6, 196
debt, 27, 72, 200, 208 see also work
crisis, 48, 52, 303 Ecuador, 3, 4, 5, 6, 11, 25, 45, 194, 196, 208
interest rates, 210 credit, 209–10
pressures, 212 dollarization, 199
servicing, 208–9, 212 industrialization, 199
see also credit informal economy, 199, 202–3, 212
decentralization, 49, 52–4, 146, 247, 334 labour code, 199
decision-making in households, 207–8, 315, social services provision, 199–200, 211
316 education, 2, 17–18
decommodification, 24 access to, 8, 54, 221, 248
defamilialization, 24, 35, 134, 145, 152, 177 for all, 296
see also family benefits of, 18
democracy, 36 curriculum design, 293–4
democratization, 2, 27, 28–9, 30, 31, 33, 51 drop-outs, 281–2
Index 343
education – continued Belgian Presidency, 2001, 176

elementary/primary, 2, 3, 18, 281, 282, governance, 176
283–4, 286, 292, 296 Portuguese Presidency, 2000, 176
enrolment, 2–6, 17–18, 123, 280–1, 282, see also individual states
283, 286–7, 292–3 export processing zones (EPZs), 221
expenditure on, 17, 280, 283–7, 292–3, 318
as filter for social policies, 1–2 fair-trade initiatives, 231
and gender, 2–6, 278–96: ‘feminine family
subjects’, 280, 294–5; performance, benefits see family benefits
281–2; single-sex schools, 18, 279, definition of, 119–20
280, 290–1, 294, 296 gender (power) relations, 26, 77, 87,
government, 287, 289 127n8, 131, 146, 184, 334
higher, 287 importance of, 24–5, 26, 34, 76–7, 116–17
management, 285, 287, 288 income pooling, 208, 243
physical infrastructure, 285 law, 88, 89–94, 98: see also Islamic law
private, 18, 280, 288–9, 291–2, 293 life, 179
research, 279 models, 90: adult-worker, 173–87; familial
resources, 283 welfare, 30–1; see also male-
secondary, 2, 3, 17–18, 278–96: worker/female-carer model
participation rates, 280, 281 planning see family planning
scholarships, 293 policies, 102, 152
and social policy, 278, 279–80 poverty, 76–7, 117: see also poverty
teachers, 285, 290: of mathematics, networks, 77, 252: see also social networks
294–5; women, 287, 289 structures, 119–20, 135, 173
in Uttar Pradesh, India, 278–96 see also care(work); children;
Egypt, 12, 13, 93 defamilialization
elderly care, 22, 31, 35, 142, 144, 145, 161 family benefits, 69–73, 74, 75–80, 112, 117,
see also care(work) 137
elderly people, 221, 250 see also maternity benefits
as care givers, 250–1 family planning, 99, 106n17, 106n21, 136,
care of see elderly care 248
in Shanghai, China, 268–9 Federation of South African Women
and health, 250, 265 (FSAW), 111
as pensioners, 273 feminism, 12, 25, 36, 53, 59, 88, 157, 153
poverty, 251 advocacy, 20, 53
social protection, 252 analysis/criticism, 7, 10, 23–4, 58, 80, 87,
El Salvador, 11, 12, 13 91, 182, 240
Elson, D., 22 of difference, 157
employment see work institutionalized, 167
Eriksson, Nancy, 160 interventions, 33–4, 47, 53, 153:
Escobar, L., 57, 58, 63n20 femocrats, 158
Esping-Anderson, G., 24, 87, 109 journals, 156
Ethiopia, 242 Marxist, 88–9
eugenics, 44 networks, 167
Europe, 93, 193 research, 25, 53, 124
see also Central Europe; Eastern Europe; rights work, 152, 324
European Union; Western Europe; in South Africa see under South Africa
and individual European countries in Sweden see under Sweden
European Commission, 175, 177 theories, 69
social policy agenda, 175–7, 180 Filgueira, C. and F., 62n7
see also European Union Finch, J., 179
European Union (EU), 21, 22, 72, 80, 164, Finland, 11
adult-worker-model family, 174, 175–8, Floor, W., 94
180 Flor, Celia, 332, 336
344 Index
Floro, M. S., 206, 208, 211 policies, 78–9, 83n6, 130

