Targeting Development

The Millennium Development Goals represent the first comprehensive collective

effort to reduce world poverty. Targeting Development is a major contribution to
assessing progress and problems in this critical global agenda.
Professor Frances Stewart,
Director, Queen Elizabeth House,
University of Oxford

A new global consensus has emerged that stresses that the objective of develop-
ment is to end global poverty. This consensus is accompanied by a bold new set of
targets: global poverty to be halved by 2015, along with universal primary educa-
tion, the removal of gender disparities in schooling, universal access to reproduc-
tive health care, specific reductions in infant, child and maternal mortality rates,
and a reversal in the loss of environmental resources. These ‘International Devel-
opment Targets’ were first adopted by the OECD in 1996, and have been suc-
ceeded by the yet more widely endorsed ‘Millennium Development Goals’
following from the UN Millennium Summit in September 2000.
The aim of this book is to provide a critical appraisal of these targets, and the
progress so far towards meeting them. The book consists of six introductory chapters
on how and why the International Development Targets and Millennium Develop-
ment Goals have become incorporated into development policy, and what their
overall value is. Each chapter in the second part analyses whether current trends
suggest that a specific target can be reached. Contributors assess the main constraints
that exist to achieving each of these targets and the resulting implications for policy.
This impressive collection, featuring an array of respected contributors and a
Foreword from Mark Malloch Brown of the UNDP, will be required reading
among development economists and those interested in development studies
more generally. Perhaps more importantly, the lessons learned from this book will
need to be understood and acted upon by policy-makers at both national and
international levels.

Richard Black is Professor of Human Geography at the University of Sussex, UK.

Howard White is a Fellow at the Institute of Development Studies, University of
Sussex, UK. He is also co-author of Program Aid and Development and Econometrics and
Data Analysis for Developing Countries, both of which are also published by Routledge.

All royalties from the sale of this book will be donated to Oxfam.
Routledge Studies in Development Economics

1 Economic Development in the Middle East

Rodney Wilson

2 Monetary and Financial Policies in Developing Countries

Growth and stabilization
Akhtar Hossain and Anis Chowdhury

3 New Directions in Development Economics

Growth, environmental concerns and government in the 1990s
Edited by Mats Lundahl and Benno J. Ndulu

4 Financial Liberalization and Investment

Kanhaya L. Gupta and Robert Lensink

5 Liberalization in the Developing World

Institutional and economic changes in Latin America, Africa and Asia
Edited by Alex E. Fernández Jilberto and André Mommen

6 Financial Development and Economic Growth

Theory and experiences from developing countries
Edited by Niels Hermes and Robert Lensink

7 The South African Economy

Macroeconomic prospects for the medium term
Finn Tarp and Peter Brixen

8 Public Sector Pay and Adjustment

Lessons from five countries
Edited by Christopher Colclough

9 Europe and Economic Reform in Africa

Structural adjustment and economic diplomacy
Obed O. Mailafia
10 Post-apartheid Southern Africa
Economic challenges and policies for the future
Edited by Lennart Petersson

11 Financial Integration and Development

Liberalization and reform in sub-Saharan Africa
Ernest Aryeetey and Machiko Nissanke

12 Regionalization and Globalization in the Modern World Economy

Perspectives on the Third World and transitional economies
Edited by Alex E. Fernández Jilberto and André Mommen

13 The African Economy

Policy, institutions and the future
Steve Kayizzi-Mugerwa

14 Recovery from Armed Conflict in Developing Countries

Edited by Geoff Harris

15 Small Enterprises and Economic Development

The dynamics and micro and small enterprises
Carl Liedholm and Donald C. Mead

16 The World Bank

New agendas in a changing world
Michelle Miller-Adams

17 Development Policy in the Twenty-First Century

Beyond the post-Washington consensus
Edited by Ben Fine, Costas Lapavitsas and Jonathan Pincus

18 State-Owned Enterprises in the Middle East and North Africa

Privatization, performance and reform
Edited by Merih Celasun

19 Finance and Competitiveness in Developing Countries

Edited by José and María Fanelli and Rohinton Medhora

20 Contemporary Issues in Development Economics

Edited by B.N. Ghosh

21 Mexico beyond NAFTA

Edited by Martín Puchet Anyul and Lionello F. Punzo
22 Economies in Transition
A guide to China, Cuba, Mongolia, North Korea and Vietnam at the turn
of the twenty-first century
Ian Jeffries

23 Population, Economic Growth and Agriculture in Less Developed

Countries
Nadia Cuffaro

24 From Crisis to Growth in Africa?

Edited by Mats Lundahl

25 The Macroeconomics of Monetary Union

An analysis of the CFA franc zone
David Fielding

26 Endogenous Development
Networking, innovation, institutions and cities
Antonio Vasquez-Barquero

27 Labour Relations in Development

Edited by Alex E. Fernández Jilberto and Marieke Riethof

28 Globalization, Marginalization and Development

Edited by S. Mansoob Murshed

29 Programme Aid and Development

Beyond conditionality
Howard White and Geske Dijkstra

30 Competitiveness Strategy in Developing Countries

A manual for policy analysis
Edited by Ganeshan Wignaraja

31 The African Manufacturing Firm

An analysis based on firm surveys in sub-Saharan Africa
Dipak Mazumdar and Ata Mazaheri

32 Trade Policy, Growth and Poverty in Asian Developing Countries

Edited by Kishor Sharma

33 International Competitiveness, Investment and Finance

A case study of India
Edited by A. Ganesh Kumar, Kunal Sen and Rajendra R. Vaidya
34 The Pattern of Aid Giving
The impact of good governance on development assistance
Eric Neumayer

35 New International Poverty Reduction Strategies

Edited by Jean-Pierre Cling, Mireille Razafindrakoto and François Roubaud

36 Targeting Development
Critical perspectives on the Millennium Development Goals
Edited by Richard Black and Howard White
Targeting Development
Critical perspectives on the
Millennium Development Goals

Edited by Richard Black and

Howard White
First published 2004
by Routledge
11 New Fetter Lane, London EC4P 4EE
Simultaneously published in the USA and Canada
by Routledge
29 West 35th Street, New York, NY 10001
Routledge is an imprint of the Taylor & Francis Group
This edition published in the Taylor & Francis e-Library, 2005.
“To purchase your own copy of this or any of Taylor & Francis or Routledge’s
collection of thousands of eBooks please go to [Link].”
© 2004 Editorial matter and selection, the editors; individual
chapters, the contributors
Typeset in Baskerville by Wearset Ltd, Boldon, Tyne & Wear
Printed and bound in Great Britain by Antony Rowe Ltd,
Chippenham, Wiltshire
All rights reserved. No part of this book may be reprinted or
reproduced or utilised in any form or by any electronic, mechanical,
or other means, now known or hereafter invented, including
photocopying and recording, or in any information storage or
retrieval system, without permission in writing from the publishers.
British Library Cataloguing in Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging in Publication Data
A catalog record for this book has been requested

ISBN 0-203-40323-1 Master e-book ISBN

ISBN 0-203-34596-7 (Adobe eReader Format)

ISBN 0–415–30376–1 (Print Edition)
Contents

List of illustrations xi
Notes on contributors xiv
Foreword by Mark Malloch Brown xviii
Preface xx
List of abbreviations xxii

1 Millennium Development Goals: a drop in the ocean? 1

HOWARD WHITE AND RICHARD BLACK

2 Heaven or hubris: reflections on the new ‘New Poverty

Agenda’ 25
SIMON MAXWELL

3 Using development goals and targets for donor agency

performance measurement 47
HOWARD WHITE

4 Using development goals to design country strategies 77

HOWARD WHITE AND DAVID BOOTH

5 Monitoring progress towards the Millennium

Development Goals at country level 96
DAVID BOOTH AND HENRY LUCAS

6 Are the Millennium Development Goals feasible? 124

JAN VANDEMOORTELE

7 Halving world poverty 145

MICHAEL LIPTON WITH HUGH WADDINGTON
x Contents
8 Towards universal primary education 166
CHRISTOPHER COLCLOUGH

9 Promoting gender equality 184

RAMYA SUBRAHMANIAN

10 Reducing infant and child death 209

HOWARD WHITE

11 Towards reproductive health for all? 235

HILARY STANDING

12 The global challenge of HIV/AIDS 256

RONALD SKELDON

13 Clean water for all 273

RICHARD JOLLY

14 Achieving sustainability in Africa 292

JAMES FAIRHEAD

15 Building a global partnership for development? 307

PETER CLARKE

16 Aid, trade and debt: how equal is the global partnership? 323
KEVIN WATKINS AND JULIANA AMADI

Index 354
Illustrations

Figures
2.1 Progress towards the Millennium Development Goals 27
2.2 Interactive dimensions of poverty and well-being 34
3.1 Overview of Department for International Development
strategy: vision statement, fifteen-year horizon 60
6.1 Incidence of income poverty in China, 1987–1999 126
6.2 Average net primary enrolment ratio (NER) and
under-five mortality rate (U5MR) in developing countries 128
6.3 Under-five mortality by wealth group in selected countries 134
6.4 Underinvestment in basic social services 136
6.5 Decline of Official Development Assistance 137
6.6 Broken promises of the 1990s 138
8.1 Scatter plot of primary gross enrolment ratio (GER)
against GNP per capita, 1990: low- and middle-income
countries 178
10.1 Trends in child and infant mortality by region 211
10.2 The Mosley–Chen framework for analysing mortality
(modified) 219
10.3 Relationship between structure and level of mortality 223
10.4 Trends in mortality rates for selected countries 224
12.1 Trends in mortality from AIDS among children under
5 years old and end-1999 adult prevalence rate, selected
African countries, 1981–1996 258
12.2 Projected population of Botswana to 2020 showing
deficits due to AIDS-related deaths by cohort 260
12.3 HIV-1 seroprevalence among Royal Thai Army conscripts
by region, 1989–2000 263

Tables
1.1 The International Development Targets 4
1.2 Target-related indicators 5
xii Illustrations
1.3 The Millennium Development Goals 7–9
2.1 Do’s and don’ts of development 43
3.1 Data coverage of baseline data for the International
Development Targets 66–67
3.2 DFID Public Service Agreement and Service Delivery
Agreement 68–73
4.1 Tabular presentation of future UK development strategy:
Malawi 84
4.1 Definitions and perceptions of poverty 89
4.2 Reporting of International Development
Targets and other indicators 90–91
4.3 Intended outcomes and monitoring procedures 92–94
7.1 Incidence of poverty in developing regions, 1990–2015 152
8.1 Trends in primary gross enrolment ratio (GER) by region
and gender, 1965–1997 169
8.2 Female enrolments at primary level in developing
countries, relative to male enrolments, 1980 and 1997 170
8.3 Sub-Saharan African countries with gross enrolment
ratios (GERs) less than 100: public expenditures on
primary schooling and related characteristics 172–173
9.1 How gender enters the Millennium Development Goals 188
9.2 Regional achievements in gender equality 193
10.1 Annual mortality reductions over 25 years (1973–1998) by
region 213
10.2 Infant mortality rates for selected countries 214
10.3 Regression results for infant mortality 216–217
10.4 Causes of death for selected regions, 2000 227
10.5 Child mortality and immunisation by asset quintile in
selected African countries 228
10.6 Benefit incidence of public spending on health in
selected countries 230
11.1 Millennium Development Goal for maternal health 236
11.2 Levels of current contraceptive use by major areas 244
12.1 Global situation of the HIV/AIDS epidemic 256
12.2 Impact of AIDS-related mortality on life expectancy at
age 20, selected countries, 2000–2005 259
13.1 Global water supply and sanitation coverage, 2000 276
13.2 Water supply and sanitation technologies considered to
be ‘improved’ and those considered to be ‘not improved’ 277
13.3 Coverage of rural, urban and total water supply by
region, 2000 278
13.4 Access to rural, urban and total sanitation by region,
2000 279
13.5 Estimated coverage of drinking water supply and sanitation
in developing countries, 1970, 1980, 1990 and 2000 279
 Illustrations xiii
13.6 Additional population required to be covered to meet the
2015 targets for water and sanitation 280
13.7 Country progress on halving the proportion of people
without sustainable access to safe drinking water by 2015 281

Boxes
1.1 Why do international targets matter? 12
2.1 Department for International Development Public
Service Agreement, 2002–2006: objectives and
performance targets 31–32
2.2 The benefits of agricultural growth 37
2.3 Assessing development agency policy reduction efforts 41
3.1 Data quality: the case of maternal mortality 53
3.2 The Department for International Development’s
strategy papers 58
4.1 Poverty Aim Marker (PAM) classifications 82
5.1 What makes a ‘good’ indicator? 104
11.1 Key moments in the development of the reproductive
health approach 239
13.1 Jim Grant’s ten commandments 288
Contributors

Juliana Amadi, Research Assistant, Institute of Development Studies, Univer-

sity of Sussex. Juliana Amadi is an economist with an interest in agricul-
tural issues. She has previously worked on international commodity
issues, including a spell at the International Rubber Study Group.
Her publications include analyses of crop-level productivity in South
Africa.
Richard Black, Professor of Human Geography, University of Sussex and Co-
Director of the Sussex Centre for Migration Research, University of
Sussex. Richard Black’s work focuses on the study of international migra-
tion, including forced migration and post-conflict return, and related
social and economic transformations. He is Director of a DFID funded
Development Research Centre on migration, globalisation and poverty,
building on work on migration, return, and post-conflict reconstruction in
the Balkans and West Africa. This work developed out of earlier work on
refugees and environment in West Africa, and his book Refugees, Environ-
ment and Development (Longman, 1998).
David Booth, Research Fellow, Poverty and Public Policy Group, Overseas
Development Institute. David Booth is a sociologist whose work focuses
on policy processes surrounding poverty reduction. Recent work has
addressed new aid modalities and the national policy process; process-
oriented monitoring and impact assessment; and the role of political
analysis in aid, especially in sub-Saharan Africa and Latin America. He
led a recent study on poverty reduction strategies in African countries,
the findings from which were published in a special issue of Development
Policy Review (March 2003).
Peter Clarke, NGO field worker specialising in local democracy and devel-
opment. He has worked for the last sixteen years in the small town of
Estelí, Nicaragua, in a changing context of revolution, hyperinflation,
structural adjustment and natural disaster. During a recent two-year leave
of absence at the Institute of Development Studies he was able to develop
his concerns about the lack of relation between aid practice, policy
rhetoric, development research and broader social theory.
 Contributors xv
Christopher Colclough, Professorial Fellow in Development Economics,
Institute of Development Studies, University of Sussex. Christopher Col-
clough’s areas of specialisation include educational planning and reform;
the costs and financing of education and training; and labour markets
and structural adjustment. Recent books include Achieving Schooling for
All in Africa: Costs, Commitment and Gender (with others), Ashgate 2003;
Public Sector Pay and Adjustment (ed.), Routledge 1997; Marketizing Educa-
tion and Health in Developing Countries (ed.), OUP 1997, and Educating All
the Children (with Keith Lewin), OUP 1993. He has worked as an adviser
to many governments and international agencies, particularly on ques-
tions of education and employment policy. He was a consultant to the
ANC and subsequently to the Department of Education in South Africa,
1994–2000, providing advice on the problems of financing the new
government’s education policies. He is currently Director of the Global
Monitoring Report Team on Education for All, based at UNESCO, Paris,
where he is responsible for producing an independent annual report
which charts progress towards the six ‘Dakar’ goals and the two Millen-
nium Development goals for education.
James Fairhead, Professor of Anthropology, Department of Anthropology,
University of Sussex. James Fairhead’s early work focused on issues of
power, knowledge and practice in African agriculture and ecology. This led
to an anthropology of farming and food systems, of development, and of
colonial and post-colonial sciences, policy and administration. Routledge
1998, two books (Misreading the African Landscape, CUP 1996 and Reframing
Deforestation, both co-authored with Melissa Leach) use anthropological and
historical methods to confront analytical traditions in African environ-
mental sciences and to expose their political, social and economic
commitments. His most recent book, Science, Power and Society, CUP 2003, has
taken an ethnographic approach to contemporary science and policy, com-
paring experiences in West Africa and the Caribbean. He is now working on
social dimensions to pharmaceutical trials and immunisation.
Richard Jolly, Honorary Professorial Fellow and Research Associate, Insti-
tute of Development Studies, University of Sussex. Sir Richard Jolly is
a development economist whose career has combined operational
involvement with research and teaching in a wide range of countries and
situations. He is currently working on a history of the UN’s contributions
to development ideas and thinking in the economic and social arena and
on long-term trends in global inequality. He has previously been Director
of IDS (1972–1981), Deputy Executive Director, Programmes of UNICEF
(1982–1995) and, from 1996 to 2000, Principal Coordinator of UNDP’s
Human Development Report.
Michael Lipton, Research Professor, Poverty Research Unit at Sussex,
University of Sussex. Michael Lipton’s research stresses poverty impacts of:
xvi Contributors
urban–rural and state–market linkages; farm technology and science; nutri-
tion economics; land reform; aid; and population change. In the 1970s he
headed a comparative analysis of village studies from developing countries,
leading to books by the team on migration, labour use and nutrition. In
the 1980s and 1990s he continued to work on urban bias, rural finance,
small-scale farming and post-harvest grain loss. Recently he has worked
mainly on demographic interactions with poverty, and contributed to UN
Human Development Reports on poverty, globalisation and technology; to
the World Bank’s 2000/2001 World Development Report on poverty; to
the Asian Development Bank’s Emerging Asia (1997); as Lead Scholar to
the International Fund for Agricultural Development’s 2001 Rural Poverty
Report; and to analyses of the impact of transgenic crops in developing
countries by the Nuffield Council on Bioethics (1999 and 2003).
Henry Lucas, Fellow, Institute of Development Studies, University of Sussex.
Henry Lucas is a statistician who has specialised in research methods,
particularly in the area of health sector analysis and more recently for
PRSP monitoring and evaluation. He has been the principal researcher in
a number of recent studies involving the combination of quantitative and
qualitative/participatory fieldwork methods. These have included a review
of the Malawi Social Action Fund, an exploration of the links between
energy, poverty and gender in poor rural areas of China, and monitoring
and evaluation aspects of a major DFID health-care programme in Nigeria.
He has many years of experience in China and has undertaken a number
of studies on both rural and urban healthcare reforms.
Simon Maxwell, Director, Overseas Development Institute. Simon Maxwell
is Director of ODI and President of the Development Studies Association
of the UK and Ireland. He worked overseas for ten years, in Kenya, India
and Bolivia, and then for fifteen years at the Institute of Development
Studies at the University of Sussex, latterly as Programme Manager for
Poverty, Food Security and the Environment. He became Director of
ODI in 1997. He is an economist with research interests in development
theory and policy; poverty; food security; economic, social and cultural
rights; and aid. He has advised many international agencies and govern-
ments on poverty reduction issues.
Ronald Skeldon, Professorial Fellow, Department of Geography, University
of Sussex, Honorary Professor, University of Hong Kong, and Adjunct
Professor at the Institute for Population and Social Research at Mahidol
University, Thailand. Ronald Skeldon’s research is based around ques-
tions of population and development, primarily in East and Southeast
Asia. Recent work has focused on the migrations of the Chinese peoples,
particularly from Hong Kong, and on irregular movements of migrants
in and through Southeast Asia. Other research has concentrated on
population mobility and HIV/AIDS in Southeast Asia and on questions
of child labour in Asia.
 Contributors xvii
Hilary Standing, Fellow, Institute of Development Studies, University of
Sussex. Hilary Standing is a social anthropologist specialising in health
research and social development. Current interests include household
level and gender aspects of health and formal and informal care systems,
gender and equity in the context of health reforms, the management of
organisational change in health sector restructuring, especially the
changing roles of providers, and improving greater accountability within
health systems. She has worked extensively in rural and urban South Asia
and convenes the international Gender and Health Equity Network.
Ramya Subrahmanian, Fellow, Institute of Development Studies, University
of Sussex. Ramya Subrahmanian is a development specialist with exten-
sive experience in the areas of gender, social development and educa-
tion. Her experience includes work on mainstreaming gender/ social
development into development policies and institutional processes in a
wide range of development agencies. Current research includes work on
education exclusion, policy processes, livelihoods and education, includ-
ing a recently co-edited volume, Child Labour and the Right to Education in
South Asia: Needs versus Rights?, Sage 2003.
Jan Vandemoortele, Leader, Socio-economic Development Group, United
Nations Development Programme. Jan Vandemoortele is an economist
whose work has focused on labour market policies, public finance, income
distribution, poverty reduction and social policy. Between 1995–2001, he
was head of Social Policy at UNICEF. Between 1991–1994, he served with
UNDP as Senior Economist in Malawi/Zambia as well as with the Regional
Bureau for Africa. Prior to that, he worked for thirteen years for the ILO,
mostly in Africa.
Hugh Waddington, Research Assistant, Poverty Research Unit at Sussex,
University of Sussex. Hugh Waddington is an economist with an interest
in poverty issues. He is currently engaged in preparing case studies of aid
effectiveness for DFID.
Kevin Watkins, Senior Policy Advisor, Oxfam, UK. Kevin Watkins has exten-
sive experience of researching and campaigning on global and develop-
ment issues. Recent publications for Oxfam include Rigged Rules and Double
Standards: Trade, Globalization, and the Fight Against Poverty and the Oxfam
Education Report.
Howard White, Fellow, Institute of Development Studies, University of Sussex.
Howard White is an economist with interests in poverty analysis, macroeco-
nomic debates (especially related to growth and distribution), the
determinants of human development and aid effectiveness (including the
recent co-authored book, Programme Aid and Development, Routledge 2003).
He has worked in a number of African countries (most recently Ghana,
Tanzania and Zambia) as well as Sri Lanka and Vietnam. He is currently on
secondment to the Operations Evaluation Department of the World Bank.
Foreword

The Millennium Development Goals have had a catalytic effect on the

global development debate, largely because of their simplicity and mea-
surability – and thus accessibility. Anybody can understand them, grasp
that they matter, judge whether or not his or her country and the wider
world is doing enough to achieve them, and take action if they are not.
They are a bottom-up, grass-roots, pocket-book development agenda,
firmly focused on the bread and butter of political life everywhere.
These Millennium Development Goals did not come out of thin air.
They were agreed at the historic UN Millennium Summit in 2000 and are
derived from the UN conferences of the 1990s. They also lie at the heart
of the Monterrey Consensus and the Johannesburg Plan agreed at the
World Summit for Sustainable Development in 2002.
As United Nations Development Programme Administrator I am
responsible for leading the United Nations system in developing a strategy
to support the achievement of the Goals. But my involvement extends
beyond this formal responsibility to my own deep commitment to the
principles behind the Millennium Development Goals. Central to reach-
ing the Goals is the task of forging partnerships: of governments of poor
and rich countries, of private corporations and foundations, of multilat-
eral agencies, of civil society organisations, and, most importantly, of the
poor themselves.
From hard experience with many past initiatives, we know that owner-
ship is indispensable to success. To ensure that ownership, the Goals need
to be customised and tailored to national circumstances and built into
national medium-term goals and strategies. For over seventy of the world’s
poorest countries, these strategies take the form of a Poverty Reduction
Strategy Paper (PRSP). The Goals can be seen as both the front and the
back end of the PRSPs: they represent the overarching objectives of the
PRSPs and the way to monitor whether they are performing as advertised.
As such, the Goals offer a real-time accountability framework – tracking
what is working and what isn’t, where progress is being made and thus
helping drive real change in a much more timely and effective manner
than the traditional five- and ten-year conference reviews.
 Foreword xix
The success or failure of this entire vision depends crucially on a global
partnership between developed and developing countries, expressed in
Goal 8. It is no exaggeration to say the outcome will hinge on the commit-
ments of rich countries to help poorer nations that are undertaking eco-
nomic, political and social reforms in good faith. While the reallocation of
domestic resources will be instrumental – alongside strengthened gover-
nance and sound social and economic policies – these measures alone are
unlikely to be sufficient to achieve the Millennium Development Goals.
Goals and targets to halve hunger and poverty will fail without a funda-
mental restructuring of the global trading system, particularly in agricul-
ture and textiles. The fight against HIV/AIDS, malaria and other diseases
will be lost without affordable essential drugs. Without steep and fast debt
relief, macroeconomic stability will remain elusive for most of the heavily
indebted poor countries. And last, but by no means least, it is important to
remember that an extra $50 billion in annual official development assis-
tance will be a minimum to meet the Goals.
Yet there is a chance the poor will get those resources – and, much
more important, the political and social momentum for real change on
the Millennium Development Goals as a simple but powerful idea whose
time has come. Already these Goals are taking discussions about develop-
ment to classrooms and coffee houses, town halls and tenements, fields
and factories, showing everyone from the president or prime minister
down to schoolchildren where more attention needs to be paid and where
things are off track, where more resources are essential and where they
are not being effectively used.
It is against this background that I warmly welcome this book. The
various chapters provide thoughtful and important contributions to for-
mulating reforms to meet the Millennium Development Goals. The con-
tributors do not suggest that meeting the Goals or implementing the
required changes will be easy – and I agree with them. But nor do they
suggest that it is impossible – again, I agree with them. It is by tackling the
issues they raise that we can make the dream of a world without poverty a
practical reality.

Mark Malloch Brown

UNDP Administrator
Preface

A new global consensus has emerged that stresses that the objective of
development is to end global poverty. This consensus is accompanied by a
bold new set of targets – global poverty to be halved by 2015, along with
universal primary education, the removal of gender disparities in school-
ing, universal access to reproductive healthcare, specific reductions in
infant, child and maternal mortality rates, and a reversal in the loss of
environmental resources. These International Development Targets
(IDTs) were first adopted by the OECD in 1996, and have been taken up
enthusiastically by the New Labour government in Britain and many inter-
national development agencies. The UN Millennium Summit in Septem-
ber 2000 endorsed the approach, setting its own Millennium
Development Goals (MDGs), which again put the elimination of poverty
at the heart of the international development agenda.
The aim of this book is to provide a critical appraisal of these targets,
the progress so far towards meeting them, and obstacles to their attain-
ment. The book consists of six introductory chapters on how and why the
International Development Targets and Millennium Development Goals
have become incorporated into development policy, their overall value,
and an assessment of whether they are feasible, followed by a chapter
addressing each goal or target (or related set of targets). Each chapter in
the second part sets out the background behind the target, and analyses
whether current trends suggest the target can be reached. Is even a global
halving of poverty – rather than its ‘eradication’ – achievable? Can the
world realistically move towards universal education and access to repro-
ductive healthcare, and does it make sense to set a time limit for the
achievement of these objectives?
Contributors assess the main constraints that exist to achieving each of
these targets and the resulting implications for policy. They also consider
the relevance of the targets that have been set in London and Washington
to the problems of the poorer countries of the world, suggesting altern-
ative approaches where appropriate. In doing this, the aim is to undertake
the first comprehensive account of the International Development
Targets and Millennium Development Goals and progress towards them.
 Preface xxi
The book is based on a series of ‘Sussex Development Lectures’ first
delivered at the University of Sussex in Autumn 2001. These lectures rep-
resent a collaborative initiative between a group of Sussex institutes and
centres concerned with international development, including the Institute
of Development Studies (IDS), the Culture, Development and Environ-
ment Centre (CDE), the Science and Technology Policy Research Unit
(SPRU), the Institute of Education (USIE) and the Economics, Inter-
national Relations and Politics, Social Anthropology and Geography
Subject Groups. The lecture series was delivered primarily by faculty based
at Sussex, and this is reflected in the origin of the majority of contributors
to this volume. However, we have also gone outside Sussex to commission
a small number of additional chapters, in order to give a fully rounded
view of the targets, and gain additional insights.
We would like to thank all those who have contributed to making this
book a reality. Special thanks go to Hugh Waddington for his careful
editing of the chapters and Julie McWilliam for producing the final ver-
sions. Part of Howard White’s input to the editing of this collection has
been funded by DFID support to research on poverty reduction policies
and we are grateful for this funding. All royalties from the production of
this book are being donated to Oxfam.

Richard Black
Howard White
Falmer, University of Sussex
Abbreviations

ACP Africa, Caribbean and Pacific

ARDE Annual Review of Development Effectiveness
ARROW Asian-Pacific Resource and Research Centre for Women
ATC Agreement of Textiles and Clothing
ATP Aid-Trade Provision
CBN cost of basic needs
CEDAW Convention on the Elimination of All Forms of
Discrimination against Women
CEO Chief Executive Officer
CHANGE Center for Health and Gender Equity
CIA Central Intelligence Agency (USA)
CIDA Canadian International Development Agency
CPR contraceptive prevalence rate
CSP Country Strategy Paper
CWIQ Core Welfare Indicators Questionnaire
DAC Development Assistance Committee (of the Organisation for
Economic Cooperation and Development)
DFID Department for International Development (UK)
DHS Demographic and Health Survey
DOTS Directly Observed Treatment Short Course
DRC Democratic Republic of Congo
DTI Department for Trade and Industry (UK)
EC European Commission
EPI Expanded Programme of Immunisation
EU European Union
FAO Food and Agriculture Organisation
FCO Foreign and Commonwealth Office (UK)
FEM food energy method
FEZ Food Economy Zone
FY fiscal year
GAO General Accounting Office (USA)
GATS General Agreement on Trade in Services
GAVI Global Alliance for Vaccines and Immunisation
 Abbreviations xxiii
GDP gross domestic product
GER gross enrolment ratio
GM genetic modification
GNI gross national income
GNP gross national product
GPRA Government Performance and Results Act (USA)
HDR Human Development Report
HIPC heavily indebted poor country
ICPD International Conference on Population and Development
IDA International Development Agency
IFAD International Fund for Agricultural Development
IMF International Monetary Fund
I-PRSP Interim Poverty Reduction Strategy Paper
ISP Institutional Strategy Paper
LDC least developed country
MCH maternal and child health
MFA Multi-Fibre Agreement
MMR maternal mortality ratio
MoD Ministry of Defence (UK)
MTEF Medium Term Expenditure Framework
NAO National Audit Office (UK)
NEPAD New Partnership for Africa’s Development
NER net primary enrolment ratio
NGO non-governmental organisation
NHP National Health Programme (Oman)
NIC newly industrialising country; National Intelligence Council
(USA)
ODA Official Development Assistance
OECD Organisation for Economic Cooperation and Development
OED Operations Evaluation Department
ORT oral rehydration therapy
PAM Poverty Aim Marker
PARIS Partnerships in Statistics for Development in the Twenty-First
Century
PEAP Poverty Eradication Action Plan (Uganda)
PER Public Expenditure Review
PET public expenditure tracking
PETS Public Expenditure Tracking Studies
PPA Participatory Poverty Assessment
PPP purchasing power parity
PRA Participatory Rural Appraisal
PRSC Poverty Reduction Strategy Credit
PRSP Poverty Reduction Strategy Paper
PSA Public Service Agreement
RAWOO Netherlands Development Assistance Research Council
xxiv Abbreviations
RBM results-based management
ROAR Results-Oriented Annual Report
RTI reproductive tract infection
SDA Service Delivery Agreement
SPA Strategic Partnership for Africa (previously Special
Partnership for Africa)
STI sexually transmitted infection
SWAP sector-wide approach
TRIMs Trade-Related Investment Measures
TRIPs Trade-Related Aspects of Intellectual Property Rights
TSP Target Strategy Paper
UCI universal coverage of immunisation
UN United Nations
UNCTAD United Nations Conference on Trade and Development
UNDP United Nations Development Programme
UNESCO United Nations Educational, Scientific and Cultural
Organisation
UNFPA United Nations Fund for Population Activities
UNICEF United Nations Children’s Fund
UNRISD United Nations Research Institute for Social Development
USAID United States Agency for International Development
VIP ventilated improved pit latrine
WDR World Development Report
WHO World Health Organisation
WIDER World Institute for Development Economics Research
WSSCC Water Supply and Sanitation Collaboration Council
1 Millennium Development Goals
A drop in the ocean?
Howard White and Richard Black

Introduction
What are the development targets, such as the International Development
Targets and the Millennium Development Goals, for? For supporters, they
are a crystallisation of what it is that international development is sup-
posed to be about. The Targets are seven quantifiable goals, against which
the performance of donors and international development agencies can
be measured. First set out in the Organisation for Economic Cooperation
and Development (OECD) document Shaping the Twenty-First Century
(OECD, 1996), they won unprecedented support and prominence. In the
UK in particular, the Department for International Development (DFID),
and its former Secretary of State, Clare Short, was vocal in promoting the
International Development Targets. They have occupied a central posi-
tion in two government White Papers, the public pronouncements of the
Secretary of State, and within DFID in developing its new anti-poverty
strategy. Meanwhile, agreement on the ‘Millennium Development Goals’
at the Millennium Summit in New York in September 2000 has extended
the number of agreed targets to eighteen, although some are not precisely
defined.1
Yet there have been many previous development goals and targets over
the decades. International development agencies and donors have sought
to promote economic growth, and then ‘growth with equity’; in the 1980s
there was then a shift towards the meeting of ‘basic needs’, before the
rhetoric of ‘sustainable development’ took over in the 1990s. Why, then, is
this set of goals and targets new? Do they justify analysis, or are they simply
a passing fad? Clearly, our argument is that they do merit attention. As a
comprehensive and measurable set of indicators, the International Devel-
opment Targets and the successor Millennium Development Goals have
established themselves as a major force in current development practice.
They help to define both the goal of development cooperation activities,
and a set of priorities to be followed within these activities. They sit along-
side other international goals and targets, notably those relating to climate
change that followed from the Earth Summit in 1992. International targets
2 Howard White and Richard Black
also stand as a basis on which development practitioners can be held
accountable for their actions.
This does not mean, however, that the targets are, or should be univer-
sally accepted as, a guide and measure of international development
performance. It is clear, for example, that they are more significant for
some donors than for others. The World Bank and the International
Monetary Fund (IMF) both adopted the International Development
Targets, and they became a reference point for the World Bank’s annual
flagship statistical publication World Development Indicators. However, for
most other donors they did not assume the importance they did for DFID,
although the Millennium Development Goals have gained greater promi-
nence. Both the United Nations Development Programme (UNDP) and
the World Bank have Web sites dedicated to the Millennium Development
Goals, with links from their home pages. The extent to which the targets
have influenced strategy and programmes is of course another matter,
one which is pursued throughout this volume. Whilst growing out of the
OECD’s Development Assistance Committee (DAC), and endorsed by the
United Nations (UN), the targets have already arguably been overtaken
for some, especially in the US administration, by new concerns to tie
development assistance to security issues in the post-11 September ‘war on
terrorism’. Moreover, although quite broad in their conception of
poverty, and ranging across several different sectors, the targets could still
be seen as quite limited and narrow in their scope – more basic even than
‘basic needs’. Some would reject the contention that the goals of inter-
national development practice can be reduced to a set of quantifiable
indicators, whilst the action required to achieve these goals could be per-
ceived as rather unambitious. In this sense, are the International Develop-
ment Targets and the successor Millennium Development Goals simply a
drop in the ocean?
It is the aim of this chapter to examine what the International Develop-
ment Targets and Millennium Development Goals are trying to achieve,
and whether it is worth achieving. Subsequent chapters then follow up
specific aspects of an approach based on targets, before focusing attention
on the extent to which progress is being made towards meeting various
individual targets. This chapter begins with a brief introduction to the
targets as measures of performance, and why they have gained in import-
ance. We then move on to discuss the extent to which the International
Development Targets and Millennium Development Goals really matter,
and some of the problems that are inherent in their adoption. Finally, we
provide a summary of how the argument is taken forward in subsequent
chapters.
 Millennium Development Goals 3
What are the International Development Targets and
Millennium Development Goals?

The International Development Targets

The International Development Targets are diverse both in nature and
provenance (Table 1.1). Divided into three fields – economic well-being,
social development and environmental sustainability and regeneration –
they collectively represent a set of goals for poverty reduction, embodying
a multidimensional conception of poverty (Baulch, 1996; White, 1999).
The targets demonstrate that in current development practice, poverty is
about more than just a lack of income. Instead, drawing on the work of
the UNDP on ‘human development’, and a series of UN conferences
through the 1990s, they extend to include deprivation with respect to
other aspects of well-being such as health and education.
The targets share a number of characteristics. First, each is designed to
be quantifiable, although in some cases this causes difficulty. In particular,
there is no agreed indicator for access to reproductive health services.
Contraceptive prevalence is commonly used as an indicator, but is
not acceptable in some cultures, whilst the United States Agency for
International Development (USAID) is forbidden by law to support
programmes that provide abortion services. Even where contraceptive
prevalence is acceptable, there is no agreed target level, since the desired
level depends on desired fertility.
Second, most of the targets define the expected outcomes of develop-
ment, rather than inputs in the form of resources. This, for example, sets
them apart from the best known of previous development targets – that
developed nations should set aside 0.7 per cent of their gross domestic
product (GDP) in development assistance. However, here too there are
exceptions. Thus the target on environmental sustainability stresses the
need for national-level planning, rather than any particular environ-
mental outcome, whilst targets for reproductive health and education
stress access to services rather than health or educational results.
Third, accompanying the targets as a whole is a statement recognising
the importance of qualitative factors related to governance. The current
consensus amongst the international community is that democratic
accountability and a lack of corruption are necessary to achieve poverty
reduction goals. Although there are numerical indicators for such things
(such as those produced by Freedom House on political freedom and civil
liberties),2 they have not been used to monitor progress. This is because
donors have been unable to reach agreement on what should go into such
indicators. In this context, ‘good governance’ is less a goal in itself, and
more a precondition for meeting the development targets.
Confusingly, the targets too have not remained fixed. An additional
target was added on HIV/AIDS, namely a 25 per cent reduction in HIV
4 Howard White and Richard Black
Table 1.1 The International Development Targets

Target Where the target comes from

Economic well-being
• The proportion of people living in • Copenhagen Declaration and
extreme poverty in developing Programme of Action (1995)
countries should be reduced by at
least one-half by 2015
Social development
There should be substantial progress in
primary education, gender equality,
basic health care and family planning,
as follows:
• Universal primary education should • Jomtien Conference on Education for
be achieved in all countries by 2015 All (1990), endorsed at Copenhagen
Summit on Social Development
(1995) and Beijing Conference on
Women (1995)
• Progress toward gender equality and • Cairo Conference on Population and
the empowerment of women shall Development (1994), also Beijing and
be demonstrated by eliminating Copenhagen
gender disparity in primary and
secondary education by 2015
• The death rates for infants and • Cairo, confirmed at Beijing
children aged under 5 years should be
reduced in each developing country
by two-thirds of the 1990 level by 2015
• The rate of maternal mortality • Cairo, confirmed at Beijing
should be reduced by three-quarters
during this same period
• Access should be available through • Cairo
the primary health care system to
reproductive health services for all
individuals of appropriate ages,
including safe and reliable family
planning methods, as soon as possible
and no later than the year 2015
Environmental sustainability and regeneration
• There should be a current national • Rio Conference on Environment and
strategy for sustainable development, Development (1992)
in the process of implementation, in
every country, by 2005, so as to ensure
that current trends in the loss of
environmental resources – forests,
fisheries, fresh water, climate, soils,
biodiversity, stratospheric ozone
and the accumulation of hazardous
substances and other major indicators
– are effectively reversed at both
global and national levels

Source: Development Assistance Committee (1996).

 Millennium Development Goals 5
infection rates amongst 15 to 24-year-olds in the worst-affected countries by
2005 and globally by 2010. Meanwhile, the rather general target on
environmental sustainability was altered to include more specific goals,
including some ‘outcomes’ in terms of protection, energy use and emis-
sions. There are also more indicators that can be used to measure progress
than those explicitly mentioned in the definition of each target. A list of
indicators has been developed around each target (Table 1.2), many of
which are also outcome indicators. In the three cases in which the target
itself does not measure outcome (education, reproductive health and
environment), some of the related indicators do so (e.g. literacy rates and
fertility rates). In other cases, the additional indicators listed do not
measure outcomes, but provide a quantifiable indicator of access to a
service that critically affects the desired outcome (e.g. attended births).
The International Development Targets were formally adopted at
the Thirty-Fourth High-Level Meeting of the DAC on 6–7 May 1996
in Paris. Given the recent prominence of ‘participatory’ approaches to

Table 1.2 Target-related indicators

Target Indicators

Poverty reduction Population below a dollar a day

Incidence times depth of poverty
Poorest fifth’s share of national consumption
Prevalence of underweight under 5
Universal primary education Net primary enrolment ratio
Survival to fifth grade of primary education
Literacy rate of adults
Gender equality Ratio of girls to boys in primary and secondary
education
Ratio of literate females to males
Infant mortality reduction Infant mortality rate
Child mortality reduction Under-five mortality rate
Maternal mortality reduction Maternal mortality ratio
Births attended by skilled health personnel
Reproductive health Contraceptive prevalence rate
Total fertility rate
Environment Existence of national strategies for sustainable
development
Sustainable environment Population with access to safe water
Biodiversity: land area protected
Energy efficiency: GDP per unit of energy use
Carbon dioxide emissions

Source: DAC ‘Methodological Note’, DCD/DAC (98)6/ADD, Paris: OECD/DAC (available

at [Link]
6 Howard White and Richard Black
development, and the fact that the targets themselves are conceived as part
of the process of making development assistance more accountable to aid
beneficiaries (see pp. 12–13), it is somewhat ironic that it was a developed
country group such as DAC that set targets for developing countries.
There are two defences against this criticism. First, developing country
governments do not have to sign up to precisely these poverty reduction
goals. Rather, the idea is that donor support should be predicated upon
recipient commitment to poverty reduction, measured against an appro-
priate target that they may choose themselves. For example, the target of
the Ugandan government is to reduce the poverty head count to 10 per
cent or less by 2016, in addition to meeting specific target figures for
infant and child mortality. Second, the International Development Targets
were based on resolutions passed at various international conferences,
and in this sense they have already been endorsed by developing coun-
tries. Nonetheless, in some cases it is the principle rather than the specific
target that was adopted at UN conferences. This is the case, for example,
with the target set for reduction in income poverty.

The Millennium Development Goals

The discontent felt in some quarters that the International Development
Targets represented an imposition by the developed country members of
the Development Assistance Committee manifested itself at the Millen-
nium Summit in New York held on 6–8 September 2000. There, an altern-
ative set of development targets, the Millennium Development Goals, was
adopted (UN, 2000). The Millennium Declaration contained a list of goals
which overlapped with the International Development Targets but were
not the same. However, a year later the UN document Road Map towards
the Implementation of the United Nations Millennium Declaration laid out a
finally agreed list of Millennium Development Goals as shown in Table
1.3. This list fully encompassed the earlier targets, whilst adding new ele-
ments of its own.3 The most notable differences between the two sets of
targets are as follows:

• There are more Millennium Development Goals than International

Development Targets, comprising eight goals with eighteen separate
targets (although seven of these targets relate to the new eighth goal
of global partnership); there is an expanded list of forty-eight indic-
ators to monitor the targets (which are still under development).
• The International Development Targets are embodied in the first
seven Goals, but with additional aspects added, notably nutrition,
shelter and diseases other than HIV/AIDS.
• Although governance indicators are still generally absent, female
representation in parliament is included amongst the expanded set of
indicators for gender equality.
 Millennium Development Goals 7
Table 1.3 The Millennium Development Goals

Goals and targets Indicators

Goal 1: Eradicate extreme poverty and hunger

Target 1: Halve, between 1990 and 1 Proportion of population below US$1
2015, the proportion of people whose per day
income is less than one dollar a day 2 Poverty gap ratio [incidence
multiplied by depth of poverty]
3 Share of poorest quintile in national
consumption
Target 2: Halve, between 1990 and 4 Prevalence of underweight children
2015, the proportion of people who (under 5 years of age)
suffer from hunger 5 Proportion of population below
minimum level of dietary energy
consumption
Goal 2: Achieve universal primary education
Target 3: Ensure that by 2015 children 6 Net enrolment ratio in primary
everywhere, boys and girls alike, will be education
able to complete a full course of 7 Proportion of pupils starting grade 1
primary schooling who reach grade 5
8 Literacy rate of 15- to 24-year-olds
Goal 3: Promote gender equality and empower women
Target 4: Eliminate gender disparity in 9 Ratio of girls to boys in primary,
primary and secondary education secondary and tertiary education
preferably by 2005 and to all levels of 10 Ratio of literate females to males of
education no later than 2015 15- to 24-year-olds
11 Share of women in wage employment
in the non-agricultural sector
12 Proportion of seats held by women in
national parliament
Goal 4: Reduce child mortality
Target 5: Reduce by two-thirds, 13 Under-five mortality rate
between 1990 and 2015, the under-five 14 Infant mortality rate
mortality rate 15 Proportion of 1-year-old children
immunised against measles
Goal 5: Improve maternal health
Target 6: Reduce by three-quarters, 16 Maternal mortality ratio
between 1990 and 2015, the maternal 17 Proportion of births attended by
mortality ratio skilled health personnel
Goal 6: Combat HIV/AIDS, malaria and other diseases
Target 7: Have halted by 2015, and 18 HIV prevalence among 15- to 24-year-
begun to reverse, the spread of old pregnant women
HIV/AIDS 19 Contraceptive prevalence rate
20 Number of children orphaned by
HIV/AIDS
continued
8 Howard White and Richard Black
Table 1.3 continued

Goals and targets Indicators

Target 8: Have halted by 2015, and 21 Prevalence and death rates associated
begun to reverse, the incidence of with malaria
malaria and other major diseases 22 Proportion of population in malaria
risk areas using effective malaria
prevention and treatment measures.
23 Prevalence and death rates associated
with tuberculosis
24 Proportion of tuberculosis cases
detected and cured under DOTS
(Directly Observed Treatment Short
Course)
Goal 7: Ensure environmental sustainabilitya
Target 9: Integrate the principles of 25 Proportion of land area covered by
sustainable development into country forest
policies and programmes and reverse 26 Land area protected to maintain
the loss of environmental resources biological diversity
27 GDP per unit of energy use (as proxy
for energy efficiency)
28 Carbon dioxide emissions (per
capita). [Plus two figures for global
atmospheric pollution: ozone
depletion and the accumulation of
global warming gases.]
Target 10: Halve, by 2015, the 29 Proportion of population with
proportion of people without sustainable access to an improved
sustainable access to safe drinking water source
water
Target 11: By 2020, to have achieved a 30 Proportion of people with access to
significant improvement in the lives of improved sanitation
at least 100 million slum-dwellers 31 Proportion of people with access to
secure tenure. [Urban/rural
disaggregation of several of the above
indicators may be relevant for
monitoring improvement in the lives
of slum dwellers.]
continued
 Millennium Development Goals 9
Table 1.3 continued

Goals and targets Indicators

Goal 8: Develop a Global Partnership for Developmenta

Target 12: Develop further an open, Some of the indicators listed below will be
rule-based, predictable, non- monitored separately for the Least Developed
discriminatory trading and financial Countries, Africa, landlocked countries and
system. Includes a commitment to good small island developing states
governance, development, and poverty Official Development Assistance (ODA)
reduction – both nationally and 32 Net ODA as percentage of DAC
internationally donors’ gross national income
(GNI) [targets of 0.7 per cent in
Target 13: Address the special deeds of
total and 0.15 per cent for LDCs]
the Least Developed Countries (LDCs).
33 Proportion of ODA to basic social
Includes: tariff and quota free access
services (basic education, primary
for LDC exports; enhanced
health care, nutrition, safe water
programme of debt relief for Heavily
and sanitation)
Indebted Poor Countries (HIPC) and
34 Proportion of ODA that is untied
cancellation of official bilateral debt;
35 Proportion of ODA for
and more generous Official
environment in small island
Development Assistance (ODA) for
developing states
countries committed to poverty
36 Proportion of ODA for transport
reduction
sector in land-locked countries
Target 14: Address the special needs of
Market access
landlocked countries and small island
37 Proportion of exports (by value and
developing states (through Barbados
excluding arms) admitted free of
Programme and Twenty-Second
duties and quotas
General Assembly provisions)
38 Average tariffs and quotas on
Target 15: Deal comprehensively with agricultural products and textiles
the debt problems of developing and clothing
countries through national and 39 Domestic and export agricultural
international measures in order to subsidies in OECD countries
make debt sustainable in the long term 40 Proportion of ODA provided to
help build trade capacity
Target 16: In co-operation with
developing countries, develop and Debt sustainability
implement strategies for decent and 41 Proportion of official bilateral HIPC
productive work for youth debt cancelled
42 Debt service as a percentage of
Target 17: In co-operation with
exports of goods and services
pharmaceutical companies, provide
43 Proportion of ODA provided as
access to affordable, essential drugs in
debt relief
developing countries
44 Number of countries reaching HIPC
Target 18: In co-operation with the decision and completion points
private sector, make available the 45 Unemployment rate of 15- to 24-
benefits of new technologies, especially year-olds
information and communications 46 Proportion of population with
access to affordable essential drugs
on a sustainable basis
47 Telephone lines per 1,000 people
48 Personal computers per 1,000 people

Source: UN (2001) [Link]/DAC.

Note
a The selection of indicators for Goals 7 and 8 is subject to further refinement.
10 Howard White and Richard Black
• The new eighth goal adds targets relating to the direct contribution of
the developed countries to meeting the Millennium Development
Goals, with mention of aid, debt relief and tariff barriers. It is signifi-
cant that such actions (e.g. debt reduction, higher aid) have been
added. However, the wording of most of these is vague (‘address . . . ’,
‘deal comprehensively with . . . ’, ‘more generous . . . ’, etc.) compared
to the precise numerical goals for developing country performance.

With the adoption of the Millennium Development Goals, the promi-

nence of the targets has increased amongst donor agencies. The extent to
which these targets represent a valuable set of measures on which to assess
international development performance is considered in more depth in
Chapter 3.

Why have targets gained importance?

Why have first the International Development Targets, and now the Mil-
lennium Development Goals, gained an importance that escaped previous
development targets? Three reasons suggest themselves. The first is the
establishment of poverty at the top of the development agenda during the
1990s. Of course, poverty has always been a concern for development
agencies, though the extent to which it has received explicit attention has
varied greatly across time, and between different agencies. Yet during the
1990s nearly all development agencies have reaffirmed and strengthened
their commitment to poverty reduction and have been searching for ways
in which to realise this commitment. The end of the Cold War in some
respects released development aid from its political straitjacket, allowing
greater autonomy for development agencies in the definition of their
objectives. Thus in the UK, DFID was able to steer an International Devel-
opment Act through Parliament in early 2002, which defined the objective
of development as the eradication of poverty.4
Adoption of the International Development Goals, and subsequently
the Millennium Development Goals, has been part of this process. For
example, for DFID, the targets have been internalised into agency prac-
tice, setting out in concrete terms the general orientation towards poverty
eradication. The new formulation has been designed not only to act as a
guide for action, but also as a bulwark against the diversion of aid that was
perceived by some in government and outside to have occurred in the
past – notably incidents such as the linking of UK development aid for the
Pergau Dam in Malaysia to the sale of arms. Nonetheless, there have been
countervailing tendencies – notably the increased role played by humani-
tarian assistance, and the rise of political–military involvement in this
sector both before, but especially after, 11 September 2001.
Second, the 1990s also saw an emerging emphasis on results-based
management (RBM), especially in North America. For example, the
 Millennium Development Goals 11
World Bank adopted this approach in 1993, whilst in 1995 the Canadian
International Development Agency (CIDA) produced an overview of its
experience (Brown, 1995).5 Of course, donors have always aimed to have
some sort of monitoring at the project level, albeit often with an input
focus. Project monitoring is carried out both by the project management
and, often drawing on that, by agency staff for their own purposes
through supervision missions, mid-term reviews and the like. For example,
the World Bank has ‘Project Performance Reports’,6 which are completed
for each project following a staff mission to the project, usually on an
annual basis. These reports include an assessment of performance on
several criteria and overall development impact as judged against the
project objectives. However, the rise of results-based management has
shifted the focus to country programmes and the agency’s overall
performance (DAC, 2000: 18).
Once again, results orientation has taken a particular hold in the UK. It
first emerged in the 1990s, notably with the introduction of the Citizen’s
Charter, and league tables for schools. Since 1997, the Labour govern-
ment that took office in that year has warmly embraced this approach,
starting with its five pre-election pledges and continuing, for example,
with the expansion of league tables to other public services.7 A recent best-
selling account of ‘new’ Labour’s first term in office in Britain put it that
‘the government had more targets – over 6,000 on one count – than
Stalin’ (Rawnsley, 2001: 292). Within UK government practice, what are
known as ‘Public Service Agreements’ (PSAs) embody this approach. The
Public Service Agreement and Service Delivery Agreement are written
commitments of a department’s objectives, related performance meas-
ures, and the activities to be undertaken in support of those objectives. In
turn, the International Development Targets are an obvious set of devel-
opment results for development agencies to attach themselves to, and
DFID has used them as the basis for its PSA targets (see Chapter 3).
The third point is that the targets have received strong support from
some key institutional actors, which has helped to propagate them else-
where. For the International Development Targets, their origin in the DAC
and subsequent championing by DFID helped them gain a foothold in the
international community, their position becoming assured once they were
adopted by the World Bank. The Millennium Development Goals came
from an even broader institutional base, the UN, but the seriousness with
which they have been taken by the international community rests with the
groundwork done for the International Development Targets.
However, whilst there is no doubt that the targets have become import-
ant, questions can be raised as to whether this is a good thing or not. This
question can be asked in two ways. First, are outcome targets of any sort a
good guide to policy? Second, are the specific targets that have been set
for development a good set of performance measures? These questions
are addressed in the next two sections.
12 Howard White and Richard Black
Do targets matter?
Targets can be seen as important for a number of reasons (Box 1.1). First,
the identification of targets can be one way to define an organisation’s
purpose, since they set out what exactly is to be achieved in terms of out-
comes. During the past decade, most development agencies have adopted
an overall aim related to poverty reduction. For DFID, this is the ‘the elim-
ination of poverty in poorer countries’, whilst for the World Bank, the aim
is ‘a world free of poverty’. Yet what does ‘the elimination of poverty’
mean in a world where it is accepted that poverty is multidimensional?
Targets make what is potentially a very general statement of purpose
much more specific, by outlining what development organisations aim to
achieve across the different dimensions of poverty. That said, there is of
course quite a gap between the aspiration to ‘eliminate’ poverty, and the
target of reducing it by half within twenty years, leading to the possible
charge that setting an exact target leads to a watering down of aspirations.
The focus on outcomes is nonetheless important. There has been a
tendency, certainly amongst development agencies, to have performance
measurement systems (monitoring and evaluation) with too strong a focus
on inputs. In such systems, often the only question is: ‘Was the money
spent?’ Worse, there is a tendency for budgetary units to spend money
quickly towards the end of a budget year, in order to meet their financial
targets. In contrast, stressing outcomes reminds us what the money is
being spent for. If the outcomes are not being achieved, then having
spent the money may be a bad thing rather than a good thing.
A second important point is accountability. It is increasingly recognised
that public bodies need to be held accountable for their activities. With
clear targets for development, international aid agencies can in principle
be held to account for their performance in achieving these targets, just as
companies are held accountable to shareholders, government depart-
ments to taxpayers or politicians to voters. The statement of a target repre-
sents a commitment to achieve that target, so that the agency can be
judged by whether it does so or not. By stating a clear target, an organisa-
tion can also show clearly what its purpose is, so, in the case of govern-
ment departments, can make a case for funding. In an age when all
government budget headings are under scrutiny, and public support for

Box 1.1 Why do international targets matter?

• They define the objectives of policy in terms of outcomes.

• They provide a basis for accountability.
• They make a case for carrying out supporting activities.
• They allow comparisons of performance to be made.
• They create a sense of common purpose.
 Millennium Development Goals 13
international development assistance cannot be assumed, aid programmes
can arguably be defended on the basis that they help to make quantifiable
progress towards reducing world poverty.
However, there are some problems with using targets to make develop-
ment agencies accountable for their performance (see also Chapter 3 for
an elaboration of such problems). First, there is the issue of how simple or
complex the target is. A simple target – eliminating poverty, for example –
is easy for the public to understand, but correspondingly difficult to
achieve. Yet as soon as it is made more concrete, it also becomes more
complex, such that there is no longer any single ‘bottom line’ on which a
government or agency can be measured. This links to a second point, con-
cerning the discrepancy between the wide range of activities across which
the public wishes a government or government agency to perform, and
the narrow range of remedies the public has in holding a government or
one of its agencies to account.
In a company, for example, it is possible for shareholders to remove an
underperforming director, or require an existing board to modify its
approach in certain ways following an annual meeting. In contrast, voters
have only one direct sanction against underperformance of government
ministers or ministries – to sack the entire government – and even then,
usually have the chance to use that sanction only once every four or five
years. This begs the question of how voters (or others) might hold a
government or development agency accountable for failure to meet one
or more of the international development targets. It is not generally pos-
sible to vote out an international development minister whilst leaving her
government in place. Nor, given the twenty-year lifespan of the inter-
national development targets, is it likely that ministers or officials respons-
ible for setting the targets (and committing themselves to achieving them)
will still be around at the time when it is clear whether or not they have
been achieved.
Of course, some progress can be made in this direction. For example, it
is possible to tell already whether certain individual development targets
are likely to be achieved – indeed, that is one purpose of this book. More-
over, some indirect pressure could be brought to bear on development
agencies, for example via the media, if progress is not being achieved.
There might be calls for the resignation of a minister or senior officials
who were seen as falling short of stated objectives. However, even if it were
possible to hold individuals or departments accountable in this way, this
still begs a number of important questions, notably the question of to
whom such individuals or departments should be accountable. In the case
of aid donors, accountability is formally to ‘Northern’ electorates, which
are not directly affected by whether development targets are met or not.
In the case of international organisations such as the UN, the lines of
accountability are less clear. In neither case is there direct accountability
to the beneficiaries of development aid.
14 Howard White and Richard Black
However, this point does lead to a third argument for setting inter-
national development targets, which is that they can be linked to a rights-
based approach. Rights-based approaches – increasingly popular with
many development agencies – focus not so much on people’s needs, but
more on what they have a right to expect – both in terms of basic human
rights such as the right to life, food, water, shelter, etc., and in terms of
their right to have their views represented to agencies that have an impact
on their lives (DFID, 2000; Crook, 2001). If governments sign up to the
targets, this means in principle that the people whose interests they are
meant to serve can lobby them to behave in the manner most consistent
with meeting the targets. In this conception, ‘beneficiaries’ of aid have a
right to expect that agencies will act fairly in pursuing stated objectives.
In the parallel field of international humanitarian assistance, a rights-
based approach stresses how people’s rights as defined by international
humanitarian law have been violated, and sees international intervention
as helping people to regain those rights. This positions the beneficiaries
of aid as claimants, rather than as beggars (Slim, 2001). Nonetheless, it
should be remembered that UN conference resolutions do not have legal
status. In this sense, for a rights-based argument to be fully convincing in
terms of the international development targets, ‘Southern’ governments
need to reflect their commitment to the targets by passing relevant legisla-
tion (e.g. making school attendance compulsory).
A further point in favour of targets is that where it is possible to disag-
gregate performance either by service provider or for different areas of
provision, then standardised performance targets allow comparisons to be
made. One clear example of this is in the compilation of league tables for
schools, universities, hospitals and social services departments in the UK.
This is seen as helping those using services to make informed choices
about where the best available provision can be found, as well as encour-
aging those using the services of ‘underperforming’ providers to put pres-
sure on the school or hospital in question to improve. Such tables are also
used to influence the provision of public money, whether through
‘rewarding’ the high achievers, or through the provision of targeted
resources to turn round those that are seen as falling below the expected
standard.
In the case of the International Development Targets and Millennium
Development Goals, the relative performance of countries and regions
can also be compared through such standardised tables. However, it is less
clear what conclusion should be drawn from this comparison. Specifically,
if a country is falling short of achieving the outcome target, should it
receive more assistance or less? Alternatively, should failure to make
progress lead to the provision of assistance of a different kind? In part, the
answer depends on why the country is ‘off track’, which hints at the fact
that outcome-based measures do not in themselves constitute an adequate
basis for performance measurement. This point is argued at more length
 Millennium Development Goals 15
in Chapter 3. At the same time, the ranking of countries in this way reveals
another confusion about targets, namely: who is responsible for achieving
them?
For example, the argument so far has implied that the targets are essen-
tially about measuring and monitoring the performance of development
agencies – especially UN agencies and the development cooperation min-
istries of Northern governments. However, the ranking of countries in
terms of their progress towards the targets implies that countries them-
selves, and especially the governments of (Southern) countries, are the
ones expected to meet them. In this sense, although publication of
‘league tables’ of countries’ progress towards targets might lead to the cit-
izens of these countries questioning why they have not made more
progress, there are limits on what they can do about it.
Interestingly, a similar argument can be made within the UK in relation
to league tables for schools and hospitals. For example, parents or
patients might question why their local school or hospital is underper-
forming, but they have relatively few avenues through which they can
promote improved performance, not least because they do not hold the
purse strings. As a result, the standard reaction is, for those who are able
to do so, to choose a different school or hospital, perhaps by moving to a
different area, effectively worsening the situation for those (mainly poorer
households) who are unable to move. In a world of increased possibilities
for mobility, it is arguably possible for wealthier individuals in developing
countries to have exactly the same reaction, with their migration to the
North leading to similar consequences.
Looked at in general, targets can have a positive motivational role both
within and across organisations. As a prominent supporter of the Inter-
national Development Targets, DFID has posters that set out each target
prominently displayed in its offices worldwide. Where several develop-
ment organisations are working in the same area, then targets can create a
sense of common purpose across these agencies. Setting targets can be
argued to have played the role of increasing the cohesiveness of the
international development community, helping to focus the renewed
attention to poverty in the 1990s. However, these targets are arguably
much less motivating for Southern governments, especially those that are
‘left behind’.
Comparison with performance targets in the UK also reveals a number
of other weaknesses and limitations of the approach of targeting develop-
ment. For example, focus on a single indicator can risk distorting pro-
grammes or sacrificing quality. Thus targets to reduce health service
waiting lists in the UK are claimed by some to have been achieved by pri-
oritising quicker and cheaper operations. The target set for a loan scheme
in London to minimise default encouraged rescheduling of ultimately
unpayable debts (Jackson, 2000). Meanwhile, targets for local government
authorities to collect more recyclable waste have sometimes been met, but
16 Howard White and Richard Black
with the consequence that these authorities have then been forced to
burn or bury the waste, as recycling capacity has been exceeded (NAO,
2001).
Similar concerns apply to some of the Millennium Development Goals.
For example, with respect to the goal of universal primary education,
getting more children into already overcrowded classrooms, with few
materials and poorly motivated teachers, might meet the target for quan-
tity, but only at the expense of the quality of education provided to the
children. To take another example, achieving the target of having a
national strategy for sustainable development does not mean that this
strategy is implemented. Indeed, the resources put into developing a
formal strategy may well divert resources from existing activities that are
actually promoting sustainability.
Moreover, as noted above, performance targets may also tend to sim-
plify problems, focusing on what is identifiable and measurable whilst
ignoring what really matters but is more complex. This is an argument
that will resound with those working in poverty analysis, where the com-
plexity and multidimensionality of poverty is well understood (see, for
example, Jodha, 1988; Chambers, 1997).8 Even if a target is well identified,
it may have adverse organisational effects by discouraging innovation.
Managers are likely to rely on tried and tested methods where there is a
clear target to be achieved, rather than risk missing that target by trying
out a new approach. Targets may also undermine other forms of account-
ability. In the bid to satisfy performance criteria, particular problems or
interests may be missed. Finally, the focus on outcomes may deflect atten-
tion from the costs borne in achieving those outcomes, so that the effi-
ciency focus of traditional management systems is lost (Cummings, 1997).

Conclusion and structure of the book

This chapter has argued that International Development Targets, launched
in 1996, have caught the attention of the development community, atten-
tion which has increased with the Millennium Development Goals. Unlike
previous targets, they have not (as yet) fallen by the wayside, but continue
to be referred to and monitored. Their dominance is explained partly by
the renewed focus on poverty, partly by the rise of results-based manage-
ment and partly by the support they have received from key actors includ-
ing the UK government. Targets can play an important role in
accountability and performance measurement, but they are not without
their disadvantages.
Chief amongst these disadvantages is a continued lack of clarity about
whose targets the International Development Targets and Millennium
Development Goals actually are. The emergence of a ‘rights-based’ agenda
for development stresses how the recipients of international aid can use
clear objectives to argue for their right to be dealt with in accordance with
 Millennium Development Goals 17
these objectives. They, and others, can, in principle, hold development
actors accountable for their actions, perhaps for the first time. But who
are ‘development actors’, and to whom are they supposed to be account-
able? Moreover, why should development actors be accountable on the
basis of a series of targets that they have devised from the ‘top down’,
rather than on the basis of criteria identified from the bottom up?
What is missing from discussion of targets for international develop-
ment agencies is any theory of accountability. For real accountability, at
the very least there needs to be more transparency as to who is responsible
for what, and more ownership of goals by those expected to meet them. In
this sense, scale is also important. Individuals and agencies need to be
held accountable for targets that are realistic and achievable at the level at
which they are working. It would be ridiculous for an individual develop-
ment worker to be held ‘accountable’ for not eradicating poverty; but it is
equally ridiculous to hold the North responsible for ensuring that every
child in Bangladesh goes to school.
Here, it is not particularly important that a range of different actors
endorse the principle of setting targets for development, or even the sub-
stance of what each individual target is. If endorsement of the approach is
all that happens, the likelihood is that targets simply bureaucratise further
the business of development, making it more remote from ordinary
people in both North and South. Rather, goals and targets need to be
something that can inspire both those working within international devel-
opment agencies, and those outside who might call aid workers to account
for their actions.
The remainder of this book is structured in two parts. Chapters 2–6
explore a number of issues relating to the setting of targets for develop-
ment, whilst Chapters 7–14 focus on individual targets, goals or sets of
targets, to assess what each target is, why it has been set, what progress has
been made towards meeting it, and what the prospects are for its being
achieved.
In Chapter 2, Simon Maxwell shows how the setting of international
targets for development forms part of the ‘new New Poverty Agenda’
which emerged in the late 1990s. This agenda has five elements: the
targets themselves; a revised strategy for poverty reduction as laid out in
the World Bank’s World Development Report, 2000/2001; Poverty Reduction
Strategy Papers (PRSPs); new tools in the form of Medium-Term Expendi-
ture Frameworks (MTEFs), sector-wide approaches (SWAPs) and Poverty
Reduction Strategy Credits (PRSCs); and a stress on performance manage-
ment. Providing a simple guide to this new maze of abbreviations, the
chapter argues that the new agenda is indeed ‘new’ and different
from what went before, but nonetheless needs to be applied flexibly. Six
risks are identified in the new way in which poverty has been constructed,
followed by six principles that could guide a more flexible approach: (1)
the need for ‘subsidiarity’; (2) the need to focus on essentials; (3) the
18 Howard White and Richard Black
importance of not forgetting difficult sectors and cross-cutting issues; (4)
the need to recognise the political nature of poverty; (5) the importance
of partnership; and (6) the need for a process-based approach, which
accepts that achievements will take time to materialise.
Next, in Chapter 3, Howard White considers further the question of
whether it makes sense to use global targets as a measure of development
performance. The chapter sets out a number of criteria that could be used
for the assessment of performance, and considers how the International
Development Targets and Millennium Development Goals match up
against these. To be most effective, performance measures should satisfy a
number of criteria. White argues that the International Development
Targets and Millennium Development Goals satisfy only some of these cri-
teria. They are relevant, mostly well defined and correspond to existing
indicators. But they are also mostly outcome oriented, with little effort
made to build a consensus around an underlying logical model of how the
targets are to be achieved. In the absence of such a model it is extremely
difficult to say anything sensible in terms of attributing changes in target
indicators to the actions of the development community.
Some donors, notably DFID, have tried to bridge this gap between
development actions and desired outcomes through the preparation of
‘Target Strategy Papers’ (TSPs) and ‘Country Strategy Papers’ (CSPs). In
these documents, DFID has tried to set out how it will achieve its targets,
especially the reduction of poverty. In Chapter 4, Howard White and
David Booth examine how CSPs analyse poverty and what the implications
are for the strategies used to halve world poverty. However, Booth and
White argue that the problem of the ‘missing middle’ remains; there con-
tinues to be a gap between the causes of poverty identified in the poverty
profile and the proposed interventions.
In Chapter 5, David Booth and Henry Lucas shift attention to the ques-
tion of how progress towards the Millennium Development Goals can be
accurately monitored by individual countries. Again the focus is on PRSPs,
and the extent to which these have involved adequate procedures and
monitoring systems to assess effectively whether poverty is being reduced.
A review of existing PRSPs and Interim Poverty Reduction Strategy Papers
in Africa suggests that a lot remains to be done in terms of improving the
quality of recording of administrative data. In response, the chapter seeks
to propose some practical ways forward, and highlights areas in which
further work is needed on the strategic selection of indicators.
Despite the limitations in approach and monitoring systems high-
lighted so far, a book on progress towards the Millennium Development
Goals would hardly be complete without an overall review of whether the
goals themselves are feasible. This is provided in Chapter 6 by Jan Vander-
moortele, who seeks to go beyond averages and aggregates to consider the
unevenness of progress towards different goals, in different places and for
different groups in society. Vandermoortele argues that the Millennium
 Millennium Development Goals 19
Development Goals are both technically feasible and financially afford-
able, and that most will be met by at least some countries. Nonetheless,
much more committed leadership, stronger partnerships, extra money
and deeper participation by the poor will be needed to ensure that they
are met overall.
In the second half of the book, attention turns to individual targets,
goals and sets of goals. In Chapter 7, Michael Lipton and Hugh Wadding-
ton address the target of reducing income poverty by half by 2015. They
argue that global trends suggest considerable progress towards this target.
However, a closer look at regional patterns suggests a more pessimistic
view. Most of the progress at the global level has been fuelled by progress
in East Asia, particularly China. This region simply cannot sustain the
same contribution to the target: doing so would require more people to
be brought out of poverty by 2015 than are actually poor at the moment!
Elsewhere, Lipton and Waddington explore the obstacles to poverty
reduction, notably the absence of growth and the persistence of inequal-
ity. They ask what factors underlie these constraints, and argue that to
achieve accelerated poverty reduction, we must address the underlying
fundamentals, tackling issues such as asset distribution and the spread of
new technologies.
The Universal Declaration of Human Rights, adopted by the United
Nations in 1948, stated that primary schooling should be free and compul-
sory in all nations. More than fifty years later, the world still has far to go
in order to achieve this goal. Yet in Chapter 8, Christopher Colclough
argues that the recent reaffirmation of the aim of achieving schooling for
all by 2015, and of gender equity in schooling by 2005, again risks failure,
especially in the poorest regions of the world. The challenge is greatest in
the countries of sub-Saharan Africa, where scarcely half the eligible chil-
dren were attending primary school at the end of the twentieth century.
Across sub-Saharan Africa as a whole, the proportion of children enrolled
was no higher than it had been in 1980. Girls were hardest hit, with 20 per
cent fewer girls attending school than boys.
Colclough examines some of the main causes of this disappointing
progress. Costs are important constraints on school participation – at both
macro and household levels. But commitment is also found to be lacking
in some countries, where higher levels of spending on schooling than
presently exist could easily be afforded. Where household demand is low,
girls tend to be last to be enrolled. Policies need to address not only
supply-side changes, but also measures to support and stimulate demand.
In Chapter 9, Ramya Subrahmanian addresses progress towards gender
equality, focusing her attention not only on the specific gender goals, but
on how gender issues permeate other targets. She argues that feminist lob-
bying and activism have successfully put gender onto the international
agenda through strategic advocacy at international conferences and a
push for greater inter-agency collaboration. As a result, several targets and
20 Howard White and Richard Black
goals have a gender component – including commitments to remove
gender disparities in education, reduce maternal mortality by three-quar-
ters, and ensure reproductive health for all. But do the Millennium
Development Goals really promote a discourse of gender equality and
empowerment? And are the new policy instruments and approaches
framed to meet the targets working in favour of gender equality? Subrah-
manian argues that the ways in which concepts of gender equality and
empowerment are used by different agencies and brought into the
framing of targets present a double-edged sword: while gender equality is
given rhetorical prominence, this is done in ways that run the risk of
depoliticising what are in essence issues of rights and redistribution.
In Chapter 10, Howard White addresses what he argues is the most
important development challenge embodied in the development targets:
ending avoidable infant and child deaths. Although there has been
progress in reducing mortality, this has slowed in recent years, especially
in the worst-affected areas (notably sub-Saharan Africa). This reversal is
partly because of HIV/AIDS and conflict, but also reflects the impact of
long-run economic decline and the resulting deterioration in health
service provision. Current trends suggest that the targets will not be met.
Yet it is not clear how the international community could or should
respond. Some argue that reduction in infant mortality is largely driven by
growth – if only poverty can be eradicated, infant and child mortality will
fall. For others, there is scope for health provision to improve things.
White argues for the latter view, suggesting that it is indeed possible for
the target to be met with a sufficient push and resources. Relevant policies
include universal immunisation – even though immunisation rates fell in
Africa in the second part of the 1990s.
In Chapter 11, Hilary Standing addresses the goal of providing access
for all to reproductive health, which was included as an International
Development Target but removed from the final list of Millennium Devel-
opment Goals. The target is placed in the context of a range of reproduc-
tive health indicators. Discussion then focuses on how reproductive heath
itself has come to prominence in international development discourses,
especially following the Cairo Conference on Population and Develop-
ment Goals, and what are the key obstacles to improving reproductive
health. A major problem is funding, but the goal of providing reproduc-
tive health for all is often tied up with wider health reforms taking place in
most low- and middle-income countries. Standing argues that it is unlikely
that the original International Development Target for reproductive
health will be met, but that nonetheless, the Cairo agenda for reproduc-
tive health provides an innovative and important way forward.
Like the targets for gender inequality and child and infant mortality,
the target to halt and reverse the spread of HIV/AIDS can also be seen as
underpinning a number of other targets embodied in the Millennium
Development Goals. In Chapter 12, Ronald Skeldon considers whether
 Millennium Development Goals 21
the specific target of a 25 per cent reduction in HIV/AIDS amongst 15- to
24-year-olds in the worst-affected countries by 2005, and globally by 2010,
might be achieved. He argues that progress has been made in many coun-
tries, but that global progress is more difficult, given the different stage of
the epidemic in different countries. Recent shifts in global geopolitical
concerns have made galvanising the international community on this issue
more and more problematic, even though reducing HIV/AIDS has been
maintained as a Millennium Development Goal after failing to achieve the
initial list of International Development Targets.
The next two chapters turn attention to targets in the field of the
environment and sustainable development. Whilst the International Devel-
opment Targets were quite specific from the start in their objectives on
poverty, health and education, the environment has provided a more diffi-
cult field to develop targets than others. Here, the UN’s Millennium
Development Goals are more specific than the OECD’s International
Development Targets: they suggest we should aim to halve the proportion
of people in the world without access to safe drinking water. In Chapter
13, Richard Jolly reviews progress towards meeting this goal, stressing that
much remains to be done if safe and clean drinking water for all is to
become a reality. He also highlights the need for a parallel indicator on
hygiene and sanitation, an issue that was a major area of debate at the
Earth Summit in Johannesburg in August and September 2002.
Then, in Chapter 14, James Fairhead considers the earlier target of
ensuring that each country has a national sustainable development strat-
egy, in the process of implementation, by 2005, as well as exploring more
generally the ethics of setting such a target. In an argument that could be
applied more broadly to the Millennium Development Goals as a whole,
Fairhead questions an approach that requires Southern countries to
adjust their policies to meet a target that deals with a problem that they
could be seen as having little or no responsibility for creating. He also
focuses on the broader need for transparency at national and inter-
national levels if targets for ‘sustainability’ as well as in specific sectors are
to be met.
In the final two chapters we turn to the rather complex goal of ‘devel-
oping a global partnership for development’. Unlike the other Millen-
nium Development Goals, this goal encompasses some eight individual
targets, yet none of these is quantified in the same way as the majority of
the other targets, making analysis of progress towards them somewhat
more difficult. Rather than seeking to measure progress, we therefore
include two chapters which look at how and why ‘partnership’ and the
associated concerns of aid, trade and debt have come to be included in
the Millennium Development Goals, and some of the questions raised by
this inclusion. In Chapter 15, Peter Clarke draws on an analysis of the
public statements of DFID to question whether a ‘global partnership for
development’ is either possible or meaningful. Clarke uses the example of
22 Howard White and Richard Black
the unequal power relations between Northern and Southern govern-
ments to illustrate how the discourse of partnership could be viewed as a
‘political technology’ which depoliticises development and creates a
veneer of ownership on the part of less powerful groups. Such an outcome
is not inevitable, however, with Clarke suggesting that it is at a local level
that we need to look for contestation and negotiation over a more
genuine ‘partnership’.
Finally, in Chapter 16, Kevin Watkins and Juliana Amadi argues that for
a genuine global partnership for development, Northern countries need
to take more decisive action in opening trade to developing countries
(including the commodities they mainly produce), increasing develop-
ment finance and reducing the burden of debt. Although aid spending
has recently risen in some countries, five of the G-7 donors have failed to
reverse a long-term decline in aid, whilst the sustainability of debt relief
can also be called into question.
Overall, the chapters of this book do not paint a rosy picture of
progress towards the International Development Targets or Millennium
Development Goals, and in some cases take a quite critical stance on the
process of their development and implementation. Nonetheless, we would
argue that the Targets and Goals do represent a valid attempt to define
the purpose of work by both international development agencies and
national governments. The UK government in particular has spearheaded
this approach in a range of international forums, and deserves recogni-
tion for its attempts to make aid more accountable. By focusing on a series
of key indicators, the targets have helped to put the fight against poverty
into the mainstream of development activity, whilst resisting the tempta-
tion to define poverty too narrowly.
Nonetheless, problems remain. First, it is clear from individual chapters
that many of the Millennium Development Goals remain quite unlikely to
be reached by their target date, if at all. Second, it is not always clear how
the activities of aid agencies might have contributed to whether they are
reached or not. Worst of all, there is a danger that the targets themselves
will not make agencies more accountable, but simply more bureaucratic.
Where should international development agencies go from here? Three
immediate options present themselves in relation to the targets themselves.
One option would be to put the best gloss possible on progress towards
reaching the targets, although this could be regarded simply as political
‘spin’. A second option would be to refine the targets, so that they are
more meaningful, or more achievable, although this could be seen as
‘fixing’ the targets. A third option might be to scrap the targets altogether,
although this could easily be seen as defeatism.
We do not advocate any of these steps, but rather set ourselves the
more modest aim of understanding the targets – including why they have
been set, and what might need to occur if some or all of them are to be
achieved. It is also important to keep in mind that the setting of targets is
 Millennium Development Goals 23
only a means to an end, rather than an end in itself. In this regard, the
real goal remains much broader: the eradication of poverty, no less. The
setting of a series of development goals is one step towards this goal.

Notes
1 For example, the second goal is to halve hunger but the benchmark to define
hunger is not given. Goals 12 to 18 do not include a quantified target.
2 Downloadable from [Link].
3 The process by which this outcome was reached has not been documented, but
DFID claims to have worked toward this end, i.e. that the Millennium Develop-
ment Goals should not conflict with the International Development Targets.
The fact that they do not has made it relatively straightforward for donors to
switch from one set of targets to the other.
4 The text of the Act is available on [Link]
[Link].
5 The US agency USAID also has extensive experience, which is drawn on in
Chapter 3. A more comprehensive review is DAC (2000).
6 Formerly known as Form 590.
7 See National Audit Office (2001: appendix 2) for a chronology of performance
measurement in UK central government.

References
Baulch, R. (1996) ‘The New Poverty Agenda – a Disputed Consensus’, IDS Bulletin,
27(1): 1–10.
Brown, B. (1995) Results-Based Management: Danger or Opportunity. Lessons from the
CIDA Experience, Ottawa: CIDA.
Chambers, R. (1997) Putting the Last First, London: IT Publications.
Crook, R. (2001) ‘Editorial Introduction’, IDS Bulletin, 32(1), 1–7, Brighton: Insti-
tute of Development Studies.
Cummings, F.H. (1997) ‘Logic Models, Logical Frameworks and Results-Based
Management: Contrasts and Comparisons’, Canadian Journal of Development
Studies, 18: 587–596.
Denery, L. and Walton, M. (1999) ‘Are the Poverty and Social Goals for the 21st
Century Attainable?’ IDS Bulletin, 30(2): 75–91.
Department for International Development (DFID) (2000) Realising Human Rights
for Poor People: Strategies for Achieving the International Development Targets, London:
DFID.
Development Assistance Committee (DAC) (1996) Shaping the Twenty-first Century,
Paris: OECD.
—— (2000) ‘Results Based Management in the Development Co-operation Agen-
cies: A Review of Experience (Executive Summary)’, Paris: OECD.
Jackson, A. (2000) ‘An Evaluation of Evaluation: Problems with Performance Mea-
surement in Small Business Loans and Grant Schemes’, Progress in Planning, 55:
1–55.
Jodha, N.S. (1988) ‘Poverty Debate in India: a Minority View’, Economic and Polit-
ical Weekly, 22(45–47): 2421–2428.
National Audit Office (NAO) (2001) Measuring the Performance of Government Depart-
ments, London: The Stationery Office.
24 Howard White and Richard Black
Organisation for Economic Cooperation and Development (OECD) (1996)
Shaping the 21st Century: The Contribution of Development Co-operation, Paris: OECD.
Rawnsley, A. (2001) Servants of the People: The Inside Story of New Labour, London:
Penguin.
Slim, H. (2001) ‘Not Philanthropy but Rights: Rights-Based Humanitarianism and
the Proper Politicisation of Humanitarian Philosophy in War’, Humanitarian
Practice Network Conference Paper, London: Overseas Development Institute,
February.
United Nations (UN) (2000) United Nations Millennium Declaration, Document
A/RES/55/2, New York: UN ([Link]
[Link]).
—— (2001) Road Map towards the Implementation of the United Nations Millennium
Declaration: Report of the Secretary-General, Document A/56/326, New York: UN.
White, H. (1999) ‘Global Poverty Reduction: Are We Heading in the Right Direc-
tion’, Journal of International Development, 11(4): 503–519.
2 Heaven or hubris
Reflections on the new ‘New
Poverty Agenda’1
Simon Maxwell

Oh! pleasant exercise of hope and joy!

For mighty were the auxiliars which then stood
Upon our side, we who were strong in poverty reduction!
Bliss was it in that dawn to be alive
But to be a poverty planner was very heaven!
(after William Wordsworth)

Introduction
Wordsworth, above and as amended, has it absolutely right: poverty plan-
ning is on a roll. This is because a new construction has been put in place
that locks together five elements of new thinking about the subject. The
five elements are:

• the Millennium Development Goals, with poverty reduction at their

heart;
• international consensus on how to reduce poverty, best summarised
in the World Bank’s World Development Report for 2000/2001, ‘Attack-
ing Poverty’;
• a mechanism for operationalising the strategy at country level, in
Poverty Reduction Strategy Papers;
• technologies for delivering aid in support of Poverty Reduction Strat-
egy Papers, notably Medium Term Expenditure Frameworks, sector-
wide approaches and Poverty Reduction Strategy Credits;
• underpinning the other four, a commitment to results-based manage-
ment.

There is no doubt that this construction is superior to previous attempts at

housing a new approach to poverty – for example, the initiative on poverty
reduction that followed the publication in 1990 of an earlier World Develop-
ment Report on the subject, described at the time as defining a New Poverty
Agenda (Lipton and Maxwell, 1992). That had a strategy, and a degree of
operationalisation (e.g. through the World Bank’s Poverty Handbook and
26 Simon Maxwell
Operational Directive (World Bank, 1992a, b)), but lacked the targets, the
aid modalities and the emphasis on results-based management.
It is too early to say whether the current and new ‘New Poverty Agenda’
will deliver the expected results in terms of real poverty reduction for real
people in developing countries. On present trends, as we see in this
volume, many of the Millennium Development Goals will not be met. We
should not, however, be churlish; as Wordsworth remarks, this is a new
strategy which

The beauty wore of promise, that which sets

The budding rose above the rose full blown.

True enough, but there are risks associated with this budding rose, and
this chapter is largely about these risks. They lie in the realms of eco-
nomics, politics and public administration. They include issues to do with
equity, consensus-building, partnership, and managing change by simple
targets. They are not insignificant risks, but nor are they immutable, and
this is the point: by being clear about the risks, we can modify strategies
and take action to maximise the probability of success. Not to do so is to
fall into the trap of hubris, of pride before a fall.

Poverty reduction: the new construction

There is much to commend in the new approach to poverty reduction.

The Millennium Development Goals

The Millennium Development Goals are summarised in Chapter 1. They
represent the latest manifestation of the International Development Targets,
agreed (mostly) at United Nations (UN) conferences in the early 1990s
and codified in 1996 by the Development Assistance Committee (DAC) of
the Organisation for Economic Cooperation and Development (OECD).
Yet progress against the targets is uneven. For the key poverty reduction
target, the data (Figure 2.1) show that developing countries taken as a
whole are more or less on track to halve the proportion of those in
absolute poverty by 2015, but that this is not true for Europe and Central
Asia, the Middle East and North Africa, South Asia, and (most markedly)
sub-Saharan Africa. In sub-Saharan Africa, the proportion of people living
below an income level of one dollar a day is rising, not falling. Global
success owes a great deal to rapid progress in poverty reduction in East
Asia and the Pacific: here, the proportion fell sharply, and the number in
absolute poverty declined during the 1990s, by 192 million people.
We will come later to some risks associated with using targets to drive
policy, but their value in clarifying policy and in providing a political
framework deserves to be noted. In particular, the targets have provided
 Reflections on the new ‘New Poverty Agenda’ 27

Poverty rate East Asia and Europe and Latin America and
(% below $1.09) the Pacific Central Asia the Caribbean
1990 30 4 20
1999 25 3
Goal 2015 15
Progress Made 20 2

Source: Global Economic Prospects

Rate of progress 10
15 1
needed to meet goals 10 0 5
1990 2015 1990 2015 1990 2015

Middle East and

North Africa South Asia Sub-Saharan Africa Developing Countries
3 50 50 30

25
2 40 40
20
1 30 30 15
0 20 20 10
1990 2015 1990 2015 1990 2015 1990 2015

Figure 2.1 Progress towards the Millennium Development Poverty Goals.

Source: World Bank website: [Link]

political impetus to poverty reduction efforts. As Wordsworth rightly

observes,

What temper at the prospect did not wake

To happiness unthought of? The inert
Were roused, and lively natures rapt away!