Folbre, N., 178 see also capabilities approach; see also
Fox, Vicente, 54 care(work); inequalities
Fraser-Moleketi, G. J., 120 gender lens, 240
Fredriksson, Gunnar, 159, 161 gender-responsive budgeting (GRB), 29–30,
322–39
G7, 14 analysis, 326
Ge, S., 259 definition, 322
Social Security Economics, 258 initiatives, 323, 324–5
Gelb, S. 122, 123 at local level, 334–5
gender special-interest allocations, 325–7
analysis, 12, 58, 69, 83n2, 237: see also Germany, 11, 222–3
under feminism Ghana, 13, 242, 246
bias, in employment see gender Gini coefficient (index), 199, 201, 306,
discrimination 307, 308
and budgeting see gender-responsive globalization, 14, 168, 195, 201, 222
budgeting (GRB) Goldblatt, B., 124, 127n4
development index (GDI) see under Gómez Sabaini, J. C., 304, 308
United Nations González de la Rocha, M., 57, 58, 63n20
and education see under education Govender, Pregs, 331, 332
equality see gender equality government budgets, 123
hierarchies, 237, 245–6, 252 see also budgeting; gender-responsive
identity, 197: see also capabilities budgeting (GRB)
approach Grindle, M., 29
justice, 184–5: see also gender equality gross domestic products (GDP), 2, 72,
lens, 240, 252 319n6
mainstreaming, 323–4 and taxation, 301, 303, 304–5, 306, 308,
patriarchal contract, 88, 93: see also 310, 317
patriarchal societies Guatemala, 45
relations, 32, 36, 57, 89, 124, 132, 196–7, Gutmann, M., 197
252: see also under family
roles and stereotypes, 7, 10, 20, 27, 59, Haggard, S., 19
61, 70, 77, 79, 89, 91, 132, 159: Hakim, C., 174
assumptions on, 2, 43, 45, 113; Hansson, Albin, 154, 156
changes, 132, 196, 197; in education, Harman, E., 124
293–5; skills/work, 6 health care, 2
sensitivity, 56–7, 113, 333: see also access to, 4, 8: see also health insurance
gender-responsive budgeting (GRB) in Africa, 245–6
see also care(work); decision-making in in China, 263–4
households; division of labour; for the elderly, 142
women expenditure on, 318
gender discrimination, 98, 165, 168n8, financing, 241–2, 252
238, 329 hierarchy, 237
gender equality, 10, 76, 79, 80, 82, 125–7, insurance schemes see health insurance
152, 326, 328 management, 247
of capabilities, 183–5: see also capabilities migration of staff, 246
approach mother and child (MCH), 237–8, 241
and care, 175, 183–4: see also care(work) preventative, 245, 249
constitutional, 109–10 primary, 249
models, 161, 181 private, 245, 246, 249: expansion, 245
indices, 130, 199 reform, 16, 22, 237: see also health sector
institutionalization, 138 reform (HSR)
pay, 6, 7, 12, 13, 72, 164, 176, 197, 266: reproductive, 241, 248–50, 253, 275
see also inequalities systems, 1, 4, 240–1
Index 345
health care – continued social policy, 232

workers, 244–5, 246, 247, 252: women as, welfare system, 230
244 widows, 221
see also social spending see also Uttar Pradesh
health insurance, 7, 16–17, 242–3 individualism, 177, 183–4
in China, 259–76 Indonesia, 326
community-based (MHI), 242–4 inequalities, 2, 25, 33, 58, 127n6, 152
in Tanzania, 242, 243, 244 access to social benefits, 17, 25, 70
in Thailand, 230, 230–1 education enrolment, 2: see also
see also health sector reform (HSR); social education
insurance gender, 127n7, 197: see also gender
health sector reform (HSR), 16, 17, 237, equality
238–9, 243, 238–53 income, 6, 7, 12, 13, 72, 164, 176, 197,
and care giving, 250–2 265, 266, 268: see also wage levels
in China, 258–76 type of work, 10–11, 197
gender implications, 250–2, 264–6 urban–rural, 8, 201
models, 247 see also women
recommendations, 252–3 inflation
and reproductive health, 248–9, 253, 275 in Argentina, 307
in sub-Saharan Africa, 239–53 in Bolivia, 200, 213n5
in the UK, 247 in Ecuador, 199
see also health insurance informal economic sector, 9, 16, 36n5, 54,
Hernes, H., 151–2 193–4, 217, 231
Hertha, 155 contribution of, 217
Hesselgren, Kerstin, 157, 158 dynamism, 194
hierarchical societies, 9 employment in, 11–12, 36n5, 46, 193,
HIV/AIDS see AIDS pandemic 202, 314
Holland see Netherlands and gender, 195, 217: see also gender
Holliday, I., 132 growth of, 9, 217–19
home-based work, 195, 198–9, 202–3, home-workers see home-based work
205–6, 222 street vendors, 222
combined with care, 206 welfare arrangements, 230, 275
full-time care, 223–4, 266 working conditions, 195
gendered, 217: see also gender women in, 250
lack of social protection, 213 see also domestic workers; labour
homosexuals, 117, 118, 167 informalization; migrant workers
Hoodfar, H., 101 insurance schemes, 16–17
housing, 8 see also health insurance; social insurance
human capital, 44, 278, 295 International Conference on Population
human development indicators, 12, 45, 130 and Development (ICPD), Cairo 1994,
see also social indicators 248–9
international financial institutions (IFIs),
ICESCR see Convention on Economic, 311, 323, 325
Social and Cultural Rights see also individual institutions
immigrants, 163, 166 International Labour Organization (ILO),
see also migrant workers 11, 83n6, 83n9, 94, 158, 194, 211, 218,
immunization, 102 226, 245
income tax see under taxation International Monetary Fund (IMF), 144,
income volatility, 204–5, 219–21 304, 309, 323
India, 2, 3, 4, 5, 6, 17, 225 Iran, 2, 3, 4, 5, 6, 26, 27, 32, 34, 87, 88–106
Gujarat earthquake, 2001, 225 1960s/1970s, 96–7, 99
informal economy, 230 and CEDAW, 92, 103
NGOs, 247: Self Employed Women’s Centre for Women’s Participation,
Association (SEWA) see SEWA 103, 105
346 Index
Iran – continued taxation, 304, 305, 308, 309, 310, 315,

Civil Code: Article 1117, 98; Article 1176, 316: reform, 309–10, 311, 317
99–100 Japan, 2, 3, 4, 5, 6, 8, 31, 33, 130, 131
Civil Pension Fund (1923), 94 Basic Law for Gender Equal Society
civil service, 97, 98 (1999), 137–8
constitution, 97 civil society groups and non-profit
development indicators, 95, 102 organizations, 142–3, 146
family law, 87–93, 95, 97, 103, 104–5: Council of Gender Equality, 137, 138
Family Protection Act (1967 and democratization models, 29
1975), 96, 99; ojrat-ol-mesl, 100–1, demographic changes, 133–6, 141, 142
106n19; reform of, 104 informalization of work, 136, 147
family planning, 99, 106n17, 106n21 political parties, 141–2: Democratic Party
family-friendly policies, 101–2 of Japan (DPJ), 142; Japan New Party,
feminists, 103 143; Komeito, 142; Liberal
health care, 102–3 Democratic Party (LDP), 140, 141;
Islamic era, 97–104: Khomeini regime, 99 Sakigake, 143; Social Democratic
Islamic revolution, 88, 91, 97: effect on Party of Japan (SDPJ), 142
women, 97 regime shift, 1993, 141–3
labour force, 96, 104: women in, 96, 97–8 social policy reforms, 132, 133, 136–8,
labour legislation, 102 141, 142, 143: Angel Plan, 136, 137,
literacy, 95, 97 142; Gold Plan, 136, 137, 147n9;
mahr (mehrieh), 99, 100, 105n12 Long-term Care Insurance (LTCI)
maternalism, 89, 90–2, 99–105 scheme, 137, 138, 142; Non-Profit
neoliberalism, 92 Organization law, 143
NGOs, 98–9 welfare provision, 131, 132–3, 148n14
oil industry, 94, 95–7: reliance on, 97 women-friendly policies, 136–7
pensions, 96 women politicians, 138, 143, 145
postwar, 95–6 Women’s Committee for the
publishing, 98 Improvement of the Ageing Society
Qajar period, 94 (WCIAS), 142, 143
Rafsanjani government, 101 Jbabdi, Latifa, 106
Reza Shah, 94–5
wage levels, 95 Kaar, Mehrangiz, 100, 103, 106n23
war with Iraq, 97 Karshenas, M., 95
and Women in Development, 101 Kaufman, R. R., 19
see also Women’s Organization of Iran Kenya, 242
Ireland, 36n3 Kerber, L., 90
Islamic law, 87, 90, 92 Kessler-Harris, A., 89
and clothing, 98 Key, Ellen, 157
enforcement of, 88, 92, 98: see also Kim Dae-Jung, 144
Islamic societies/states Kim Yong Sam, 143
family, 87, 88, 89–94, 98, 103, 104 kin networks see social networks
inheritance, 93, 98 Kittay, E., 183
marriage contract, 90, 93–4 Koivusalo, M., 15
schools, 105n11 Koven, S., 91, 101, 104
Islamic societies/states, 26, 32, 37n8, 87 Korea, 2, 3, 4, 5, 6, 8, 9, 31, 130
see also Iran; Middle East and North childcare, 140, 144, 145, 147n11
Africa (MENA) region civil society groups, 144, 146
Italy, 11 democratization, 132, 143
demographic changes, 133–6, 144
Jamaica, 2, 3, 4, 5, 6, 23, 319n8 economic crisis, 144, 145, 147n5
Jamaica Labour Party ( JLP), 311 family planning, 136
People’s National Party (PNP), 311 gender equality, 140, 147: Ministry of
structural adjustment, 309 Gender and Family, 140
Index 347
Korea – continued life stages, 219, 220

informalization of work, 147 loans see debt
Kim Dae-Jung administration, 144 Loewe, M., 93
Kim Yong Sam administration, 143–4 lone parents, 44, 70, 134, 135, 136, 151,
military rule, 143–4 166, 173, 187, 219, 266, 316–17
NGOs, 140 Lund, Francie, 110, 111, 114, 116, 123, 124,
political parties, 144 125, 127n2
social insurance schemes, 138–9 Lund Committee see under South Africa
social policy reforms, 136–7, 138–40:
implications for women, 145–7; Mackintosh, M., 15
Long-term Care Insurance (LTCI), macroeconomic stabilization see structural
145; adjustment
welfare provision, 132–3, 138–40: reform, male-breadwinner model see male-
143–5 worker/female-carer model
woman-friendly policies, 136–7 male-worker/female-carer model, 7, 9, 12,
women politicians, 145–6 34, 45, 89, 90, 102, 119, 141, 146, 160,
Korpi, W., 19 173, 186, 219
marginalization of groups, 153
labour force see workforce marketization, 198
labour legislation, 9, 89, 102 marriage
labour informalization, 9, 109, 136, 147, contracts, 93–4
193–5, 199–200, 211–12 dower, 94, 99
degrees of, 193, 202–3, 212 Islamic see under Islamic law
home-based work see home-based work and pensions, 162
piece-work, 11, 202, 206, 222 polygamous, 98
subcontracting, 202, 206, 212 and right to work, 155–6
see also informal economic sector; and temporary, 98, 106n15
individual states Marshall, T. H., 87, 105n4
labour markets, 2, 69 Maslow’s hierarchy of needs theory, 268
gender and, 2, 9–13, 69, 72, 131, 152, Mason, J., 179
173: see also women maternal mortality rates, 2, 4, 5, 97, 102,
incentives to enter, 152 121, 248
inflexibility, 9 maternalism, 9, 32, 45–6, 57, 58, 77, 78, 87,
informalization see labour 90, 91, 101, 105n10
informalization and religion, 91, 99, 104, 106n23: see also
risks, 6 Catholicism; Islamic societies/states
structure of, 2 maternity benefits, 70, 71, 83n10, 83n14,
Latin America, 8, 28, 29, 32, 43–53, 54, 197 112, 221, 269
CEPALISTA developmentalism, 44, 45 see also maternity leave
collectivization, 31 maternity leave, 6, 7, 70, 71, 73, 79, 83n9,
debt crisis, 303 83n19, 102, 112, 140
informalization, 9, 54, 194, means-testing, 56, 72, 74, 76, 77, 78, 110,
New Social Policy (NSP), 50–3, 54, 61 112, 118, 119, 227, 228
poverty, 198, 301 see also targeted social benefits
social policy provision, 8, 43, 44–5, 49, men
198 caring role, 161, 165, 173, 252
taxation reform, 23, 43–4, 301–19 as nurses, 245
women’s roles, 197, 207 parental leave, 10, 35, 79, 162
work and employment, 197–9 working hours, 165
see also individual states see also gender; male-worker/female carer
lesbians see homosexuals model
liberalization, 1, 7 Menem, Carlos, 312
Liebenberg, S., 118 Mensah, K., 246
life expectancy, 19, 45, 265 Messier, J., 208, 209, 213n1
348 Index
Mexico, 2, 3, 4, 5, 6, 8, 53 New Economic Model, 48, 51