A strategy to reduce poverty

The World Development Report (WDR) 2000/2001 laid out a strategy for
reducing poverty, resting on three legs: opportunity (meaning growth),
empowerment and security (World Bank, 2000). Other overviews of policy
to reduce poverty, such as the DAC Poverty Guidelines (DAC, 2001), cover
similar territory. There are, as we shall see, some rough edges, but the
World Development Report strategy comes close to capturing an international
consensus. It has some notable innovations, at least for the World Bank.
These have been summarised as follows (Maxwell, 2001c: 144–145):

• Methodological innovation, in the shape of participatory poverty

assessments in sixty countries, collectively articulating the ‘Voices of
the Poor’ (see Narayan et al., 2000).
• Partly inspired by ‘Voices’, but also drawing on an extensive literature,
explicit adoption of a multidimensional model of poverty, which sets
28 Simon Maxwell
low incomes alongside access to health and education, vulnerability to
shocks, and, importantly, voicelessness and powerlessness.
• Emphasis on the value of growth in reducing poverty, as one might
expect, but also considerable emphasis on redistribution. This is,
admittedly, largely for instrumental reasons, rather than as an end in
itself. It represents a way of raising the poverty elasticity, but is also
included because more equal societies grow faster.
• Predictably again, a commitment to markets and to openness,
but strong statements throughout about the dislocations associated
with market reform and market processes, the need for strong and
prior institutional underpinning of markets, and the ‘obligation’
(World Bank, 2000: 76) to protect losers and those excluded from the
benefits.
• Empowerment as a major theme – not just participation in a narrow
sense, but a focus on making state institutions responsive to poor
people and on building social institutions.
• Security promoted from being half a leg (as in the 1990 World Develop-
ment Report) to a full leg, with a typology of risks, and a review of miti-
gation, coping and response strategies for natural disasters, economic
crises, and many kinds of idiosyncratic risk facing individuals.
• Finally, a discussion of international actions around the core themes
of opportunity, empowerment and security, which touches on well-
worn themes (protectionism in the North, debt relief, more and
better aid), but also recognises why countries are nervous about
capital market liberalisation and calls for democratisation of global
governance institutions.

It is important to say about the WDR strategy that many of its elements
could be found in earlier efforts to synthesise an overall approach to
poverty reduction. This includes earlier WDRs on the subject, in 1980 and
1990, but most notably a series of United Nations Development Pro-
gramme (UNDP) Human Development Reports (HDRs) launched in 1990,
especially the 1997 report, which dealt specifically with poverty (UNDP,
1997). Thus the 1990 HDR introduced the multidimensional model of
poverty in its exploration of ‘human development’; it also talked about
the need for greater participation and greater equity, and emphasised the
importance of social subsidies (UNDP, 1990). The 1997 HDR called for
greater accountability in government and urged that globalisation be
managed to protect and benefit the poor.
That the World Bank should assume part of UNDP’s mantle on poverty
reduction is, of course, a cause of wonder; but it is also cause for congratu-
lation. The WDR strategy marks the closest we have yet come to an inter-
national consensus on poverty reduction. The recent DAC Guidelines on
Poverty Reduction reflect this on the donor side. In an approach consistent
with that of the WDR, they adopt a capabilities approach to understanding
 Reflections on the new ‘New Poverty Agenda’ 29
poverty, incorporating ideas about influence, freedom, status and dignity,
as well as income and assets; and they cover similar policy areas, including
growth, empowerment, social services and social protection (DAC, 2001:
37ff.). Certainly, international discourse has come a long way in the past
decade.

Poverty Reduction Strategy Papers

Poverty Reduction Strategy Papers (PRSPs) were introduced in the
context of the enhanced settlement for debt relief in heavily indebted
poor countries (HIPCs) in 1999 (DFID, 2001; Gunter, 2002). Their key
features are that they are:

• in principle, country owned and led;

• based on a participatory process, leading as far as possible to a
national consensus;
• accepted by donors on the basis of conditionality on process rather
than substance;
• designed to evolve over time, providing ‘road maps’ rather than blue-
prints for poverty reduction.

In order to qualify for relief under this process, countries must have
at least an Interim Poverty Reduction Strategy Paper (I-PRSP). Full relief
is dependent on having a full PRSP. By mid-2002 there were fifty-nine
countries involved in the PRSP process. Of these, fourteen had completed
a full PRSP that had been endorsed by the boards of the World
Bank and the International Monetary Fund (IMF), six had completed a
full PRSP that had yet to be submitted, twenty-nine had completed an
I-PRSP, and ten were yet to complete an Interim PRSP (PRSP connec-
tions, 2002).2

More abbreviations: MTEFs, SWAPs and PRSCs

Various ‘technologies’ have been put in place to help improve the
management of public expenditure and aid flows. Governments are
encouraged to produce Medium Term Expenditure Frameworks (MTEFs).
Donors are encouraged to work together in support of sector-wide
approaches (SWAPs). And the World Bank, in particular, is expected to
replace structural adjustment lending with new Poverty Reduction Strat-
egy Credits (PRSCs).
Sector-wide approaches are a particularly important feature of the new
apparatus. As described by Foster, the defining characteristic of a SWAP is
that ‘all significant funding for a sector supports a single sector policy
and expenditure programme, under Government leadership, adopting
common approaches across the sector, and progressing towards relying on
30 Simon Maxwell
Government procedures to disburse and account for all funds’ (Foster,
2000: 9). The stress on government leadership in SWAPs should reinforce
partnership and accountability.

Results-based management
Results-based management (RBM) can be characterised as an approach to
managing organisations and programmes by focusing on outputs and out-
comes rather than inputs and activities: reduction in poverty, say, rather
than expenditure on poverty reduction programmes. A good example of
its use is in public expenditure management in the UK, where the Treas-
ury negotiates output-oriented Public Service Agreements (PSAs) or con-
tracts with individual ministries. The PSA for the DFID for the period
2002–2006 is reproduced by way of illustration in Box 2.1. It contains six
objectives, mostly specifying the outcomes expected, along with specific
targets for poverty reduction or other elements of the Millennium Devel-
opment Goals.
It is important to note that RBM is not just used for setting targets. It
can also be used to structure rewards, for individuals, teams, organisa-
tions, local government departments and the like. The way in which DFID
has used the Millennium Development Goals as the basis for judging its
results is discussed in Chapter 3.

Six risks with the new construction

The new construction on poverty reduction has undoubted strengths.
Here, however, are six risks, which cut across the various levels of the new
construction. We will come later to remedies.

Targets and performance indicators may oversimplify and distort

development efforts
The use of targets is central to the new construction. They are the main
feature, of course, of the Millennium Development Goals; but they also
feature strongly in PRSPs, which are about how to reach goals at country
level, in SWAPs, which tie public expenditure to targets, and in results-
based management.
We know why targets are useful. They clarify objectives. They rally
support. And they provide an instrument with which to reform public ser-
vices. These are valuable benefits. But we also know why targets pose risks:
they can encourage a reductionist approach to complex problems, privi-
lege quantitative indicators at the expense of qualitative indicators, distort
resource allocation, and undermine professional motivation and responsi-
bility (Maxwell, 1998).
Many examples have been cited in the literature, in both developed
 Reflections on the new ‘New Poverty Agenda’ 31
Box 2.1 Department for International Development Public Service Agreement,
2002–2006: objectives and performance targets

Objective I: Reduce poverty in sub-Saharan Africa

1 Progress towards the Millennium Development Goals in sixteen key
countries demonstrated by:
• A sustainable reduction in the proportion of people living in
poverty from 48 per cent across the entire region.
• An increase in primary school enrolment from 58 per cent to 72
per cent and an increase in the ratio of girls to boys enrolled in
primary school from 89 per cent to 96 per cent.
• A reduction in under-five mortality rates for girls and boys from
158 per 1,000 live births to 139 per 1,000, and an increase in the
proportion of births assisted by skilled birth attendants from 49
per cent to 67 per cent; and a reduction in the proportion of 15- to
24-year-old pregnant women with HIV from 16 per cent.
• Improved effectiveness of the UK contribution to conflict pre-
vention and management as demonstrated by a reduction in
the number of people whose lives are affected by violent conflict
and a reduction in potential sources of future conflict, where the
UK can make a significant contribution. (Joint target with Foreign
and Commonwealth Office (FCO) and the Ministry of Defence
(MoD.)
• Effective implementation of the G-8 Action Plan for Africa in
support of enhanced partnership at the regional and country
level.

Objective II: Reduce poverty in Asia

2 Progress towards the Millennium Development Goals in four key coun-
tries demonstrated by:
• A sustainable reduction in the proportion of people living in
poverty from 15 per cent to 10 per cent in East Asia and the Pacific
and 40 per cent to 32 per cent in South Asia.
• An increase in gross primary school enrolment from 95 per cent to
100 per cent and an increase in the ratio of girls to boys enrolled
in primary school from 87 per cent to 94 per cent.
• A reduction in under-five mortality rates for girls and boys from 92
per 1,000 live births to 68 per 1,000; and an increase in proportion
of births assisted by skilled birth attendants from 39 per cent to 57
per cent.
• Prevalence rates of HIV infection in vulnerable groups being below
5 per cent; and a tuberculosis case detection rate above 70 per cent
and cure treatment rate greater than 85 per cent.

continued
32 Simon Maxwell

Objective III: Reduce poverty in Europe, Central Asia, Latin America, the
Caribbean, the Middle East and North Africa
Objective IV: Increase the impact of key multilateral agencies in reducing
poverty and effective response to conflict and humanitarian crises
3 Improved effectiveness of the international system as demonstrated by:
• A greater impact of European Commission (EC) external pro-
grammes on poverty reduction, including through working for
agreement to increase the proportion of EC official development
assistance to low-income countries from 38 per cent to 70 per cent.
• Ensuring that three-quarters of all eligible heavily indebted poor
countries (HIPCs) committed to poverty reduction receive irrevo-
cable debt relief by 2006 and work with international partners to
make progress towards the United Nations 2015 Millennium Devel-
opment Goals. (Joint target with Her Majesty’s Treasury.)
4 Secure agreement by 2005 to a significant reduction in trade barriers
leading to improved trading opportunities for the UK and developing
countries. (Joint target with Department for Trade and Industry (DTI)
and FCO.)

Objective V: Develop evidence-based, innovative approaches to

international development
Objective VI: Value for money
5 Increase the proportion of DFID’s bilateral programme going to low-
income countries from 78 per cent to 90 per cent and a sustained
increase in the index of DFID’s bilateral projects evaluated as success-
ful.

Source: UK Treasury 2002, at [Link]

psa02_ch11t.pdf

and developing countries. Some of the best come from the developed
country literature: health targets set centrally as reduction of waiting lists,
which encourage doctors to treat less urgent but easy cases and neglect
more urgent but more difficult cases (Chapman, 2002); or testing regimes
for schools, again set centrally, which distort teaching priorities, and
provide incentives to teachers to manipulate results (Davies, 2000). In
international development the main debates have been about the reduc-
tionist nature of the dollar-a-day target for poverty reduction (Maxwell,
1998), but also about the difficulty of attribution, when the phenomena
with which targets are concerned are subject to many different influences
(White, 2002). To take an easy example of the latter, the UK’s DFID is
committed in its PSA to a reduction of poverty in South Asia from 40 per
 Reflections on the new ‘New Poverty Agenda’ 33
cent to 32 per cent by 2006. This is a region in which aid is relatively
insignificant, accounting for less than 3 per cent of public expenditure;
and in which donors can have relatively little influence on the prospects
for growth and poverty reduction. Hubris? Or, a cynic might argue, commit-
ment to a target which is so likely to be achieved that there is little risk of
being exposed?
It might be thought that the right reaction to these problems is to set
better targets or manage them differently, but critics argue that the prob-
lems lie deeper, in the treatment of professional standards or in the
nature of the organisations concerned. Thus Onora O’Neill (2002: 3)
rages against the undermining of professional integrity, arguing that
‘central planning may have failed in the former Soviet Union but it is alive
and well in Britain today. The new accountability culture aims at ever
more perfect administrative control of institutional and professional life.’
Chapman sets out the argument about organisations. He argues that
public service organisations are complex, adaptive systems, which can be
expected to respond poorly to centralised targets. Targets will (1) maxi-
mise the likelihood of adverse unintended consequences, (2) increase
administrative overheads, (3) make institutions more fragile, (4) demoti-
vate staff throughout the system, and (5) cause disillusion among clients.
He concludes that ‘the current approach to policy-making and implemen-
tation can be expected to fail ’ (2002: 52, original emphasis).

A preoccupation with poverty reduction may detract from the importance of

citizenship as an intrinsic component of development
Poverty reduction is a good objective, particularly if it can be interpreted
in the wider human development sense, rather than narrowly in income
terms – and if the wider interpretation can be remembered when it comes
to monitoring achievement of the relevant Millennium Development
Goal. In addition, as is discussed in Chapter 1, poverty reduction is not the
only Millennium Development Goal.
There are gaps, however. The Millennium Development Goals are
strong on material aspects of deprivation, but not so strong on non-
material aspects – which may therefore be neglected in PRSPs and SWAPs.
This is about more than the limitations of targets.
Compare the Millennium Development Goals with the DAC framework
on poverty reduction, reproduced in Figure 2.2. The boxes on economic,
human and protective dimensions broadly correspond to the Millennium
Development Goals (though protection issues are not well handled in
the Goals). The items covered in the socio-cultural and political boxes,
however, are missing from the Goals. These correspond to the empower-
ment theme of the WDR, and might broadly be interpreted as ‘cit-
izenship’ (Gaventa et al., 2002).3
One interpretation might be that rights, for example, are instrumental
34 Simon Maxwell

PROTECTIVE
Security
Vulnerability

ECONOMIC POLITICAL
Consumption Rights
GENDER
Income Influence
Assets Freedom

ENVIRONMENT

HUMAN SOCIO-
Health CULTURAL
Education Status
Nutrition Dignity

Figure 2.2 Interactive dimensions of poverty and well-being.

Source: DAC, 2001: 39.

to achieving the other aspects, and therefore of a lower priority. In this

interpretation, the achievement of income, health or education is what
matters, and action to secure achievement is merely the means to the end:
citizen action over the right to jobs, health services or schools, for
example, or initiatives to make rights available under the constitution or
through the legal system.
In fact, few would take this line. As the DAC Guidelines make clear,
rights, influence, freedom, status and dignity are all components of well-
being, part of ‘the good life’ (Christie and Nash, 1998). They deserve
attention in their own right. Moser and Norton describe this as the
‘maximum’ scenario and make the important point that the achievement
of rights is not just about legal structures, but also about social mobil-
isation. Thus, in a volume entitled To Claim Our Rights, they argue that

a growing culture of rights strengthens the degree to which indi-

viduals relate to state structures as citizens with rights and responsibil-
ities. In turn, this weakens the extent to which people expect to
extract benefits from the state through relations of clientilism and
 Reflections on the new ‘New Poverty Agenda’ 35
patronage. The citizenship model fosters the capacity for collective
action across traditional divisions of class, ethnicity and caste, thereby
increasing the capacity of social mobilisation to favour (or at least
include) the marginalised.
(Moser and Norton, 2001: 39)

This is not simply a theoretical perspective. As Norton and Elson (2002)

make clear, there are practical implications for the design of poverty pro-
grammes and public budgets.
A more controversial question is whether equality should also feature as
an objective. The treatment in the 2000/2001 World Development Report is,
as noted above, largely instrumental: lower inequality is valued because it
increases the poverty elasticity (the amount by which a given amount of
growth reduces poverty), and also because unequal societies are charac-
terised by violence and insecurity, which undermine social capital, disrupt
orderly economic life, and divert resources from productive investment
into consumption on security (World Bank, 2000: 52ff.). All this is true
(Killick, 2002; McKay, 2002; Naschold, 2002; White, 2001), but it can also
be argued that lower inequality is a necessary condition for social inclu-
sion, an intrinsic good in its own right. As I have argued elsewhere,

Intellectual foundations for this view can be found in the literature on

social exclusion (de Haan, 1998), particularly in the French variant
which is based on a solidarity paradigm of social inclusion, stressing
the importance of social bonds (Silver, 1994). There are connections,
too, to the literature on relative deprivation, in which poverty is
defined, not in absolute terms, but as a standard of living which is below
that ‘customary, or at least widely encouraged and/or approved, in the
societies to which [people] belong’ (Townsend, 1979): a certain degree
of equality is implied by this definition.
(Maxwell, 2001b: 335)

If these arguments hold, should there not be an international target for

inequality? Building on work by Cornia, it has been suggested that a Gini
coefficient below 0.45 should be the target (ibid.: 339). On the figures
given in the 2000/2001 World Development Report, thirty out of 105 coun-
tries would miss this target.

The desire to maximise participation and build a national consensus on

poverty reduction may obscure important trade-offs and conflicts of interest
Although citizenship does not feature as prominently as it might in the
Millennium Development Goals, it is certainly true that participation is a
prominent thread of national poverty reduction processes. This begins with
Participatory Poverty Assessments (PPAs) (Norton, 2001), but extends to
36 Simon Maxwell
discussion about the content of PRSPs, and to measures which increase
the accountability of government agencies (World Bank, 2000: chapter 6).
There is an increasing wealth of experience with participatory methods,
measures to increase accountability, and, more generally, with democratic
decentralisation (Johnson, 2001).
This is entirely admirable, but it should not be concluded that conflicts
will thereby disappear. A national consensus which obscures the fact that
there will be losers as well as winners, probably among the poor as well as
between the poor and the non-poor, will do the poor no service.
Many examples could be chosen to illustrate this point, but here is one:
the contribution of agriculture to poverty reduction. It is a frequent obser-
vation that agricultural growth helps reduce poverty, but it does this
through many different pathways, with different effects on different
groups of poor people, many of them contradictory. Irz et al. (2001) sum-
marise the pathways, in Box 2.2: agricultural growth can reduce poverty by
increasing farm income or employment, by creating jobs or livelihoods in
related industries or by means of consumption linkages, by means of lower
food prices, or through increased tax revenue being spent on services
which benefit the poor. The interests, then, are contradictory: between
food suppliers and food buyers; between the producers and consumers of
industrial raw materials; and between the urban poor, who favour lower
food prices, and the rural poor, many of whom will benefit as net food
buyers from lower prices, but many of whom will benefit from greater
employment in farm or farm-related industries if prices are high.
The implication of this example is that PRSPs and SWAPs, in particular,
need to articulate choices and adjudicate between them. In the case
of agricultural development, key choices are between crops, farm sizes
and degrees of labour intensity, between degrees of liberalisation, and
between investment in high- and low-potential areas (Maxwell, 2001a: 35).
No one should pretend that these choices are not highly contested in
most poor countries, and unlikely to be the subject of a national consen-
sus. Indeed, one of the main conclusions of a recent review of the institu-
tionalisation of PRSPs is that ‘politics matters’ (Booth, 2003). Politics, of
course, is partly about the art of achieving change, even when there is no
consensus.

A focus on public expenditure may distract attention from the importance

of macroeconomic policy
The issue of choices also arises in the context of overall macroeconomic
policy, a topic often neglected in countries where the management of
public expenditure seems to have become the main instrument of public
policy. Yet the fact is that choices made with respect to financial or mone-
tary policy, or trade policy, can have big effects on poor people. Poverty
strategies are not good, on the whole, at making these choices clear.
 Reflections on the new ‘New Poverty Agenda’ 37
Box 2.2 The benefits of agricultural growth

Farm economy
• Higher incomes for farmers, including smallholders.
• More employment on-farm as labour demand rises per hectare, the
area cultivated expands, or frequency of cropping increases. Rise in
farm wage rates.

Rural economy
• More jobs in agriculture and food chain upstream and downstream of
farm.
• More jobs or higher incomes in non-farm economy as farmers and farm
labourers spend additional incomes.
• Increased jobs and incomes in rural economy allow better nutrition,
better health and increased investment in education amongst rural
population. Lead directly to improved welfare, and indirectly to higher
labour productivity.
• More local tax revenues generated and demand for better infrastruc-
ture: roads, power supplies, communications. Leads to second-round
effects promoting rural economy.
• Linkages in production chain generate trust and information, build
social capital and facilitate non-farm investment.
• Reduced prices of food for rural inhabitants who buy in food net.

National economy
• Reduced prices of food and raw materials raise real wages of urban
poor, reduce wage costs of non-farm sectors.
• Generation of savings and taxes from farming allows investment in non-
farm sector, creating jobs and incomes in other sectors.
• Earning of foreign exchange allows import of capital goods and essen-
tial inputs for non-farm production.
• Release of farm labour allows production in other sectors.

Source: Adapted from Irz et al. (2001).

An example is trade policy. It hardly needs to be stated that trade policy

is among the most highly contested areas of development policy. On the
one hand, trade liberalisation is seen as a rapid route to growth and
poverty reduction; on the other hand, it is criticised for being risky at best
and often damaging to the poor (Bussolo and Solignac Lecomte, 1999;
McCulloch et al., 2001; Morrissey, 2002; Oxfam International, 2002). Very
different prescriptions follow. Some are to do with public expenditure:
investment in education and infrastructure to facilitate trade, for example,
or expenditure on safety nets and other measures to compensate the
38 Simon Maxwell
losers from trade liberalisation. Many others have to do with the overall
development stance of the country. The list could include the commit-
ment to state involvement in productive sectors, the regulation of utili-
ties, land reform, labour market issues and core labour standards – all
trade related at some level. As pointed out in Oxfam’s recent trade
report,

It had been hoped that the new framework for poverty reduction
developed by the IMF–World Bank would help to integrate poverty
reduction into all aspects of government policy. . . . Unfortunately . . .
experience to date has not been encouraging – especially in the case
of trade. No national PRSPs to date have even provided a credible
analysis of the potential impact of trade liberalisation on the poor,
and none has reviewed existing commitments on trade reform in the
light of such an analysis.
In view of the enormous impact of trade liberalisation – for better
or for worse – on the livelihoods of the poor, it is essential that its
implications are subjected to a proper assessment in advance, rather
than a retrospective justification on the basis of pre-conceived theory.
The timing, sequencing, and coverage of liberalisation all need to be
carefully reviewed. For example, it may make sense to liberalise
imports for a particular agricultural good after the implementation of
an investment programme to develop the capacity of small farmers,
but not before. Above all, trade liberalisation should be made part
of an informed national public debate about poverty-reduction strat-
egies.
(Oxfam International, 2002: 246)4

Focusing on sectors where sector-wide approaches work well may over-

emphasise social sectors at the expense of growth policies and cross-cutting
themes like rural development
SWAPs are proving to be a powerful instrument for improving the man-
agement of aid, delivering coordinated donor support through the
budget and in pursuit of agreed targets. This is certainly a big improve-
ment over a world where donors cherry-pick individual projects, often
outside the budget. There are, however, problems in applying the approach
to cross-cutting sectors such as rural development, say, or nutrition.
Early experience shows that there are conditions under which SWAPs
are likely to work, and others that will be more difficult. Foster (2000: 10)
lists the requirements as being the following:

• public expenditure as a major feature of the sector;

• a substantial donor contribution;
• basic agreement on strategy between government and donors;
 Reflections on the new ‘New Poverty Agenda’ 39
• a supportive macro budget environment;
• manageable institutional relationships; and
• incentives such as to garner support in sectoral ministries.

Several of these present problems from the point of view of cross-cutting

issues. For example, the first requirement has been cited as the reason
why SWAPs work well in sectors such as health, education and roads,
where the private sector is relatively small, and where expenditure (rather
than, say, pricing policy) is the main determinant of outcomes, but not so
well in agriculture, where public expenditure is less important and policy
more important (Foster et al., 2001). Nutrition is another sector where
public expenditure certainly has a role to play, but also where private
expenditure (not least on food) is important, and where policy decisions
(e.g. on exchange rates or food safety regulation) can have a major
impact.
The fifth requirement, about institutional relationships, has been inter-
preted to mean that SWAPs work well when an identifiable ministry con-
trols a sector, and not so well otherwise. Foster observes that

sector programmes have worked most effectively where they are defined
in terms of the area of budget responsibility of a single sector ministry,
programmes in education or health have proved more manageable
than sector programmes for cross-cutting themes such as the environ-
ment.
(2000: 10)

This would appear to place a multi-sectoral area at some disadvantage.

The sixth requirement, about incentives, may also be relevant to cross-
cutting programmes in some circumstances, especially if large realloca-
tions of funds between programmes (and probably between ministries)
are involved. The argument is that public officials will respond well to
SWAPs only if they are likely to receive increased funding as a result.
Winners will support SWAPs; losers will not.

Commitment to partnership may degrade into a form of covert

conditionality
‘Partnership’ is another term that runs through the new construction of
poverty. It features centrally in the Millennium Development Goals and is
especially important in the discussion about new aid modalities, for
example the Comprehensive Development Framework (Wolfensohn,
1999). Key concepts include shared ideals, trust, transparency, dialogue
and frequent review. The main fault lines in the debate are about how to
achieve genuine, reciprocal accountability, and about the extent to which
partnership arrangements should be contractual.
40 Simon Maxwell
On reciprocal accountability, the problem arises from the inequality
built into the relationship between a rich and powerful donor and a
poorer, less powerful ‘beneficiary’. Donor countries have expectations of
their developing country partners, like good governance; but the real flow
of resources and concessions, for example aid and trade access, is from
the donor to the beneficiary. In these circumstances, the relationship is
most easily understood as one of ‘asymmetric accountability’ (Maxwell
and Riddell, 1998).
One way to counter this is for the partnership to be contractual, with
obligations written down and procedures put in place for independent
review. A prototype example is the Cotonou Convention between the Euro-
pean Union (EU) and seventy-seven countries of the African, Caribbean
and Pacific (ACP) Group. This specifies standards with respect to human
rights, democratic principles, the rule of law, and corruption. Joint institu-
tions of the ACP, particularly the joint Council of Ministers, monitor the
partnership and adjudicate disputes. The system has yet to be tested fully,
and at this stage applies more to the developing countries than to the EU.
But it would be perfectly possible to imagine future Conventions specifying
standards to which the EU should adhere (concerning aid flows, for
example, or trade access).
An alternative, though weaker, approach is that of peer review. The
approach has long been used in DAC, whereby donor countries monitor
each other’s performance. It is now being taken up by developing coun-
tries, for example in the context of the New Partnership for Africa’s Devel-
opment (NEPAD, 2001). An innovation proposed by NEPAD is that peer
review be reciprocal, involving review of donor policies by African govern-
ments. This is certainly a step on the road to full reciprocal accountability.
The DAC Guidelines on Poverty Reduction emphasise the importance of
partnership, and give some indication as to what might be required on
the donor side. The guidelines stress that ‘working in partnership
means giving serious attention to assessing agency performance in
measuring up to agreed responsibilities and obligations’ (DAC, 2001: 47)
and suggest the indicative criteria set out in Box 2.3. The use of the
word ‘commitment’ in respect of resource transfers suggests that the
DAC has moved some way from an earlier and more cautious position, in
which the term ‘compact’ was preferred to the idea of ‘contract’ (OECD,
1996).

Ways forward
We have listed problems, but the intention is not to undermine new
approaches. Remember, these are risks, and the way to deal with risks is to
recognise them in advance and take appropriate action.
What, then, would be appropriate? There are two alternative approaches,
 Reflections on the new ‘New Poverty Agenda’ 41
Box 2.3 Assessing development agency policy reduction efforts

Working in partnership means giving serious attention to assessing agency

performance in measuring up to agreed responsbilities and obligations. The
following indicative criteria could be useful in this regard:

• Is the development agency’s country strategy based on the partner

country’s own assessment and strategy for addressing poverty?
• To what extent does the agency’s country strategy address the multidi-
mensional aspects of poverty?
• To what extent have the agency’s co-operation activities been carried
out jointly or in co-ordination with other bilateral and multilateral
development agencies (e.g. missions, appraisals, data collection, analyses)?
• Allowing for agency constraints, to what extent have agency admin-
istrative and financial requirements been adjusted to, or harmonised
with, the partner country’s existing procedures or with those of other
external partners, where these procedures are deemed appropriate?
• To what extent has the agency implemented its support in a manner
which respects and fosters partner country ownership?
• Has the agency supported and strengthened country-led planning,
implementation and co-ordination processes?
• Has the agency helped to facilitate civil society’s participation (at local,
national and international level) in debating and deciding the contents
of the country’s poverty reduction strategy in ways that respect govern-
ment efforts and concerns?
• Has there been a clear, serious commitment of resources to poverty
reduction?
• Has a commitment been made to provide predictable resources over a
medium-term planning time-frame?
• Has sufficient care been taken to avoid duplication of effort and to
build on complementarities across the external development com-
munity?
• Have efforts been made to improve policy coherence within the agency
and, more broadly, across the full range of Development Assistance
Committee member government ministries and departments, and has
progress been achieved?

Source: DAC (2001: 47).

42 Simon Maxwell
and again Wordsworth had his finger on the button, identifying both a top-
down approach:

The playfellows of fancy, who had made

All powers of swiftness, subtilty and strength
Their ministers, – who in lordly wise had stirred
Among the grandest objects of the sense,
And dealt with whatosever they found there
As if they had within some lurking right
To wield it;

and a bottom-up approach:

. . . they, too, who, of gentle mood,

Had watched all gentle motions, and to these
Had fitted their own thoughts, schemers more mild,
And in the region of their peaceful selves;

A bottom-up approach seems more consonant with current approaches,

which stress developing country ownership and process rather than
the wielding of power by the ‘Lords of Poverty’ (Hancock, 1989) and
other outside agencies. What might it mean in practice? There are six
principles:

• Practise subsidiarity. Donors need to be flexible in encouraging coun-

tries to set their own targets and design their own strategies, and
central governments need to be equally flexible in dealing with local
government and professional bodies. Conditionality on process rather
than substance needs to be maintained, in the implementation of pro-
grammes as well as in setting targets.5 This might mean that targets
and programmes set locally bear little relation to the Millennium
Development Goals. So be it.
• Focus on the essentials. The last thing PRSPs should be is documents
which touch every base, cover every point, mention every sector,
design every programme, set every target, and establish every monitor-
ing procedure. A PRSP should help to establish the big picture and
adjudicate on the key choices: state or market, institutions before
liberalisation, open economy or closed economy. If the big picture is
right, then the sectors will begin to take care of themselves.
• Do not forget the ‘difficult’ sectors and the cross-cutting issues. Remember
that the current technologies of aid delivery favour social sectors
because these meet the criteria for successful SWAPs. Yet it is import-
ant to include sectors such as agriculture, which do not meet the cri-
teria, and to tackle issues such as rural development which are not
usually ministry-led sectors at all.
 Reflections on the new ‘New Poverty Agenda’ 43
• Recognise the political nature of poverty reduction. Ownership of the
process is important, and so is participation. Empowerment is rightly a
major theme of the new approach. However, it would be naive to
expect a national consensus on poverty reduction policy: the poor are
not a homogeneous group, and will not have identical interests.
• Build partnerships based on reciprocal accountability. It is important not to
forget that partnership is a two-sided coin. Mutual accountability is a
cardinal principle.
• Follow a process approach. Interim and even full PRSPs are just the first
step. They are being revised at regular intervals. Donors should not
ask too much of the PRSPs, especially in the early stages. Partnerships
take time to build.

These principles are not theoretical. They can be summarised opera-

tionally (Table 2.1), for donors, as a list of do’s and don’ts. If we imple-
ment these principles, then there is no reason to worry about hubris. We
can be, indeed, in very heaven:

Not in Utopia, subterranean fields

Or some secreted island, Heaven knows where!
But in the very world, which is the world
Of all of us, – the place where in the end
We find our happiness, or not at all!

Table. 2.1 Do’s and don’ts of development

Do Don’t

Reinforce government leadership Impose rigid conditionalities

Encourage a broad-based debate Assume consensus is possible
Expect a strategic vision which Expect all to agree with the vision
identifies and adjudicates between
Focus only on social sectors amenable to
key macroeconomic and public
sector-wide approaches (SWAPs)
expenditure choices
Insist on international targets
Ensure that productive sectors and
cross-cutting issues are properly dealt Suggest using results-based management
with in a narrow way
Encourage subsidiarity in setting Insist on the perfect plan before starting
targets
Make unrealistic demands for data
Encourage process approaches to
Set performance standards for one side
managing public services
only
Disburse quickly
Revise frequently
Build two-way accountability

Source: Adapted from Maxwell and Conway (2000a: 21).

44 Simon Maxwell
PS: Wordsworth, of course, was not a poverty planner in the modern
sense. All the quotations come from his poem of 1809, ‘The French
Revolution as it appeared to enthusiasts at its commencement’. The full
text can be found at [Link]

Notes
1 This chapter has grown out of lectures for the Netherlands Development Assis-
tance Research Council (RAWOO) in The Hague; the Von Hugel Institute,
Cambridge; the London School of Economics; the Development Studies Associ-
ation, Manchester; the Royal Holloway College, University of London; and the
International Training Centre, the International Labour Organisation, Turin.
Thanks to participants for insights and advice. It also draws on earlier papers
dealing with different aspects of the new construction (Maxwell and Riddell,
1998; Maxwell and Conway, 2000a, b; Maxwell and Ashley, 2001; Maxwell,
2001b, c; Maxwell and Christiansen, 2002). I apologise in advance for the
degree of self-referencing which this entails. An earlier version of the paper
appears on the RAWOO Web site at [Link] [Link]. This
version also appears in Development Policy Review 21(1). Responsibility is mine.
Special thanks to David Sunderland.
2 The fourteen countries with PRSPs endorsed by the boards were Albania,
Bolivia, Burkina Faso, The Gambia, Guinea, Honduras, Mauritania, Mozam-
bique, Nicaragua, Niger, Tanzania, Uganda, Vietnam and Zambia.
3 It is notable, however, that rights did not feature in WDR 2000/2001. This was a
notable gap in WDR’s attempt to forge an international consensus, not least since
UNDP’s HDR took rights as its theme in the same year (Maxwell, 2001c: 146ff.).
4 These findings are broadly consistent with a recent World Bank/IMF review
which concluded that ‘None of the PRSPs has dealt systematically with part
experience of trade reforms, but several PRSPs (including Albania, Honduras
and Mozambique) have included specific measures in support of trade promo-
tion and liberalization. Only in Honduras and Mozambique was there an
attempt to clarify the link between these reforms, and growth and poverty reduc-
tion’ (IDA and IMF, 2002: para. 43).
5 Chapman (2002) develops the idea of a ‘soft systems’ approach to managing
public-sector organisations, building on the theory of learning organisations.
See Maxwell (forthcoming) for more detail and a discussion of the application
to results-based management.

References
Booth, D. (2003) ‘Are PRSPs Making a Difference? The African Experience’, Devel-
opment Policy Review, 21(2), special issue, March.
Bussolo, M. and Solignac Lecomte, H.-B. (1999) Trade Liberalisation and Poverty,
ODI Poverty Briefing 6, London, December.
Chapman, J. (2002) System Failure: Why Governments Must Learn to Think Differently,
London: DEMOS.
Christie, I. and Nash, L. (1998) The Good Life, Demos Collection: 14, Demos: London.
Davies, N. (2000) ‘Fiddling the Figures to Get the Right Results’, Guardian, 11 July,
UK.
de Haan, A. (1998) ‘Social Exclusion: an Alternative Concept for the Study of Depri-
vation’, IDS Bulletin, 29(1): 10–19, Brighton: Institute of Development Studies.
 Reflections on the new ‘New Poverty Agenda’ 45
Development Assistance Committee (DAC) (2001) The DAC Guidelines on Poverty
Reduction, OECD: Paris.
Foster, M. (2000) New Approaches to Development Co-operation: What Can We Learn
from Experience with Implementing Sector Wide Approaches?, ODI Working Paper 140,
London.
Foster, M., Brown, A. and Naschold, F. (2001) ‘Sector Programme Approaches:
Will They Work in Agriculture?’, Development Policy Review, 19(3): 321–338.
Gaventa, J., Shankland, A. and Howard, J. (eds) (2002) ‘Making Rights Real:
Exploring Citizenship, Participation and Accountability’, IDS Bulletin, 33(2):
1–11, Brighton: Institute of Development Studies, April.
Gunter, B. (2002) ‘What’s Wrong with the HIPC Initiative and What’s Next?’,
Development Policy Review, 20(1): 5–24, March.
Hancock, G. (1989) Lords of Poverty, London: Macmillan.
International Development Agency (IDA) and International Monetary Fund
(IMF) (2002) Review of the Poverty Reduction Strategy Paper (PRSP) Approach: Main
Findings, at [Link]
Irz, X., Lin, L., Thirtle, C.G. and Wiggins, S.L. (2001) ‘Agricultural Productivity
Growth and Poverty Alleviation’, Development Policy Review, 19(4): 449–466,
December.
Johnson, C. (2001) ‘Local Democracy, Democratic Decentralisation and Rural
Development: Theories, Challenges, and Options for Policy’, Development Policy
Review, 19(4): 521–532, December.
Killick, T. (2002) Responding to Inequality, DFID Inequality Briefing Paper 3,
London: DFID, March.
Lipton, M. and Maxwell, S. (1992) The New Poverty Agenda: An Overview, IDS Discus-
sion Paper 306, August, Brighton: Institute of Development Studies.
McCulloch, N., Winters, L.A. and Cirera, X. (2001) Trade Liberalisation and Poverty:
A Handbook, London: Centre for Economic Policy Research.
McKay, A. (2002) Defining and Measuring Inequality, DFID Inequality Briefing Paper
1, London: DFID, March.
Maxwell, S. (1998) ‘International Targets for Poverty Reduction and Food
Security: A Mildly Sceptical but Resolutely Pragmatic View with a Call for
Greater Subsidiarity’, Canadian Journal of Development Studies, Special Issue, 19:
77–96.
Maxwell, S. (2001a) ‘Agricultural Issues in Food Security’, in Devereux, S. and
Maxwell, S. (eds) Food Security in Sub-Saharan Africa, London: ITDG Publishing.
—— (2001b) ‘Innovative and Important, Yes, but Also Instrumental and Incom-
plete: The Treatment of Redistribution in the “New Poverty Agenda” ’, Journal of
International Development, 13(3): 331–341.
—— (2001c) ‘WDR 2000: Is There a New “New Poverty Agenda”?’, Development
Policy Review, 19(1): 143–149, London: ODI.
—— (forthcoming) ‘Lost in Translation? Implementation Constraints to Results-
Based Management’, in UNDP/DFID (provisional title/publisher) Enhancing
Development Effectiveness: A New Focus on Managing for Results, New York: Earthscan.
Maxwell, S. and Ashley, C. (2001) ‘Rethinking Rural Development’, Development
Policy Review, 19(4): 395–425, December.
Maxwell, S. and Christiansen, K. (2002) ‘ “Negotiation as a simultaneous equa-
tion”: building a new partnership with Africa’, International Affairs, 78(3):
477–491.
46 Simon Maxwell
Maxwell, S. and Conway, T. (2000a) New Approaches to Planning, World Bank
Operations Evaluation Department (OED) Working Paper series 14. Summer,
Washington, DC: World Bank.
—— (2000b) Perspectives on Partnership, World Bank Operations Evaluation Depart-
ment (OED) Working Paper series 6, Summer, Washington, DC: World Bank.
Maxwell, S. and Riddell, R. (1998) ‘Conditionality or Contract? Perspectives on
Partnership for Development’, Journal of International Development, 10(2),
March–April.
Morrissey, O. (ed.) (2002) ‘Investment and Competition Policy: Issues for Devel-
oping Countries’. Development Policy Review, 20(1): 63–73, March.
Moser, C. and Norton, A. (2001) To Claim Our Rights: Livelihood Security, Human
Rights and Sustainable Development, London: ODI.
Narayan, D. with Patel, R., Schafft, K., Rademacher, A. and Koch-Schulte, S.
(2000) Voices of the Poor: Can Anyone Hear Us?, New York: Oxford University Press
for the World Bank.
Naschold, F. (2002) Why Inequality Matters for Poverty, DFID: Inequality Briefing
Paper 2, London: DFID, March.
NEPAD (2001) Policy Document, October 2001 (English version) at [Link]
[Link]/Documents/[Link].
Norton, A. (2001) A Rough Guide to PPAs: Participatory Poverty Assessment – An Intro-
duction to Theory and Practice, London: ODI.
Norton, A. and Elson, D. (2002) What’s Behind the Budget? Politics, Rights and
Accountability in the Budget Process, London: ODI.
O’Neill (2002) A Question of Trust, Reith Lectures 2002: 3, at [Link]
uk/radio4/reith2002/lecture3_text.shtml.
Organisation for Economic Cooperation and Development (1996) Shaping the 21st
Century: The Contribution of Development Co-operation, Paris: OECD.
Oxfam International (2002) Rigged Rules and Double Standards: Trade, Globalisation,
and the Fight against Poverty, Oxford: Oxfam.
PRSP Connections (2002) PRSP Connections: Monitoring and Synthesis Project, Issue 5,
London: ODI, May.
Silver, H. (1994) Social Exclusion and Social Solidarity: Three Paradigms, IILS Discus-
sion Papers 69.
Townsend, P. (1979) Poverty in the UK, Harmondsworth, UK: Penguin.
UNDP (1990) Human Development Report, New York: Oxford University Press.
—— (1997) Human Development Report, New York: Oxford University Press.
White, H. (2001) ‘National and International Redistribution as Tools for Poverty
Reduction’, Journal of International Development, 13(3): 343–352.
—— (2002) ‘A Drop in the Ocean? The International Development Targets as a
Basis for Performance Measurement’, appendix 2 in National Audit Office
(2002) DFID: Performance Management – Helping to Reduce World Poverty, Report by
the Controller and Auditor General, National Audit Office, April.
Wolfensohn, J. (1999) ‘A Proposal for a Comprehensive Development Framework
(a discussion draft)’, memo to the board, management and staff of the World
Bank Group, 21 January, Washington, DC: World Bank.
World Bank (1992a) Poverty Handbook, Washington, DC: World Bank (discussion
draft).
—— (1992b) Poverty Reduction Operational Directive, Washington, DC: World Bank.
—— (2000) World Development Report, New York: Oxford University Press.
3 Using development goals and
targets for donor agency
performance measurement1
Howard White

Introduction
In Chapter 1 it is suggested that the dominance of the International
Development Targets and subsequently the Millennium Development
Goals in current development policy is partly a consequence of a stress on
results-based measurement, which seeks to judge development agency
performance against outcomes. However, whilst such an approach is intu-
itively appealing, there are a number of potential problems with it, which
are addressed in this chapter. The chapter starts by setting out a list of the
desirable properties of performance measures, and then considers how
these measures can be applied using logic models. In the light of this dis-
cussion, the International Development Targets and Millennium Develop-
ment Goals are assessed. It is argued that although they are reasonable
measures of progress, they cannot easily be used to monitor the perform-
ance of particular agencies, as is illustrated by the attempts of UK aid
agency the Department for International Development (DFID) to inter-
nalise the development targets in developing its own strategy.