1917 constitution, 46 see also neoliberalism
Commission on Gender Equality, 324 New Poverty Agenda see under World Bank
corporatism, 46–7 Nicaragua, 45
civil codes, 46 Nigeria, 241
Fox administration, 54, 56 non-governmental organizations (NGOs),
informal sector, 53–4 92, 116, 118, 142, 247, 332, 334, 336–7
one-party rule, 54 and women, 92, 98–9, 140, 330–1, 336–7
Oportunidades programme see see also individual NGOs
PROGRESA/Oportunidades Nordic countries, 21, 35,
Pártido Accíon Nactional (PAN), 54, welfare states, 87, 133, 158, 159
63n17 see also individual states
Pártido Revolucionario Institutional (PRI), North, the (global), 1, 21, 27
54 compared with the South, 9, 36
social and political rights, 46, 47, 54 nurses, 245–6
Mexico City, 197 importance of, 246
Michel, S., 91, 101, 104 private practice, 246, 253n2
micro-credit, 20, 48, 210 recruitment by rich countries, 246
Middle East and North Africa (MENA) training, 246
region, 26, 87–8, 90, 91, 103, 105n5 nurseries, 83n19, 96, 102, 106n20, 137, 140
use of Sharia, 88–92 Nussbaum, Martha, 174–5, 180, 182, 184, 186
see also Iran
migrant workers, 8, 117, 222–3 O’Connor, J., 176
in China see under China OECD countries, 10, 11
female, 275 see also individual states
gender inequalities, 271 oil-rich states, 89, 91
remittances by, 124–5 see also Middle East and North Africa
self-employed, 271 (MENA) region; rentier economies
social security, 222–3, 259, 269, 270–1, 275 Oliff, M., 249
Millennium Development Goals (MDG), 15, Open Door International, 156
49, 296 Orloff, A. S., 24, 89, 90, 127n8
Mkandawire, T., 20, 133 outsourcing, 206, 222
Moberg, Eva, 160 see also subcontracting
Moghadam, V. M., 26
Molyneux, M., 23–4, 25 Palme, J., 19
Morocco, 106n23 Palme, Olof, 161
Mozambique, 242, 245 parental leave, 7, 10, 31, 71, 79, 83n19,
motherism see maternalism 137, 146, 161, 162, 176
Museveni, Yoweri, 32, 332 ‘daddy leave’, 10, 35, 79, 162, 186, 221
Muslim family law see under Islamic law for women see maternity leave
Myrdal, Alva, 157, 158 see also family
Myrdal, Gunnar, 157 Patel, L., 114
Pateman, C., 93
Naidoo, Ravi, 116 paternalism, 46
Nairobi, Kenya, 241 patriarchal societies, 26–7, 33, 89–91
National Commission on the Role of ideology, 248
Filipino Women (NCRFW), 327 see also male-worker/female-carer model
National Iranian Oil Company, 96 pensions, 18–19, 22, 102, 125, 139, 319
Nelson, J. A., 178, 183 crisis, 221
neoliberalism, 36, 52, 193, 198, 213 earnings-related, 159
economic model, 1, 48, 68 and female dependency, 7, 46
policies, 217–18, 311, 312 and life expectancy, 19
Netherlands, 11 pooling, 124
Netshitenzhe, Joel, 122 privatization, 18–19
Index 349
pensions – continued poverty alleviation, 1, 15, 19, 21, 28, 48, 50,
racialization, 113 52, 54–5, 117, 123, 198, 218, 311, 317
reforms, 18–19, 227 Attacking Poverty (World Development
widow’s, 161–2 Report 2000/1), 50
see also social benefits; and individual states see also anti-poverty programmes
Personal Status Code in North Africa and poverty line, 45, 72, 119, 123, 201, 205, 213n4
the Levant, 92 Povey, E. R., 98
see also Islamic law privatization, 2, 15, 72
Peru, 45 of education, 288–9, 291–3
Peter, F., 184 of health care, 245
Philippines, 97, 232, 326–7, 332, 335 of pensions, 18–19
Development through Active Women of social services, 2, 48, 113–14, 119, 198
Networking (DAWN) Foundation, PROGRESA/Oportunidades programme,
332, 336 Mexico, 20, 25, 53–8, 61–4
piece-work, 11, 202, 206, 222 accountability, 56
Pichetpongsa, A., 206 concept of co-responsibility, 55, 56
Poland, 2, 3, 4, 5, 6, 22, 26, 30, 34, 68–9, evaluations of, 57–8
70–1, 73–80, 83n6 gender impacts, 56–7
economic reform post-socialism, 71–2 success of, 55
family benefits, 70–2, 73: compared with targeted benefits, 54–6
Czech Republic, 75–82 public expenditure, 29, 322, 329
Plenipotentiary for the Equal Status of see also social spending
Women and Men, 80 public sector
Solidarity movement, 71, 78 Africa, 245–6
Polanyi, K., 1 as care provider, 249: see also care(work)
political liberalization, 30 in China, 260
see also democratization employment, 19, 29, 92, 165, 200, 245:
political participation, 2, 5, 6, 31, 32–4, 36, decline, 194, 246; low morale, 246;
80, 90 preferred by women, 165–6, 315
by women, 2, 5, 31, 32–4, 36, 80, 99, 110, funding, 29, 246
145–6, 152–3, 155, 167, 168n12, 244, gendered hierarchy, 245–6
332–7: as MPs, 110, 130, 143, 145–6, institutions, 2, 29, 252–3
244, 326, 330, 331 public works, 115, 120, 121, 133
political parties, 32–4, 36, 311
see also under individual states Qavam, Ahmad, 95
population growth, 44 Quirós Viquez, A., 59
postcolonial states, 28 Quito, Ecuador, 210
post-socialist countries, 68–82, 83n1, 160
see also Central Europe; Eastern Europe racism, 111–12
poverty, 33, 45, 198, 199 apartheid regime see under South Africa
alleviation see poverty alleviation Ranson, M. K., 225, 230
children in, 119 recession, 1980s, 47
coping strategies, 208 religion
definitions, 213n4 Islam see Islam
discourse on, 1, 15, 26–7 Roman Catholicism see Catholicism
as enduring phenomenon, 48 and the state, 92
extent of, 62n15 ultra-conservative, 249
family, 76–7: see also under family rentier economies, 89–90
and gender, 33, 121 reproductive health see under healthcare
increase, 10 reproductive rights, 78, 79, 248, 249
of single parents, 166 see also abortion
urban, 208–9 rights, social and political, 32, 44, 46–7, 51,
variable see earnings variability 87, 100, 110, 117
see also inequalities legislation, 43, 46
350 Index
Robeyns, I., 182, 183 Sharp, Rhonda, 326