Desirable properties of performance measures

The literature identifies several desirable features of performance meas-
ures (e.g. Hakes, 2001; Jackson, 2000; National Audit Office, 2001). In
particular, it can be argued that measures should be:

• Relevant and balanced. Balance breaks down into three areas: (1) the
range of measures should cover all elements of an organisation’s acti-
vities – and they should not include things which are not a part of
them (i.e. indicators should be relevant); (2) they should focus on
both short- and long-term performance (which may correspond to the
distinction between outcomes and impact); and (3) they should cover
the whole process that leads to a measure being met (including
inputs, activities and outputs). This last point is further discussed
below. A further question of balance is the different weight given to
48 Howard White
different measures; some can be more important than others, and this
fact should be explicitly recognised.
• Known, understood and trusted. For performance measures to have their
desired organisational impact, they must be known by the members of
the organisation and their meaning must be understood. But it is also
necessary that they are trusted. That is, members must believe they are
useful and well defined and that the quality of the underlying data is
good.2 A related issue is that of ownership. Ideally, measures should
be owned by those responsible for delivering the performance,
which means they are involved in developing the targets. In practice,
performance measurement systems are often passed down from the
top or designed by outsiders.
• Affected and attributable. Changes in performance measures should be
affected by the activities of the organisation, and the extent of the
effect should be measurable (attributable). This is also referred to as
the problem of additionality. The UK Treasury defines additionality as
‘the amount of output from a policy as compared with what would
have occurred without intervention’ (quoted in Jackson, 2000: 11) For
outcome indicators, attribution can be problematic, and this is a
major issue in relation to development targets, as we shall see.
• Achievable. Targets should be achievable, but not too easily.
• Linked to existing management systems. Organisations will already have
in place management information systems. The rise of results-based
measures was in part a response to the fact that existing systems
focused on inputs and internal activities, e.g. spending and staffing,
rather than achieving outcomes. But new systems should not be
separate from or parallel to these existing systems. Rather, they should
be integrated, hence encompassing the logic of how inputs lead to
outcomes.

Using performance outcome measures

Performance outcome measures serve two main functions: accountability,
and improving organisational performance. The accountability function is
in principle straightforward. An organisation commits itself to achieving
certain outcomes and it either does or does not deliver. For taxpayers,
politicians or shareholders, this may theoretically be sufficient, since they
can respond by withholding votes or funds – though problems with such
an approach are highlighted in Chapter 1. Further problems arise in the
use of performance measures as a management tool. Managers need to
understand why outcomes have or have not been achieved. There are
three questions here, but only one answer. The questions are: (1) how can
performance measures help change organisational practice to enhance
the likelihood that targets will be met; (2) how can measures be inter-
preted to understand how inputs have or have not led to the desired out-
 Donor agency performance measurement 49
comes; and (3) how can the organisation’s activities be linked to the out-
comes? The one answer is the importance of logic models.3
Ideally, performance measures affect organisational culture. Simply
shifting to performance-based measurement is itself a substantial change
in practice: ‘the cultural change required to achieve performance-based
management in any public or non-profit agency presents an enormous
challenge’ (Schemer and Newcomer, 2000: 63). But the process of plan-
ning how targets will be achieved opens up substantial possibilities for
changing working practices. This planning process should be systematic
and broadly based within the organisation. This is where logic models
come in.
Logic models and the logical framework are schemes for linking inputs
to outcomes. Terminology varies somewhat, but all capture the same basic
idea. The stages recognised in the model used by the NAO are resources,
inputs, processes (often called activities), outputs and outcomes (also
called impacts) (NAO, 2001: 2). It is a rather obvious truism that resources
should be utilised in such a way as to achieve the desired outcomes. But a
failure to do this has been commonly observed. For the government sector,
the most substantial evidence base comes from the USA, since the 1993
Government Performance and Results Act (GPRA) requires all govern-
ment agencies to have outcome-based targets against which their perform-
ance is to be judged. The General Accounting Office (GAO) makes annual
assessments of plans and outcomes. Both the GPRA itself and GAO (1998)
suggest that performance measures should span inputs, outputs and out-
comes. However, in practice such logic models have been applied in few
GPRA plans (Scheirer and Newcomer, 2000). The absence of indicators
reflects the fact that there is too little conscious analysis as to how plans
and programmes will affect outcomes (GAO, 1999).
The attraction of logic models is partly that they should force the
agency to examine programmes to see if they really will achieve the
desired outcomes (Millar et al., 2000). However, logic models are a start-
ing point for planning, rather than the end point. It is an easy matter to
put a poverty-related goal at the top of a log-frame as a goal, but a rather
more complex one to be clear how the inputs will help achieve that goal
and to monitor progress accordingly. For example, DFID’s budget support
to Kenya in 2000, which was paid in support of the public-sector reform
programme (i.e. to finance retrenchment), included a reduction in the
maternal mortality rate as a goal indicator. It is very difficult indeed to
trace any links between this intended outcome and the activities under-
taken by DFID or the Kenyan government, particularly given their short-
term nature. It is all too easy to explain away discrepancies between what
is actually achieved and the initial targets with reference to problems of
attribution or external factors. Nonetheless, if used properly, logic models
should be able to help overcome these problems, although the evidence
base for this being so in practice is quite thin.
50 Howard White
Logic models provide a basis for understanding performance. That is,
they potentially solve the problem of attribution. Various types of model-
ling can be undertaken to demonstrate the determinants of outcomes and
thus the contribution made by an organisation’s activities. Yet in practice
such modelling is too costly, cumbersome and frequently contentious to
provide a basis for regular performance monitoring. Instead, less rigorous
strategies have been proposed to establish ‘plausible association’ between
programme efforts and performance (Scheirer and Newcomer, 2000: 68).
These should include attempts to account for the influence of external
factors, which may of course be positive or negative.
Attribution becomes harder as we move along the causal chain. It is
easy to attribute responsibility for delivering inputs, and usually for carry-
ing out activities, although external factors may play a part. These activi-
ties should lead to desired outputs which deliver the target outcomes,
again subject to external factors. If the underlying model is correct, then
indicators should capture whether the organisation is doing what it needs
to do to achieve the outcomes – which may well be the case even if the
targets are not met. Conversely, targets may be met ‘by accident’ (i.e.
because of external factors) rather than on account of the conscious
efforts of the responsible agency. Whilst logic models can uncover this
fact, it is unlikely to matter very much.

The International Development Targets and Millennium

Development Goals as performance measures4

Relevance and balance

The Millennium Development Goals and International Development
Targets score highly on relevance. They capture some of the main aspects
of poverty in the developing world. Maxwell (1998) argues that the Goals
and Targets simplify the complex phenomenon of poverty by reducing it
to a dollar a day. But this argument misses the point that every one of the
targets is poverty related, not just the one for income poverty. Rather, the
Goals and Targets should be praised for capturing the multidimensional
nature of poverty.
It can further be argued that the Goals and Targets have made them-
selves relevant. Their prominence has resulted in a consensus around a
particular set of indicators for monitoring poverty reduction, paving the
way for an unprecedented degree of co-ordination amongst donor agen-
cies, and facilitating a harmonisation of performance monitoring which
has proved difficult to achieve in the past (DAC, 2000: 21).
But criticisms may be made regarding what is excluded. First, some
important dimensions of development are missed out. Neither food nor
shelter appeared in the International Development Targets, though these
have been added as the second and eleventh goals in the Millennium
 Donor agency performance measurement 51
Development Goals. A more pertinent critique may be the focus on the
measurable. Although the Millennium Development Goals also mention
the importance of qualitative aspects of development, such as the import-
ance of governance, most discussion of the indicators themselves focuses
on those which can be measured. However, some other agencies are active
in promoting the qualitative aspects.
Nor are the Goals balanced, in that for most of them the target date is
2015, with one for 2005. Fifteen years (twenty when they were set) is defi-
nitely a long-term goal. Although progress can be measured with respect
to being on track, it is preferable to set explicit short-term goals. These
short-term goals may be either interim targets for the same variables, or
targets for outputs that will help achieve the desired outcome. This brings
us to another lack of balance in the indicators.
The Goals are also not balanced in that they are mostly outcome meas-
ures, as are most of the wider set of indicators associated with them. Some
are output indicators, but not in a consistent manner of a set of indicators
capturing inputs, process, outputs and outcomes for a single target, where
outputs are the products of the ‘investment’ such as immunised or edu-
cated children, and outcomes are direct measures of welfare such as mor-
bidity, mortality or higher expenditures. Hence, they do not provide a
basis for monitoring performance or taking the steps necessary to achieve
the outcomes they contain. This problem is not because such indicators
are not amenable to international agreement. For example, the Copen-
hagen Social Summit of 1995 launched the 20:20 initiative, by which 20
per cent of government spending and 20 per cent of aid monies should be
devoted to providing basic services to the poor. Data are available to
monitor progress in meeting this target, and this progress has been
dismal, especially on the part of donors. Another example is immunisa-
tion, an important factor for child survival, which has suffered setbacks in
recent years. A further example would be the establishment of adequate
food security systems. Although nutrition is not amongst the main target
indicators, it is in the wider variable list and is an important determinant
of some of the outcomes which are listed, most notably child survival.

Are the targets well defined and can progress be measured?

Most of the Millennium Development Goals are clearly defined. Four
exceptions may be noted. First, access to reproductive health is not meas-
ured by any existing indicator and the proxy of contraceptive prevalence
is problematic. Second, the targets are loose, with child survival terminol-
ogy used in a way which would upset demographers. Infant mortality is the
probability of dying before the first birthday, child mortality that of dying
between the first and fifth birthdays and under-five mortality that between
birth and fifth birthday. Under-five mortality is thus a weighted average of
infant and child mortality, but is not liked by many demographers as the
52 Howard White
factors underlying the two mortality rates vary (see Chapter 10). The most
appropriate indicators would therefore be infant and child mortality, but
the targets are for infant and under-five mortality (although some sources
give child mortality instead). Third, the environmental target speaks of
reversing current trends in resource use. Taken literally, this means that
the available quantity of environmental resources should begin to increase
rather than decrease. Aside from being unattainable, this target does not
fit with the general consensus that it is all right to use environmental
resources but they must be used in a way which is consistent with overall
sustainable development. DFID, which uses the expression ‘managing
environmental resources’, recognises this fact. Finally, it may be ques-
tioned whether equality in school enrolments is an adequate proxy for
gender equality. A defence of the measure is that equality of education is a
necessary starting point for achieving other forms of equality.
Knowledge of the Goals and Targets amongst DFID staff is high.5 This
does not imply that they are aware of the precise technical definition of
the variables. For example, understanding of the purchasing power parity
(PPP) dollar a day used for income poverty is mainly restricted to econo-
mists,6 and understanding of the relevant age groups for infant and child
mortality to health advisers. But it is doubtful whether lack of this precise
knowledge impedes work on the targets.
The chosen indicators are all ones for which data were already being
collected. They are now published in a variety of places, notably the Devel-
opment Assistance Committee (DAC) Web site, the World Bank’s
World Development Indicators and DFID’s International Development Statistics.
However, data quality varies by indicator. An issue for all the indicators is
coverage, meaning for how many countries data are available. The Organi-
sation for Economic Cooperation and Development (OECD) document
Methodological Note reports the baseline data for the various indicators. For
some indicators coverage is low (see Appendix 3.1, pp. 66–67, for a
summary). For example, income-poverty data were missing for about 30
per cent of the population of the developing world, and nearly two-thirds
of those in sub-Saharan Africa. For net primary enrolments there is only
12 per cent coverage for South Asia and 69 per cent globally.
A further issue is the frequency and timeliness with which data
are available. Income poverty data come from income and expenditure
surveys, which are not conducted annually in many developing coun-
tries but rather every three to four years. Vital registration systems and
health facility-based reporting have inadequate coverage of the popu-
lation in developing countries to be a reliable source of health data.
Good-quality data on child health, including infant and child mortality,
are provided by Demographic and Health Surveys (DHSs), but these are
conducted about every four years and not in all countries. It is difficult to
respond to performance data which are available only with a three-year
lag.
 Donor agency performance measurement 53
Even where data are available, they may be of poor quality. From the
targets themselves, maternal mortality data are the most problematic area.
These data are notoriously unreliable (see Box 3.1), to the extent that it is
surprising that the indicator was deemed suitable for inclusion amongst
the targets. In measuring progress on this target, A Better World for All (IMF
et al., 2001) reported the percentage of attended births rather than mater-
nal mortality itself.
But the setting of the Goals and Targets has in itself provided an
impetus to the improving of data quality. As attention is focused on these
outcomes, there is an awareness that efforts need to be made to ensure
that the underlying data are sound: ‘a more concerted effort by the donor
community is needed to support partner countries’ capacity to collect
data and monitor progress towards the international goals over the
coming years’ (DAC, 2000: 22).

Are the targets achievable?

Whether or not the targets can be achieved has been the area of analysis
that has attracted most attention. Chapter 1 reviews some of the existing
evidence and Chapter 6 presents a new discussion of this issue. The main
point to note is that there is common agreement that most of the targets
will not be met: ‘on current trends, none of the international develop-
ment goals on health and education are likely to be achieved at the global
level’ (IMF et al., 2001: 12). For example, on current trends there will still

Box 3.1 Data quality: the case of maternal mortality

Maternal mortality for Ghana jumped from 400 to 1,000 per 100,000 from
one issue of the World Bank’s World Development Report (WDR) to the next.
Maudlin (1994) showed that although they both used the same source, the
WDR reported data for fifty-six developing countries, whereas the UNDP’s
Human Development Report (HDR) did so for fifty-five of these fifty-six and
forty-eight further countries (for which the WDR indicated that data were
unavailable). The Development Assistance Committee baseline data sheets
state that coverage for this indicator includes practically every country in the
world.
Counting differences of less than 50 points as the same, HDR gave
higher values than WDR for twenty-six countries, lower for twelve and about
the same for seventeen. Some differences are substantial, for example Benin
at 800 and 161 in the two sources. The correlation coefficient between the
two sets of figures is only 0.7, dropping to 0.4 for high-mortality countries. A
comparison of WDR and DHS data for 1997 also shows substantial discrep-
ancies.

Source: White et al. (2001).

54 Howard White
have been 100 million children of school-going age out of school in 2001.
These pessimistic conclusions arise even from ‘base run’ predictions
which are typically based on the usually rather optimistic growth forecasts
produced by the World Bank and IMF. Forecasts based on historical
growth performance show even greater divergence from the targets, and
current economic performance suggests that growth from the 1990s may
overestimate that for the new decade.
The second point is that performance varies greatly by region. The
targets themselves are defined at various levels of aggregation. Some apply
globally (e.g. income poverty), some are country specific (e.g. infant and
child mortality) and some necessarily apply to each country (e.g. universal
primary education will be achieved only when it is achieved in all coun-
tries). But even the targets stated as global figures are intended to be met
at the disaggregated level – that is, for each country and region: ‘while
expressed in terms of their global impact, these goals must be pursued
country by country’ (DAC, 1996: 2). Here performance varies: ‘the educa-
tion and gender equality goals are likely to be achieved in some regions
and many countries. And a few countries are on track to achieve large
reductions in infant and under five mortality’ (World Bank, 2001: 12). In
general, Africa is performing worst with respect to the different targets
and East Asia the best. Some of these differences are explained by con-
tinued high growth in East Asia (so that the income-poverty target has
already been met) compared to low growth in Africa. But African coun-
tries also suffer from much higher rates of HIV/AIDS, which has helped
reverse the long-run decline in mortality rates in several countries, and
from extensive conflict, which undermines the ability of countries to
attain any of the targets. Experience varies for the other regions. The
transition economies of Eastern Europe and the former Soviet Union saw
very sharp increases in poverty in the early 1990s so that the income-
poverty target is unlikely to be met, and in some countries, notably Russia,
social indicators have also worsened.
Whether or not the targets will be achieved is not the same thing as
whether they are achievable. The papers reviewed offer limited advice on
this point. The main argument made is the importance of a good policy
environment (see Collier and Dollar, 2002; Demery and Walton, 1999;
IMF et al., 2001): ‘whether or not poverty incidence will be halved by 2015
depends in part on how well economies are managed’ (Demery and
Walton, 1999: 83). The argument is that better policies promote growth,
which reduces income poverty and other forms too, such as mortality
rates. Most papers on the subject do not discuss an explicit role for devel-
opment agencies. The exception is that by Collier and Dollar, who build
on the suggestion that aid works best at promoting growth, and so redu-
cing poverty, when the policy environment is right. Hence, the prospect of
meeting the Goals and Targets is enhanced if aid is concentrated on poor
countries with good policies. This piece of research has had an influence
 Donor agency performance measurement 55
on several development agencies, though its theoretical and empirical
basis is highly contestable (see Lensink and White, 2000).

Attribution
The Millennium Development Goals do least well with respect to attribu-
tion. It is impossible (or at very best, virtually impossible) for an individual
agency to isolate its impact on global, or even country-level, trends in the
relevant indicators. This fact is demonstrated by the experience of the
United States Agency for International Development (USAID).
In 1997, USAID laid out six strategic development goals (e.g. ‘broad-
based economic growth and agricultural development encouraged’), and
for each of these defined a set of outcome indicators at both country and
global levels (e.g. ‘average annual growth rates in real per capita income
above one per cent’). With respect to the growth goal, the fiscal year (FY)
2000 performance report states that ‘nearly 70 per cent of USAID-assisted
countries were growing at positive rates in the second half of the 1990s,
compared with 45 per cent in the early part of the decade’ (USAID, 2001:
v). However, that same performance report noted that ‘one cannot rea-
sonably attribute overall country progress to USAID programs’ (ibid.: viii).
Commenting on the previous year’s USAID Performance Report, GAO
had similarly observed that the goals were ‘so broad and progress affected
by many factors other than USAID programmes, [that] the indicators
cannot realistically serve as measures of the agency’s specific efforts’
(2000: 1–2). In response to these criticisms, the FY 2000 performance
report (USAID) announced that the indicators related to the strategic
goals will no longer be used to measure USAID’s performance (but they
will be reported as being of interest in their own right, being referred to
as ‘Development Performance Benchmarks’). Rather, performance will be
measured against the strategic objectives of the individual operating units
(e.g. country programmes).
The difficulties of attribution are further illustrated by a GAO report
on USAID’s child survival programme, launched in the 1980s (GAO,
1996). Entitled Contributions to Child Survival Are Significant but Challenges
Remain, it lists the uses of funds under the child survival budget line and
reports progress on reducing child mortality and immunisation rates in
recipient countries (an activity supported by a number of donors). It
makes pertinent observations on some of the activities financed by child
survival programmes (a bridge in Mozambique and a water tower in
Egypt) and points out that many countries with high mortality do not get
funds, whereas those with low mortality continue to do so. But no attempt
is made to link the inputs described to the outputs and outcomes
reported. To do so would not be an impossible task: there are models of
the determinants of infant and child mortality which could be used to
look at the trends in these determinants and how they have been affected
56 Howard White
by USAID’s activities. But it would be a major undertaking – not one that
can be accommodated within routine performance measurement. The
best that performance measurement can do is have a set of indicators
spanning inputs to outcomes based on such an underlying model. It is for
these reasons that evaluators have turned to the logical framework and
approaches such as theory-based evaluation.

Integration with existing management information systems

The DAC report on results-based management (DAC, 2000) identifies
three approaches to agency-wide monitoring:

• aggregating project- and programme-level outputs;

• aggregating project- and programme-level outcomes;
• reporting country-level trends.

This classification may be simplified to ‘bottom-up’ versus ‘top-down’

systems. Bottom-up systems take individual activities as the primary unit of
analysis and aggregate performance across countries, sectors and the
agency as a whole. Top-down systems report on outcome indicators for a
country (or at least sector in a country) and whole regions. Top-down
systems suffer from attribution problems – can a link be made between
the observed outcome and the agency’s activities? If not, then the
information is of limited, if any, value, in guiding management systems.
Bottom-up systems, which are more strongly rooted in traditional manage-
ment information systems, face problems of aggregation and linking to
relevant outcome indicators.
All donor agencies have some sort of monitoring and evaluation system
at the project and programme level, which should provide a basis for both
feedback at the project level and ‘feed-up’ to management. For our pur-
poses here we are interested in the two questions suggested in the previ-
ous paragraph. First, are the data that are collected of a suitable form to
be aggregated to give an overall indication of agency performance (and
broken down at the country, regional and sectoral levels)? Second, if
there is such an aggregation, does it yield information on outcomes in
relation to the development goals?
The World Bank’s rating system is an example of a system which does
yield agency-wide results. All activities are rated under a number of cri-
teria on a regular basis and upon completion. These ratings include an
overall rating of whether the activity has been satisfactory. Hence overall
portfolio performance can be judged by the percentage of projects
deemed satisfactory. Such analysis is published in the Annual Review of
Development Effectiveness (ARDE) produced by the Operations Evaluation
Department of the World Bank. This system does yield information of
use to management, which has a systematised approach to identifying
 Donor agency performance measurement 57
‘problem areas’. In the decade from 1991 the percentage of ‘problem pro-
jects’ rose from 11 to 20 per cent, resulting in the creation of the Portfolio
Management Task Force, whose report (known as the Wapenhans Report
after the lead author) advocated a number of changes and contributed to
the adoption of results-based management within the World Bank.7
Whilst a ‘satisfactory’ project is defined as one which is substantially
meeting its development objectives, this does not allow us to say anything
about the contribution of the World Bank to the meeting of the goals.
The information collected is simply not the right sort to provide that
information. There is a misalignment between the data collected from the
‘bottom up’ and the sort of outcomes being monitored in ‘top-down’
International Development Target-oriented systems. The same is true of
all other agencies which collect data of a form suitable for bottom-up
aggregation.
There are two possible responses to this problem of misalignment. The
first is to say that it is inevitable. The problem of attribution is not going to
be solved for routine monitoring purposes, so no attempt should be
made to link agency performance as measured by bottom-up systems
with agency impact on the Millennium Development Goals and Inter-
national Development Targets. That answer does not seem satisfactory for
agencies, like DFID, which have pinned their performance to the Inter-
national Development Target mast. So the second response to the mis-
alignment problem is to resort to logic models. The bridge must be made
between observing satisfactory activities and presumed impact on develop-
ment outcomes. Whether this is feasible is an issue I return to in the next
section.

Summary
The Millennium Development Goals and International Development
Targets are relevant as a measure of development progress, are well
defined and understood, and data are mostly available, if of variable
quality. But they have two important shortcomings, and one perhaps less
important one. First, the indicators themselves are output and outcome
oriented. There has not been international agreement on the underlying
logic model to produce the indicators required to monitor inputs and
processes necessary to achieve these outcomes. The partial exception is
the consensus on the need for market-led economic growth, though
whether this is the best way to achieve the targets is debatable. Second, it
is not possible to attribute changes in these measures to the actions of
development agencies, either individually or collectively. Both these argu-
ments point to the need for a more holistic performance measurement
system, one which integrates the monitoring of inputs and process with
that of outputs and outcomes. Finally, the targets seem unlikely to be
attained.
58 Howard White
Whilst the Millennium Development Goals and International Develop-
ment Targets are mostly a satisfactory measure of development progress,
this does not mean that they are suitable indicators with which to measure
the performance of any single development agency. Indeed, a short
period of reflection shows that the problem of attribution alone means
that outcome indicators are by themselves unlikely ever to be suitable,
especially ones of a global nature. Hence some modification is neces-
sary. To what extent has DFID’s approach to the targets tackled these
problems?

The use of targets by donor agencies: the Department for

International Development
DFID has embraced the International Development Targets and then the
Millennium Development Goals more strongly than any other bilateral
donor. The targets have been given pride of place in both the 1997 and
2000 White Papers on International Development, they have featured
strongly in the public pronouncements of the former Secretary of State,
Clare Short, and have been promoted in various ways such as through
posters. DFID staff view them as highly relevant to their work. But more
important than these changes in external presentation have been the
internal efforts to accommodate the Goals and Targets, principally through
strategy papers, the Public Service Agreement (PSA) and the Service Deliv-
ery Agreement (SDA).

From target to strategy

At a workshop held to discuss the first White Paper, many commentators
argued that the Paper was full of good intentions but rather silent on how
these intentions were to be fulfilled (see the papers in White, 1998). But
since that time, DFID has put in place a process for elaborating a strategy
to this end. Central to this work have been the strategy papers: Target
Strategy Papers, Institutional Strategy Papers and Country Strategy Papers
(see Box 3.2). The discussion here focuses on the first two, since Country
Strategy Papers are the subject of Chapter 4.

Box 3.2 The Department for International Development’s strategy papers

Target Strategy Papers (TSPs) address a single development target (e.g.

income poverty) or group of targets (health).
Institutional Strategy Papers (ISPs) have been, or are being, prepared for
each of DFID’s main partners amongst international organisations.
Country Strategy Papers (CSPs) outline the main challenge of poverty
reduction in each of the partner countries.
 Donor agency performance measurement 59
The intended relationship between the different documents is shown
in Figure 3.1.8 Target Strategy Papers have a central role, defining the
long-term strategies to meet the development targets. These strategies
should inform the content of the Institutional Strategy Papers and Country
Strategy Papers. CSPs should also be informed by the relevant ISPs.
Various processes have been put in place to ensure these links, starting
with the collaborative way in which strategy papers were prepared. The
geographical desks were asked to draw up a response to the TSPs, outlin-
ing the implications for their work. In practice, most of the ‘new genera-
tion’ of CSPs currently in use were prepared prior to either TSPs or ISPs
being available, so that the links shown in Figure 3.1 will not have oper-
ated in practice. But the question is the extent to which CSPs prepared
since the beginning of 2001 are consistent with the TSPs (and indeed the
PSA, which is discussed below).
TSPs might be faulted for their sectoral bias, which can overlook cross-
sectoral linkages – for example, health matters for education and vice
versa, whilst lack of access to water close to home may be the main thing
keeping girls from school. Hence preparing papers on each target sepa-
rately misses the importance of a multidimensional approach. Examining
the papers shows this criticism to be only partially justified, since several
papers discuss these cross-sectoral linkages. For example, the environment
TSP makes much of public health arguments.
All strategy papers have a common structure which corresponds to a
logic model. The first section sketches out the challenge, e.g. the relevant
International Development Target(s) in the case of a TSP, how an organi-
sation relates to the poverty agenda for an ISP and the poverty situation in
a country for a CSP. The next section lays out the strategy necessary to
meet this challenge. This is a strategy for all partners, not DFID alone.
The paper then reviews the role of different actors and, finally, the part
that DFID can play.
However, analysis of CSPs shows that they suffer from a ‘missing
middle’ (see Chapter 4). Figure 3.1 shows also the PSA/SDA and the
annual performance plans, which are important in bringing the Millen-
nium Development Goals and International Development Targets down
to a realistic management time-frame. Fifteen years is too long for man-
agement targets. The PSA (see Box 2.1, pp. 31–32) and SDA bring these
down to targets for 2006, though sometimes making them more ambitious
than their longer-run counterparts. For example, the target for under-five
mortality requires an annual reduction of around 6 per cent, which over
twenty-five years would imply a nearly 80 per cent reduction as against the
International Development Targets of two-thirds. Since most countries
are not on track for the longer-run target, the chances of meeting the
shorter-run one must be doubted. The PSA has five development object-
ives and one management one (value for money). Each objective has a
number of performance targets, which are a mixture of process, output
 1997 White Paper: Eliminating Poverty –
A Challenge for the 21st Century

2000 White Paper: Making Globalisation work

for Poor People

Target Strategy Papers

Strategy and Policy/Strategy Organisational

planning
statements

3–5 Knowledge and Partnership Policy Country Strategy Institutional Public Service Medium Term
Year Research Strategy Statement with Papers Strategy Papers Agreement/ Framework
Horizon Paper Non-Governmental Service Delivery
Organisations Agreement
End-of-period evaluation informing new strategy

Annual strategy reviews and

resource allocation Annual Plan and Institutional Policy and Resource
Performance Performance Plans

Annual Personal
Development Plans

Operational
Engaging with Partners,
Programmes, Projects and
Technical Co-operation

Figure 3.1 Overview of Department for International Development strategy: vision statement, fifteen-year horizon.
 Donor agency performance measurement 61
and outcomes. Some objectives (health and education) remain largely
focused on outcome performance measures, whereas others (income
poverty) are far more process oriented. The SDA is firmly focused on
process issues linked to each of the PSA objectives.
The PSA and SDA deal to some extent with problems of affect and attri-
bution in two ways. First, the targets are defined in relation to a smaller
number of countries rather than all developing countries. For example,
the health and education targets are set with respect to the top ten recipi-
ents of UK health and education sector support respectively. Second, the
PSA and SDA, at least to some extent, provide the logic model which
is missing from the Millennium Development Goals and International
Development Targets taken by themselves.9 However, there are also prob-
lems in the approach, principally that of data availability: as mentioned
earlier, data of the sort required are often collected on a three- to four-
year cycle at best. Hence there may well be gaps in monitoring fulfilment
of many of the PSA targets. The outcome indicators given in the PSA are
also not suitable for judging agency performance. But these outcomes can
be observed, along with the contribution that DFID may or may not have
made to their fulfilment, as judged by the input and process indicators
contained in the PSA and SDA.
The PSA shown (in Appendix 3.2) is the second produced by DFID
and differs from the first in that it no longer specifies targets for the
top thirty recipients of the various types of aid.10 The shift to a smaller
number of targets reflects a more realistic objective in terms of attribu-
tion, though it by no means solves that problem. On the other hand, it
may be thought difficult to obtain representative data for a smaller
number of countries. Though the target relates to specific countries,
proper monitoring requires data on all countries. As indicated above,
these data are not likely to be readily available for several indicators.
Moreover, it is not for DFID alone to develop the monitoring systems to
collect these data. The Millennium Development Goals and International
Development Targets potentially provide the basis for harmonised moni-
toring procedures across donors, although this has not been achieved in
the past.
The PSA does solve the problem of how to judge agency performance
by aggregating across the agency. And it probably will do so without creat-
ing burdensome reporting procedures which have weighed down other
agencies:

In USAID, for example, operating units and implementing partners

are beginning to complain that there is no time left for implementing
programs, and that much of the higher-order results data collection is
not considered directly relevant or useful to them, but is only being
used to ‘report upward’.
(Development Assistance Committee, 2000: 23)
62 Howard White
But others might argue that the disjuncture between project- and
programme-level monitoring and judging DFID’s performance is a bad
move. This aggregation is not based on activity-level performance. Indeed,
the performance measures explicitly exclude many DFID-financed activities.
So wouldn’t the Department’s performance best be judged by aggregat-
ing the performance of different activities? In practice, the data do not
exist to do so, and trying to collect them would indeed be onerous. In
recent years a database, PRISM, has been developed to contain data on
project activities, though the rules which are applied mean that only 30
per cent of projects should be included. And of the over 700 projects on
which information should be held, the output to purpose review (not
available for all projects, but a major part of monitoring) was available for
just over fifty.11 Even if PRISM were to have full coverage, it is not clear
that it will be able to generate aggregate data on agency performance (in
the manner of the World Bank described above), and certainly not to give
any information relating to the Millennium Development Goals and Inter-
national Development Targets.
At present, one must wonder on what data DFID management do base
their decisions. There is no ‘bottom-up’ system to indicate overall perform-
ance. And the goal-related indicators embodied in the PSA are of little
operational use.12
However, DFID is one of the foremost donor agencies in the related
developments of increased budget support and Sector Programmes (sector-
wide approaches, SWAPs, in DFID terminology). These are consistent with
both harmonised procedures and monitoring based on country-wide
performance indicators. To the extent that these are developed and DFID
harnesses the information they provide, then there will be some sort of
systematised feedback, though not in a form that can be readily aggregated.
However, important issues remain as to (1) the extent to which the PSA
and SDA manifest themselves in the daily work of DFID (e.g. by being
incorporated in CSPs), and (2) whether the underlying model is ‘right’.
On the first of these questions, DFID staff readily point to changes
brought about by the renewed focus on poverty. For example, DFID’s
programme to China has changed from being focused on infrastructure
in the relatively affluent seaboard provinces to social sectors in poorer
western provinces. It is debatable how much the change comes from
adopting the International Development Targets per se rather than an
increased poverty focus. Such a focus has already been there since 1990.
And, for example, during the 1990s the Zambia programme shifted from
secondary schools and support to hospitals to primary education and
health clinics. Attitudes as to how important the goals themselves are to
arguing for the poverty agenda vary between staff – but their effect is seen
as either neutral or positive; there was no suggestion that they have
detracted attention from ‘main issues’. Those who see a positive role
argue that stressing the Millennium Development Goals and International
 Donor agency performance measurement 63
Development Targets has enabled internal policy changes (abolishing Aid
and Trade Provision, more aid for South Asia) and can be a useful tool in
discussions with partners.
It is striking that DFID staff and outside experts stress that a major part
of DFID’s contribution to achieving the targets comes through its influ-
ence on partner-country policies and the actions of other actors. Yet mea-
suring impact through influence is an under-researched area.
Whether the model is right is not a matter of objective fact. Targets do
not in themselves contain the strategy as to how they should be attained,
so that competing strategies may be proposed. The changes that DFID is
making, e.g. towards selectivity, are in line with the consensus amongst
donor agencies, if disputed by some critics (Lensink and White, 2000).
Other areas of contention include the nature and depth of debt relief and
cost recovery schemes for basic services.

Data quality
DFID staff are aware of data quality issues. Whilst operational staff express
appropriate scepticism as to the quality of the data, various initiatives are
under way to promote use of the data and improve them. Indeed, an
advantage of the Millennium Development Goals and International Devel-
opment Targets is said to have been to draw attention to data quality
issues. The publication formerly called British Aid Statistics has become
International Development Statistics and includes data on the Millennium
Development Goals and International Development Targets. CSPs are
required to report a country’s performance with respect to the Inter-
national Development Targets indicators (though this has been done in
an uneven way – see Chapter 4). Staff of the DFID’s statistics department
are well informed on data quality and are active in supporting initiatives
for them to be improved, notably DAC’s PARIS initiative (Partnerships in
Statistics for Development in the Twenty-First Century).13
DFID staff do not feel that the Millennium Development Goals or
International Development Targets stress quantitative at the expense of
qualitative aspects of development since they believe the importance of
the latter is well understood. In addition, DFID has supported work by
DAC on governance indicators to quantify the governance target. When
the working group was unable to reach agreement, DFID support shifted
to ongoing efforts to develop acceptable indicators by the World Bank.14
The difficulty in selecting measures of ‘governance’ revolves in part
around disagreements as to what constitutes ‘good governance’. With
respect to human rights, European countries may wish to include the
absence of the death penalty, but this is still applied in other developed
countries. The extent of restrictions on individual freedom for reasons of
national security is another murky area. The indicators being developed
by the World Bank attempt to sidestep these issues to some extent by
64 Howard White
proposing a set of process indicators – such as civil service wages and the
nature of elections (e.g. whether they use proportional representation or
not) – which have no normative content. Performance measures on the
quality of governance – such as corruption and the predictability of policy-
making – are included separately.

Summary
The PSA and SDA have potentially provided a vehicle to enable DFID to
travel the road travelled in rather more time by USAID. That is, there is a
move away from the outcome indicators as the measure of performance.
DFID staff are indeed generally, although not universally, sceptical that the
Millennium Development Goals or International Development Targets can
be used to assess the DFID’s performance. The PSA and SDA spell out a
number of process indicators which may be used to capture DFID’s contri-
bution toward meeting the Millennium Development Goals and Inter-
national Development Targets. This role is appreciated by DFID staff. One
commented that the Millennium Development Goals and International
Development Targets could not be used to assess DFID’s performance but
that DFID should ‘track changes and make sure that their work is consis-
tent with the Millennium Development Goals and International Develop-
ment Targets’ (adding that the PSA played this role). This scepticism also
appeared in that no respondents thought that the fact that indicators are
currently not on track to meet the targets called for major changes in
DFID’s strategy. It is also shown by the ease with which the DFID Depart-
mental Review explains away deviations from the targets. However, the
PSA also does contain outcome indicators, so it is worth emphasising that
development outcome indicators of this sort are not a suitable vehicle for
judging the performance of individual agencies. And there are problems
of incomplete coverage of DFID’s activities and misalignment with the
reporting system being developed using PRISM.

Conclusions
The International Development Targets, launched in 1996, have caught
the attention of the development community, which has become stronger
with the successor Millennium Development Goals. Unlike previous targets,
they have not fallen by the wayside, but continue to be referred to and
monitored. Targets can play an important role in accountability and
performance measurement, though they are not without their disadvant-
ages. To be most effective, performance measures should satisfy a number
of criteria. The Millennium Development Goals satisfy only some of these
criteria. They are very relevant, are mostly well defined and correspond to
existing indicators. But they are mostly outcome oriented, with little effort
made to build a consensus around an underlying logic model of how the
 Donor agency performance measurement 65
targets are to be achieved. In the absence of such a model it is extremely
difficult to say anything sensible as to whether changes in target indicators
can be attributed to the actions of the development community. To put it
bluntly: the Millennium Development Goals are not suitable for judging
the performance of individual development agencies.
Amongst bilateral donors, the UK’s Department for International Devel-
opment has been a prominent supporter of the targets and has made sub-
stantial steps to internalise them. Nonetheless, this chapter has not been
able to assess the extent to which the model contained in the Public
Service Agreement and Service Delivery Agreement has influenced DFID’s
work on the ground. The targets have been central to the two White
Papers produced since 1997 and strategy papers have been produced that
seek to show how the targets may be achieved. The PSA and SDA contain
performance measures which are related to the international goals, but
also include inputs, process and outputs. Hence they contain a model of
how to achieve these interim targets and so, implicitly, the Millennium
Development Goals themselves. To the extent that performance is judged
by these intermediary indicators, the PSA and SDA represent an improve-
ment. However, worries remain over the misalignment between these top-
down targets (which have incomplete coverage of DFID’s activities) and
the nascent bottom-up system in PRISM. A key area of further investiga-
tion is the extent to which the SDA affects the work of the various parts of
DFID.
Appendix 3.1 Data coverage of baseline data for the International Development Targets

East Asia Europe and Latin Middle East South Asia Sub-Saharan World
and Pacific Central Asia America and and North Africa
Caribbean Africa

Number of countries 20 27 35 15 8 48 200

Poverty head count Countries 6 9 9 5 4 11 44
% of pop. 83 71 65 52 78 36 71
Poverty depth Countries 4 16 10 3 4 11 48
% of pop. 86 71 65 52 78 36 57
Poor’s consumption share Countries 8 16 15 7 5 18 69
% of pop. 86 72 73 57 86 58 86
Underweight Countries 12 10 25 1 8 45 120
% of pop. 98 66 99 28 100 100 89
Net primary enrolment Countries 9 16 24 13 2 30 127
% of pop. 88 73 90 97 12 56 69
Completion fourth grade Countries 11 8 17 9 4 32 100
% of pop. 84 28 81 50 78 61 68
Literacy Countries 8 12 25 13 7 36 120
% of pop. 97 68 100 89 100 83 83
Girls to boys enrolments Countries 14 25 25 13 7 42 161
% of pop. 94 99 66 97 100 98 96
Female to male literacy Countries 28 8 6 10 21 4 77
% of pop. 78 95 94 63 98 3 84
Infant mortality rate Countries 20 27 35 15 8 48 200
% of pop. 100 100 100 100 100 100 100
Under-five mortality rate Countries 20 27 33 15 8 48 187
% of pop. 100 100 99 100 100 100 100
Maternal mortality rate Countries 12 26 23 11 5 38 144
% of pop. 100 99 99 97 98 97 100
Attended births Countries 14 18 14 12 8 39 137
% of pop. 42 99 62 97 100 88 86
Contraceptive prevalence Countries 5 6 15 8 3 29 69
% of pop. 90 43 61 76 95 84 75
Process for sustainable Countries 34 7 16 5 18 6 86
development
Access to safe water Countries 21 9 33 13 7 45 151
% of pop. 100 26 100 85 100 89 89
Land area protected Countries 27 12 23 13 5 39 146
% of pop. 100 100 99 100 98 98 100
GDP per unit of energy use Countries 6 24 21 12 5 18 116
% of pop. 95 97 96 100 98 76 95
Carbon dioxide emissions Countries 20 26 35 13 9 45 ..
% of pop. 100 100 100 100 100 99 ..

Source: OECD Methodological Note DCD/DAC (98) 6/ADD, Paris: OECD/DAC (available at [Link] [Link]).
Appendix 3.2 DFID Public Service Agreement and Service Delivery Agreement. Aim: the elimination of poverty in poorer countries

Objectives Performance targets Delivery

I To reduce poverty 1 An increased focus by DFID on poor Successful delivery depends on DFID and multilateral
through the provision countries, particularly those with effective institutions becoming more selective, and focused on
of more focused and governments pursuing high growth and poverty reduction. DFID will therefore:
co-ordinated pro-poor economic and social policies, as • Deliver a more effective and focused bilateral programme
development assistance demonstrated by: by: (a) working with partners in poor countries to deliver
by the international • An increase in the percentage of country strategies which support poverty reduction; and
community to low- and DFID’s bilateral programme going to (b) allocating DFID support, taking account of numbers
middle-income poor countries, particularly those with of poor people, the effectiveness of country programmes,
countries. favourable policy environments; and partners’ progress in developing and implementing
• An increase in the percentage of sound pro-poor policies.
European Commission (EC) • Seek to improve the effectiveness of EC development
development assistance going to poor assistance and the European Development Fund by
countries and; working with other government departments (especially
• Adoption and implementation of Foreign and Commonwealth Office, and Her Majesty’s
effective Poverty Reduction Strategies (HM) Treasury) and European Union (EU) member
by 2004 in all countries accessing states to: (a) establish better organisation of EC
International Development Agency programme delivery, by end-2001; (b) gain agreement in
(IDA) high-impact or adjustment Council and Commission to redirect allocations and
lending. spend towards programmes which reduce poverty by 2003;
(c) increase the proportion of EC country-specific Official
Development Assistance (ODA) going to poor countries
from 50 per cent in 1998 to 70 per cent in 2006.
Where poor countries have demonstrated a clear
commitment to developing and implementing
comprehensive Poverty Reduction Strategies, donors should
respond by supporting delivery of these Strategies. DFID will
therefore:
 • Provide support to at least twelve partner countries by
2004 to develop and implement Poverty Reduction
Strategies in co-ordination with other donors.
II To promote sustainable 2 To promote the integration of Relief of unsustainable debt burdens is essential if poor
development through developing countries into the global countries are to harness the resources they need for
co-ordinated UK and economy through co-ordinated UK and economic growth and development, to reduce poverty and
international action. international action, including by: reap the benefits of globalisation. DFID will therefore:
• Relief of unsustainable debt by 2004 for • Work to secure faster, wider and deeper debt relief for the
all Heavily Indebted Poor Countries poorest countries, through effective implementation of
(HIPCs) committed to poverty reduction, the Heavily Indebted Poor Country (HIPC) Initiative.
building on the internationally agreed Effectiveness will be determined by the involvement of all
target that three-quarters of eligible creditors (including securing the necessary financing for
HIPCs reach decision point by end-2000 multilateral creditors), the extent of front-loading
(joint target with Her Majesty’s (meaning that more of the benefit of debt relief is felt in
Treasury); and the early years (including from Decision Point)), the
• Gaining international agreement on the speed of the process and the strength of the link to
integration of social, economic and poverty reduction, ensuring that debt relief assists
environmental aspects of sustainable countries to implement their national poverty reduction
development into poverty reduction strategies and achieve the international development
programmes. targets.
To ensure that development is sustainable over the long
term and benefits future as well as current generations, poor
countries need to integrate sustainable development into
their policies and programmes. DFID will therefore work
towards:
• Developing guidance on the principles of sustainable
development, securing OECD Development Assistance
Committee (DAC) agreement to it by mid-2001, and work
to secure wider international agreement by end-2001;

continued
Appendix 3.2 continued

Objectives Performance targets Delivery

• Successful integration of these principles into

government, multilateral and DFID policies and
programmes in ten key DFID partner countries by early
2004, including agreed approaches to water resources
management, and capacity-building for environmental
management.
Successful integration of poor countries into the global
economy will depend on the creation of a supportive
environment in which trade and enterprise can flourish, and
contribute to poverty reduction. DFID will therefore:
• Work with the public and private sectors to improve the
business environment, especially access to finance and
other business services for enterprises that employ or
benefit the poor.
• Promote increased private-sector foreign investment in
poor countries by turning the Commonwealth
Development Corporation (CDC) into a public–private
partnership, when business conditions are right, with
majority private capital. CDC is required to make 70 per
cent of its new investments in poor developing countries
and seeks to make 50 per cent of its new investments in
sub-Saharan Africa and South Asia.
• Work with the EU and other partners for multilateral
trade negotiations to improve trading opportunities for
poor countries, whilst working with other donors to
deliver more effective capacity-building support for poor
countries so that they are equipped to participate fully in
the international trading system.
 Effective action to tackle HIV is essential if poor countries
are to sustain economic growth, development and poverty
reduction. DFID will therefore:
• Work with partners in countries with high, or increasing,
HIV prevalence to develop and implement strategies
which intensify multi-sector and co-ordinated
international action on HIV/AIDS.
3 Improved effectiveness of the UK Successful delivery depends on governments, donors,
contribution to conflict prevention and international bodies, civil society groups, the private sector
management, as demonstrated by a and others co-operating closely in the design and delivery of
reduction in the number of people coherent, complementary policies and interventions in
whose lives are affected by violent conflict order to defuse tensions, reduce violence, tackle the factors
and by a reduction in potential sources that underlie armed conflict, and build governments and
of future conflict, where the UK can institutions capable of sustaining peaceful and democratic
make a significant contribution (joint societies.
target with FCO and MoD).
Where the UK can make a significant contribution, DFID,
FCO and the Ministry of Defence (MoD) will work in
partnership with others to:
• Strengthen international and regional systems and
capacity for conflict prevention, early warning, crisis
management, conflict resolution/peacemaking,
peacekeeping and peace-building.
• Contribute to global and regional conflict prevention
initiatives, such as curbing the proliferation of small arms
and the diversion of resources to finance conflict.
• Promote initiatives in selected countries, including
indigenous capacity building, to help avert conflict,
reduce violence and build sustainable security and peace.

continued
Appendix 3.2 continued

Objectives Performance targets Delivery

III Improved education 4 Improved education systems in our top Successful delivery depends on donors and poor-country
outcomes in key ten recipients of DFID education partners working together to design and deliver effective
countries receiving support demonstrated by: policies and support for education. DFID will therefore work
DFID education support. • an average increase in primary school in partnership with others to support:
enrolment from a baseline established •Implementation of the agenda agreed by the
in 2000 of 75 per cent to 81 per cent on International Community at the Dakar World Education
the basis of data available in 2004; and Forum in April 2000 through the provision of focused
• improvements in gender equality in support by relevant multilaterals, partner countries,
education, particularly primary bilateral donors and non-governmental organisations.
education • Successful adoption and implementation of education-
sector strategies which include explicit objectives on
equitable access for girls and boys by 2004, in at least eight
of our top ten recipients of bilateral education assistance.
• Development of basic monitoring and evaluation
mechanisms and their integration into education sector
strategies by 2004 in at least eight of our top ten recipients
of bilateral education assistance.
IV Improvements in 5 Improvements in child, maternal and Successful delivery depends on donors and poor-country
health outcomes in key reproductive health in our top ten partners working together to design and deliver effective
countries receiving recipients of DFID health care assistance policies and support for health. DFID will therefore work in
DFID health care demonstrated by: partnership with others to support:
assistance. • a decrease in the average under-five • Development and implementation of strategies focused
mortality rate from 132 per 1,000 live on improving access to safe water and sanitation and
births in 1997 to 103 on the basis of reducing levels of child mortality in at least eight of the
data available in 2004; top ten recipients of bilateral health assistance by 2004.
 • an increase in the proportion of births • Development and implementation of health-sector
assisted by skilled attendants from a strategies by 2004 in at least eight of the top ten recipients
baseline established in 2000 of 43 per of bilateral health assistance which: (a) aim to improve
cent to 50 per cent on the basis of data child health outcomes and include actions to strengthen
available in 2004; and immunisation and prevention, and the treatment of
• improved access to reproductive health childhood illnesses, including malaria where endemic;
care. and (b) include explicit policy and operational
frameworks to strengthen the capacity of health systems,
improve the quality and coverage of maternal health care,
and ensure universal access to reproductive health
services.
• Strengthened multilateral initiatives to combat HIV/AIDS
in Africa (UNAIDS) and roll back malaria (WHO)
demonstrated through national strategies, with jointly
agreed milestones, in at least five of the top ten recipients
of DFID health-care assistance.
Value for money. 6 Improved value for money and Successful delivery of improved value for money will be
effectiveness of projects in DFID’s measured by the index of evaluated success. This depends
bilateral programme, as demonstrated on annual project scoring and risk labelling of projects;
by a year on year improvement in the roll-out and full use of Performance Reporting Information
index of their evaluated success. System for Management by 2001; effective quality control
and monitoring.

Source: [Link]
74 Howard White
Notes
1 This chapter is based on a report prepared for the National Audit Office’s
(NAO’s) review of performance measurement in the Department for Inter-
national Development (NAO, 2002). Thanks to Robert Owen of NAO for support
to that work which has informed my views. Useful comments on an earlier draft
were received from Richard Black. The usual disclaimer applies.
2 For a discussion of data quality issues, see Divorski and Scheirer (2000).
3 ‘Logic models’ is the preferred term in the literature; these are very similar to
the concept of the logical framework (log-frame) familiar in development
agencies. Cummings (1997) distinguishes logic models, the logical framework
and results-based management only to conclude that the three are closely
related.
4 This and the following sections draw on a questionnaire completed by a
number of key informants within the DFID.
5 This statement is based on the assessment of the DFID staff interviewed and on
the author’s experience of leading a poverty training programme targeted at
all DFID staff.
6 The income-poverty line is a dollar a day, but a dollar buys much more in, say,
Kampala or Delhi than it does in New York. The poverty line is equal to the
local cost of purchasing the goods which could be purchased for one dollar in
the USA.
7 See Carvalho and White (1996: 9–17) for further discussion.
8 The figure is from the DFID Poverty Guidance, Bridging the Gap, at [Link]
dfi[Link]/Pubs/files/poverty_bridgegap_guidance.pdf.
9 Sketching the logic models underlying the various objectives shows them to be
somewhat patchy – that for health is the most complete. They are particularly
weak on outputs.
10 Comments by NAO on an earlier version of this chapter asked what was the
view of UK academics of this change. I sent an e-mail question to ten academics
who ‘do aid’ (there are not so many who do). Of the seven replies, three had
not heard of the PSA, one had but was unsure what it was, and the other three
had heard of it but were unaware of the change. DFID does not generally draw
on the academic community for these ‘management tools’ (policy documents
are a different matter), and even academics with a close relationship with
DFID, or a good working knowledge of aid, are usually vague on the Depart-
ment’s internal workings.
11 Personal communication from Michael Flint, main author of the DFID’s first
development effectiveness report.
12 The questionnaire sent to some DFID staff in preparing this chapter asked
what the Department’s response should be to the fact that A Better World for All
showed that none of the Millennium Development Goals or International
Development Targets would be met at the global level. Virtually all respon-
dents thought no response was necessary – clearly illustrating that data on the
goals do not yield information of operational significance.
13 See [Link]
14 For a summary of this work see [Link]
[Link]. This site includes a discussion of available indicators including
those from other sources such as Freedom House.
 Donor agency performance measurement 75
References
Carvalho, S. and White, H. (1996) ‘Implementing Projects for the Poor: What Has
Been Learned?’ Directions in Development, Washington, DC: World Bank.
Collier, P. and Dollar, D. (2002) ‘Aid Allocation and Poverty Reduction’, European
Economic Review, 26: 1475–1500.
Cummings, F.H. (1997) ‘Logic Models, Logical Frameworks and Results-Based
Management: Contrasts and Comparisons’, Canadian Journal of Development
Studies, 18: 587–596.
Demery, L. and Walton, M. (1999) ‘Are the Poverty and Social Goals for the 21st
Century Attainable?’, IDS Bulletin, 30(2): 75–91.
Development Assistance Committee (DAC) (1996) Shaping the Twenty-first Century,
Paris: OECD.
—— (2000) Results Based Management in the Development Co-operation Agencies: A
Review of Experience (Executive Summary), Paris: OECD.
Divorski, S. and Scheirer, M.A. (2000) ‘Improving Data Quality for Performance
Measurement: Results from a GAO Study of Verification and Validation’, Evalu-
ation and Program Planning, 24(1): 83–94.
General Accounting Office (GAO) (1996) Contributions to Child Survival Are Signifi-
cant but Challenges Remain, Washington, DC: GAO.
—— (1998) The Results Act: An Evaluator’s Guide to Assessing Agency Performance,
GAO/GGD-10.1.20, Washington, DC: GAO.
—— (1999) Managing for Results: Opportunities for Continued Improvements in Agen-
cies’ Performance Plans, GAO/GGD/AIMD-99-215, Washington, DC: GAO.
—— (2000) Observations on the US Agency for International Development’s Fiscal Year
1999 Performance Report and Fiscal Years 2000 and 2001 Performance Plans, Wash-
ington, DC: GAO.
Hakes, J.E. (2000) ‘Can Measuring Results Produce Results: One Manager’s View’,
Evaluation and Program Planning, 24: 319–327.
International Monetary Fund, Organisation for Economic Cooperation and Devel-
opment, United Nations and World Bank (2001) A Better World for All, Washing-
ton, DC.
Jackson, A. (2000) ‘An Evaluation of Evaluation: Problems with Performance Mea-
surement in Small Business Loans and Grant Schemes’, Progress in Planning, 55:
1–55.
Lensink, R. and White, H. (2000) ‘Assessing Aid: A Manifesto for Aid in the 21st
Century?’, Oxford Development Studies, 28(1): 5–17.
Maudling, W.P. (1994) ‘Maternal Mortality in Developing Countries: A Compari-
son of Rates from Two International Compendia’, Population and Development
Review, 20(4): 413–421.
Maxwell, S. (1998) ‘International Targets for Poverty Reduction and Food Secur-
ity: A Mildly Sceptical but Resolutely Pragmatic View with a Call for Greater Sub-
sidiarity’, Canadian Journal of Development Studies, 19, Special Issue 1998.
Millar, A., Simeone, R. and Carnevale, J. (2000) ‘Logic Models: A Systems Tool for
Performance Management’, Evaluation and Program Planning, 24(1): 73–81.
National Audit Office (2001) Measuring the Performance of Government Departments,
London: The Stationery Office.
—— (2002) Performance Management: Helping to Reduce World Poverty, London: The
Stationery Office.
76 Howard White
Scheirer, M.A. and Newcomer, K. (2000) ‘Opportunities for Program Evaluators to
Facilitate Performance-Based Management’, Evaluation and Program Planning,
24(1): 63–71.
United States Agency for International Development (USAID) (2001) FY 2000
Performance Review, Washington, DC: USAID.
White, H. (ed.) (1998) Journal of International Development: Special Issue, on 1997 UK
White Paper on International Development.
White, H., Killick, T., Kayizzi-Mugerwa, S. and Savane, M.-A. (2001) African Poverty
at the Millennium, Washington, DC: World Bank.
World Bank (2001) ‘Progress toward the International Development Targets: Sub-
mission by the World Bank to UK Select Committee on International Develop-
ment’, Washington, DC: World Bank.
4 Using development goals to
design country strategies1
Howard White and David Booth

Introduction
As described in Chapter 3, the UK’s Department for International Devel-
opment (DFID) has embraced the International Development Targets,
and subsequently the Millennium Development Goals, as central to its
mission. However, if targets are to mean much in practical terms, they
need to be pursued in a concrete fashion at the country level. Chapter 3
described DFID’s overall strategic approach, in which strategy papers play
a key part. This chapter focuses on DFID’s Country Strategy Papers (CSPs)
as an example of country-level priority setting. It reviews the first CSPs
completed under the ‘new regime’ following the 1997 White Paper on
International Development and the extent to which they promote a new
approach to poverty reduction. Our main argument is that there appears
to be a ‘missing middle’ in the CSPs. Whilst the poverty reduction object-
ives, linked to international targets, are clearly stated, and planned spend-
ing for the coming years is laid out, there is little to connect the two.
Most donor agencies define a strategy for each major partner country
on a three- to five-year cycle. DFID is no exception, so that CSPs are not
new to the agency. However, a new set of guidance was issued for these
papers following the 1997 White Paper, which attempted to make them
more strategic in nature in order to mainstream poverty reduction into
DFID’s activities. This chapter assesses how successfully this has been
done. In the next section we review the first thirteen of the ‘new CSPs’,
discussing the contents of their poverty analysis, what they say about
partnership and how all this informs the future DFID strategy proposed
for the country. In doing this we use the DFID guidance on CSPs as a
benchmark for what they should contain. The subsequent section deepens
the discussion, with reference to a set of particular problems that are
known to pose particular challenges to country programme design and
management within the current policy framework.
78 Howard White and David Booth
DFID’s Country Strategy Papers
The CSP is the key document which outlines DFID’s intended programme
in the country concerned for the coming five years. According to the guid-
ance for these papers, a CSP must comprise the following five sections: A)
Summary; B) The Challenge (outlining the nature of poverty and the
main poverty reduction objectives); C) Partnerships (other stakeholders
involved in poverty reduction efforts); D) Current UK Development Port-
folio; E) Future UK Development Strategy (overview of strategy); F)
Implementing the New Strategy (details of programmes and projects to be
supported); G) Programme Resources (table of the allocation of resources
for the coming period). This structure can be labelled a logical one,
embodying the logic model which Chapter 3 argues to be important to
good strategy. In addition, CSPs are intended to be produced in a collabo-
rative manner and the process of CSP preparation has to be summarised
in an annexe to the paper. In this section we discuss the thirteen ‘new’
CSPs available at the time of our review.2
Before we proceed, a caveat is in order. The following analysis is based
entirely on the content of the CSPs. The comments are thus not to be
taken as judgements on the programmes, which may or may not have
shortcomings that are suggested by the CSPs. However, the logic of the
exercise being carried out here implies that if a CSP is deficient, so is the
programme to which it refers. There are clearly cases when something
probably has been done but was missed by us simply because the CSP does
not mention the fact. We realise that there is a tight space constraint on
CSPs and that the demands on what they must do in this space are high.
In this section we first look at partnerships and the process of prepar-
ing a CSP. We then turn to the adequacy of the poverty analysis and the
strategy laid out in the papers.

Partnerships and process

DFID alone will not achieve the poverty reduction goals in any of these
countries. It will, however, contribute to a joint effort from a variety of
stakeholders. Hence partnership is central to the strategy, which implies
that the process by which the strategy is developed should be a collabora-
tive one. A note of caution should be expressed on the problem of donor
proliferation. If every donor wishes to have a collaborative process on ‘their’
strategy for a country, this will greatly consume the time and resources of
government and other stakeholders. The ideal is of course for a govern-
ment-led process defining the overall strategy which allocates roles to the
various partners, thus defining the strategy for each of them. In principle,
this is what Poverty Reduction Strategy Papers (PRSPs) are intended to
do. In practice, it is too early for most of them to have greatly influenced
what is actually being done. And there are varying degrees to which donor
 Designing country strategies 79
agencies take the PRSP as their reference point. More progress is being
made in at least some countries with government-led initiatives at the
sector level with what are called sector programmes, or the sector-wide
approach (SWAP).
The appropriate partners are those engaged in poverty reduction, with
differing degrees of involvement with government depending on the
seriousness of that government’s anti-poverty programmes. There is no
systematic basis for assessing whether a particular government is to be con-
sidered pro-poor or not, hence there is little discussion in the CSPs of the
different partnership arrangements that are appropriate under different
circumstances. Guidance suggests that there can be ‘high’ and ‘low’ scenarios,
with the extent and nature of DFID support varying according to the
government’s stance. This option is adopted only in the Kenya CSP.
However, the discussion of government does usually at least state whether
or not the government is committed to poverty reduction (sometimes also
gender equality) and whether its objectives are consistent with the Inter-
national Development Targets.
The CSPs adopt a far less critical stance with respect to the donor
community. The discussion of partnerships with other bilateral donors
and multilateral agencies is largely a description of their areas of involve-
ment, often with a note that more donor co-ordination would be desir-
able. The idea that other donors might be an important channel for
influence appears only sometimes, and the issues on, and means by, which
it might be desirable to influence other donors are hardly considered.
There is thus no serious discussion of whether the programmes of the
various agencies, of which the World Bank and International Monetary
Fund (IMF) are invariably key players, are consistent with poverty reduc-
tion priorities.
Similarly, civil society is invariably seen as a good thing, whereas one
can imagine possibilities for a more nuanced discussion. In the real world,
organisations within civil society have strengths and weaknesses that often
closely reflect those of government.
Each CSP reports on the country strategy preparation process. This is
done in varying degrees of detail, ranging in length from half a page to
two pages. The consultation process can be divided into the part that
happens in-country and the part that takes place elsewhere. Within each
part we distinguish three stages: (1) initial consultation (in-country only);
(2) feedback on/formulation of the actual strategy; and (3) review of the
draft CSP. Although the distinction between these different stages is not
always watertight, it is clear that opportunities to influence the general
nature of the programme are greater the earlier in the preparation
process that consultation begins.
In general, there were initial consultations with, and limited to, govern-
ment. There were, however, exceptions, where a wider range of stakehold-
ers was involved from the start, and others (two) in which initial strategic
80 Howard White and David Booth
thinking was internal to DFID and based on commissioned studies. Con-
sultations continued and widened in most countries as the strategy
evolved, though in rather fewer cases were actual drafts subject to this
process. In-country consultations appear to have been most systematically
and comprehensively approached in South Africa.
Consultation with the private sector and academics during strategy
formulation in the UK was common,3 and in some cases this included non-
governmental organisations (NGOs) active in the country. Other donors
were also consulted, though this was most usually in-country.

Poverty analysis
As discussed in previous chapters, international development goals are
designed, at least in part, to focus the attention of both donor and devel-
oping country on the issue of poverty reduction. CSPs are expected to
crystallise this focus at the country level: ‘CSPs set out the strategy by which
DFID will achieve its objectives, particularly the elimination of poverty.’4
The first, and most obvious, question is: do the country strategies
address poverty? In practice, poverty emerges as a key issue in all the CSPs
reviewed, especially in the ‘Challenge’ section. Explicit mention of poverty
issues, or identification of links to poverty reduction targets, is less common
in discussions of the future UK development strategy. In addition, and as
will be clear from what follows, the extent of the poverty analysis under-
taken varies quite widely between CSPs.
The foundation of the poverty analysis comes in ‘The Challenge’,
which ‘should analyse the economic, social, political and environmental
situation and recipient Government policies. It should assess the causes,
characteristics and consequences of poverty in the country and should
review the country’s progress with respect to the main International Devel-
opment Targets.’5 This is quite a demanding task, requiring both a
poverty profile and identification of the main causes of poverty in a
couple of pages, along with policy analysis of the recipient government’s
position.
A second set of questions relates to how poverty is defined in the
country strategies and programmes. What concepts and indicators of
poverty are used? How have poor people’s own perspectives been incorpo-
rated into the analysis?
The tables in the appendices to this chapter provide the details on
which our discussion is based. For example, Appendix 4.1 presents data
on the definition of poverty used and incorporation of the perspectives of
the poor, whilst Appendix 4.2 records which indicators are reported
(characterised as International Development Target-related and others).
Although the majority of CSPs also report social indicators, it seems,
despite adoption of the multidimensional International Development
Targets, that the notion of income poverty as the ‘real meaning’ of
 Designing country strategies 81
poverty is quite resilient. A locally defined poverty line is used most fre-
quently, sometimes accompanied by the ‘dollar a day’ line.6 Of course, a
head count based on a locally defined line in itself tells us little, but it does
provide a benchmark against which to measure progress.
The four Southern African countries give the most comprehensive treat-
ment. The South Africa CSP reports the current status of all the Inter-
national Development Target-related indicators in a single table, whereas
other papers give them section by section. But overall, less than half the
countries meet the requirement to report progress in meeting the Inter-
national Development Targets; in the majority of the cases even the current
levels are not reported. None of the CSPs for Asian countries reports more
than one or two International Development Target-related indicators. The
most commonly reported non-International Development Target indicators
are life expectancy and literacy, with gender disaggregation of the latter
being common. Explicit mention of the results of participatory analysis is
rare and only in one case does the discussion lead on participatory results.
Moving from the poverty profile to the causes of poverty, the relevant
questions are: (1) how has poverty been analysed?; and (2) what balance is
evident between social and economic perceptions of poverty? The causes
of poverty are addressed only indirectly in many CSPs. Most commonly, a
leading section on poverty pays little attention to causes. On the other
hand, sections follow on the economy, human development, politics and
the environment, which may be seen as indirectly discussing causes of
poverty. But such links are only rarely explicit.
Although some CSPs give greater emphasis to economic causes of
poverty (including the distribution of income and wealth, which is given a
particularly central role in South Africa and Vietnam, but mentioned in
many other cases), they all give treatment of social issues, the most com-
monly mentioned being gender discrimination. It is notable that the CSPs
do not shy away from mentioning political constraints to poverty reduc-
tion, and problems in government (most usually corruption) may even be
identified as a central constraint on reducing poverty (e.g. in the high-
lighting of patronage in the Nepal CSP).
That said, neither this section nor that on partnership can be seen as
representing a systematic attempt to identify the extent of genuine govern-
ment commitment to poverty reduction, gender equity or the achievement
of the International Development Targets. Within DFID there has been
work by both the Caribbean and the Africa regions in developing checklists
to assist in identifying pro-poor governments (covering macro policy
stance, social service provision, accountability and respect for human
rights). But these still have some way to go in tackling the important con-
ceptual issues posed by the development of such checklists.7
82 Howard White and David Booth
Strategy: the missing middle
Whilst poverty reduction objectives are linked to the International Devel-
opment Targets, how well are these objectives followed through in the
proposed strategy? A general question here is ‘What is a strategy?’ The
CSPs themselves appear somewhat confused on this subject, and the guid-
ance tends to tiptoe around the topic rather than confronting it directly.
Thus in many CSPs it is difficult to work out what ‘the strategy’ is, in the
sense of the intended linkage between the goals/overall objective/purpose
and DFID-supported activities. The guidance requirement that explicit
attention be paid to linking activities to poverty reduction is often disre-
garded. We found it difficult to find explicit links between the poverty
analysis and the proposed future programme, partly on account of the
weakness noted earlier, that explicit consideration of the causes of poverty
is missing from the poverty analysis.
These observations lead us to identify a general problem of ‘the missing
middle’ in CSPs. The middle that is insufficiently developed (or at least,
under-reported in CSPs) includes (1) distinguishing and ranking the
causes of poverty corresponding to the poverty profile in ‘The Challenge’,
and (2) being clear and precise about how the DFID programme proposes
to break into the circles of causation that have been identified, and why
the activities chosen are the best from this point of view.
An important distinction can be made between enabling, inclusive and
focused (or targeted) interventions (see Box 4.1 for the definition of
these terms). This is partly because these are collections of activities,
which individually could be more readily classified. In only one case could
the classifications be quantified. However, it is clear that focused/targeted
interventions are very much in the minority.8
In principle, this last observation does not say anything about the degree
to which country programmes are rigorously constructed for maximum

Box 4.1 Poverty Aim Marker (PAM) classifications

Enabling actions which underpin pro-poor economic growth and the broader
policies and context for poverty reduction and elimination, and will lead to
social, environmental and/or economic benefits for poor people.
Inclusive actions (e.g. sectoral development programmes) which aim to
benefit broad-based population groups, including poor people, but also
address issues of equity and barriers to participation or access of poor
people.
Actions focused predominately on the rights, interests and needs of poor
people.

Source: DFID Statistics Department, Policy Information Marker System; Policy Aim Marker;
Policy Objective Marker
 Designing country strategies 83
impact on poverty, and are in that sense poverty focused. Logic alone does
not compel any particular balance between the three types of action.
However, we suspect that the actual balance reflects the fact that there has
been quite a lot of ‘re-dressing’ of current programmes that are skewed
towards the enabling and inclusive approaches, a suspicion confirmed by
looking at the CSPs prepared before the 1997 White Paper.

More on the missing middle

The problem of the missing middle suggests that a logical framework (log-
frame) approach might usefully be applied. Indeed, a log-frame is actually
required by the guidance, but it does not constitute a part of the published
report (and so, we understand, has not been produced in all cases – only
one has been available to us). Table 4.1 shows a log-frame based on the
Malawi report.9 The Malawi example is not being singled out as an example
of bad practice. To the contrary, it is a case where such a treatment may
easily be applied and where useful steps have been taken which help to illus-
trate the challenges facing country programme design generally.
In the Malawi case a logic does seem to flow from goal through to activi-
ties. But this is done without reference back to the causes of poverty (the
goal is said to be set that way, but not the selection of activities). The main
causes of poverty are seen as low growth with high inequality, poor human
resource development (and the problem of AIDS), and a legal system
which does not meet the needs of poor people. Questions that should
thus be raised include: (1) do the planned activities conform to these pri-
ority areas?; (2) do the planned activities confront the main concerns in
the priority areas?; (3) do the planned activities engage government and
other actors in a way which prioritises poverty reduction?; and (4) how
can the link between these activities and poverty reduction be monitored?
Underlying several of these questions is the degree to which the shape
of the UK development programme is determined by inertia. Some role
for inertia is inevitable, as ongoing commitments must be honoured.
However, that being the case, it seems to us that it would be better to be
honest about this and distinguish between the elements of the pro-
gramme that are justified in strict log-frame terms, and those that will be
allowed during a period of restructuring the programme.10

Outcomes, impacts and monitoring

It is also important to consider what the intended outcomes of the strategy
are. Relevant questions are: (1) how will poverty impacts be identified?;
(2) what are the indicators for assessing achievements?; and (3) how will
this impact be measured?
Most CSPs do not give any explicit outcome targets (Appendix 4.3),
and the goals they do have are not usually quantified. Some are explicitly
Table 4.1 Tabular presentation (modified log-frame) of future UK development strategy: Malawi

Level Goal Purpose Specific objectives Impact areas Activities/interventions

Content To contribute To build Policies and actions Rethinking the role of Advice on restructuring; financing
to the partnerships that which promote government, including redundancy package; co-finance
elimination promote poverty sustainable livelihoods decentralisation the United Nations Development
of poverty in elimination, equity Programme’s (UNDP’s) pilot
Malawi and human rights decentralisation programmes;
in Malawi influence government and World
Bank to protect key services during
restructuring
Economic and Budget support; debt relief; advice
financial management on public financial management
(including customs); support to
Poverty Policy Unit and Anti-
corruption Bureau
Human rights Support to police force and prison
service; radios to women’s groups
Sustainable rural Advocacy and assistance with land
livelihoods reform; advice and support to enhance
soil fertility; rural pubic works; credit
and training
Better education, Better access to and Support health sector strategy and
health and quality of basic health projects in contraceptive supply,
opportunities for and education TB care and reproductive health.
poor people Supporting Policy Investment
Framework for education, and
community schools project and
support for teacher development
Protection and better Safeguarding the Support for Forestry Action Plan
management of the environment and for community initiatives for
natural and physical biodiversity protection
environment
Determined by: Analysis of the From White Paper Poor people’s priority Not stated
challenge, and needs, DFID’s
consultations comparative advantage;
with civil society and linkages to goal
and government and purpose
 Designing country strategies 85
vague. Discussion of measuring poverty impact or of monitoring proce-
dures is the exception rather than the rule. There is a major difficulty in
monitoring DFID impact with reference to national indicators such as
infant mortality or the poverty head count. Even when sound, these indic-
ators change slowly and respond to myriad factors that are impossible to
disentangle. These difficulties (which may anyhow be insoluble) are exac-
erbated when funds are channelled through sector programmes, or when
working through partnership arrangements (so that influence is a major
objective). But these difficulties should be confronted rather than ignored.
Guidance has commented that CSPs tend to be rather grandiose and long
term in their targets and proposed monitoring, whereas a more day-to-day
set of indicators (say, on involvement with government) would be more
appropriate. We agree.

Discussion
The points raised in the previous section highlight some priority problems
in country programme design that call for attention. We do not think the
more important of these issues are technical deficiencies arising from a
particular document format. They are at the heart of what it means to be
more strategic about reducing poverty and meeting DFID’s other central
objectives.
Several issues for further discussion flow directly from the above discus-
sion of the CSPs. Others that we want to suggest are a selection of espe-
cially timely and difficult specific examples of the same central challenges
to country programme design. Some brief additional comments are made
in this section to suggest why these issues seem to call for closer attention.
Key issues concern the requirements of a strategic approach to facili-
tate the following:

• Pro-poor partnerships – by developing the criteria for a selective approach

to partners and spelling out the implications for the content of
country programmes.

Selectivity is a key feature of the White Paper’s concept of partnership for

poverty reduction. The distribution of DFID resources is supposed to
reflect an assessment of the level of commitment of government and
other potential partners to achieving the International Development
Targets or equivalent national objectives. Government commitment is
notoriously hard to assess, for both conceptual and empirical (evidence)
reasons. For these and other reasons, new CSPs do not travel at all far
down this road. Yet it is hard to see how it is possible to be strategic about
poverty without at least some guiding principles for such an assessment
(including implications for the CSP preparation process) and clear ideas
about the implications of different conclusions for programme content.
86 Howard White and David Booth
• Overall country programme design – by filling in the missing middle
between the poverty profile and the proposed activities.

As we have seen, this is a weakness in the explicit content of CSPs, if not in

the country programmes themselves. There is more vagueness than there
needs to be about general and country-specific causes of poverty. It is
often not clear that instruments have been selected because they are
judged to be the most effective way of breaking into the chain of causa-
tion. Other criteria (honouring existing commitments, etc.) are operating
but are not made explicit. Potentially important instruments which are
also key partnership issues, such as influencing other donors and agen-
cies, do not seem to be taken as seriously as they might be. Monitoring of
effectiveness is discussed in ways that reflect the missing middle rather
than helping to solve the problem.

• In-country arrangements for attracting and using heavily indebted poor country
(HIPC) debt relief – by articulating the form and content of DFID influ-
ence on government in this area.

Influence strategies have been identified by recent evaluations as a typ-

ically under-articulated and under-appraised component of country pro-
grammes. Yet providing clear, consistent and timely advice to a
government partner is potentially one of the most powerful ways DFID
can contribute to poverty reduction. The shaping of a government’s
arrangements for attracting and deploying new forms of external bud-
getary support, including debt relief, is an important current area where
influence of this kind could yield important results.
But how and in what direction should influence be exercised in this
case? The answers are by no means obvious. There is an ongoing debate
amongst economists about whether it is strictly true that all aid is fungible
– that is, impossible to tie down to particular uses because of the way
external finance releases domestic funds for other uses. The idea that aid
is fungible has recently gained common currency, with a push coming
from the World Bank’s Assessing Aid report. This bears on whether it is
sensible to channel new money into earmarked ‘poverty funds’ (the
‘Uganda model’). Though it may assist in the international politics of debt
relief and the domestic politics of poverty reduction, earmarking also may
conflict with building up a country’s capacity for outcome-oriented bud-
geting and resource management. Public finance issues at this point
merge with social development and governance questions. These include
the conditions for improving morale, effectiveness and accountability in
the civil service, and the likelihood of any of these problems being more
readily solved under a decentralised government system.

• Increasing the likelihood of ‘pro-poor growth’ – by influencing the design of

structural reforms.
 Designing country strategies 87
This is another area in which DFID influence could have been more delib-
erate and carefully thought out than it has been. There is clear evidence
from IMF and World Bank sources on a number of countries in Africa
that the design of structural adjustment packages in the 1990s had serious
shortcomings. As a result, there was less economic growth, and what there
was had fewer benefits for the poor, than might otherwise have been the
case. Influencing the policies of the international financial institutions in
this regard does not seem to have been a priority for many DFID country
programmes. Meanwhile, the downturn in Asia has brought the issue of
alternative approaches to economic stabilisation and adjustment back to
the forefront of the international agenda. How might country programme
managers (and DFID advisers placed in the multilateral financial institu-
tions) go about defining a new approach to these questions?

• New sector-wide partnerships, or SWAPs – by taking a hard look at the dif-

ficulties of reconciling the demands of partnership at this level and
those of meeting the International Development Targets.

One of the commonest kinds of ‘missing middle’ or strategic fuzziness in

country programmes is a set of loose and unchallenged assumptions about
the value of social-sector spending and aid for poverty reduction. Pro-
gramme components that have a fairly strong ‘inertial’ element are re-
dressed to suit the requirements of a poverty-focused approach. But if a
strategic approach means what it says, programme design should surely
begin with what it would take to meet the relevant International Develop-
ment Targets in this particular country context, working backwards to
identify the most suitable modes of intervention.
The adoption of a sector-wide approach has the important potential of
increasing the overall rationality of resource allocation across a sector. In
health and other fields, it is a welcome departure from the tendency of
donor funding to produce unsustainable islands of excellence in seas of
mediocrity and deprivation. On the other hand, it has been pointed out
that sector-based partnerships do not necessarily have a positive impact on
the rationality of resource allocation between sectors. Also, it could be
argued that making a prior commitment to a SWAP on partnership
grounds places yet another obstacle in the way of a genuinely strategic
approach to the International Development Targets, since it cannot be
assumed that the interventions that are indicated by this method will have
any particular sectoral location.
Last but not least, it is notorious that SWAPs do not necessarily improve
services to the poor in the short term. This may be because the poor have
limited access to the relevant services and because there are no resources
with which to significantly improve access. It may also be because the
sector has its own political economy of interest groups and patronage net-
works that tends to override the pro-poor policy commitments of the
88 Howard White and David Booth
sector. Either way, a strategic orientation on DFID’s part calls for an effort
to reconcile these opposing principles within an overall approach that is
driven by poverty analysis and informed by a subtle understanding of the
country context.

Conclusions
The DFID Country Strategy Papers are a departure from earlier such
papers, in that they have been redefined to mainstream poverty reduction
in the agency’s work. The papers have a logical structure flowing from an
analysis of the poverty situation and the causes of poverty, laying out a
strategy to address poverty and locating DFID’s position within that strat-
egy. However, in practice there is a missing middle, by which we mean
that the interventions to be supported by DFID are not clearly located in
the context of the main causes of poverty which have been identified. This
argument has strong parallels with that made in Chapter 3. The only
other similar analysis of which we are aware for another agency finds
similar shortcomings: the country strategies of the World Bank in Africa
were found to be strong on anti-poverty in the discussion of high-level
strategy, but far less so when it came to what was actually being done on
the ground (World Bank, 1997). Agencies need to develop clear ideas of
how it is that they can affect poverty outcomes – that is, adopt a logical
framework approach. They then need the institutional flexibility to take
on board the implications of such analysis.
Appendix 4.1 Definitions and perceptions of poverty

Country Definition of poverty/poverty indicators Incorporation of the perspective of the poor

Bangladesh Locally defined poverty line Not explicitly mentioned

China Locally defined and dollar-a-day poverty lines Not explicitly mentioned
Ghana Social indicators and locally defined poverty line Yes, leads with perceptions of poor before quantitative
(notes that dollar-a-day data unavailable) poverty profile
India Locally defined poverty line and social indicators Not explicitly mentioned
Kenya Dollar-a-day poverty line Mentions that Participatory Poverty Assessment (PPA)
‘corroborates’ quantitative results
Malawi Income-poverty trend mentioned (no line indicated) Very brief mention (‘Malawians feel they are getting
and discussion of social indicators under International poorer’)
Development Target headings
Mozambique Sections on economic well-being (locally defined Not explicitly mentioned
poverty line), human development (social indicators),
environment and qualitative elements
Nepal Income poverty line and social indicators Not explicitly mentioned
South Africa Dollar-a-day poverty line and social indicators Not explicitly mentioned
Tanzania Locally defined poverty line, and education and health Yes, features prominently in poverty profile to list
indicators. Also mentions that poverty is complex and dimension/causes of poverty
discusses Participatory Rural Appraisal (PRA) results
Uganda Locally defined poverty line and social indicators Yes
Vietnam Income poverty line and social indicators Not explicitly mentioned
Zimbabwe Locally defined poverty line and minimum food Not explicitly mentioned
requirements
Appendix 4.2 Reporting of International Development Targets and other indicators

Country Income Primary Ratio of Infant Under- Maternal Reproductive National Comment/other indicators
poverty school male to mortality five mortality health strategy for
enrolments female rate mortality rate services sustainable
enrolments development

Bangladesh x * .. .. .. .. .. .. Ratio of male to female

literacy
China x .. .. .. .. .. .. .. Life expectancy, Human
Poverty Index
Ghana x .. .. .. .. .. .. .. Ratio of male to female
literacy, child labour
India x .. .. x .. .. .. x Literacy, life expectancy
(some indicators reported at
state level)
Kenya x * .. .. .. .. .. .. Indicators mentioned but
values not given
Malawi .. x x x x x x x Life expectancy, female
literacy, secondary school
enrolments; target-specific
sections
Mozambique x x x x x x x x Ratio male to female literacy,
secondary school
enrolments, access to safe
water; target-specific sections
Nepal x .. .. .. .. .. .. .. ..
South Africa x x x x x x x x Table of International
Development Targets and
current status
Tanzania x x x x x x x x Total fertility rate (TFR),
child poverty rate (CPR),
literacy
Uganda .. .. .. x x x .. x Ratio male to female literacy,
life expectancy, access to
water, total fertility rate; list
of government health targets
(some overlap International
Development Targets)
Vietnam x .. .. x .. .. .. .. Life expectancy
Zimbabwe x x x x x x x x Life expectancy; target-
specific sections

Notes
x, current level of indicator reported; *, no level reported but trend indicated; .., indicator not mentioned.
Appendix 4.3 Intended outcomes and monitoring procedures

Intended outcomes How will Indicators How measured or

poverty impacts monitored
be identified?

Bangladesh Long-term objective: help Bangladesh achieve Not discussed None listed One individual in
sustainable reductions in poverty. Six thematic (implicitly the DFID Bangladesh
objectives: (1) sustainable improvements in International responsible for each
livelihoods and basic services for the poor, extreme Development objective
poor and those vulnerable to poverty; (2) sustainable, Targets?)
broad-based and pro-poor growth; (3) better
governance and more effective institutions;
(4) improved realisation of human rights;
(5) improvements in the position of women in society;
and (6) consistency in DFID and broader UK and
Bangladeshi government policies in support of the
elimination of poverty in Bangladesh
China Goal: poverty elimination Not discussed None listed Not discussed
Ghana Overall objective: support goals in Vision 2020 Not discussed Given implicitly by Not discussed
(improve quality of life, generate employment, and support for
reduce poverty) government
programmes
India Goal: make progress towards the elimination of Not discussed Poverty target, but Not discussed
poverty. Purpose: work effectively with partners to indicator not
reduce poverty significantly over next 10 years identified
Kenya Goal: reduce proportion living in extreme poverty Not discussed Goal gives income- Not discussed
by 2015. Purpose: sustainable improvements in the poverty target
livelihoods of poor people in Kenya over the next
5 years
Malawi Goal: to contribute to the elimination of poverty in Not discussed Goal gives (vague) Not discussed
Malawi. Purpose: to build partnerships that promote income-poverty
poverty elimination, equity and human rights in target. No others
Malawi. Six impact areas (see Table 4.1) are classified listed
under DFID’s specific objectives. Specific outcomes
not indicated
Mozambique Four impact areas (see Table 4.1) are classified Not discussed None listed Not discussed
under the first two of DFID’s specific objectives.
Specific outcomes not indicated.
Nepal Goals: significant reduction in poverty in 10–20 years; Not discussed Goals give (vague) Not discussed
measurable improvements in key indicators, references to
nationally and in selected areas within 5–10 years indicators
South Africa Support to government strategy (see Table 4.1) Not discussed Some project/ Not discussed
activity-specific
targets (e.g. water
supply services)
Tanzania Goal: a 50 per cent reduction in number of absolute Not discussed Goal gives income Strengthen capacity to
poor by 2015. Purpose: sustainable improvements in poverty target monitor impact of
the livelihoods of poor people in Tanzania over the sector programmes.
next 5 years. Five impact areas (see Table 4.1) ‘to Increased attention to
make a measurable difference in partnership’ monitoring impact,
using evaluations.
Annual reviews
considering impact

continued
Appendix 4.3 continued

Intended outcomes How will Indicators How measured or

poverty impacts monitored
be identified

Uganda Goal: to reduce the proportion of Ugandans living Not discussed Goal gives income Monitoring mentioned
in absolute poverty to less than 10 per cent of the poverty target. but system not
total population by 2017. Purpose: to contribute to Government health discussed
sustainable improvements in the livelihoods and targets listed (infant
conditions of poor people in Uganda. Britain seeks mortality rate,
measurable impact in the five areas identified in under-five mortality,
Table 4.1 immunisation,
malnutrition,
maternal mortality;
contraceptive
prevalence, fertility
rate). None listed
for other impact
areas
Vietnam Long-term goal: reduction in extreme poverty Not discussed Goal gives income- Not discussed
(International Development Target-based); and poverty target
immediate goal to promote pro-poor growth.
Purpose: promote and support government
policies for reform and provision of public services
and investment
Zimbabwe Purpose: to reduce poverty in Zimbabwe Monitor International Annual progress
poverty impact Development review with
of programmes Target-based government and civil
and progress society
towards
International
Development
Targets
 Designing country strategies 95
Notes
1 This chapter is based on an issues paper prepared for the session ‘How do
CSPs address poverty?’ at the joint DFID Economists’/Social Development
Advisors’ retreat on 23 September 1999 at Eynsham Hall, UK. We are grateful
to Richard Black for comments on an earlier draft of this chapter. The usual
disclaimer applies.
2 See the tables in Appendix 4.1 for a list of countries. This list includes nine of
the top ten recipients of UK aid in 1999–2000. The exception is Zambia, for
which a prototype ‘new CSP’ was prepared in advance of the White Paper and
so is not covered by this review.
3 The Guidance emphasises the importance of consultation with the private
sector, although it is unclear if this means the domestic private sector or UK
business interests. Generally, CSPs try to cover both of these.
4 Quotations are from the various memos of guidance on CSPs unless otherwise
indicated.
5 The Guidance, and this review, preceded the adoption of the Millennium
Development Goals, so the International Development Targets are referred to
in this chapter.
6 The Guidance suggests that both of these should be given.
7 Critical issues are whether to judge on past performance or future plans (and
whether to use the level or the change in commitment), the extent to which
outcome indicators can be used to measure policy stance, and how to deter-
mine acceptable threshold levels when these may vary from country to country.
8 This fact is consistent with reviews of aid programmes which generally find that
the amount directly targeted to poverty reduction is low, probably somewhere
in the region of 10–15 per cent (see White, 1996).
9 The Malawi log-frame is one of the few available to the authors, though the
presentation here is our own modification.
10 The Guidance requirement that changes between future and current strategies
be explicitly addressed is met scantily if at all in most cases; the CSP for
Vietnam is a notable exception.