Roh Tae Woo, 143 Si, Andrea, 332
Rothschild, D., 28 Singapore, 266
Rubery, J., 176–7 single people, 141, 148n13
rural–urban migration, 201 as household heads, 173, 318, 331
as parents see lone parents
Santa Cruz, Bolivia, 207 Skocpol, T., 91, 105n9
Saudi Arabia, 90 Skweyiya, Zola, 123
Seaga, Edward, 313 Smith, T., 328, 329
Sector-Wide Approaches (SWAps), 247 social benefits, 8
self-employment, 193, 202, 209, 217, 269 family see family benefits
and taxation, 310, 313 maternity see maternity benefits
see also informal economic sector means-testing see means-testing
self-help, 50 and moral judgements, 119–20, 124
Sen, Amartya, 174–5, 180, 182–3, 184 racially discriminated, 112, 117: see also
co-operative–conflict bargaining, 184 racism
Development as Freedom (1999), 181 targeted see targeted social benefits
Seong, K.-R., 144 work-related, 2, 6–7, 8, 9, 159
SEWA, India, 17, 198–9, 225–6 see also social spending
Academy, 226 social capital, 49, 60
Bank, 225 social contract, 7
credit, 225 ‘social hygiene’ movements, 44, 50
grant from GTZ, 226 social indicators, 2, 3, 11–12, 198
health care, 225, 230 Gender Development Index see under
influence of, 232 United Nations
literacy training, 226 see also under individual states
social insurance (IIS), 225–6, 229, social institutions, 1–2, 31, 152, 240
230, 243 see also family
as trade union, 225, 229 social insurance, 7, 12, 16, 34, 73, 113, 225–6
sexual harassment, 185–6 see also social spending; welfare state; and
sexual rights see reproductive rights individual states
sexually transmitted diseases (STDs) Social Investment Funds, 51
clinics 241–2 social networks, 24–5, 35, 179, 209–11
see also AIDS pandemic and financial support, 243–4
sex workers, 223 and loans, 209, 210
Shah Mohammad Reza Pahlavi of Iran, 95 and migrant workers, 222
Shanghai region of China, 259–76 and social provision, 210–11
elderly people, 268–9 social policy, 15, 323
‘green card’ residence permits, 263, as electoral issue, 33, 36, 78, 141–2, 306
276n2 in Iran, 87–106: see also Iran
health insurance, 262–76: contributions, in Japan and Korea, 130–48: see also
267; exclusions, 266, 270, 275, 276; Korea; Japan
gender equality, 271–2, 276; making, 2, 7, 48, 68–9, 78, 89, 114, 117,
individual accounts, 266–7, 268; 146, 174, 276, 279, 295: women in,
maternity, 270; reforms, 262–4; 152–3; see also brokering
shared responsibility, 267–8; neoliberal, 48, 88
supplementary, 271–2, 276 as political struggle, 30–2, 44, 152, 159
life expectancy, 265 and poverty relief, 48
migrant population, 261, 262–3, 271, reform, 33, 44, 48, 72, 78
274–5, 276 role in development, 14, 15, 21, 28, 133
retired workers (pensioners), 263 targeting see targeted social benefits
Trade Union, 262, 269 universal redistributive model, 15, 19–21,
see also China 69, 78, 133
Sharia see Islamic law see also social spending
Index 351
social protection see social security Community Based Public Works