References
White, H. (1996) ‘How Much Aid Is Used for Poverty Reduction?’, IDS Bulletin,
27(1): January.
World Bank (1997) Taking Action for Poverty Reduction in Sub-Saharan Africa: Report
of an Africa Region Task Force, Washington, DC: World Bank.
5 Monitoring progress towards the
Millennium Development Goals
at country level1
David Booth and Henry Lucas

Introduction
How can countries monitor progress towards meeting the Millennium
Development Goals? This chapter draws lessons from a review of the process
of establishing Poverty Reduction Strategy Papers (PRSPs) for the poorest
countries of the world. Two main areas are covered. First, the chapter con-
siders the progress made towards establishing effective poverty monitoring
systems, drawing on evidence from Africa. We then analyse the process of
choosing realistic indicators to monitor whether Poverty Reduction Strat-
egy Papers have moved countries closer towards meeting the Millennium
Development Goals.
The chapter first focuses on institutional and procedural questions
about monitoring, then moves on to addresss the selection of indicators.
After that there follow two sections that adopt a more forward-looking
perspective, making practical suggestions about how to meet some of the
biggest challenges facing those concerned with monitoring progress in
reducing poverty. A final section presents conclusions.

Roles of monitoring and information in a Poverty

Reduction Strategy Paper context
The PSRP is intended to put poverty at the centre of a government-led
policy process. The approach taken to monitoring needs to be corre-
spondingly innovative. There is much to be learned from the established
fields of project planning and sector programming. But the point of
departure needs to be a clear understanding of the change of gear that
the initiative is meant to facilitate. We should start from what a PRSP is
meant to be, not simply from the accumulated wisdom of the monitoring
and evaluation profession, or indeed from the more recent field of
poverty monitoring. We identify five aspects of the approach.
First, the advent of PRSPs means that it is no longer ‘business as usual’.
PRSPs have brought new concerns to the attention of policy-makers in the
South. The form of conditionality of aid is changing, with emphasis shifted
 Monitoring progress at country level 97
from traditional forms to a focus on in-country processes. Such ‘process
conditionality’ is a means of opening up discussion among stakeholders
about the ways and means of addressing poverty reduction goals. Several
points are included in this new form of conditionality. For example, the
processes that lead to poverty reduction are expected to be:

• more ‘owned’ by the country – that is, more rooted in national

processes of policy dialogue and accountability;
• more comprehensive, both in terms of the sectors covered and in co-
ordinating the full range of national and international resources; and
• more performance based and outcome oriented in the way they allo-
cate resources.

Second is the need for a more realistic view of the policy process. The
framework for monitoring the plans needs to be realistic about the way
the different elements fit together. Attention has to be paid to the sub-
stantial literature on the nature of the policy process (involving both
policy formulation and policy implementation), which may be general, or
country or region specific (see, for example, Hill, 1993, and Turner and
Hulme, 1997, respectively). It cannot be assumed that formal commit-
ment to a set of objectives on the part of senior government officials
implies an ability, or even a willingness, to deliver all of the consequent
actions.
Third, monitoring and evaluation is not simply a technical matter, but
is also about politics. Most of the conceptual vocabulary of monitoring
and evaluation reflects a rationalistic model of the policy process, not the
realistic one just described. This does not mean that current thinking on
monitoring and evaluation is irrelevant in the real world of policy. But it
does imply the need for a shift in emphasis on the role played by monitor-
ing and evaluation. Special attention should be given to the parts of the
literature that address the issue from a more realistic perspective, such as
that concerned with ‘process monitoring’ (Mosse et al., 1998).
Fourth, contrary to the impression that is sometimes given, an outcome-
oriented approach to monitoring does not imply an exclusive focus on
final outcomes/impacts2 or poverty monitoring (see Chapter 3 of this
volume). Improved poverty outcome data are important for several pur-
poses, including basic analysis of the causes of poverty, which is essential
to good policy design. This issue – what might be called ‘poverty monitor-
ing’ as opposed to ‘PRSP monitoring’ – has a certain importance. But in
most countries it is more important to focus on the process of implement-
ing policy rather than on the final outcome, if learning and accountability
are to be achieved. Policy is likely to improve, and/or become more
outcome oriented, only if new incentives come into play. Information on the
final outcomes or impacts of policy rarely has practical implications, since
(1) it arrives too late, and (2) it has too many difficulties of attribution.
98 David Booth and Henry Lucas
Paradoxically, this means that information on the performance of policy
implementation may be more powerful in making policy more oriented
towards outcomes than final outcome monitoring can hope to be. This is
a key issue in deciding the scope and balance of monitoring systems for
PRSPs. In addition, reliable data on intermediate output and outcome
issues are very hard to come by in most countries, even on an untimely
basis.
Fifth, the above assumes that the principal role of monitoring policy is
to ensure that policy-makers learn and are held accountable to domestic
stakeholders. This assumption neglects the fact that policy-makers are also
held accountable to donors. The role of a monitoring system in providing
for accountability to donors is not unimportant, because nothing discour-
ages donors more from pooling their funds in sector programmes or
general budget support than the perception that accountability require-
ments will not be satisfied. On the other hand, the best bet for enhanced
accountability to donors is undoubtedly one that also enhances respon-
siveness to domestic stakeholders.

Five areas of interest in monitoring Poverty Reduction Strategy Papers

This section goes on to consider current documentation on PRSPs, and
what it has to say about the following five areas: input monitoring; the
monitoring of implementation and intermediate outputs and outcomes;
the measurement of poverty outcomes and impacts; measures to make
information more available; and steps to enhance the use of information.
A strong overall impression of the current situation is that thinking and
practice are at quite an early stage. Some topics, such as the financing of
sustainable monitoring systems, are hardly covered at all. On the range of
activities and the allocation of institutional responsibilities, the details
given in the PRSPs vary from thin to comprehensive but still incomplete
(e.g. Uganda). For this reason, the review was largely concerned with
rather elementary questions about each activity, such as, is monitoring
mentioned at all, are steps being taken to initiate activity, and do these
steps make sense?

Input monitoring and budget reform

Guidance and training materials for implementing PRSPs identify input
monitoring as important. For example, it is usual to refer to indicators of
expenditure on particular items such as primary education. However, the
usefulness of this type of measure depends on the way budget line items
are defined, and thus on the nature of the prevailing budget system.
Other questions are the degree to which actual releases of funds are deter-
mined by initial allocations, and what percentage of funds reach their
final destinations.
 Monitoring progress at country level 99
A key step in outcome-oriented budget reform is the establishment of a
Medium Term Expenditure Framework (MTEF). In Uganda and Tanza-
nia the MTEF has laid the basis for progress in the preparation of budgets
on a programme basis in line ministries and local government. However,
while the Ugandan authorities see the MTEF as providing the framework
for their Poverty Eradication Action Plan,3 Tanzania makes little of this
and other improvements in public expenditure management systems in its
PRSP. In Burkina Faso, significant headway has been made in linking
funding to performance, providing the context for discussions about
reforming the budget process. In Mauritania the introduction of pro-
gramme budgeting was scheduled to start in 2002, though the country has
some way to go before the gap between the budget and the bulk of tar-
geted project expenditure on poverty begins to close.
Countries not yet in a position to prepare a PRSP have been able to
submit an Interim Poverty Reduction Strategy Paper (I-PRSP), which pro-
vides an overview of a poverty reduction strategy and the process by which
a full PRSP will be prepared. An I-PRSP is sufficient to qualify for HIPC
funds. Most I-PRSPs in Africa mention budget reform, either confirming
that it is happening (in Benin, Ghana, Guinea, Kenya, Malawi, Mali,
Rwanda and Senegal) or stating it as an objective (in The Gambia, Niger
and Zambia). Assessments of PRSPs frequently emphasise the impossibility
of setting overall priorities in the absence of the realistic expenditure ceil-
ings that would be provided by an MTEF. Whilst not strictly a monitoring
issue, this shows how reform of public expenditure management may be a
precondition for improvements in monitoring.

Public expenditure tracking

Without budget reform and the associated technical improvements in
public expenditure accounting, it is not usually practical to take the step
of comparing budget allocations with releases according to sectoral and
sub-sectoral priorities. Moreover, even in countries where budget reform
is taking place, there is little discussion of this issue in PRSPs.
Public Expenditure Tracking (PET) Studies, which report how much
spending reaches the final destination (e.g. how much of the education
budget is spent on items appearing in schools), have been undertaken in
a number of countries. In Uganda, such exercises are now a routine part
of monitoring the country’s Poverty Eradication Action Plan (PEAP).
Tracking exercises have been undertaken as part of Tanzania’s rolling
Public Expenditure Review (PER), although this fact is not highlighted in
the country’s PRSP. Other countries are at very different stages. Some
declare an intention to undertake tracking studies (e.g. Guinea, Rwanda),
but others have not completed the more basic step of systematically
reviewing public expenditure, and many report nothing on the subject.
The Kenyan I-PRSP includes a summary of perhaps the ideal poverty
100 David Booth and Henry Lucas
monitoring system. It locates the national poverty reduction effort squarely
within high-level arrangements for monitoring and tracking prioritised
public expenditures. A stakeholder committee would meet monthly, and
transmit its concerns through a committee of Permanent Secretaries to
the Cabinet. In the way it integrates financial and implementation issues
and guarantees a hearing for stakeholder assessments of monitoring data
at the highest policy level, the Kenyan proposal provides a model of what
might be done in all countries. However, few countries could achieve such
an arrangement at this stage. Even in Kenya, whether this can be imple-
mented is open to doubt.

Implementation monitoring with administrative data

A key dimension of performance monitoring is monitoring intermediate
outputs, including process-based aspects of policy implementation. African
PRSPs cannot be accused of neglecting this aspect of monitoring if the crite-
rion is the volume of indicators identified for the purpose. However, this
would clearly not be an appropriate criterion.
The indicators identified are numerous and rather unselective, raising
doubts as to the feasibility of obtaining reliable data at reasonable cost. Do
PRSPs include steps for bringing administrative data and/or sectoral man-
agement information systems closer to the required quality standards, and
do they envisage alternative means of acquiring quick feedback? Overall,
these concerns are very striking by their absence. PRSPs almost invariably
include a commitment to make arrangements for monitoring plan imple-
mentation using official statistics. Occasionally, data deficiencies are men-
tioned as a problem, and the institutional and technical arrangements for
co-ordinating data from different sources are frequently discussed. However,
the possibility that there might be fundamental obstacles to using routine
data to monitor progress on account of severe problems of unreliability is
not acknowledged at all.

Alternative feedback mechanisms

In the light of the above comments, it is not surprising that the coverage
of other forms of feedback on implementation is also slight. There are two
reasons why methods such as participatory beneficiary assessments, imple-
menter self-assessments using focus-group methods, ‘exit polls’ and simple
service delivery surveys might be considered. One is that they provide a
check on information reported, slowly and unreliably, through official
channels. They can also provide a more dynamic type of input into the
political process of the PRSP, highlighting problems while there is still
time to act on them and to mobilise public interest and pressure at the
same time.
There are hints of such possibilities in some Papers. In general, they
 Monitoring progress at country level 101
are not fleshed out sufficiently to justify confidence that they will be
pursued in the absence of strong donor pressure and offers of funding.
But a different picture may emerge from new initiatives mooted in a
number of countries, such as the regular stakeholder opinion polls that
are proposed in Burkina Faso, or the participatory monitoring arrange-
ments that have been suggested in Tanzania and The Gambia.
Where service-quality enquiries and self-assessments have been used to
set benchmarks for public service reforms before the advent of PRSPs,
these have not always been considered relevant by those drafting the
Papers. In Senegal’s PRSP a mechanism of this sort is mentioned; but
the equivalent arrangements in Ghana do not figure in the initial moni-
toring proposals.

Measurement of final poverty outcomes/impacts

The implementation of PRSPs will produce a dramatic improvement in
the quality and general availability of survey-based household consump-
tion data. Many new surveys are currently under way. These will not over-
come all the problems of data shortage and comparability that have
prevented serious analysis of poverty trends in recent years. However,
the coverage is set to experience the same sort of qualitative leap in
household income and expenditure survey as was caused by the wave of
support for the ‘Social Dimensions of Adjustment’ process a decade
earlier.4
Of course, the problems of sustainability that eventually affected this
earlier round of surveys will also affect this one. However, this problem is
anticipated in some documents, with several countries experimenting with
light surveys (using short questionnaires, perhaps with smaller samples)
for more frequent use, allowing a sensible spacing of large-scale surveys
and censuses.
The strengths and limitations of household survey data for understand-
ing national poverty profiles are more widely appreciated than they were a
decade ago. Many of the country plans, with firm declarations of the
importance of non-income dimensions of poverty, state that Participatory
Poverty Assessments (PPAs) will be undertaken, though details are gener-
ally lacking.

Increased access to information by stakeholders

PRSPs are intended to empower a range of actors to engage in a construc-
tive debate about why poverty reduction has proved so difficult and
what can be done about this. This is not an easy thing to achieve. Even in
highly institutionalised democracies, governments share information with
political rivals only when compelled to do so by law or convention. Non-
governmental actors often lack the expertise to make intelligent use of
102 David Booth and Henry Lucas
official statistics. In sub-Saharan Africa in particular, political and civil
society is poorly equipped to engage with the PRSPs.
Regarding monitoring, a few Papers include a continuing role for the
stakeholders mobilised for the design process. For example, the details of
Uganda’s Plan, which entails an ongoing dialogue across political and
civil society on poverty reduction priorities, are quite well known in the
country. This example probably represents the apex of current African
achievement in this area, at least in terms of the openness of the process
and the willingness of the government to make relevant information avail-
able and reasonably accessible. That said, the arrangements are much
better for information concerning final outcome (whose content is gener-
ally encouraging) than for intermediate performance indicators.
The Ugandan model depends on the centralisation of the analysis and
dissemination of poverty-related information in a unit within the Ministry
of Finance. The leadership of this Ministry has been strongly committed
to openness and is not averse to the use of official information for advo-
cacy purposes. In other countries, however, the institutional framework
may be less favourable. Nevertheless, PRSPs give some support to ongoing
monitoring by stakeholders. One issue to be confronted is that in several
cases the stakeholder monitoring committees that are proposed sound
like bilateral forums for government and donors. Increasing bilateral dia-
logue alone is obviously not the point of developing a PRSP.

Use of information for policy improvement

The focus of most of the documents at this point is on improving the avail-
ability of raw data. However, discussion is needed on the analytical uses to
which good data might be put. The question then arises as to which institu-
tions in the country actually have an incentive to use the data for purposes
that serve policy improvement. This has been raised as an issue in Uganda.
Existing survey data would, it is claimed, support specific studies that could
improve the targeting and effectiveness of the programmes of a number of
line ministries. However, under prevailing conditions (an incomplete
transition to performance budgets for line ministries, and an incomplete
results-based public service reform), line ministries do not have strong
incentives to commission the necessary analytical work. Incentives for data
use are, of course, a long-standing issue in poverty monitoring in Africa. It
appears that this remains a big problem that is not even recognised in most
PRSPs.
 Monitoring progress at country level 103
Choosing indicators: rationale, credibility and realism

What are indicators for?

It is important not to detach the choice of indicators from the aims of the
planning exercise of which they are a part. In even a preliminary assess-
ment of a country’s approach to indicator choice, the purpose of each of
the proposed indicators needs to be a primary consideration. The quality
of the indicators can be assessed only in terms of the role(s) they are
expected to play. Indicators are supposed to track progress towards
certain objectives. This presupposes both that the objectives are clear and
that the intermediate steps necessary to achieve them have been identi-
fied. However, initial experience with PRSPs shows that whilst setting
objectives is easy enough, identifying credible intermediate steps is much
more difficult.
Most PRSPs have a ‘missing middle’ (see Chapter 4).5 They do not spell
out how the identified activities can be expected to result in the achieve-
ment of the identified goal. This is not surprising. To the extent that
improvements are possible, they will arise from the social and political
dynamics of the planning process in the medium term. It is unrealistic to
expect PRSPs to bring immediate improvements.
The ‘missing middle’ problem explains a lot about current approaches
to the monitoring of PRSPs, such as the focus on final outcome/impact
measurement. This leads us to expect a lack of direction in selecting indic-
ators. If the strategy for reducing poverty is weak at the ‘action plan’ level,
the rational basis for selecting indicators will also be limited. The choices
will reflect other considerations, such as the question of which targets can
be met before a debt relief package is completed, and what the corres-
ponding indicators are.

What is a ‘good’ indicator?

Box 5.1 summarises what is considered a ‘good’ indicator in the ‘monitor-
ing and evaluation’ chapter of the World Bank Poverty Reduction Strategy
Sourcebook (Tikare et al., 2001). While few would disagree that the qualities
mentioned are desirable, reflection on the reliability of the indicators
which are ‘available frequently’ in much of sub-Saharan Africa suggests
that data quality may be of primary importance. It is not useful to track
over time variations in indicators whose margin of error is greater than
the expected changes. The need to trade off between reliability and other
qualities may lead to the adoption of ‘second-best’ indicators in many
instances.
104 David Booth and Henry Lucas
Box 5.1 What makes a ‘good’ indicator?

Good indicators:
• are direct and unambiguous measures of progress: more (or less) is
better;
• measure factors that reflect objectives;
• vary across areas, groups, over time, and are sensitive to changes in pol-
icies, programmes, institutions;
• are not easily blown off course by unrelated developments and cannot
be easily manipulated to show achievement where none exists; and
• can be tracked (better if already available), are available frequently,
and are not too costly to track.

Source: Prennushi et al. (2001, box 2).

Alternative data sources: reliability and cost

The World Bank Sourcebook also promotes the need for disaggregated
indicators, in terms of location, gender, income level and social group,
without which ‘it is hard to design good policies and programmes’ (Pren-
nushi et al., 2001: 9). Such disaggregation is also essential for effective
project and programme management. This requirement, coupled with
those for timeliness and affordability, would seem to imply a need to focus
on indicators derived from administrative sources. While surveys may in
principle provide better indicators in terms of the above criteria, their use
for the frequent generation of reliable estimates, at the level of disaggre-
gation proposed, would make excessive demands on national statistical
resources.
But what about the quality of administrative data? Routine data sources
in most countries suffer from well-known limitations. This implies the need
for expectations to be limited, and second-best options to be explored. For
example, while such basic indicators as service utilisation, access and cost
are not ideal, they may provide a reasonable basis for predicting beneficial
final outcomes and be either usable at present, or at least susceptible to
improvement in the short run at minimal resource cost.
The use of such indicators is likely to be unsatisfactory in the absence
of supporting information on the quality of services available. Knowledge
of satisfactory performance on both types of indicator – for example, high
levels of utilisation of good-quality reproductive health services at low cost –
would be a sound basis for expectations that programme objectives in this
area would be met. Absence of any one of these indicators might give
cause for concern. Regular quality assessments, using qualitative and par-
ticipatory approaches, could play an important role in delivering this
 Monitoring progress at country level 105
information. As a minimum, reliable audit indicators assessing the ade-
quacy of supervisory activities could provide some degree of quality assur-
ance.

The current situation

Rationale for indicator selection

A clear rationale for indicator selection is often lacking in practice. The
authors of many of the PRSPs have listed a wide range of traditional indi-
cators in a fairly undiscriminating way. Selection seems to have operated
on the basis of relevance to the various projects and programmes which
have been included within the Paper. As a result, it is often difficult to see
how the indicators could be effectively used to consider broader strategic
issues.
In many instances, the indicators cover economic growth, macro-
economic stabilisation, human resource development and other general
measures of development performance, alongside indicators specifically
related to poverty reduction. Without denying the importance of macro-
level stability and growth for sustained poverty reduction, there would be
a strong case across much of sub-Saharan Africa for measures such as the
growth rate in the agricultural sector, or price movements and interest
rates that particularly affect poor people, to be substituted for the broader
measures chosen.
One problem is that the targets set in PRSPs and the conditions set for
debt relief do not coincide closely. Nevertheless, it seems that some indic-
ators have been selected primarily because they are also conditions for
debt relief. In our view, such indicators should be clearly identified and
distinguished from those for the poverty reduction strategy. For every indi-
cator, it should be standard practice to specify the intended uses and
users. Such an approach would facilitate the categorisation of indicators
by purpose and allow the designation of a limited number of ‘core indic-
ators’ to monitor overall performance of the strategy. In Burkina Faso,
donors have stated their interest in a small core set of performance indic-
ators and agreed to limit their attention to that set. However, the degree
to which this process involved wider stakeholders in a national dialogue
remains an important issue.

The quality problem in administrative data

Many of the indicators proposed in PRSPs derive from routine administra-
tive data. Given that such sources are generally agreed often to be at best
highly unreliable, the PRSPs often appear to be highly optimistic as to the
possibilities for measuring short-term indicator movements from such
data. The problem is particularly serious where regional disaggregation is
106 David Booth and Henry Lucas
required. As a general rule, the quality of administrative data depends on
the quality of administrators, and both tend to be correlated with incomes
per capita. The poorest areas typically have the least reliable data. For
example, rural health workers in poor areas (given that their government
salaries are sometimes barely sufficient to purchase basic food and
clothing) have become very adept at providing information that satisfies
higher levels of administration while not limiting their alternative income-
generating activities.
Variations in the quality of administrative data between regions may
influence national estimates, as these are often based on partial coverage.
Poorer regions not only tend to provide less reliable data, but often fail to
provide data on time. As national estimates are sometimes based on ‘gross-
ing up’ the information available when estimates are required, biases that
tend to underestimate poverty indicators may be introduced.

The denominator problem

Many of the selected indicators relating to education, health and more
general access to services require age-specific estimates, sometimes at
regional level. These will reflect the well-known ‘denominator problem’ of
indicator construction: the fact that the base populations are not known.6
Changing population structures, particularly via migration, may need to
be considered in the interpretation of trends over time. The influence of
such changes on enrolment, access and utilisation measures can be sub-
stantial. Again, poor regions may be particularly affected by migration.

Community involvement in indicator choice?

The need for participatory approaches to the design of PRSPs is stressed in
the Poverty Reduction Strategy Sourcebook (Tikare et al., 2001) and by donors in
general. However, it is very difficult to identify any evidence of community
involvement in the list of proposed indicators. In general they follow stan-
dard guidelines, not only in the areas of economic growth and stabilisation,
but in education, health and other areas of social policy. The inclusion of a
number of ‘client satisfaction’ indicators appears to be the only diversion
from this norm. Even in this case there is a tendency to suggest a simplistic
‘opinion poll’ approach, which may not be the most useful way of tapping
the views of stakeholders. One important role for community involvement
that is underexplored in the documents is identifying factors relating to
the failure of programmes to deliver intended benefits.

Data improvement versus data on improvement

Almost all PRSPs for countries in sub-Saharan Africa stress the need to
build statistical capacity and increase the quality of information available.
 Monitoring progress at country level 107
However, there is a practical problem that needs to be taken into account:
it is often difficult to distinguish between the effect of improved measure-
ment and real trends in economic and social variables. For example, a
more systematic approach to determining all sources of income or non-
market consumption may result in artificial increases in related indicators.
Similarly, improved disease surveillance systems usually lead to higher
reported prevalence rates.

What to monitor and why

Monitoring designs cannot be expected to solve the problems of weak
planning. PRSPs should contain decisions about what needs to be done in
order to achieve poverty-reduction goals. Deciding what to monitor and
how to do it should be a next step, rather than being defined in advance.
At best, thinking about monitoring can provide a way back into an unfin-
ished debate about strategy.7 This is particularly feasible if stakeholders
who have been mobilised in the design of Papers remain active within the
institutional arrangements for monitoring and see this as part of their
job. However, that only means that monitoring processes may prompt
some revisiting of the substance of poverty reduction strategies. It does
not imply that monitoring is the same as planning, or can be a substitute
for it.
It follows that our discussion in this section has to touch on what ought
to be included in the Papers themselves, as well as on what should be
monitored. We need to set some definite limits, otherwise the task would
become impossibly broad. We do this by largely limiting the discussion to
the challenges facing the most promising strategy, the Ugandan Poverty
Eradication Action Plan.

What kinds of final outcomes/impacts?

Handling multidimensionality
Most PRSPs have some commitment to goals additional to a reduction in
the percentages under the poverty line. With respect to the monitoring of
Papers, this implies paying attention to Demographic and Health Surveys
(DHSs) and national Human Development Reports as well as household
expenditure surveys. There is also normally mention of the need for a par-
ticipatory poverty assessment exercise.
However, despite frequent references to multidimensionality, income
poverty is invariably a central focus in the Papers. In spite of the fre-
quently stated concern to move away from income poverty measures and
give greater weight to participatory assessments and qualitative informa-
tion, the traditional poverty line-based head count, poverty gap and
intensity indicators predominate. As Thin et al. (2001) point out, income
108 David Booth and Henry Lucas
is typically presented in Papers not as a means to improve welfare but as
an end in itself: ‘paradoxically . . . lack of education and lack of adequate
nutrition are seen as less basic to the definition of poverty than lack of
income’ (p. 8). This would be of less concern if the level of expenditure
per capita were a good predictor of nutritional status, social condition,
empowerment, or other factors. However, the tendency in the literature is
increasingly to find relatively low associations between different measures
of deprivation (e.g. Sahn, 2001).
Experience of monitoring the final outcomes of PRSPs in a balanced
multidimensional fashion is as yet limited. However, for a number of
years Uganda’s Poverty Monitoring and Analysis Unit has been
working on Poverty Status Reports, and frequent briefings weave different
qualitative and quantitative poverty information into a single fabric.
Although the activity of the Unit did not, until recently, draw the Uganda
Bureau of Statistics and those conducting participatory assessments into a
close relationship with each other, it has the capacity to move across the
relevant areas of expertise, and this has ensured that these different actors
have not inhabited completely different worlds. Poverty monitoring units
or Observatoires in a number of other countries, including Rwanda, may
develop a similar capability.

Why collect final outcome data?

An excessive focus on the final outcome/impact level may downgrade the
essential role of input and intermediate output and outcome indicators.
As we have said, final outcome data are largely useless for providing the
sort of quick feedback on performance that is most needed for learning
and accountability purposes. The speed with which survey data become
available is improving fast (data from the Rwanda survey were incorpo-
rated in the PRSP within months). However, results are likely to remain
relatively slow to appear in generally usable form, and problems in
attributing trends or patterns to specific policy measures will remain.
Knowledge of final outcomes/impacts is less necessary than imagined
from the point of view of impact evaluation. Theory-based evaluation
(Weiss, 1998) requires the specification of a chain of theoretical ‘cause
and effect’ linkages, which allows the likelihood of beneficial outcomes
and impacts and sustainability to be assessed. For example, it is very diffi-
cult and expensive to demonstrate the impact, or even the outcome, of a
given health project or programme. Even in the simplest case, that of
immunisation, because both morbidity and ‘cause of death’ statistics are
so difficult to obtain, it is usually impossible to infer in a particular
instance that a given expenditure on measles vaccination has led to an
identifiable decline in under-five morbidity or mortality. However, it is
often perfectly reasonable to rely on past evidence of such a causal link.
Effective use of donor and government resources (measured in terms of
 Monitoring progress at country level 109
output indicators), on activities mutually agreed (based on previous
experience) to be causally linked to increased welfare of the poor, would
seem to be a rational basis on which to assess performance.
Final outcome or impact indicators retain an important strategic
role, either as confirmatory or as warning signals. In the latter case,
they may indicate either that assumed causal links were not as expected
or that previously unconsidered external factors needed to be taken
into account. The key quality of such indicators would be an ability
reliably to determine trends over time and differences between localities
and groups.
Expenditure surveys, DHSs and PPAs will remain essential in providing:

• information on who the poor are, and what their priority concerns
seem to be; and
• policy learning of a deeper sort: a better understanding of how
poverty sometimes gets reduced, why it very often does not and, thus,
what the entry points are for reducing poverty.

The construction of ‘poverty profiles’ (cross-tabulations of poverty and

other household characteristics) is likely to remain the principal focus in
most countries. In some cases more ambitious diagnostic work may be
appropriate. However, it is the quality of analysis and interpretation that
must be paramount, not the quantity or apparent sophistication. Closer
integration of the Integrated Household Survey and participatory appraisals
as currently being attempted in Uganda may be of more value than
advanced econometric analysis.

What kinds of intermediate variable? Learning from Uganda

One of the features of Uganda’s Plan, especially in its revised (2001–2003)
form (Uganda, 2001), is its serious effort to fill in the ‘missing middle’.
For each of the plan’s overarching goals, the document discusses relevant
evidence on what is working (and what is not). There is a recognisable
effort to diagnose policy failures and identify corresponding actions. Sug-
gestions on how progress might be monitored follow immediately, and
are reproduced in a summary matrix. Particular attention is devoted to
the middle columns of the matrix, headed respectively ‘Outcomes’ and
‘Outputs/access/proximate determinants of outcomes’.
The Ugandan document is a model in terms of intentions. The institu-
tional arrangements are also encouraging. But the intentions are not
entirely realised, and some quite significant gaps still need to be filled.
This is a central claim in one study of Uganda’s monitoring and evalu-
ation: Hauge (2001) argues that a gap remains at the level of expected
intermediate results. For example, in the Budget Framework Paper for
Education,
110 David Booth and Henry Lucas
Goals are expressed as increases in the pupil:teacher/classroom/book
ratios. There is little discussion, and no targets, pertaining to the crit-
ical dimensions of the quality issue: such as drop-out rates, years of
educational completion or examination attainment standards. One is
left with no answer to the question: what difference would we like
improvements in [these] ratios to make, in terms of educational
quality?
(ibid.: 9)

Hauge notes the danger that, with output-based performance orientation,

managers become motivated to establish goals they know they can

attain, with little regard for whether they make a difference on the
ground or contribute to longer-term goals. . . . Without a clear and
common set of first order goals and targets cascading through a
national development management system, it is not given that there is
congruence between planning and management activity or that every-
body is pulling in the same direction.
(ibid.: 9, 17)

Hauge concludes: ‘emphasis must be placed on distillation of clear and

consistent poverty goals, targets and performance indicators pertaining to
the reach and outcome levels of change – covering a medium term time-
frame such as 2, 5 and 10 years’ (ibid.: 24).
These conclusions were reached before the last revision of the Plan.
However, they reflect a reality that has certainly not gone away, even if
some headway has been made in some sectors. Uganda, like many other
countries, is in the middle of a reform of public management that
includes an outcome-oriented or programme-based approach to budget-
ing, and a results-oriented reform of human resource management in the
civil service. The country has its share of slow or stalled implementation in
these areas. However, it is distinguished by an unusually vigorous use of
existing instruments by the Ministry of Finance, Planning and Economic
Development to challenge line ministries and local government. Cur-
rently, these focus on the medium-term Budget Framework Papers, and a
number of carrots and sticks connected with the operation of the Poverty
Action Fund (Bevan and Palomba, 2000; Foster and Mujimbi, 2001).
Under these arrangements, line ministries are offered better de facto access
to resources if they can demonstrate plausible linkages between proposed
programmes and the goal of poverty eradication.

Input monitoring: its scope and importance

An outcome-oriented approach should not imply neglecting improve-
ments in input monitoring. One danger is that this will be regarded too
 Monitoring progress at country level 111
narrowly, as limited to budget allocations and financial inputs only.
Experience suggests a number of issues that need tracking on the finan-
cial side, and also that some non-financial inputs are worth watching
closely. Monitoring the effectiveness with which inputs are delivered to
different levels of government, and to service-providing institutions, has
an extremely important place in monitoring systems. Areas that need to
be covered include:

• the execution, as distinct from the formulation, of the budget; and

• the extent to which funds reach their specific intended destinations,
such as schools or clinics.

Uganda provides a classic example of what can be gained from tracking

inputs more effectively. A series of surveys of 250 public primary schools
carried out during 1991–1995 found that as little as 13 per cent of the
central government’s contributions to the schools’ non-wage expenditure
was reaching them. As a result of these surveys, a campaign to publicise
the funds sent to districts for schools led to over 90 per cent of an
increased allocation reaching its destination in subsequent years (Reinikka
and Svensson, 2001). Moreover, it is not only financial inputs that can be
missing. In the regional consultations around the PRSP in Benin, Ministry
of Finance officials were surprised to be told that teacher absences repre-
sent a serious and chronic problem in rural schools (Bierschenk et al.,
2001). Similarly, many studies of rural health care have highlighted the
widespread practice whereby trained staff use untrained ‘assistants’ to
provide clinic services while they engage in more remunerative private-
sector activities (see, for example, Assiimwe et al., 1997).

How to monitor: obtaining valid and reliable information

Deciding what to monitor has some immediate implications for how to do
it; some instruments are inherently unsuitable for obtaining the desired
information. For example, if the reach of essential services is the key ques-
tion, it is important to survey whole populations rather than particular
facilities. Even if appropriate sources exist, there is also the question of
how well they perform, and whether that performance can be improved.
Emphasis on the monitoring of intermediate outputs and outcomes
suggests a large role for administrative data and management information
systems. However, these are subject to well-known problems of reliability.
What to do about such problems is as important as getting the right com-
bination of different instruments. These form the two major concerns of
this section.
112 David Booth and Henry Lucas
Snags and new developments in the monitoring of final outcomes

Comparative reliability of participatory assessments and surveys

The battle to get the multidimensional concept of poverty accepted for
operational planning purposes is not entirely won. The status that tends to
be given to the traditional, survey-based approach is well illustrated by the
discussion by McGee and Brock (2001: 25–26) of the controversy in
Uganda about ‘contradictions’ between the findings of a PPA and house-
hold survey results. A principal finding from the former was that the poor
saw themselves as getting poorer while the rich were getting richer. The
survey results, on the other hand, were said to demonstrate that ‘if any-
thing, growth in living standards has been strongest among the poorest
households’. For many, the immediate reaction was to ask, ‘Why does the
PPA not reflect the true situation?’
The subsequent analysis and discussion focused mainly on the partici-
patory assessment findings, pointing out that they should not be treated as
directly comparable with the survey results. Changing levels of consump-
tion expenditure should not be expected to coincide with perceptions of
changing levels of poverty. The two methods of assessment should rather
be seen as complementary, offering alternative perspectives that could
jointly provide greater insight.
While this point is well taken, it is also relevant to question the implicit
assumption that the use of poverty lines to assess changes in income
poverty is always reliable and robust. If great care is not taken about
methods and assumptions, household surveys can get it badly wrong, as a
recent example from The Gambia illustrates. Three supposedly compara-
ble household surveys suggested that the proportion of the population
falling below a food poverty line halved over one three-year period, and
then trebled over the following six years. But these apparent trends
turned out to be largely explained by technical issues in the definition and
measurement of poverty.

Divisions of labour between surveys and participatory assessments

PPAs raise questions that lead to a re-examination of the methods used in
survey analysis. But the comparative advantage of these assessments is not
in challenging surveys on their own ground. Although there is some scope
for methodological triangulation between the two methods – that is, for
using data from the one to check those from the other – the areas of
direct comparability have been exaggerated. That being the case, it is
more important to develop other kinds of complementarity between the
two approaches. This implies an iterative, puzzle-solving relationship,
focused less on ‘what?’ and more on ‘why?’ (Appleton and Booth, 2001).
The second PPA in Uganda has taken up these conclusions in its
 Monitoring progress at country level 113
design. The fieldwork is being prepared to ensure that questions arising
from the panel element in the survey are pursued in the Participatory
Assessment study sites, and that any findings feed back into the design and
analysis of the survey. It has also been agreed that PPA will become less
focused on exploring poverty perceptions and other final-outcome issues,
and more on investigating known problems in implementing the PEAP.
Household surveys and participatory poverty assessments have come to
symbolise the ‘quantitative’ and ‘qualitative’ approaches to poverty moni-
toring. However, this traditional distinction has proved problematic.8
Booth et al. (1998) suggest that it may be more useful to think in terms
of ‘contextual’ and ‘non-contextual’ information. The former must be
treated as requiring interpretation within its ‘social, economic and cul-
tural context’, the latter as ‘untainted by the particularities of the context
in which it is collected’. For example, an observation that households
below the poverty line in a given country tend to have high dependency
ratios might be an example of the latter; complaints that a corrupt local
official was disrupting access to health services, of the former.
In purely practical terms, stressing the importance of ‘context’ has
proved useful in advocating the value of participatory techniques in
poverty assessment and monitoring, as it is more readily accessible to
senior policy-makers than the quantitative/qualitative dichotomy. It also
has a natural affinity with the focus on geographical locality in poverty
monitoring.

Geographical information systems and poverty targeting

In most countries, geographical targeting is a key policy instrument in the
poverty reduction strategy. The motivation is often self-evident. Remote,
inaccessible areas with limited access to markets and public services are
associated with high rates of poverty. At the same time, programmes
designed to reduce poverty can be relatively easily targeted at ‘poor areas’,
particularly if these have well-defined administrative boundaries.
Such policies have been criticised in terms of both their low ‘sensitivity’
– failure to identify poor households living outside these areas – and their
low ‘specificity’ – leakage of benefits to the non-poor living in them.9
These problems obviously increase with the size of the targeted areas.
Geographical targeting would be much more cost-effective if it could be
undertaken at the level of local districts or even individual villages
(Bigman and Fofack, 2000). Unfortunately, the information required to
work at this level is rarely available. If income poverty lines are used for
resource allocation, for example, the expenditure surveys used for area
classification will typically be based on sample sizes of around 2,000–4,000
households. This will usually not allow disaggregation below the level of
very broad regions, which may be above the primary administrative divi-
sions of the country (Hentschel et al., 2000). A possible way round this
114 David Booth and Henry Lucas
problem is to combine household survey and census data, as is being done
to generate estimates of poverty incidence for each of Vietnam’s sixty-one
provinces (Minot and Baulch, 2001).
Geographical targeting is traditionally based on administrative areas,
given that national data-collection systems are organised on this basis.
However, as Devereux (2001) points out, disaggregation by administrative
area may not be very useful in terms of identifying vulnerable population
sub-groups. A district, for example, though it may be the lowest administra-
tive level in a given country, may still contain a highly heterogeneous popu-
lation, particularly in terms of the range of livelihood systems adopted. A
number of agencies have addressed this problem. For example, the
food economy approach developed by Save the Children Fund–United
Kingdom divides a country into ‘Food Economy Zones’ (FEZs), based on
dominant livelihood systems. These zones can be characterised using both
secondary data sources, for example by mapping census or survey enumer-
ation areas onto the zones, or primary data collection, for example using
participatory techniques with communities within the zones. In statistical
terms this technique can be seen as an attempt to define strata that are
homogeneous in terms of livelihood strategies and thus likely to display
homogeneity in terms of policy impact.

Combining Geographical Information Systems and Participatory Poverty

Assessments?
Geographical Information Systems may have a role to play in combining
the results from participatory assessments and household surveys. For
example, in The Gambia, wet- and dry-season participatory assessments
are being undertaken as part of a three-year Canadian-funded project.
The areas included were selected from the enumeration areas sampled for
the 1998 National Health and Population Survey, and the PPA gathered
qualitative information relating to income sources and expenditure items
from households included in that survey. There are thus possibilities for
combining data at various geographical levels: providing basic survey
information on specific poverty target groups in particular regions, and
supporting this with qualitative information on those same populations
from participatory work.

Process monitoring: reforming and challenging administrative systems

The practical need for intermediate process monitoring

Poverty monitoring in the narrow sense is not only of limited use for
accountability and immediate learning purposes, but also in some respects
unnecessary. A case in point is the enormous difficulty and expense of
accurately measuring short-run declines in maternal mortality, one of the
 Monitoring progress at country level 115
primary Millennium Development Goals. The health non-governmental
organisation (NGO) Options is among those stressing the value of ‘process
indicators’ based on routinely collected facility data to monitor the situ-
ation of pregnant women. This example illustrates the wider challenge
posed by the tracking of the key intermediate steps in implementing a
PRSP.
Such indicators have been found potentially useful in areas of health
monitoring. In an analysis of trends in infant mortality rates in Zambia,
Simms et al. (1998) found that the most highly correlated variable was
attendances at antenatal clinics. This probably simply indicated the exist-
ence of a reasonably functional local health service. The ratio of clinic
births to antenatal clinic visits is also a useful local indicator of women’s
ability to afford maternal health services (whilst visits may be free or very
low cost, giving birth at a clinic is expensive compared to having a tradi-
tional birth attendant).
Whatever their merits, ‘For process indicators to be successfully used,
projects need to invest time and resources in building the capacity of facil-
ity staff to understand, collect and use routine data’ (Options, 2001: 1).
Similar sentiments have been expressed repeatedly over the years, not
only in relation to health staff, but with reference to teachers, extension
workers, local government administrators and other actors in local service
delivery. It is difficult to find much evidence of the considerable ‘time and
resources’ which have indeed been allocated to this task.

Confronting incentive issues in administrative systems

The response to incentive problems should clearly not be to abandon the
attempt to measure progress. Process indicators are central to monitoring
poverty reduction strategies, and relatively low-level service delivery and
administrative staff will be key actors in delivering the required data.
However, more innovative strategies are required, beyond the established
approaches based on information systems design and training programmes.
One seldom-addressed issue in poverty monitoring is that many of
those charged with gathering data and reporting on the poor are them-
selves living very close to the poverty line. A qualified nurse in a public
village health station in Nigeria has a salary equivalent to US$1 per day. A
graduate teacher in The Gambia earns around 80 US cents. Less qualified
staff, for example agricultural or health extension workers, may have
incomes below the poverty line. Moreover, the lowest-paid staff are com-
monly found in precisely those areas that have the highest concentration
of poor households.
What are the implications for monitoring? Two key issues are relevant.
First, making additional demands on those who perceive themselves as
inadequately rewarded is unlikely to be met with much enthusiasm.
Second, poorly paid staff typically look for ‘livelihood strategies’ to
116 David Booth and Henry Lucas
increase their incomes. Such strategies usually involve at least non-
observance of their working codes of conduct, and often illicit use of the
resources or status provided by their position. In many countries, central
administrations have limited capacity to regulate such behaviour. Those
behaving in this fashion will tend to regard improved monitoring with
considerable suspicion, if not open hostility. Control over information –
for example, about fee rates or official opening hours of health facilities –
may be a valuable ‘livelihood asset’, and one not willingly surrendered.
One of the few projects to directly address this issue has recently been
started in Cambodia by Médecins sans Frontières (van Damme and Meessen,
2001). This scheme is heavily subsidised, but is not without wider interest.
It offers a ‘new deal’ to local health workers and administrators, as a way of
breaking a downward spiral linking low basic salaries with poor service
quality, low utilisation and minimal fee income from which to pay bonuses.
A related example of the effective use of contracting was observed in
recent evaluation work in poor rural areas of China (Yu et al., 1998).
Service providers (‘village doctors’) were contracted under a limited pre-
payment scheme. Claims for payment from the village health care fund
required that the provider return a simple patient diagnosis and treatment
record to the fund manager. The file of such records provided a basic but
effective information system that could in principle be used to monitor
both health service utilisation and quality of care. The simple existence of
this system appears to have been sufficient to improve provider behaviour.

Communities versus providers?

In recent years there has been increased emphasis on community partici-
pation in the design and implementation of development projects. It
might therefore seem reasonable that actual and potential users of
services should be encouraged to play a larger role in monitoring the
delivery of those services. However, detailed consideration of possible
mechanisms raises many difficult questions. Why should communities take
on such activities? What benefits might they gain? Do suitable community
groups exist, or could they be created? How should such groups be consti-
tuted and what training and resources would they need? What should they
monitor and how could monitoring be undertaken? What relationship
would they have with providers and how would providers respond? How
should they relate to existing service managers, to other local government
officials and to NGOs?
The ‘balance of power’ between providers and users must be taken into
account. Contracts work best when the services to be delivered are relat-
ively easy to measure and monitor. There is a need for effective penalties
for default and for both parties to have equal recourse to enforcement.
Qualified staff are in short supply, particularly in poor areas. This fact
gives them considerable status and may allow them to dictate conditions
 Monitoring progress at country level 117
of service. Even when community monitoring identifies inappropriate or
even illegal behaviour, local administrators may side with extension
workers, teachers or health providers, to avoid losing them.
Monitoring strategies that fail to address the concerns and interests of
providers stand little chance of success. An alternative is the development
of ‘partnership’ models – supporting providers and user communities to
negotiate jointly determined priorities, establish common objectives and
agree how best to use their joint resources to pursue those objectives. In
Bolivia the community health information system pools data collected by
community health promoters and health service providers. These are pre-
sented in accessible graphical format and are used to stimulate joint
decision-making, progress monitoring and advocacy to higher levels
of government (Howard-Grabman, 2000). Other examples of providing
information to communities on education outcomes in Cambodia, The
Gambia and Ghana demonstrate how both local officials and parents can
become more actively involved through obtaining even incomplete data
on performance.

Beyond administrative data: parallel systems, special surveys and non-

survey instruments
There are sound reasons not to rely on the reform of routine systems but
to develop information sources that run parallel to them. One reason is
that such reforms may take some time. The other is the problem of inher-
ent limitations mentioned above. We need sources of information that are
not facility based because reach is a crucial issue, and facility-based data
cannot tell us much about reach. Not relying on administrative data does
not necessarily mean creating new structures. In many countries there are
relatively cheap and simple, yet reasonably reliable, data collection instru-
ments such as those set up for famine early warning purposes. Four spe-
cific kinds of instrument merit special attention: service delivery surveys
(and household surveys collecting data on service use and quality);
integrity and business climate surveys; commissioned studies; and qualitat-
ive impact monitoring or participatory process monitoring.
Service delivery surveys have been used to good effect in a number of
countries, including Bangladesh, Tanzania and Uganda (Foster, 2001). A
typical survey combines interviews with representative samples of house-
holds, interviews with service providers and key informants, schedules
completed by enumerators giving details of facilities and services, and, in
some cases, user ‘exit polls’. The information generated includes the pro-
portion of the population using government and other services, differ-
ences in patterns of use across social categories, and reasons for use and
non-use. Such surveys cover the key gap in administrative data, namely,
the reach of official provision and the factors responsible for limiting
access. Whilst no doubt subject to some methodological imperfections,
118 David Booth and Henry Lucas
these surveys effectively sidestep the problem of motivating service providers
to report on themselves. Other standard surveys contain underexploited
information on service use, including integrated household surveys and,
in a more focused way, Core Welfare Indicators Questionnaires (CWIQs).
These sources should be used more intensively as means of tracking the
performance of PRSPs.
Integrity surveys and surveys that investigate the climate of business
confidence in a country are also worthy of attention. For example, official
corruption is often ignored in PRSPs. However, this may change as
national dialogue on the Paper develops through the review and revision
phases. The same goes for a somewhat broader range of issues in gover-
nance and the rule of law. Existing examples include Uganda, where use
has been made of both Integrity Surveys and business climate surveys in
the biennial Poverty Status Reports. Although they combine focus-group
work and exit poll surveys, these studies are subject to the well-known
weaknesses of ‘attitude’ surveys and could no doubt be strengthened with
research with a more ‘behavioural’ emphasis (Appleton and Booth, 2001:
sec. 2.4). Nonetheless, they provide a very useful complement to other
survey-based and administrative information.
Commissioned studies may take the form of surveys, or be based on
one-off participatory assessment exercises. There are also studies commis-
sioned to investigate a specific ‘missing middle’ issue, such as the pros and
cons of alternative approaches to meeting final outcome goals. Thus
Foster and Macintosh-Walker (2001: 5) report:

In the face of a disappointing public response to the expansion of

primary health services, Ghana and Bangladesh have researched the
causes of unequal access and are developing more specific strategies
for reaching the poor. Zambia and Cambodia have focused basic edu-
cation interventions on understanding the barriers to enrolment by
the poor and introducing specific policies to address them. The
problem of cost to parents was identified as a major barrier in all but
one of our education cases (most dramatically in Uganda), and a key
intervention has been to reduce costs to parents.

Qualitative impact monitoring and participatory process monitoring cover a

rather broad category including a range of technically different but substan-
tially similar traditions and techniques. In several countries there are long-
established arrangements for conducting regular participatory ‘beneficiary
assessments’ in connection with social funds and other large projects. In
Zambia the group originally set up for this purpose was subsequently
involved in the World Bank PPA, and has since contributed to a range of
commissioned sectoral policy studies. In Kenya, Malawi and Benin, poverty
monitoring arrangements are being upgraded and mainstreamed within
the Papers (Bierschenk et al., 2001; Gomonda, 2001; GTZ-SPAS, 2001).
 Monitoring progress at country level 119
There are a number of challenges. What is needed is to draw fully on
the extensive experience of official and NGO project monitoring and
impact assessment using learning-process and participatory methods (e.g.
Brown et al., 2001; Estrella, 2000; Mosse et al., 1998; Roche, 1999), while
adjusting for the very different purpose and scope of PRSPs. Another is to
achieve the same balancing act with respect to the recorded experience of
traditional participatory assessments and their linkage to policy processes
(e.g. Holland with Blackburn, 1998; Norton et al., 2001; Robb, 1999).

Conclusions
This chapter has explored existing Poverty Reduction Strategy Papers in
sub-Saharan Africa from a particular angle. It has deliberately taken a
robustly realist approach in two respects. First, the criteria applied to
describing and assessing the documents in respect of monitoring systems
reflect not just established monitoring and evaluation principles, but a
vision of the policy process that is more realistic and less rationalistic than
the norm in this field. Second, our discussion of indicators and data
sources is equally stringent in not ignoring what is known about the real
condition of African countries’ information systems. In our view, anything
less than this would do poor service to the cause of more effective anti-
poverty action in the region. The thinking reflected in the PRSPs on the
topic of monitoring is very patchy. This is partly because most countries
only have Interim Papers that are designed in part to get access to debt
relief and further international loans. However, our purpose is not to criti-
cise, but to identify topics on which action might be taken, or further
inquiries justified.
All concerned are turning a blind eye to the poor quality of administra-
tive data. This matters in that intermediate output/outcome monitoring
(in addition to input monitoring and tracking) is likely to be the
most fruitful for generating information capable of changing behaviour
and procedures. The enthusiasm for household surveys, and monitoring
final outcomes/impacts, is in many ways justified. But it will be a pity if
it provides an excuse for not tackling the issue of quick feedback on
implementation.
Two key questions arise: how can improvement of administrative
reporting and management information systems be addressed, given the
limited achievements of numerous previous attempts; and how should this
activity be balanced against development of other monitoring procedures?
There are various alternatives to the management information systems
approach, some already fairly well institutionalised within the better
public service reform programmes, others reflecting a decade of work by
participation specialists at the World Bank and elsewhere, and yet others
pioneered by non-governmental organisations. NGO experience on
impact assessment may have additional clues as to worthwhile shortcuts in
120 David Booth and Henry Lucas
monitoring. Further systematic reviews of these alternatives would be
worthwhile.
More strategic selection of indicators is an obvious topic for further
work. However, this cannot be pursued as a mere monitoring question.
The appearance of randomness in current indicator listings arises in good
part from the ‘missing middle’ in poverty reduction strategies. Most
Papers developed to date fail to identify which critical changes need to
occur for the identified actions to produce the desired results. Ideally, a
monitoring system should focus on detecting quickly whether such key
changes are occurring.
A point of entry into this topic from the perspective of good practice in
the development of monitoring systems is the question of the continuing
involvement of a range of stakeholders in monitoring activities. If non-gov-
ernmental stakeholders remain mobilised after debt relief has been
secured and can receive feedback on implementation issues, fresh think-
ing on strategic bottlenecks and priority actions may be stimulated.
Further work to document the lessons of early experience on this point
would be justified.

Notes
1 This chapter derives from a report commissioned by the Department for Inter-
national Development (DFID) (Booth and Lucas, 2001a, b) on behalf of the
Strategic Partnership with Africa’s (SPA’s) Poverty Monitoring Task Team. Its
description of the content of PRSPs refers to the situation in 2001.
2 The language conventions in this field are a mess. The main issue is that differ-
ent meanings are given in different contexts to the words ‘outcome’ and
‘impact’. The Development Assistance Committee (DAC) and the monitoring
and evaluation profession convention, in the field of poverty reduction policy, is
that outcomes are ‘specific results and the utilisation of means/services by bene-
ficiaries’. Movements in measures of poverty are referred to as ‘impacts’.
However, in the broader social science fields concerned with poverty reduction
strategies and poverty information, it has been conventional to speak of the final
goal of policy as to influence poverty ‘outcomes’, or ‘outcomes for the poor’.
There is also a tendency to associate ‘impact’ with the activity of evaluation,
implying that an impact is not just a final result, but one that can be attributed
to a specific intervention. We have tried to avoid misunderstanding by qualifying
everything. Thus we distinguish intermediate outcomes, closely linked to inter-
mediate outputs, and final outcomes or poverty outcomes. In deference to the
DAC convention, we often write ‘final outcomes/impacts’.
3 This is the Ugandan equivalent of a Poverty Reduction Strategy Paper.
4 The perhaps misleadingly named ‘Social Dimensions of Adjustment’ was a joint
World Bank–United Nations Development Programme (UNDP) initiative, sup-
ported by several bilateral agencies, primarily aimed at improving data collec-
tion.
5 Uganda is a partial exception.
6 Population estimates in years removed from that in which the census is taken
are derived from demographic models, often based on parameters estimated
from Demographic and Health Surveys (DHSs). The procedure is reasonably
 Monitoring progress at country level 121
reliable at the national level, but the estimates are not intended for sub-national
estimation and provide little evidence on internal migration. Adjusting such
demographic models to allow for the unprecedented impact of the AIDS pan-
demic is also a relatively new and uncertain methodological exercise.
7 See the PRSP Institutionalisation Study (Booth and associates, 2001: chapter 1).
8 For example, PPAs have been widely used to generate ‘quantitative’ findings on
specific communities.
9 These concepts are used in a number of areas including medicine and engin-
eering. They also relate to the traditional Type I and Type II errors of hypothe-
sis-testing.

References
Appleton, S. and Booth, D. (2001) ‘Combining Participatory and Survey-Based
Approaches to Poverty Monitoring and Analysis’, Uganda Workshop Back-
ground Paper, second draft, 27 May.
Assiimwe, D., Mwesigye, F., McPake, B. and Streetland, P. (1997) ‘The Private-
Sector Activities of Public-Sector Health Workers in Uganda’, in Bennett, S. and
McPake, B. (eds) Private Health Providers in Developing Countries: Serving the Public
Interest? London: Zed Books.
Bevan, D. and Palomba, G. (2000) ‘Uganda: The Budget and Medium Term
Expenditure Framework Set in a Wider Context’, background paper for Poverty
Reduction Support Credit with DFID finance, October.
Bierschenk, T., Thioléron, E. and Bako-Arifari, N. (2001) ‘Institutionalising the
PRSP Approach in Benin’, chapter 2 in Booth D. and associates, PRSP Institution-
alisation Study: Final Report, submitted to the SPA, October.
Bigman, D. and Fofack, H. (2000) Geographical Targeting for Policy Alleviation:
Methodology and Applications, Washington: World Bank Regional and Sectoral
Studies.
Booth, D., Holland, J., Hentschel, J., Lanjouw, P. and Herbert, A. (1998) Participa-
tion and Combined Methods in African Poverty Assessment: Renewing the Agenda,
London: DFID, Social Development Division, February.
Booth, D. and Lucas, H. (2001a) Desk Study of Good Practice in the Development of
PRSP Indicators and Monitoring Systems: Initial Review of PRSP Documentation,
report commissioned by DFID for the SPA, May.
Booth, D. and Lucas, H. (2001b) Desk Study of Good Practice in the Development of
PRSP Indicators and Monitoring Systems: Final Report, report commissioned by
DFID for the SPA, November.
Booth, D. and associates (2001) PRSP Institutionalisation Study: Final Report, submit-
ted to the SPA, October.
Brown, D., Hussein, K. and Longley, C. with Howes, M. (2001) Review of Participa-
tory Rural Appraisal in The Gambia, London: ODI.
Devereux, S. (2001) ‘Food Security Information Systems’, in Devereux, S. and
Maxwell, S. (eds) Food Security in Sub-Saharan Africa, ITDG.
Estrella, M. (ed.) (2000) Learning from Change: Issues and Experiences in Participatory
Monitoring and Evaluation, London: Intermediate Technology Publications/ IDRC.
Foster, M. (2001) Use of Surveys to Improve Public Expenditure Management, London:
CAPE/ODI, June.
Foster, M. and Mackintosh-Walker, S. (2001) Sector Wide Programmes and Poverty
Reduction, London: CAPE/ODI.
122 David Booth and Henry Lucas
Foster, M. and Mujimbi, P. (2001) How, When and Why Does Poverty Get Budget
Priority? Uganda Case Study, London: Overseas Development Institute, April.
Gomonda, N. (2001) ‘Qualitative Impact Monitoring of the Poverty Alleviation
Policies and Programmes in Malawi’, Second Forum of Poverty Reduction Stra-
tegies, Dakar, 10–13 September.
GTZ-SPAS (2001) An Introduction to KePIM: Kenya’s Participatory Impact Monitoring
Exercise, Nairobi: Ministry of Finance and Planning and Poverty Eradication
Unit, Office of the President.
Hauge, A. (2001) Strengthening Capacity for Monitoring and Evaluation in Uganda: A
Results Based Management Perspective, Washington, DC: World Bank OED, Evalu-
ation Capacity Development Working Paper 8, January.
Hentschel, J., Lanjouw, J., Lanjouw, P. and Poggi, J. (2000) ‘Combining Census
and Survey Data to Trace the Spatial Dimensions of Poverty: A Case Study of
Ecuador’, World Bank Economic Review, 14(1): 147–165.
Hill, M. (ed.) (1993) The Policy Process: A Reader, New York: Harvester Wheatsheaf.
Holland, J. with Blackburn, J. (1998) Whose Voice? Participatory Research and Policy
Change, London: Intermediate Technology Publications.
Howard-Grabman, L. (2000) ‘Bridging the Gap between Communities and Service
Providers’, in Cornwall, A., Lucas, H. and Pasteur, K. (eds) Accountability
through Participation: Developing Workable Partnership Models in the Health
Sector, IDS Bulletin, 31(1): 88–96.
McGee, R. and Brock, K. (2001) ‘From Poverty Assessment to Poverty Change:
Processes, Actors and Data’, IDS Working Paper 133, Brighton: Institute of Devel-
opment Studies, July.
Minot, N. and Baulch, B. (2001) ‘The Spatial Distribution of Poverty in Vietnam
and the Potential for Targeting’, research funded by DFID Poverty Analysis and
Poverty Support Trust Fund and World Bank Development Economics Research
Group, mimeo, August.
Mosse, D., Farrington, J. and Rew, A. (eds) (1998) Development as Process: Concepts
and Methods for Working with Complexity, London: Routledge/ODI.
Norton, A. with Bird, B., Brock, K., Kakande, M. and Turk, C. (2001) A Rough
Guide to PPAs, London: Overseas Development Institute.
Options (2001) Options News, newsletter no. 7, July 2001.
Prennushi, G., Rubio, G. and Subbarao, K. (2001) ‘Monitoring and Evaluation’,
World Bank Poverty Reduction Strategy Paper Sourcebook (draft for comments, April
2001) ([Link]/poverty).
Reinikka, R. and Svensson, J. (2001) ‘Explaining Leakage of Public Funds’, paper
prepared for the World Institute for Development Economics Research
(WIDER) Development Conference on Debt Relief, Helsinki, 17–18 August.
Robb, C. (1999) Can the Poor Influence Policy? Participatory Poverty Assessments in the
Developing World, Washington, DC: The World Bank.
Roche, C. (1999) Impact Assessment for Development Agencies: Learning to Value Change,
Oxford: Oxfam/Novib.
Sahn, D.E. (2001) ‘Strengthening Quantitative Methods Through Incorporating
Qualitative Information’, in Kanbur, R. (ed.) Qualitative and Quantitative Poverty
Appraisal: Complementarities, Tensions and the Way Forward, Cornell University
Workshop, 15–16 March.
Simms, C., Milimo, J.T. and Bloom, G. (1998) ‘Reasons for the Rise in Childhood
Mortality during the 1980s in Zambia’, IDS Working Paper 76.
 Monitoring progress at country level 123
Thin, N., Underwood, M. and Gilling, J. (2001) Sub-Saharan Africa’s Poverty Reduc-
tion Strategy Papers from Social Policy and Sustainable Livelihoods Perspectives, Report
for DFID, Oxford: Oxford Policy Management, March.
Tikare, S., Youssef, D., Donnelly-Roark, P. and Shah, P. (2001) ‘Organizing Partici-
patory Processes in the PRSP’, World Bank Poverty Reduction Strategy Sourcebook
(draft, April 2001) ([Link]/poverty).
Turner, M. and Hulme, D. (1997) Governance, Administration and Development:
Making the State Work, London: Macmillan.
Uganda (2001) Poverty Eradication Action Plan (2001–2003), vol. 1, Kampala: Min-
istry of Finance, Planning and Economic Development.
van Damme, W. and Meessen, D. (2001) ‘Sotnikum New Deal, the First Year:
Better Income for Hospital Staff; Better Service to the Population’, mimeo,
Cambodia: Médecins sans Frontières.
Weiss, C. (1998) Evaluation, second edition, Englewood Cliffs, NJ: Prentice Hall.
Yu, H., Lucas, H., Gu, X.-Y. and Shu, B.-G. (1998) ‘Financing Health Care in Poor
Rural Counties in China: Experience from a Township-Based Co-operative
Medical Scheme’, IDS Working Paper no. 66, Brighton: Institute of Develop-
ment Studies.
6 Are the Millennium Development
Goals feasible?1
Jan Vandemoortele

As long as you travel to a goal, you can hold on to a dream.

(Anthony de Mello)

Introduction
The Millennium Development Goals are a set of numerical and time-
bound targets that express key elements of human development. They
include halving income poverty and hunger; achieving universal primary
education and gender equality; reducing under-five mortality by two-
thirds and maternal mortality by three-quarters; reversing the spread of
HIV/AIDS; and halving the proportion of people without access to safe
water. These targets are to be achieved by 2015, from their level in 1990
(United Nations, 2000).
It is often said that global targets are easily set but seldom met, which
poses the question as to whether the Millennium Development Goals are
feasible. Progress in over 130 developing countries regarding the many
dimensions of human development – such as education, health, nutrition
and income – is difficult to summarise. The 1990s saw many success
stories, including those relating to education in Guinea and Malawi;
HIV/AIDS in Senegal, Thailand and Uganda; child mortality in Bangladesh
and the Gambia; nutrition in Indonesia, Mexico and Tunisia; and income
poverty in China. Globally, the number of polio cases dropped from nearly
250,000 in 1990 to less than 3,000 in 2000, making eradication of the
disease by 2005 a realistic goal.
But for each success story there have been setbacks. The under-five
mortality rate increased in Cambodia, Kenya, Malawi and Zambia – an
unprecedented trend after decades of steady decline. The primary school
enrolment ratio dropped in Cameroon, Lesotho, Mozambique and Tanza-
nia. The gender gap in primary education widened in Eritrea, Ethiopia
and Namibia. Instead of decreasing, malnutrition increased in Burkina
Faso and Yemen. Access to water became more difficult for millions of
people; Bangladesh faced a major problem with arsenic water poisoning.
In the 1990s, countless countries saw their HIV prevalence rate double,
 Are the Millennium Development Goals feasible? 125
triple, quadruple, even increase tenfold – severely undermining the feas-
ibility of most Millennium Development Goals, in health and beyond.
Monitoring can be done at different levels, from the global to the local.
The level of assessment will influence the outcome regarding the feas-
ibility of the Millennium Development Goals. If Millennium Development
Goals appear feasible at the global level, it does not necessarily imply that
they will be feasible in all nations or at all locations. Averages are com-
monly used at each level to measure Millennium Development Goal
progress. While they give a good sense of overall progress, averages can be
misleading. The failure to understand that the average is an abstraction
from reality can lead to unwarranted conclusions that are based on deduc-
tion from abstractions, not on real observations.
A good assessment of progress towards the Millennium Development
Goals must, therefore, go beyond averages and aggregates. The failure to
disaggregate according to gender, for instance, easily leads to the fallacy of
‘misplaced concreteness’ (Daly and Cobb, 1994). Average household in-
come is very much an abstraction for women who have little or no control
over how it is spent; it may exist in the mind of economists but it does not
necessarily correspond with the reality faced by millions of poor women.2
This chapter reviews global progress towards the Millennium Develop-
ment Goals during the 1990s. The picture that emerges shows a very
uneven pattern across regions and countries and between different socio-
economic groups within the same country. Although the picture is mixed,
the overall conclusion is that none of the agreed targets for 2000 were met
at the global level. If the 1980s were the ‘lost decade for development’, the
1990s should go down in history as the ‘decade of broken promises’. If
current trends prevail, only one Millennium Development Goal will be
reached by 2015.

Is progress on track?
This review is based on the best data that are currently available. It focuses
on indicators for which global information is reasonably reliable, compa-
rable and up to date.3 However, it must be kept in mind that global trends
are estimates; they are never precise or actual values. Therefore, different
sources often give different estimates, without necessarily being inconsis-
tent. Indicators without trend data or with inconsistent data have been
omitted, hence the review does not include all Millennium Development
Goals. Subsequent chapters fill some of these gaps, as well as amplifying
the argument for particular indicators.

Income poverty
In developing countries the average proportion of people living on less
than $1 per day4 decreased from 32 per cent in 1990 to 25 per cent in
126 Jan Vandemoortele
1999, according to the latest estimates (World Bank, 2002).5 The simple
extrapolation of this trend suggests that the world is on track to halving
income poverty by 2015. Unfortunately, the reality is more complicated
and decidedly less satisfactory. Most of the global progress was due to a
rapid decline in Asia, particularly in China. Progress in Latin America and
the Caribbean, sub-Saharan Africa, and the Middle East and North Africa,
combined, was merely a tenth of what was required to meet the agreed
target.
In addition, poverty estimates for China show large discrepancies,
which seriously undermine the reliability of global poverty data (see also
Chapter 7). Figure 6.1 shows a steep decline in China’s income poverty
between 1993 and 1996, when the head-count index reportedly declined
from 29 per cent to 17 per cent. This implies that the number of people
in China struggling to survive on less than a $1 per day dropped by a stag-
gering 125,000 people per day for three years running. This remarkable
achievement came to a sudden – and mysterious – end in 1996. Actually,
the number of poor people reportedly increased slightly between 1996
and 1999.
National poverty estimates, on the other hand, show a less dramatic
decline in China’s poverty level. Poverty estimates reported by the Ministry
of Agriculture show a decrease by less than one percentage point per year
between 1993 and 1996 (Khan and Riskin, 2000), considerably less than
the four percentage points suggested by the World Bank estimates.6 More-

Figure 6.1 Incidence of income-poverty in China, 1987–1999 (percentage of the

population below the poverty line).
Source: Based on data from the World Bank (2002) and data supplied by the Ministry of
Agriculture, China.
 Are the Millennium Development Goals feasible? 127
over, if demographic change has been a major force behind China’s
success in reducing income poverty – as some analysts have documented
(Gustafsson and Zhong, 2000) – then it would be unwise to assume that its
rapid decline will continue until 2015.
In short, global poverty trends cannot be taken at face value. Given the
inherent weaknesses associated with the fixed and static poverty line of $1
per day and given the inaccuracy of purchasing power parity (PPP) conver-
sion rates, global poverty estimates are not a reliable source of information.
Global poverty data are not robust; therefore, it cannot be argued that the
world is on track to reach the target for halving income poverty by 2015.
Dozens of countries experienced a decline in average living standards in
the past decade. Moreover, the simple extrapolation of global poverty
trends to 2015 is invalid because large countries – such as China and
Indonesia – will gradually become less powerful in pulling global poverty
down as they reach lower levels of poverty. Global poverty projections will
be meaningful only if they are based on country-specific projections.

Education
In 1990 the goal was set to provide basic education for all children by
2000. The sad truth is that the 1990s saw only about a fifth of the global
progress needed. For developing countries, the average net enrolment
ratio for primary education increased from 78 per cent in 1990 to 83 in
2000. Not surprisingly, the goalpost was moved to 2015; but this promise
will not be kept either if progress does not accelerate twofold between
2000 and 2015. At the current rate, the global education target will not be
reached until 2030 (see also Chapter 8).
Figure 6.2 shows that progress was significantly slower in the 1990s than in
the preceding three decades, when the average enrolment ratio increased by
approximately ten percentage points per decade – compared with only five
percentage points in the 1990s.7 In 2000, an estimated 120 million school-age
children were not enrolled – about the same as a decade earlier. They joined
the ranks of the nearly 1 billion adults who cannot read or write – most of
them women. Globally, the world is not on track to meeting the education
target.
Failure to meet the education target will reduce the chances of reach-
ing other Millennium Development Goals because basic education is key
to unlocking positive externalities and synergies. Basic education em-
powers a young woman and enhances her self-confidence; an educated
mother is likely to marry later, space her pregnancies better, and seek
medical care for her child and herself when needed (see also Chapter 11).
Evidence shows that babies born to mothers without formal education are
at least twice as likely to suffer from malnutrition or die before the age of
5 than are babies born to mothers who completed primary school (Bicego
and Ahmad, 1996). An educated girl is also the best guarantor that her
128 Jan Vandemoortele

250
223

Under five mortality rate

NER (%), U5MR (per 000 live births)

200

166

150
132

103
100 91

Net enrolment rate 83

78
70
50 59
48

0
1960 1970 1980 1990 2000

Figure 6.2 Average net primary enrolment ratio (NER) and under-five mortality
rate (U5MR) in developing countries.
Source: Based on UNICEF (2001) and data supplied by United Nations Education, Scientific
and Cultural Organisation (UNESCO).

children attend school – thereby ending the inter-generational transmis-

sion of poverty. Health investments are more efficient when the people
are better educated, in large part as a result of the adoption of good
hygienic behaviour. In short, girls’ education is key to achieving the Mil-
lennium Development Goals.
The good news is that the gender gap in primary enrolment narrowed
in the 1990s. For developing countries, the number of girls per 100 boys
enrolled in primary school increased from 83 in 1990 to 88 in 2000.
However, this will be insufficient to reach gender equality by 2005, as
agreed in the Millennium Development Goals. Progress would have to
accelerate more than fourfold in the period 2000–2005 if this target were
to be achieved. Globally, the world is not on track to reaching gender
equality in primary education by 2005.8 At the current rate, the target will
not be met until 2025. Gender discrimination in primary school enrolment
remains a concern in several countries, particularly in sub-Saharan Africa,
South Asia and the Middle East and North Africa (see also Chapter 9).
 Are the Millennium Development Goals feasible? 129
Child mortality
In 2000, more than 10 million children under the age of 5 died, mostly
from preventable causes such as pneumonia, diarrhoea, measles, malaria,
HIV/AIDS and malnutrition. For developing countries, the average
under-five mortality rate decreased from 103 to 91 deaths per 1,000 live
births between 1990 and 2000. The rate of progress was less than half that
achieved in the previous three decades, as shown in Figure 6.2 (see also
Chapter 10). For several countries, slow progress was due to the mother-
to-child transmission of HIV, which is contributing to an unprecedented
increase in infant and child mortality. In Zimbabwe, for example, some 70
per cent of deaths among children under the age of 5 are due to AIDS
(see also Chapter 12).
There is no marked difference between girls and boys when it comes to
the average under-five mortality rate but Demographic and Health Surveys
(DHSs) show that in virtually all countries, baby boys experience higher
levels of infant mortality – sometimes 50 per cent higher – than baby girls,
mostly because of biological factors. After infancy, however, the gender
gap in terms of mortality gradually reverses, switching from being pro-
female in infancy to being pro-male in childhood (ages 1–4). Such a rever-
sal is observed in Bangladesh, Bolivia, Brazil, Cameroon, the Dominican
Republic, Gabon, Guatemala, Haiti, India, Kenya, the Kyrgyz Republic,
Nepal, Niger, Nigeria, Peru and Vietnam.
In the absence of a biological explanation, environmental factors must
be examined to understand the causes for this reversal. DHS surveys con-
sistently show that baby boys are more likely to be vaccinated and breast-
fed than baby girls.9 Both indicators suggest a greater commitment on the
part of parents and service providers to the health and development of
boys vis-à-vis girls.
If the global trend of the 1990s were to continue at the same rate until
2015, the reduction in the under-five mortality rate would be about one-
quarter – far less than the agreed target of a two-thirds reduction. Meeting
the global target will require that the rate of reduction increases more
than fivefold between 2000 and 2015 – an extremely unlikely scenario.
Almost half of the under-five deaths occur in sub-Saharan Africa, so that a
sudden and dramatic improvement in child mortality in that region must
come about if the global target is to be achieved. Globally, the world is not
on track to reach the target for child mortality.
Immunisation is essential to reducing child mortality. Measles is among
the leading causes of child mortality that are vaccine-preventable; but
immunisation coverage stagnated in the 1990s at about 70 per cent. The
coverage has to reach at least 90 per cent to reduce measles deaths effect-
ively. That level was reached only in Latin America and the Caribbean and
in East Asia, whereas coverage actually decreased in sub-Saharan Africa to
about 50 per cent in 2000, down from over 60 per cent in 1990.
130 Jan Vandemoortele
Child malnutrition
Deaths among children under the age of 5 are often associated with mal-
nutrition, mostly with moderate malnutrition; only one-quarter of the
deaths result from severe malnutrition. The crisis is, therefore, largely
invisible as the young victims seldom show outward signs of under-nour-
ishment.
In 1990 the target was set to halve the proportion of children suffering
from malnutrition by 2000. Data show that the proportion of moderately
and severely underweight children in developing countries declined from
32 per cent to 28 per cent, respectively. Thus only one-quarter of the
promise was kept. As part of the Millennium Development Goals, the goal-
post was pushed to 2015; but the current rate of progress will have to
increase threefold if malnutrition in developing countries in 2015 is to be
half the level that prevailed in 1990. Globally, the world is not on track to
meet the nutrition target.
The largest decline was observed in East Asia, especially in China; sub-
stantial improvements were made in Latin America and the Caribbean.
Less progress was made in South Asia, where underweight prevalence
remains very high. Sub-Saharan Africa saw little or no change over the
decade. Overall, the number of malnourished children in developing
countries fell by approximately 25 million – or 15 per cent – decreasing
from 174 million to 150 million. However, their numbers increased in sub-
Saharan Africa and South Asia.
Data from over 100 countries do not suggest that girls are more likely to
be malnourished than boys. Except for South Asia, most regions actually
show a slightly higher rate of malnutrition for boys. However, gender grad-
ually becomes a greater liability as girls grow older, and by the time they
reach reproductive age, many suffer from anaemia. In almost all countries,
rural children are more at risk of malnutrition than their urban counter-
parts. In some countries the percentage of rural children who are under-
weight is more than 50 per cent higher than that for urban children.

HIV/AIDS
Two decades after it was first reported, AIDS is the most serious threat to
human development in a growing number of countries. It is the leading
cause of death in sub-Saharan Africa; worldwide it is number 4 in the
league of major killers. The pandemic – raging in Africa and spreading
fast in other regions – is perhaps the greatest impediment to achieving the
Millennium Development Goals by 2015 (see also Chapter 12). Even
countries with a relatively low national HIV prevalence rate can have clus-
ters of people or specific locations where the prevalence rate is as high as
20 per cent or more; but these pockets of crises are hidden in national
statistics owing to the relatively small populations affected.
 Are the Millennium Development Goals feasible? 131
About one-third of those currently living with HIV/AIDS are aged
between 15 and 24 years. Adolescent girls are at particularly high risk,
owing to a mix of biological and social factors. HIV/AIDS is a disease for
which gender could not be more central: women represent a growing pro-
portion of people living with HIV/AIDS. In countries with high HIV
prevalence, young women with little or no education – that is, those
without much power in society – are at the greatest risk of infection (Van-
demoortele and Delamonica, 2000). Studies in Africa show that teenage
girls are five to six times more likely to be infected by the HIV virus than
boys their age (UNAIDS, 2000). New HIV infections are disproportion-
ately concentrated among poor and illiterate adolescent women.
After a strong public information campaign, Uganda saw the number
of new cases of HIV/AIDS drop from 239,000 in 1987 to 57,000 in 1997.
But even in this exceptional case, the impact on the poor – that is, those
with little or no education – was the least. The HIV infection rate amongst
educated women dropped by almost half in the 1990s, whereas it did not
show a significant decrease for women without formal schooling.
Millions of young people do not know how to protect themselves
against HIV. In the late 1990s, surveys in sub-Saharan African countries
found that half the teenagers did not know that a healthy-looking
person can be HIV-positive. The proportion of young people who do not
know that HIV/AIDS cannot be transmitted by mosquitoes is over 80 per
cent in Albania, Azerbaijan, Chad, Niger, Somalia, Tajikistan and Uzbek-
istan. Out of a sample of twenty-three countries, that proportion is less
than half in only two: Cuba (35 per cent) and Kenya (45 per cent). In
many countries, open and frank discussions about HIV transmission face a
wall of silence. Four allies make the virus so prevalent in many societies:
silence, shame, stigma and superstition. These four Ss thrive in a climate
of ignorance and illiteracy, making education a key to defeating this
deadly alliance.
But several countries face a Catch-22: education is important to reverse
the pandemic but HIV/AIDS undermines the education system. Absen-
teeism among teachers is high, owing to AIDS-related illness and deaths,
care for sick family members, attendance at funerals, and increased moon-
lighting. In Zambia, for instance, 1,300 teachers died in the first ten
months of 1998 – twice the number of deaths reported in the previous
year. In the Central African Republic, 300 teachers died in 2000, 85 per
cent as a result of AIDS. Several African countries are reportedly losing
more teachers than the number of new recruits. HIV/AIDS also reduces
the demand for basic education, because of the family’s inability to pay for
schooling, concerns about sexual activity at school as they are not always
sanctuaries and safe havens for children, and the declining quality of edu-
cation that makes many children and parents lose interest in school. Glob-
ally, no progress has been made towards the target of reducing the HIV
prevalence among young people.
132 Jan Vandemoortele
Maternal mortality
Complications during pregnancy and childbirth cause the death of approx-
imately 500,000 women each year – about one every minute. But measur-
ing maternal mortality is notoriously difficult, owing to under-reporting
and incorrect diagnoses. Countries with a comprehensive vital registration
system represent less than one-quarter of the world population.
Together with income poverty, the maternal mortality ratio is among
the most difficult indicators to monitor. But there is consensus that
the proportion of births attended by skilled health personnel – a doctor,
nurse or midwife – is very closely correlated with maternal mortality.
Access to care by a skilled health provider at childbirth – when obstet-
ric complications are most likely to occur – greatly reduces maternal mortal-
ity.
In 1990 the target was set to cut maternal mortality in half by 2000. In
developing countries the proportion of births attended by skilled health
personnel increased from 42 per cent to 53 per cent between those two
dates. This was just over a third of the agreed target.10 Not surprisingly, the
goalpost was changed to reducing the maternal mortality ratio by three-
quarters by 2015, which is slightly less ambitious. But the current rate of
progress will have to increase more than threefold if the target is to be
met by 2015. Globally, the world is not on track to reach the target for
maternal mortality.
Progress differed across regions. Sub-Saharan Africa and the Middle
East saw little or no change, whereas North Africa and East and South Asia
observed considerable progress. Latin America and the Caribbean, with
the highest percentage of births attended by skilled health workers, saw
moderate progress. High fertility, combined with high maternal mortality
risk, make a woman in sub-Saharan Africa face a one in thirteen chance of
dying in childbirth over her lifetime, compared with one in 160 in Latin
America and the Caribbean, and one in 280 in East Asia. In industrialised
countries the risk is one in 4,100.

Safe water
Safe sources of drinking water include piped water in the house, public
standpipe, borehole, protected dug well, protected spring, and rainwater
collection. In developing countries, coverage of improved drinking water
sources rose from 71 per cent in 1990 to 78 per cent in 2000 – leaving an
estimated 1.1 billion people without access to safe water.
Progress fell far short of the goal set in 1990 to reach universal access to
safe water by 2000. Not only was the goalpost moved to 2015, but the new
Millennium Development Goal target was lowered from universal cover-
age to halving the proportion of people without access to safe water. Thus
the new target is nearly five times less ambitious than the initial one. At
 Are the Millennium Development Goals feasible? 133
the current rate of progress, the world is on track to reach the new target
for safe water by 2015, but far from reaching its initial goal.
The fastest progress was made in South Asia; little or no progress was
made in the world’s poorest nations – the least developed countries. Rural
areas lag far behind; the rural–urban gap in terms of access to safe water is
greatest in sub-Saharan Africa, where only 45 per cent of the rural popu-
lation have access – against 83 per cent for their urban counterparts (see
also Chapter 13).

Do the poor benefit from ‘average’ progress?

There are different ways of reaching a global or national target. At one
extreme, it can be achieved by improving the situation of the already
better-off segments of society – that is, a top-down approach. At the other
extreme, a target can be achieved by improving the situation of the worse-
off population – that is, a bottom-up approach. Many combinations are
possible in between. The evidence suggests that most countries come
closer to following the top-down rather than the bottom-up approach.
Frequently, the poor are not fully taking part in national progress; evid-
ence suggests that disadvantaged groups are often bypassed by ‘average’
progress.
Different groups in society usually have very different levels of social
and economic well-being, based on characteristics such as gender, age,
rural/urban location, region, ethnicity, religion, wealth, and any combina-
tion thereof.11 Disaggregated data confirm that social indicators vary enor-
mously across groups within the same country. Thus national indicators
hide wide disparities.
Data from over forty DHSs show that a child from a poor family is
invariably more likely to die before age 5 than her counterpart from a rich
family – on average about twice as likely. Similarly, children from poor
families are less likely to complete primary education than children from
rich families. Data for twelve countries in Latin America show that over 90
per cent of the children in the top income decile complete primary
education. The share falls to two-thirds for children in the middle decile
and drops below 40 per cent for children in the bottom decile (Inter-
American Development Bank, 1998).
Given these significant differences in the absolute value of social indic-
ators across groups, progress too is likely to be very different for different
groups. Indeed, an increase in a national indicator does not necessarily
mean that all groups will see their situation improve at the same rate.
DHSs for 1994 and 1997 in Bangladesh, for instance, show that improve-
ments in access to basic education benefited foremost the children from
better-off families; while children from poor families saw little or no
improvement.12 In Peru, where access to primary education worsened in
the 1990s, only the poor bore the consequences; the non-poor were not
134 Jan Vandemoortele
affected. Data for over forty countries indicate that poor children repre-
sent a growing proportion of the ‘education queue’ as the national indica-
tor for education improves, suggesting that the poor often find themselves
at the end of the queue and do not always benefit from ‘average’ progress.
There are twenty-four countries with at least two DHSs between the late
1980s and the late 1990s, which makes it possible to track progress in child
mortality across different groups in the same country. They show that dis-
parities across wealth groups widened in the majority of them. The gap
between the bottom and top quintiles increased most significantly in
Brazil, Colombia, the Dominican Republic, Ghana, Indonesia, Kazakhstan,
the Philippines and Zimbabwe (Figure 6.3).13
In Indonesia, for instance, children in the bottom quintile witnessed a
reduction in under-five mortality rate by one-fifth between 1987 and 1997;
those in the top quintile saw a reduction by one-half. Thus the ratio
between the bottom quintile and the top quintile rose from 2.3 to 3.8. The
trend in Zimbabwe was even starker: a decline in average under-five mor-
tality masked a rise in the number of deaths of children in the poorest
fifth of the population. Between 1988 and 1999 the national under-five
mortality rate decreased by a modest four percentage points, but that for
the bottom quintile actually increased by some twenty percentage points.
By 1999, children in the poorest quintile had an under-five mortality rate
that was four times higher than that for their counterparts in the richest
quintile. Thus the average trend had little to do with the reality faced by

Figure 6.3 Under-five mortality by wealth group in selected countries (ratio of

average under-five mortality rate for bottom/top quintile).
Source: Based on Minujin and Delamonica (2002), using DHS data.
 Are the Millennium Development Goals feasible? 135
poor Zimbabwean children during the 1990s. Indeed, averages can be
deceiving.
On the income front, disparities are also on the rise. A growing body of
data suggests that income disparities are widening, both between and
within countries (see Cornia, 1999; Galbraith et al., 1998; Milanovic, 1999;
United Nations Conference on Trade and Development, 1997; United
Nations Development Programme, 1997). No matter how it is measured, it
is increasingly difficult to dismiss the evidence that inequality is on the
rise. Disparities are increasing not only between the rich and the poor, but
also among the poor. Nigeria, for instance, saw the poverty head-count
index decline by nine percentage points between 1985 and 1992; but
the incidence of extreme poverty increased by three percentage points
(Demery and Squire, 1996). This led to a paradoxical situation in which
the number of poor declined, yet the number of destitute people
increased. A similar story emerges for rural Kenya and rural Tanzania.
In sum, averages do not tell the full story. Groups for which social
progress has been fastest seldom represent the disadvantaged people.
Some countries appear to be on track to reach a particular target on the
basis of ‘average’ progress; yet the situation for disadvantaged groups in
those countries is stagnant or deteriorating. As disparities are widening for
a range of indicators, such as income, mortality and education, the infor-
mational value of national averages is gradually decreasing, thereby aug-
menting their potential to induce misleading conclusions.14

Are the Millennium Development Goals affordable?