social provisioning see social spending Programme, 115, 120, 121
social rights see rights, social and political corruption, 118–19
social risk management (SRM), 15, 50 Democratic Party, 127n5
social-sector restructuring, 2, 12, 44, democratization, 109, 110–11, 331: and
47–8, 49 citizenship, 115
and gender, 14–23 development indicators, 110, 121
see also health sector reform Directorate of Social Planning, 112–13
social security, 2, 218, 219, 224, 225, disability assistance, 229
231, 258 family structures, 228
innovations, 225–31 feminism, 114, 125
and taxation, 306, 308 gender-responsive budgeting (GRB), 330,
see also health care; pensions; social 331
benefits; social insurance; social health care staff, 246
services; social welfare legislation, 114
social services, 15, 22, 34 Lund Committee, 22, 117–18, 119, 120
access to, 12 migrant remittances, 124–5
commercialization of, 2, 12, 15–16, 18, National Party, 113
319n1: see also under privatization NGOs, 118, 336–7
delivery by NGOs, 92 pensions, 110, 111, 113, 120, 225,
see also education; health care 228–9, 230
social spending, 1, 8, 14, 28, 52, 75, 88 racial groups, 228
on education see under education rural households, 124–5, 228
inequalities in, 8 Taylor Committee, 122
and macroeconomics, 1, 14, 115, 198, 323 trade unions, 121: Congress of South
oil-financed, 88 African Trade Unions, 121–2;
reduced, 217–18 PATAMABA, 232; Self Employed
targeted see targeted social benefits Women’s Union, 232
and taxation, 306–7, 318–19 taxation, 331
see also social benefits traditional leadership, 118
social welfare, 19–20, 258 unemployment, 121, 224
familialization of, 15, 23, 68–9 urbanization, 111
privatization of, 113–14, 198 welfare provision, 126–7, 224, 244: Basic
see also social services; social spending; Income Grant (BIG), 120–6; child
welfare state; welfare-to-work support grant, 119–20, 123;
policies; and individual states Department of Welfare, 117, 118,
socialist economies, 19, 152 119; devolution, 113; institutions,
see also state socialism 110, 111; racial bias in, 112–13,
South Africa, 3, 4, 5, 6, 13, 21, 31, 33, 117–18; 1996 White Paper, 114,
109–27, 228–9 115–17
Afrikaner nationalist movement, 111 Women’s Budget Initiative (WBI), 328,
African National Congress, 109, 111, 115, 336–7
122: women MPs, 120 Women’s Charter of Demands, 112
AIDS, 229, 251 Women’s Charter for Effective Equality,
Alliance for Children’s Entitlement to 125, 331
Social Security (ACESS), 119 Women’s National Coalition (WNC), 331
apartheid regime, 110–13, 120, 228 women’s organizations, 109, 111,
budget arrangements, 123, 244, 333 125, 328
bureaucracy, 335 women parliamentarians, 331, 335
childcare grants, 324 South Korea see Korea
civil rights, 111, 112, 117, 125 South, the (global), 1
commitment to gender equality, 109–10 compared with the North, 9, 36
Community Agency for Social Spain, 265
Enquiry, 119 stabilization see structural adjustment
352 Index
state, role of the, 1, 14–15, 28–9, 48, 51, 52 Tanzania, 2, 3, 4, 5, 6, 239, 246, 251
state institutions, 28–30 feminist organizations, 337
state socialism, 7, 30, 33–4, 79, 81, 83n1 health care, 239, 241, 245: AIDS, 249, 251
Standing, H., 240, 241, 249 poverty, 239, 251: reduction, 244
Stephens, J., 159 public expenditure review, 244
structural adjustment, 9, 14, 48, 305, 309 public-sector workers, 246: low morale,
impact of, 14, 28, 47, 53, 198, 201 246
subcontracting, 202, 206 social health insurance, 242, 243, 244:
Subrahmanya, R. K. A., 230 community-based, 16, 242, 243
sub-Saharan Africa (SSA), 10–11, 224, 248 women politicians, 244
AIDS, 224, 248, 251 Tanzania Gender Networking Programme
democratization, 29 (TGNP), 329, 337
health sector reform, 16, 248 Tanzi, V., 304, 305
see also individual states targeted social benefits, 14, 15, 19–21, 50,
subsidies, 47, 112, 125, 127n2, 319n9 54, 200, 244, 249
Sweden, 2, 3, 4, 5, 6, 7, 10, 34 effectiveness, 19
Commission on Family and Marriage, 162 as a means of redistribution, 47
feminism, 152–3, 154, 155, 156–8, for women, 20, 53, 325
167 taxation, 23, 54, 138, 160, 199, 301, 329
folkhemmet, 154–5 avoidance/evasion, 302–3, 307, 310
gender equality, 152, 161, 164, 168n8 and development, 310–11
Group 222, 159–60 direct, 301–2, 304, 307, 308, 313, 315,
immigrants, 163, 166 318, 319n3
marginalization, 153 domestic, 305
Morgonbris journal, 154, 156 enforcement, 310–11, 318
parent-worker-citizen model, 161–2 and gender, 312–17, 329
political history, 153–64: Cossack on incomes, 307–8, 309, 313
Election, 154; Slatsjöbaden indirect, 304, 305, 307, 308, 309, 310,
agreement, 158 315, 316, 317–18: see also value-
political participation by women, 155, added tax (VAT)
167, 168n12 legislation, 315, 318
political parties, 160, 167–8: Centre Party, for married couples, 313
162; Conservative Party, 156; progressive, 317–18
Farmers’ Party, 156; Liberal Party, rates, 309, 315
156, 157, 160, 162; Social Democratic reform, 301, 305–19
Party, 153–5, 156, 160, 161, 163 regressive, 317
social policies, 152–3: Women’s Life and social security, 304, 308, 310, 313–15,
Work, 161 318–19
trade unions, 156, 163–4 structures, 302–5
unemployment, 155–6, 166 tax credits, 324, 329
wage structures, 151 and total revenue, 302, 303, 305, 306, 307
welfare state, 151–68: expansion, 153, on trade, 305, 309, 315
159–60; retrenchment, 168 under-, 301–2
as women-friendly state, 151–3, 163, 164 Taylor, Viviene, 122
women’s organizations, 154, 156, 162, teenage pregnancy, 248
167: Fogelstad Group, 155; Fredrika Teichman, J., 311
Bremer Society, 168n1; National Tendler, J., 30, 53
Association for Housewives, 156, Thailand, 3, 4, 5, 6, 194, 205–6, 211
168n1; Professional Women’s ‘30 baht’ health insurance scheme, 230,
Organization, 157; Redstockings, 167; 231–2
Social Democratic Women’s Homenet Thailand, 231
Organization, 168n1 trade unions, 9, 33, 46, 95, 121
women’s right to work, 155–7 health workers, 247
worker state, 156 industrial action, 199
Index 353
trade unions – continued voluntary sector, 26, 45

and informal workers, 9, 17 see also care(work)
links with political parties, 311
providing welfare, 269 wage levels, 95, 199, 264, 265–6, 272
and women, 46, 156, 163 Washington consensus, 217
see also SEWA, India welfare models, 36, 175
Tsikata, D., 33 see also individual models
Turkey, 106n18 welfare state(s), 7, 21, 22, 36n2, 131
in developing countries, 7, 22, 87, 279:
Uganda, 32, 242, 244, 249, 251, 330 see also individual states
Forum for Women in Democracy development, 7, 8–9, 36n22, 43–4,
(FOWODE), 330–1, 332, 334 90, 109
new constitution, 331 exclusions and inequalities, 7, 8, 20:
unemployment, 6, 17, 72, 113, 200 see also inequalities
benefits, 83, 263 institutionalized, 159
in the Depression era, 156 Islamic arrangements, 94
gendered, 74, 98, 126, 218, 273, 313 literature, 109
racialized, 113 models, 24, 87–8
rates, 72, 73–4, 98, 199, 200, 274 restructuring, 140, 168
UNICEF, 47 in Sweden, 151–68
United Kingdom (UK), 11, 36n3, 247, 324 woman-friendly, 89, 151–68
Women’s Budget Group, 324, 328, 329, 330 welfare-to-work policies, 175
United Nations Westerberg, Bengt, 162
Economic Commission for Latin America Western Europe, 7
and the Caribbean see ECLA(C) White, G., 28–9
Gender Empowerment Measure women, 26
(GEM), 130 access to benefits, 6–7, 253
Gender-related Development Index access to labour markets, 126: see also
(GDI), 110, 121, 130, 147n1 labour markets
United States (USA), 11, 89, 91, 105n9, 193 access to training, 57, 60, 221, 246
see also USAID breastfeeding, 221, 227
Urban Poor Home-based Worker Survey bureaucrats, 335
(UPHWS), 203, 205, 207, 208, 209, 211, with children, 10, 20–1, 33, 54–5, 91,
213n1, 213n6 102, 116, 152, 160, 316–17: see also
urbanization, 44, 111 maternalism; work-family balance
Uruguay, 8, 43, 62n8, 308 citizenship, 87, 110, 112, 114, 124,
USAID, 309, 332 127n8: economic, 89, 92
Uttar Pradesh, India, 18, 278–96 cooperatives, 17
education, 279, 280–96: single-sex domestic work, 251: see also domestic
schools, 18, 279, 280, 290–1, 294 workers
fertility decline, 279 dual role, 68, 77, 159, 173, 250: see also
gender politics, 279, 280–1, 292 home-based work; work-family
political structure, 290 balance
public expenditure, 283–7, 289–90 economic dependence, 106n18, 124,
Secondary Education Commission 207–8: see also male-worker/female-
(Acharya Narenda Dev Committee), carer model; patriarchal societies
296n8 economic vulnerability, 60, 61, 198
Uttaranchal, 288 educated, 197, 295
employment patterns, 8, 69, 72, 176–7,
value-added tax (VAT), 303, 305, 306, 307, 203, 274, 313
308, 315, 317 empowerment, 49–50, 57: index, 130
van der Berg, S., 122 as focus of GRB initiatives, 325–6:
Vietnam, 97 see also gender-responsive
Vimo SEWA see SEWA, India budgeting (GRB)
354 Index
women – continued Women’s National Coalition, South Africa,

freedom of choice, 184–6: see also 109–10, 331
capabilities approach; care(work) Women’s Organization of Iran (WOI), 96
health, 250, 252; see also healthcare women’s organizations, 31, 33, 36, 47,
as household heads, 329: see also under 52, 79–80, 83n22, 142, 146, 153,
single people 162, 332
and HIV/AIDS, 121: see also AIDS see also individual organizations
pandemic women’s rights see feminism
and marriage, 46–7, 221, 316, 319n8: see Wong, J., 143
also marriage work
MPs, 6, 110, 130, 143, 145–6, 244, 326, and benefits, 2, 6–7, 16, 17, 34, 54,
330, 331, 333 56, 70, 82, 175, 230, 264, 269–70,
in paid work, 60, 97–8, 121, 160, 165: 318
access to, 6–7, 98; balance with care see care(work)
caring see work-family balance; casual flexible practices, 175
workers, 121, 202; health workers, force see workforce
245, 247; increased participation, 2, gendered, 2, 6, 313: see also gender
9–11, 34, 141, 160, 165, 197; home-based see home-based work
informal sector, 250; part-time see paid, 2, 6, 21: combined with unpaid,
under work; rights, 157 206; and poverty, 21; see also wage
pregnant, 102, 221, 238, 275 levels
political participation see under political part-time, 10, 36n4, 147n7, 173, 193,
participation 217, 273
reproductive role, 89: see also and social inclusion, 175, 176
reproductive rights unpaid, 2, 21, 104, 327–28: see also
social indicators see social indicators care(work)
social and political rights see rights, social as welfare, 182: see also welfare-to-work
and political policies
in social-sector employment, 9, 245 women’s see under women
reproductive health, 241, 248–50, 253, see also labour markets; self-employment
275: see also AIDS pandemic, family work-family balance, 76, 79, 82, 109, 141,
planning; health care; sexually 173, 174, 177, 195, 232
transmitted diseases see also women
roles and stereotypes see under gender workforce, 7
in traditional societies, 26–7, 32, 34, feminization, 9, 10
58–9, 77: see also Islamic flexible, 199
societies/states gendered, 242, 244–5, 313
and unpaid care(work) see under restructuring, 17
care(work) rural, 17
urban areas, 8 women in see under women
violence against, 63n28, 124, 126: see also labour; work
legislation, 138, 167 working conditions, 195
widowed, 221 hazardous, 221–2, 245
see also inequalities; social benefits; for health workers, 245
women’s organizations; work-family improvement, 252
balance; women-friendly policies World Bank, 14, 29, 36n1, 48, 50, 227,
Women and Development approaches, 53 286, 309
see also Women in Development (WID) advocating tax reform, 302
Women in Development (WID), 101–2 Implementation Completion Reports,
women-friendly policies, 136–7, 296n3
151–68 and gender-responsive budgeting, 328
Women’s Committee for the Improvement of and neoliberalism, 311
the Ageing Society (WCIAS), Japan, 142 New Poverty Agenda, 43, 49, 53
Women’s Environment and Development support for MHI, 243
Organization (WEDO), 32 World Development Reports, 50, 237–8
Index 355
World Conference on Women, Beijing Zambia, 8, 242

1995, 270 Zanan, Iranian women’s magazine,
World Health Organization, 100, 103
World Health Reports, 238 Zhenjiang City, Jiangsu Province, China,
World War II, 89 262, 264, 273
Wóycicka, I., 76 Zimbabwe, 241, 242, 246, 251, 334
Zimbabwe Women’s Resource Centre
young people, 63n30, 74, 141, 248 Network (ZWRCN), 334, 335

Gender and Social Policy in A Global Context: Shahra Razavi and Shireen Hassim

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Gender and Social Policy in A Global Context: Shahra Razavi and Shireen Hassim

Uploaded by

Copyright:

Available Formats

Gender and Social Policy

Social Policy in a Development Context

Shahra Razavi and Shireen Hassim

List of Tables vii

List of Acronyms and Abbreviations x

Preface and Acknowledgements xv

1 Gender and Social Policy in a Global Context: Uncovering

Part I Historical and Regional Trajectories in

3 Gender and Post-socialist Welfare States in Central

4 Maternalist Policies versus Women’s Economic

5 Gender Equality and Developmental Social Welfare in South Africa 109

7 The Evolution of the Women-friendly State: Opportunities

8 The Adult-worker-model Family and Gender Equality: Principles

Part II Labour Market Informality and the Search

Part III Gender Dimensions of Social-sector Restructuring

13 Secondary Education in the Indian State of Uttar Pradesh:

Part IV Financing Social Provisioning and

15 Expectations versus Realities in Gender-responsive Budget

11.1 Selected African countries: percentage of the labour force

1.1 Ratio of female to male gross primary school enrolment rates 3

BpfA Beijing Platform for Action

Debbie Budlender is a specialist researcher with the Community Agency for

Maria S. Floro is an Associate Professor in the Economics Department at American

Susanna Giullari was a BA Fellow and a Research Officer employed on the

Shireen Hassim is Associate Professor in Political Studies at the University of the

Barbara Hobson is a Professor of Sociology at Stockholm University and the

Evelyne Huber (formerly Evelyne Huber Stephens) is Morehead Alumni

Francie Lund trained as a sociologist and community worker, and is Associate

Maureen Mackintosh is Professor of Economics at the Open University, UK. Her

Valentine M. Moghadam is currently Chief of Section in the Social and Human

Ito Peng is Associate Professor of Political Sociology at the Department of

Shahra Razavi is Research Coordinator at the United Nations Research Institute

Silke Steinhilber is a researcher and consultant on women’s economic and social

Ramya Subrahmanian is a Research Fellow at the Institute of Development

Paula Tibandebage is Training Co-ordinator for Research on Poverty Alleviation

A gender perspective on social policies in middle- and low-income countries, as

Shahra Razavi and Shireen Hassim

Japan ($26,940) 1.00

Czech Rep. ($15,780) 0.99

Japan ($26,940) 1.01

Czech Rep. ($15,780) 1.03

Japan ($26,940) 0.86

Figure 1.3 Ratio of female to male gross tertiary enrolment rates

South Korea ($16,950) 0.97

Figure 1.4 Ratio of female to male adult (15⫹) literacy rates

Country (per capita GDP, $US, PPP adjusted)

Figure 1.6 Percentage of parliamentary seats held by women, 2004

Japan, 2003 ($26,940) 25.7%

South Korea, 2003 ($16,950) 26.0%

The Economy as a Gendered Construction: Labour and Care

benefits. Furthermore, the gender-segmented features of the labour market have

Paid labour and social protection: the historical legacy

Many developing countries, by contrast, have been characterized by weak state

Labour market informalization: the demise of the

Gender and labour markets: continuity and change

Men Women (employed

Unpaid Paid Unpaid Paid Childcare All unpaid

Canada (1998) 4.0 6.3 5.1 5.9 1.4 1.3

100% 93.9% 95.1%

Figure 1.8 Informal employment as a percentage of total employment (15⫹)

household expenditure data effectively ignores the longstanding feminist

Table 1.2 Women’s hourly earnings as a percentage of men’s hourly earnings

Costa Egypt El Ghana South