Why are the promises not being kept? Why are hundreds of millions of
people struggling to overcome the daily grind of hunger, disease and
ignorance when the global economy is experiencing unprecedented pros-
perity? Many reasons account for this apparent paradox, and they are
often country specific. However, two reasons stand out in virtually all
countries: (1) under-investment in basic social services; and (2) public
action that frequently fails to take advantage of cross-sectoral synergies.15
Recognising the fact that global goals will require extra money as well
as new approaches, the United Nations proposed the 20/20 Initiative
(UNDP et al., 1998). It embodies the principle of shared responsibility for
the Millennium Development Goals by encouraging developing countries
to allocate about 20 per cent of their national budget to basic social ser-
vices; and developed countries to devote about 20 per cent of their devel-
opment assistance to the same services. Experience shows that once access
to an integrated package of basic social services of good quality becomes
universal, social progress can be dramatic and economic growth can be
sustainable and equitable.
But instead of a ‘20/20’ deal, the reality comes closer to a ‘12/12’ ratio
(Figure 6.4) (UNICEF and UNDP, 1998). Governments in developing
136 Jan Vandemoortele

Based on 40 country
studies, mid-1990s

Basic education Basic health and nutrition Water and sanitation

Figure 6.4 Underinvestment in basic social services (spending on basic social ser-
vices as a percentage of national budget and Official Development Assis-
tance). Based on forty country studies, mid-1990s.
Source: Based on UNICEF and UNDP (1998).

countries spend, on average, between 12 and 14 per cent of the national

budget on basic social services. Donor countries allocate, on average,
about 10–12 per cent of their aid budget on these services. Both shares
have shown an upward trend in recent years, but they fall far short of the
share of 20 per cent that is accepted as a minimum – based on the
experience of high-achieving countries such as Botswana, Costa Rica and
the Republic of Korea.
Reaching the Millennium Development Goals will require universal
access to basic social services of good quality. The financial cost of reach-
ing universal coverage is modest, whereas the benefits that beckon are
enormous. Global public spending on basic social services falls short by
about US$80 billion per year (at 1995 prices) of the level required to
ensure universal coverage (UNDP et al., 1998).16 The full implementation
of the 20/20 initiative would generate enough resources to close this
financial gap. Although large in absolute terms, US$80 billion represents
about one-third of one per cent of global annual income. Indeed, achiev-
ing the Millennium Development Goals is more about setting priorities
than about mobilising extra resources or making technological break-
throughs.17
Although the Millennium Development Goals appear affordable at the
global level, many governments will be hard-pressed to meet the financial
 Are the Millennium Development Goals feasible? 137
requirements for achieving the targets by 2015. Sub-Saharan African and
South Asian countries, in particular, will need to expand their budgetary
outlays on basic social services at a rate that will not be sustainable without
additional assistance, in terms of aid, debt relief and trade. It would be
unrealistic to expect that low-income countries can meet the Millennium
Development Goals without additional and concerted international
support.
Official development assistance and debt relief will be indispensable,
especially for the least developed and low-income countries. A steady
decline in ODA characterised the 1990s, when the relative aid effort fell
by one-third – dropping from 0.33 per cent of the combined gross
national income (GNI) of developed countries in 1990 to 0.22 per cent in
2000 (Figure 6.5) (OECD and DAC, 2001). It now stands at less than one-
third of the agreed target of 0.7 per cent. The total shortfall vis-à-vis that
target amounts to about US$125 billion per year. Aid efforts vary consider-
ably among donor countries,18 and none of the G-7 countries is a member
of the ‘G-0.7’ group, which comprises Denmark, the Netherlands, Norway
and Sweden, and more recently Luxembourg. An extra US$50 billion per
year in donor resources, as called for by the UK Chancellor of the Exche-
quer (Brown, 2001), would go a long way towards reaching the Millen-
nium Development Goals at the global level.
A study of budgetary spending in thirty developing countries found
that two-thirds of them spend more on debt servicing than on basic social
services (UNICEF and UNDP, 1998), with some spending three to five
times more on debt. In sub-Saharan Africa, governments spend about

Figure 6.5 Decline of Official Development Assistance (total Official Development

Assistance as percentage of combined gross national income of donor
countries).
Source: Based on OECD and DAC (2000).
138 Jan Vandemoortele
twice as much to comply with their financial commitment vis-à-vis external
creditors than to comply with their social obligation vis-à-vis the people.
Debt servicing often absorbs between one-third and one-half of the
national budget, making macroeconomic stability an elusive goal. To
spend more on external debt than on basic social services, when tens of
millions of people see their fundamental human rights denied, is ethically
wrong and makes no economic sense. The Heavily Indebted Poor Coun-
tries (HIPC) Initiative is a first attempt to resolve the debt problem com-
prehensively but its implementation is painfully slow, while declining
commodity prices are making it increasingly ineffective.19 For many coun-
tries, slow debt relief will mean slow Millennium Development Goal
progress.

Conclusion
The Millennium Development Goals are technically feasible and finan-
cially affordable. Yet the world is off track to meet them by 2015. The Mil-
lennium Development Goals are ambitious, but each and every target will
be met by some countries, including a few low-income countries. If these
countries can achieve the Millennium Development Goals, there is no
reason why others cannot.
Figure 6.6 summarises global Millennium Development Goal progress
so far. Of the eight targets listed, only one is on track. Monitoring income

Poverty Data are not reliable or not compatable across countries

HIV/AIDS

Child mortality

Basic education

Gender equality

Child malnutrition

Maternal mortality

Safe water

1990 2000 2015

Achieved To be achieved

Figure 6.6 Broken promises of the 1990s (share of the target covered in the 1990s,
unfinished agenda for 2000–2015).
 Are the Millennium Development Goals feasible? 139
poverty at the global level is subject to serious conceptual and measure-
ment constraints. Global estimates based on the norm of US$1 per day
tend to underestimate global poverty and overstate poverty reduction
(Reddy and Pogge, 2002). Current data on global poverty are simply not
robust enough to make an informed judgement as to whether the world is
on track towards the 2015 target.20
Little or no progress was achieved in reversing the HIV/AIDS pan-
demic; HIV prevalence rates continue to rise in numerous countries. Only
a few succeeded in reducing the spread of HIV, including Cambodia and
Uganda. The HIV/AIDS pandemic is a major obstacle on the road towards
the Millennium Development Goals.
Progress was slow for child mortality, basic education, malnutrition,
maternal mortality and gender discrimination in primary school enrol-
ment. They all recorded about one-quarter or less of the agreed target,
leaving three-quarters or more to be covered in the next fifteen years.
Since the Millennium Development Goals are to be achieved between
1990 and 2015, 40 per cent of the road should have been covered by 2000.
Instead, just over half that level was recorded.
Only a fifth of the education target was achieved in the first ten years,
leaving 80 per cent to be covered in 60 per cent of the time period
(between 2000 and 2015). No matter what the challenge is – HIV/AIDS,
child mortality, malnutrition, income poverty, maternal health, gender
discrimination or environmental degradation – basic education is invari-
ably at the centre of the solution. Failure to keep the promise to give every
child a good basic education will undermine the chances of reaching the
other Millennium Development Goals.
There is no good reason why universal primary education should not
yet be a practical reality. Progress toward that goal is truly discouraging: its
global cost is perfectly affordable (Delamonica et al., 2001); no new
technology breakthroughs are needed to get all children in school; there
is consensus that it makes good economic sense;21 and basic education is a
fundamental human right that must not be denied to any child.22 If these
conditions are not enough to ensure success, then the question arises as to
what it will take to meet the other Millennium Development Goals.
Only one Millennium Development Goal is on track, the one of halving
the proportion of people without access to safe water by 2015. However,
the current rate of progress may not be sustainable: countless countries
face acute water shortages in the near future if no swift and decisive action
is taken soon. Wastage, population growth, urbanisation and desertifica-
tion are gradually leading to looming water scarcity in many parts of the
world; industrialisation and modern agriculture are adding to the risk of
more water pollution. Future conflicts over the allocation of fresh-water
resources are likely to slow down progress in the years to come.
Not only was global progress inadequate in the 1990s, but much of it
bypassed the poor. Slow ‘average’ progress was compounded by limited
140 Jan Vandemoortele
progress for the poorest and most disadvantaged groups within countries.
Global goals are primarily meant to help improve the situation of the poor
and the disadvantaged, not only that of better-off and privileged people.
Unfortunately, the poor have benefited proportionately little from ‘average’
progress, as evidenced by widening disparities in terms of income, educa-
tion and mortality.
In sum, the world is not on track to meeting the Millennium Develop-
ment Goals by 2015. In 2001, Nelson Mandela asked, ‘Will the legacy of
our generation be more than a series of broken promises?’ In opening the
Children’s Summit in May 2002, Kofi Annan, UN Secretary-General,
stated, ‘We the grown-ups must reverse this list of failures.’ In 1993 the
late James Grant, then UNICEF Executive Director, said, ‘The problem is
not that we have tried to eradicate global poverty and failed; the problem
is that no serious and concerted attempt has ever been made.’ Sadly, these
words still ring true today.
But while the Millennium Development Goals remain unfulfilled, they
also remain feasible and affordable. Committed leadership, stronger part-
nerships, extra money and deeper participation by the poor can bring the
world back on track towards the Millennium Development Goals. It is not
too late to realise the dream by 2015.

Notes
1 The views expressed in this chapter do not necessarily reflect those of the
United Nations Development Programme. The usual disclaimers apply.
2 Cost recovery in a water project in western Kenya, for example, was low despite
seemingly high average household income. The cause was traced to the fact
that women were responsible for this expense but had little or no control over
household income. Affordability studies often target the wrong group and fre-
quently produce misleading policy advice.
3 Based on the latest information from a variety of sources, mostly from within
the United Nations system, and particularly the United Nations Children’s
Fund (UNICEF) (2001).
4 That is, US$1.08 per day purchasing power parity 1993 prices.
5 Global poverty estimates are often presented with a decimal point, which may
give a false sense of sophistication and accuracy. Given their approximate
nature, rounded figures are more appropriate.
6 The data of the Ministry of Agriculture relate to rural poverty, whereas World
Bank data refer to total poverty. Since poverty in China is overwhelmingly
rural, the discrepancy cannot be explained by the difference in geographical
coverage. The national poverty line is lower than the international one based
on $1 per day. Therefore, the national poverty estimate falls below the $1
poverty trend. It should be noted, however, that the difference between the
two poverty trends was halved, narrowing from 20 percentage points in the
early 1990s to about 10 percentage points in the late 1990s.
7 Another way of assessing progress is by measuring it against the remaining gap;
but this interpretation of the data also shows that progress slowed down in the
1990s. It should be noted that global targets set for reducing under-five mortal-
 Are the Millennium Development Goals feasible? 141
ity and net primary enrolment rate by 2015 are not out of line with historic
trends.
8 Faster progress was made towards gender equality in secondary and tertiary
education, but here too it will not be enough to close the gender gap by the
agreed date.
9 Lower levels of vaccination among baby girls explain, in part, their increased
child mortality rate vis-à-vis that of boys. Countries where a significant shift
occurs from a pro-female bias in infant mortality to a pro-male bias in child
mortality are often those with the largest gender gap in terms of immunisation.
Another factor that is likely to add to explaining the reversal in gender-specific
mortality rates is nutrition. Food allocation within the family is frequently
determined by gender, and disparities in feeding practices between girls and
boys appear early in life. Data on breastfeeding reveal that in most countries a
baby boy is more likely to be breastfed than a baby girl.
10 Assuming a proportionate change in the maternal mortality ratio and the per-
centage of births attended by skilled health personnel.
11 Therefore, generalisations about the feminisation of poverty should be used with
caution. Statements such as ‘70 per cent of the poor are women’ are not always
backed by hard evidence. Aggregates and averages can be used – and abused – to
back either side of an argument. The 1997 Human Development Report points out
that not all evidence supports different poverty levels between male- and female-
headed households (United Nations Development Programme, 1997). A recent
World Bank report on rural poverty in China states that ‘available evidence does
not suggest that women are greatly over-represented among the poor’ (World
Bank, 2001). Gender discrimination does not occur indiscriminately, but is often
mediated through a multitude of factors. Gender, for instance, is more a liability
to a poor girl than to her non-poor counterparts, to a girl from an ethnic minor-
ity than to one from the majority, to a rural girl than to an urban one.
12 DHSs do not collect income or consumption data, but the information on
household assets such as a bicycle, radio, size of the dwelling, type of construc-
tion materials, and source of drinking water makes it possible to cross-tabulate
social indicators by socio-economic groups, as has been done by Filmer and
Pritchett (1999) and Gwatkin et al. (1999).
13 Only two countries – Guatemala and Togo – reported a significant improve-
ment over time in child mortality for the poorest quintile vis-à-vis the richest
quintile.
14 An example of ‘misplaced concreteness’ is the well-publicised finding that
income of the poor rises one-for-one with overall income per capita (Dollar
and Kraay, 2001). Although the finding may be statistically correct, it is not
necessarily true. When the same statistical analysis is applied to random values,
the same results are obtained. The argument that a one-for-one relationship
exists between the income of the poor and average income per capita is more
the result of using aggregates and averages, than of actual behavioural relation-
ships. It is striking how a simple analysis based on a set of aggregate averages
can exert so much influence on so many people: policy-makers, researchers
and journalists alike.
15 Too often, health goals are pursued through health interventions alone; educa-
tion goals are pursued through education programmes alone. Basic social ser-
vices comprise an integrated package of basic education, primary health care,
nutrition, reproductive health, water and sanitation.
16 Estimating the cost for reaching the Millennium Development Goals is a
complex task which is beset by serious methodological problems. Unit costs in
absolute terms vary greatly from country to country because salaries dominate
the cost structure, which relate to the level of economic development. In
142 Jan Vandemoortele
addition, expressing them in a common currency, such as the US dollar, intro-
duces more inaccuracies. Unit costs vary much less when expressed in relative
terms (as a percentage of income per capita). Using average regional unit costs
unduly inflates the cost of reaching the target; relative country-by-country unit
costs are preferable. However, average costs may lead to considerable under-
estimation because reaching the last 10 or 20 per cent of the population with
basic social services is likely to be more costly (i.e. marginal costs will exceed
average costs). On the other hand, efficiency gains can reduce overall costs,
but improving quality may offset these gains. HIV/AIDS has significant cost
implications but its quantification is virtually impossible, given current know-
ledge and available data (see also Chapter 12). Furthermore, calculating the
costs of the each Millennium Development Goal separately will unduly inflate
the total cost because synergies and complementarities will be missed. Last, but
not least, dividing total cost between domestic sources and Official Develop-
ment Assistance (ODA) flows is subject to judgement, and can always be chal-
lenged. Thus, costing the Millennium Development Goals is not as straight-
forward as it seems at first glance. Approximating it by using the 20/20 rule is
unlikely to be less ‘scientific’ than other approaches.
17 Reaching the Millennium Development Goals will require not only extra
spending, but also better spending. Money alone will not solve the problem;
human and institutional capacities need improvement too. However, the argu-
ment that existing resources must be used more efficiently first before invest-
ing extra money in basic social services misses the point that insufficiencies
often create inefficiencies. For example, when teacher salaries absorb 98 per
cent of the budget for primary education but fail to provide a living wage,
there is little scope for improving the quality of education. Indeed, inefficien-
cies and insufficiencies are not independent but very much interdependent.
18 In 2000 they ranged from a high of 1.06 per cent of GNI for Denmark to a low
of 0.10 per cent for the USA. The aid effort of G-7 members is considerably
lower than that of non-G-7 countries. In 2000 the respective averages were 0.19
per cent and 0.45 per cent of their combined income.
19 The criteria for eligibility under the HIPC initiative are mostly trade related,
despite the obvious fact that it is government, not exporters, that repays exter-
nal debt. The ‘enhanced’ HIPC initiative does not take into account the fiscal
burden of debt servicing in determining a country’s external debt sustainabil-
ity.
20 Progress towards the poverty target is best measured on the basis of national
poverty lines; although this will not readily produce internationally comparable
poverty data. However, the quest for comparable data in this area has been
rather elusive so far.
21 Numerous empirical studies have documented the critical role of basic educa-
tion in the economic and social development in Europe, North America, Japan
and, more recently, East Asia (e.g. World Bank, 1993).
22 Perhaps it is the most ‘human’ of all human rights since it is reading and
writing that sets Homo sapiens most apart from all other species.

References
Bicego, G. and Ahmad, O. (1996) Infant and Child Mortality, Demographic and
Health Surveys Comparative Studies 20, Calverton, MD: Macro International.
Brown, G. (2001) ‘Globalisation’, address to Press Club in Washington, DC on 17
December, at [Link]/Newsroom_and_Speeches/Press/2001/
press_146_01.cfm.
 Are the Millennium Development Goals feasible? 143
Cornia, A. (1999) ‘Liberalization, Globalization and Income Distribution’, World
Institute for Development Economics and Research (WIDER/UNU), Working
Paper 157. Helsinki: WIDER.
Daly, H. and Cobb, J. Jr (1994) For the Common Good: Redirecting the Economy Towards
Community, the Environment and a Sustainable Future, Boston: Beacon Press.
Delamonica, E., Malhotra, S. and Vandemoortele, J. (2001) ‘Is EFA Affordable?
Estimating the Global Minimum Cost of “Education for All’ ’, Division of Evalu-
ation, Policy and Planning, Staff Working Paper 01–01, New York: UNICEF.
Demery, L. and Squire, L. (1996) ‘Macroeconomic Adjustment in Africa: An
Emerging Picture’, World Bank Research Observer, 11: 39–59.
Dollar, D. and Kraay, A. (2000) Growth Is Good for the Poor, Washington, DC: World
Bank, Development Research Group.
Filmer, D. and Pritchett, L. (1999) ‘The Effect of Household Wealth on Educa-
tional Attainment: Evidence from 35 Countries’, Population And Development
Review, 25(1): 85–120.
Galbraith, J., Darity, W. and Jiaqing, L. (1998) ‘Measuring the Evolution of
Inequality in the Global Economy’, mimeo, Austin: University of Texas at
Austin.
Gustafsson, B. and Zhong, W. (2000) ‘How and Why Has Poverty in China
Changed? A Study Based on Microdata for 1988 and 1995’, China Quarterly,
164(1): 983–1006.
Gwatkin, D., Rustein, S., Johnson, K., Pande, R. and Wagstaff, A. (1999) Socio-
economic Differences in Health, Nutrition and Population, Country Notes, Health,
Nutrition and Population Department, Washington, DC: World Bank.
Inter-American Development Bank (1998) Facing Up to Inequality in Latin America,
Washington, DC: Inter-American Development Bank.
Khan, A. and Riskin, D. (2000) Inequality and Poverty in China in the Age of Globalisa-
tion, New York: Oxford University Press.
Milanovic, B. (1999) True World Income Distribution, 1988 and 1993: First Calculation
Based on Household Surveys Alone, Washington, DC: World Bank, Development
Research Group.
Minujin, A. and Delamonica, E. (2002) ‘Socio-economic Inequalities in Mortality
and Health in the Developing World’, mimeo, New York: UNICEF.
Organisation for Economic Cooperation and Development (OECD) and Develop-
ment Assistance Committee (DAC) (2001) Development Co-operation Report 2000,
Paris: OECD.
Reddy, S. and Pogge, T. (2002) How Not to Count the Poor, New York: Columbia Uni-
versity, at [Link].
UNAIDS (2000) Aids Epidemic Update: December 2000, Geneva: UNAIDS.
United Nations (2000) Millennium Declaration, New York: United Nations.
United Nations Children’s Fund (2001) Progress since the World Summit for Children:
A Statistical Review, New York: UNICEF.
United Nations Children’s Fund and United Nations Development Programme
with contributions from the World Bank and the United Nations Fund for Popu-
lation Activities (1998) ‘Country Experiences in Assessing the Adequacy, Equity
and Efficiency of Public Spending on Basic Social Services’, document prepared
for the Hanoi meeting on the 20/20 Initiative, New York: UNICEF.
United Nations Conference on Trade and Development (1997) Trade and Develop-
ment Report, 1997: Globalization, Distribution and Growth, Geneva: UNCTAD.
144 Jan Vandemoortele
United Nations Development Programme (1997) Human Development Report 1997:
Eradicating Human Poverty, New York: Oxford University Press.
—— (1999) Human Development Report 1999: Globalisation with a Human Face, New
York: Oxford University Press.
United Nations Development Programme, UNESCO, UNFPA, UNICEF, WHO and
the World Bank (1998) Implementing the 20/20 Initiative: Achieving Universal Access
to Basic Social Services, New York: UNICEF.
Vandemoortele, J. and Delamonica, E. (2000) ‘The “education vaccine” against
HIV’. Current Issues in Comparative Education (online), 3(1), at [Link].
edu/cice/vol03nr1.
World Bank (1993) The East Asian Miracle: Economic Growth and Public Policy, New
York: Oxford University Press.
—— (2001) China: Overcoming Rural Poverty, a Country Study, Washington, DC:
World Bank.
—— (2002) World Development Indicators, Washington, DC: World Bank.
7 Halving world poverty1
Michael Lipton with Hugh Waddington

Introduction
There has been considerable disagreement about whether, or to what
extent, the world is on track to meet the United Nations (UN) Social
Summit target of halving ‘incidence of dollar-a-day (1993 purchasing-
power parity or PPP) poverty in developing and transitional countries’
(hereafter called ‘world’ or ‘global’ poverty) in 1990–2015. The Inter-
national Fund for Agricultural Development’s (IFAD’s) Rural Poverty
Report (2001) argued that big changes were needed to do so. However, far
from arguing that this target was ‘doomed to failure’, the Rural Poverty
Report showed that sufficient poverty reduction to meet the target had
been achieved in many places and at many times – when, and where, there
was rapid, poverty-oriented, employment-intensive rural development,
usually based on rising yields of staple foods. Unfortunately, progress in
yields of staple foods and land reform slowed right down between the
1970s and the 1990s, as did poverty reduction. Between the late 1980s and
the late 1990s, the real value of aid to agriculture fell by almost two-thirds.
Many people are fairly cynical about UN targets. There have been so
many declarations with no enforcement mechanisms, no visible means of
support and little or no result. However, some people are rather enthusi-
astic about the world poverty targets and the process put into place to
meet them. Certainly there are plenty of issues that the targets and the
process do not address. Some of the serious concerns are explained in this
chapter. However, it is something to have a set of internationally agreed
targets, not only with a policy commitment by developing countries at the
UN Social Summit in Copenhagen in 1996, but also endorsed as the
guiding principle of aid by the Organisation for Economic Co-operation
and Development (OECD) donors in 1996 and by the G-7 finance minis-
ters in 2001. Country-specific Poverty Reduction Strategy Papers (PRSPs)
(while unfortunately linked to flawed Heavily Indebted Poor Country
(HIPC) procedures) nevertheless are intended to embody government
programmes, discussed through civil society procedures, to achieve and
monitor progress towards the target of halving 1990 levels of dollar-a-day
146 Michael Lipton with Hugh Waddington
poverty by 2015 – and, increasingly, to act as a gateway to, and a structur-
ing device for, concessional financial flows from the World Bank. In other
words, the country-specific poverty-halving target has standing, a body of
knowledge and support, some implementing and monitoring procedures,
and a considerable amount of supportive international aid. It is increas-
ingly possible for critics to hold both developing-country governments and
aid donors to poverty targets, and to implementing procedures, to which
they are publicly committed.
This chapter analyses progress towards meeting the world poverty
target. The definition of the target is discussed in the next section. It is
argued that, whilst there are several objections to reviewing progress in
reducing the proportion of persons below purchasing power parity
poverty lines, and quite substantial ways of improving this definition, it is
manageable. The section after that discusses the process through which
the targets are to become realities. Several important principles have been
recognised in this respect: not only is the process ‘country-led’, but it
recognises specifically that it is both essential for effectiveness and morally
right that decisions about how to reach a particular target, such as poverty
reduction, have to involve participation by the people affected by it.
There then follows an assessment of the feasibility of the target. Much
past evidence suggests that it can be met: very large parts of the world in
recent history have reduced poverty for quite long periods and at rates
sufficiently fast to reach the target of halving world poverty by 2015.
However, despite some assertions to the contrary, we are very far
below the rate of progress needed to achieve the target. If the regions of
the world continue to reduce poverty in 1998–2015 no faster than in
1990–1998 (the longest period after 1990 with a wide spread of evidence
globally), then we shall achieve only 40 per cent of the target. In other
words, we shall underachieve by 60 per cent.
Getting poverty reduction back on track – the subject of the final
section – needs changes in policy both in developing countries and in
developed countries. The most important change is to reverse the collapse
of support for labour-intensive farming of staples in smallholdings, and to
help revive the performance of smallholders in the developing world, in
particular in the area of food production for local consumption.

Definition
Although this chapter is entitled ‘Halving world poverty’, we discuss some-
thing cruder: halving, between 1990 and 2015, the proportion of the
population in developing and transitional economies in ‘dollar poverty’,
i.e. with daily consumption per person below US$1 in 1993 purchasing
power parity (PPP).2
In 1990, 29 per cent of the population of the ‘world’ consumed less than
US$1 1993 PPP per day – purchasing power, that is, over a global average
 Halving world poverty 147
1993 bundle of goods.3 Thus the target is to bring that percentage down to
14.5 in 2015. This is narrow private consumption. It is not participatory, as
advocated by Robert Chambers (1997), who argued persuasively for letting
people define their own needs and translating those into policy (even at
the unavoidable cost that it is hard to track or compare self-assessed
poverty: one cannot discriminate between the effect of changes in the
values received by the poor, and changes in their standards or expecta-
tions). Nor is it multidimensional. The PPP bundle of goods is the average
bundle consumed by an average citizen of the world; the PPP exchange
rate measures how much people have got to spend to purchase this basket
of goods in some real purchasing power equivalent sense – a dollar is
simply used as the unit of account, what economists call a ‘numéraire’.4
Why review progress in reducing the proportion of persons below PPP
poverty lines? There are four main objections, elaborated on here.5
First, broader poverty measures, taking account of other determinants
of welfare or capabilities than private consumption, may be desirable.
However, there is a case for having a simple, readily measurable, narrow
conventional definition of poverty, because targets can be set and moni-
tored; governments, civil societies and donors can check progress for a
country, or a region (or group or gender) within it, and can compare
countries with one another and with past performance. Additionally, 95
per cent of the population of the developing and transitional countries of
the world is now covered by representative surveys, and in most cases
several surveys, allowing measurements of dollar PPP poverty to be tracked
across time.
Second, however, even if a fixed absolute measure of private consump-
tion poverty is used, a nutrition-linked measure is preferable to a dollar
PPP line. ‘A dollar per person per day of command over a PPP bundle for
a given year’, while much better than nothing as a poverty line permitting
international comparisons of poverty and of progress towards targets, is
highly problematic technically,6 not intuitive, and above all, unlike nutri-
tion-based measures, not linked to human capabilities. Among nutrition-
based measures, a ‘cost of basic needs’ (CBN) indicator is often used; this
estimates the cost of a basic or ‘typical poor person’s’ diet, adds either a
percentage allowance or a similarly costed bundle of commodities to meet
other basic needs, and defines the outcome as the poverty line (given
household size and composition). The CBN approach, however, ignores
differences among persons, even of a given age and activity level, in their
needs for calories, other nutrients, shelter and other basic needs; CBN
simply imposes standard requirements from without. Better in this and
other respects is a poverty line for private consumption per equivalent
adult (allowing for age-specific needs and more modest economies of
scale within the household) set by the food energy method (FEM). FEM
sets the poverty line at the level of consumption at which the expectation is
that persons just fulfil minimum calorie requirements.
148 Michael Lipton with Hugh Waddington
A further benefit of hooking a narrow consumption measure on to
food poverty by FEM relates to the different environments in which
people live. For example, food needs are larger in cold places than in hot
ones. Also, the level of total consumption necessary to ensure that food
needs are met is higher if other needs such as housing are very expensive.
A purchasing power parity dollar-a-day measure will thus underestimate
poverty in cold climates relative to hot ones, as food requirements, as well
as shelter and heating costs, are greater in the former. Great differences
amongst places in the cost of meeting food and other needs are taken into
account indirectly if a food-based poverty line is used.7
Third, even if we use poverty comparisons based on PPP measures,
there is compelling reason to doubt some important country-specific esti-
mates and trends8 based on the currently available data sets. For example,
the risk of poverty in Bangladesh in the late 1990s is estimated as well
below two-thirds of the risk in India, which is out of touch with both nutri-
tional data and most observers’ judgements.
Fourth, it would be very much better to use a measure that takes
account of how far below the poverty line people are. Two things have
been suggested which are preferable to the commonly used measure of
incidence: (1) a measure of poverty intensity (e.g. the poverty gap index;
that is, incidence multiplied by ‘depth’ – the shortfall of the average poor
person behind the poverty line – as a proportion of that line); (2) a
measure of the severity of poverty, which accords greater weight to those
further below the poverty line. Poverty intensity is fairly intuitive because
there is a very close relationship between the intensity of poverty and what
it would cost to rid society of poverty if society could target resources accu-
rately on its poor. There are various standard ways of measuring severity
(e.g. the squared poverty gap index; see Foster et al., 1984), and although
in principle it would be a good idea to find a measure that is intuitive, it is
difficult to do so. Certainly, a widening of poverty measures from inci-
dence to intensity is entirely desirable.
All four objections have force. They should be used to develop improved
measures of poverty and its reduction. However, except for the poverty
gap index of intensity,9 such measures are not widely available. Further-
more, the UN targets for progress in 1990–2015 were set and accepted
using the ‘dollar poverty’ measure. It would be confusing and perhaps
evasive to change the measure in the middle of the implementation and
monitoring period. Some widening of the definition is strongly desirable
from a 2015 base, however. Work to agree, measure and monitor a
widened definition should commence now.
On the other hand, the popularly advocated widening to multidimen-
sional poverty is misguided. It is infeasible to get everyone to agree on the
exact indicators that should be used to measure poverty multidimension-
ally, or on how to value (weight) them against each other. Further, factors
such as low access to health or education, low self-esteem and low esteem
 Halving world poverty 149
from others, and vulnerability to very high downward fluctuations and
risks affecting standards of living, are all extremely important and are
separate elements of policy, but to lump them together with consumption
poverty as a single policy objective will not assist rational policy-making.
These different dimensions may be causes, one of the other, or mutually
reinforcing, and they may interact in complex ways, but they are not the
same. Nor is it operational to take a multidimensional indicator, which is
some arbitrary mix of, for example, literacy, longevity and lack of con-
sumption poverty, and to try to monitor that; there are all sorts of differ-
ent reasons why that indicator might be moving in any direction. It is not
possible to make sensible policy, or apply sensible pressure from below, on
the basis of such an indicator.
In sum, a narrow definition has advantages, for both intellectual and
operational reasons. The narrow definition used now is not the best that
could be used. We have suggested ways in which it could be improved, but
monitoring of the Millennium Development Goal for poverty (halving
country-by-country 1993 dollar poverty incidence in 1990–2015) should
continue on the present consistent track. A shift to an FEM intensity
target (perhaps with supplementary poverty-linked measures) should be
put in place in 2015. The work needed to make the improvements needs
to start soon. However, we should not shift to a multidimensional index of
poverty, which is inevitably arbitrary in its exclusions, inclusions, weights
and hence interpretations. Instead we should monitor, and analyse sepa-
rately, progress against the main policy-related components of human
misery and impaired functioning – notably, but not only, infant mortality,
malnutrition and illness; illiteracy and educational deprivation; FEM con-
sumption poverty; vulnerability; and lack of esteem and self-esteem.10

Process
The internationally agreed process for meeting the poverty target is in
principle good. Each developing and transitional economy is asked to
work out its own Poverty Reduction Strategy and to incorporate this into a
Poverty Reduction Strategy Paper (PRSP), which sets a series of goals
aimed at the halving of poverty over the period 1990–2015. Those targets
are arrived at not by government in isolation, but by interplay between
each government and its own civil society. Subsequently, each PRSP is
assessed by the international donors who are supposed to support it in two
ways: (1) by shifting aid to countries that have a serious strategy and away
from countries that do not; and (2) by attempting to support sector-level
or national-level activities within the ‘serious’ countries rather than tying
aid to projects and specific activities.
In practice, although there is progress, much has gone wrong. For
instance, in practice the PRSP process has been tied up with the pro-
cedure of giving debt relief through the Heavily Indebted Poor Country
150 Michael Lipton with Hugh Waddington
(HIPC) Initiative; in fact, the great majority of PRSPs, though not all, are
for HIPCs. The aim of forgiving debt for countries which will use that
money for extra, cost-effective efforts to cut poverty is sensible, but the
process as implemented is doing little to reduce world poverty: despite the
panoply of conditions for HIPC-related debt relief, not all the countries
receiving, or scheduled for, HIPC relief are giving priority to poverty or
can be persuaded to do so.
Furthermore, poor countries that have not become highly indebted do
not, by definition, qualify for HIPC assistance. India and China together still
contain over 40 per cent of the world’s PPP dollar-poor, despite large reduc-
tions due in part to the world’s most substantial and serious (albeit, of
course, flawed) anti-poverty programmes. They have met their international
debt, and are duly punished by exclusion of their cash-strapped anti-poverty
programmes from the formal PRSP procedures and aid benefits. Indeed,
they together received only 9.2 per cent of 1999 net aid disbursements
(UNDP, 2001: 192–194). In theory, all HIPC flows are additional to other
official aid and thus cost India’s and China’s poor nothing, but in practice
this is not the case. Additionally, substantial aid donors – possibly the Inter-
national Development Agency (IDA) (the division of the World Bank giving
concessional aid to so-called low-income countries) and almost certainly the
International Fund for Agricultural Development (IFAD) – are finding that
their low-interest loans are being effectively, or in part, written off, and the
rich countries have not responded by paying those aid donors back the full
amount that they lose as a result of HIPC relief.
Even more worryingly, the PRSPs contain very few sector policies. The
Papers follow the prevailing development ‘dialogue’ (perhaps more accu-
rately termed ‘narrative’) that whether a country develops depends on its
macroeconomic policy, particularly on determined liberalisation and
globalisation. Whether or not that is part of the story – and, on the whole,
evidence suggests that countries which liberalise do enjoy faster sub-
sequent growth and poverty reduction – it is not the whole story; yet the
design of the HIPC process and its PRSPs is so heavily concentrated on
macroeconomic policies that it suggests the contrary. There are some
PRSPs in which the words ‘agriculture’ and ‘rural’ (and indeed ‘industry’
and ‘urban’) cannot be found from cover to cover. There is little hard
content in many PRSPs apart from trade, macroeconomic policy, and
perhaps social sectors and safety nets. Especially if, as is usually the case,
these emphases require retrenchment (‘cuts’) in public infrastructural
spending in the short term, they may well conflict with any commitment
to increase public resources, or improve private incentives, for activities
and areas likely to benefit poor people. Anyway, the PRSPs seldom have
much detail on the production implications of this – that is, on efficient
means of shifting scarce public resources, or private incentives, to under-
pin sectors employing or feeding the poor. In short, most PRSPs need to
be far more production sector specific if they are to have much poverty-
 Halving world poverty 151
reducing impact. (This in no way implies large-scale state involvement in
production or regulation.)
Given these criticisms, however, the process contains three big steps
towards making the world poverty target more feasible:

• The first is the consensus that the success or failure of the develop-
ment process is going to be assessed through poverty reduction. That
does not imply that nothing else matters – not that economic growth
does not matter, because you cannot reduce poverty if you do not
grow, not for very long anyway; nor that environmental sustainability
does not matter. What is implied is that poverty reduction is the main
thing to be monitored when we assess the cost-effectiveness of planned
economic progress, or claims that it has been achieved.
• The second is that there is a country-led presence – it is the develop-
ing country itself that prepares the PRSP; its needs assessment does
not come from outside. In practice, sometimes in the first round the
country will bring in a consultant to write the Paper, but this is rarely
the case. Even if it is, and a PRSP genuinely affects policy or mobilises
resources, domestic politics make it unlikely to happen thereafter!
• The third is that civil society is recognised. In practice, ‘civil society’ is
usually reduced to ‘non-governmental organisations’, which, excellent
though they often are, are seldom directly accountable downwards,
especially when they are not domestic but foreign or international.
‘Citizen-based organisations’ are accountable downwards, but in many
countries they are thin, weak, or hamstrung by authority.

Feasibility
To determine whether the halving of world poverty by 2015 is feasible, we
must look back to where we thought we were, and what has been achieved
in the past. In 1990, the base year for the target, there were about 1.3
billion dollar-poor in the world, mainly residing in China (370 million),
South Asia (495 million, mostly in India) and sub-Saharan Africa (240
million). Of the remaining, 70 million were in Latin America and rather
small numbers were in transitional economies and in the Middle East (see
Table 7.1, p. 152).
There is a huge concentration of poverty in rural areas – worldwide, 70
per cent of the dollar-poor are rural, and the percentage is much higher
in sub-Saharan Africa (Ravallion, 2000). Of the dollar-poor’s consumption
(in cash and kind), some 65–75 per cent by value comprises food. Indeed,
about half comprises staple foods (mostly rice, wheat and maize; also,
especially in Africa, sorghum, millets, cassava and yams). Over 60 per cent
of the dollar-poor in the world are still either smallholders or farmworkers
in their main occupation, though most of them obtain additional income
from other non-farm sources.
152 Michael Lipton with Hugh Waddington
Table 7.1 Incidence of poverty in developing regions, 1990–2015 (assuming regional
rate of change in 1990–1998 and 1998–2015 is identical)

Incidence of poverty Change in Population

incidence of
poverty, 1990–1998

1990 1998 2015* R(1990–1998)** 2015

China 31.6 17.2 4.7 0.9268 1417.7

Other East Asia and 18.5 11.3 4.0 0.9402 717.5
Pacific
India 45.7 44.2 41.2 0.9958 1211.7
Other South Asia ** ** 15.5 ** 473.8
Europe and Central Asia 1.6 5.1 59.9 1.1559 478.3
Latin America and 16.8 15.6 13.3 0.9908 624.9
Caribbean
Middle East and North 5.7 5.5 5.1 0.9955 402.5
Africa
Sub-Saharan Africa 47.7 46.3 43.5 0.9963 873.8
Total developing and 29 24 23.2 6200.3
transitional countries

Sources: World Bank (2000: 23, 334–335); UNDP (2000: 223–226, 285–286); South Asian
data, pers. comm., Shaohua Chen and Martin Ravallion.
Notes
* Projection of poverty incidence to 2015 if each country or region changes its annual inci-
dence of dollar poverty at the same rate in 1998–2015 as in 1990–1998. As explained in
the text, if instead the global level changes in dollar poverty 1990–1998 were used for the
projections, it would understate global poverty in 2015 at 16.1% (rather than the 23.2%
reported here).
** r(1990–1998): ratio, to its level in previous year in 1990–1998 assuming steady change, of
proportion of people consuming below $1.08 in 1993 constant purchasing power per day.
This proportion is ‘incidence of dollar-a-day private consumption poverty’ ($1.08 in 1993
purchasing power is approximately equal to $1.00 in 1985 constant purchasing power).
This was calculated separately for Bangladesh (incidence in 1990, 33.7%; in 1998, 21.7%;
in 2015, 13.7%; 2015 population, 161.5 million), Nepal (42.2%, 31.0%, 16.0%; 32.7
million), Pakistan (47.8%, 32.5%, 14.4%; 222.6 million), Sri Lanka (3.8%, 5.4%, 11.5%;
21.9 million) and Afghanistan plus Bhutan (assuming ‘South Asian poverty incidence’ for
1990 and 1998; 32.7% of 35.1 million in 2015).

Therefore, big reductions in poverty are likely to be achieved by

addressing rural poverty, in large part by reducing local food deprivation.
Experience in many parts of the world has shown that rapid success in
poverty reduction proceeds through employment-intensive rural develop-
ment, usually based on rising yields of staple foods, and commonly associ-
ated either with initially not-too-unequal access to farmland or with land
reform (IFAD, 2001). The urban poor benefit from this type of progress
as well as the rural sector, since increases in food production lead to
reductions in the price of food for urban food consumers, the share of
whose income spent on food is also large. The pressure of migrants on
urban labour markets is also less because it pays people to stay in rural
 Halving world poverty 153
areas,11 and therefore the rate of increase in urban real wages rises in the
formal as well as the informal sectors.
Moreover, rapid falls in poverty have been achieved, principally on the
basis of rural agricultural growth, in large parts of Asia – notably the great
river valleys of China, the deltas of India and the irrigated areas of Punjab
and some other states in India – and in large parts of Latin America. In
these areas between 1975 and 1990 there was very fast growth in the yields
of staple foods, particularly rice, wheat and maize, and associated with that
was very fast growth in the income and consumption of smallholders and
farmworkers; the rural poor in affected areas enjoyed much higher
incomes, for both small farmers and landless labourers, and the urban
and the rural poor had cheaper food. Dollar poverty fell at the target rate
in South Asia in 1975–1990; it fell faster in East and South-East Asia. In
China in particular, extensive evidence suggests that over the period
1977–1985, large areas of the country experienced the fastest period of
poverty reduction ever seen over such a large population of the world.
Food production grew every year for six years by more than 6 per cent. At
the same time there was massive land redistribution from formerly state-
or commune-held land to family farmers under the Household Responsi-
bility System; reduced price extraction by food quotas; and, as a result, a
big shift in incentives towards food production, first for own and local
consumption, later for the towns. This was not simple liberalisation: irriga-
tion and irrigation management was provided by an effective state system,
as were research-improved varieties, principally of rice and wheat. In con-
trast, poverty incidence has stagnated in sub-Saharan Africa, where there
is little research in staple foods and only around 5 per cent of total crop
land is irrigated (most of that being for wealthier farms in a few selected
locations).
However, since 1990 there has been a sharp slowing-down in the rates
of growth of production and yields of staple foods, of research inputs, and
of irrigation facilities in the developing world. For instance, since the
1970s the rate of growth in output per hectare of staple foods has been 3
per cent per year; through the 1990s it was only 1.3 per cent (Lipton,
1999). To revive the unprecedented poverty reduction of 1975–1990, and
to spread it to neglected areas, its basis in the growth of yields of labour-
intensive staples needs revival also.
One very important development in agricultural technology in 2001
was the announcement by Dr Gurder Singh Khush, the great plant
breeder at the International Rice Research Institute, of the New Plant
Type, which has the potential to make a substantial difference to rice
yields. However, getting progress comparable to that during the Green
Revolution will require a substantial amount of work on water control and
water management – both on pricing and technology – as well as a revival
of the rate of improvement of seed development, particularly in semi-arid
and non-water-secure areas (IFAD, 2001). It is difficult to see how that can
154 Michael Lipton with Hugh Waddington
be done without genetic modification (GM), which will require refocusing
of the GM industry away from products (and traits such as herbicide resis-
tance) of interest mainly to capital-intensive large farmers and food
processors, towards staple foods (and traits enhancing yield and robust-
ness to moisture stress) of interest mainly to the world’s poor (see Lipton,
1999).
So what progress has been made in halving world poverty in 1990–2025
so far and, given these trends, are we on target? The agreed method of
checking progress on world poverty reduction is from the 1990 and 1998
estimates of ‘share of developing and transitional population living on less
than $1 a day’12 poverty in World Bank (2000: 23), derived from Chen and
Ravallion (2000). This section reports these rates and asks: what progress
towards the target of halving poverty by 2015 will be made not if the world
changes substantially, but if 1990–1998 rates of poverty reduction con-
tinue in 1998–2015? It might be thought better to use information about
distribution of consumption per person below the poverty line, at least
in major countries or regions, and to estimate how a continuation of
1990–1998 country/region-specific growth rates of mean GDP would affect
mean consumption and hence, assuming unchanged distribution, the pro-
portions of persons below the dollar PPP poverty line.13 Also, some sort of
global or regional/national economic modelling might be attempted,
seeking to estimate likely changes in rates of economic growth, income
distribution, and hence dollar PPP poverty incidence. These methods
have been tried by others; Hanmer and Naschold (2000) and Collier and
Dollar (2001) use similar methods to estimate 2015 poverty incidence of
developing regions on the basis of existing forecasts of annual GDP
growth of these regions. However, such methods require heroic assump-
tions and/or pose forbidding requirements for information – and for
understanding of quantified, durable, multi-country growth and distribu-
tion changes and transmission mechanisms. The results of these methods
often produce projections of poverty considerably lower than those given
here. However, it would seem unduly optimistic to assume that without
substantial changes in policy, the climate for poverty reduction – in regard
to global economic growth, trade liberalisation for developing-country
exports, changes in labour intensity of production methods, etc. – will be
much more favourable for the poor in 1998–2015 than was the case in
1990–1998. Will global growth be faster? Will its impact on global poverty
incidence be systematically more? One could hardly rely on either,
without substantial policy changes.
So, this section tries only to answer the question: if (1993 PPP dollar)
poverty incidence changes in 1998–2015, in main countries and regions,
at the same rate as in 1990–1998, what will the incidence be in 2015, and,
therefore, how will the fall in 1990–2015 compare with the world poverty
target? This is not a silly question, since it is not obvious that poverty
trends in 1990–1998 globally were somehow ‘special’, distorted or misre-
 Halving world poverty 155
ported, in ways that invalidate the naive assumption that the global path of
poverty reduction in 1998–2015 will be much the same as in 1990–1998,
barring major policy shifts. The 1990–1998 national surveys used, which
cover more than 90 per cent of the exposed population (i.e. that of devel-
oping and transitional economies), were carefully screened for coverage
and method. The 1990–1998 trends do, however, contain two cases where
extrapolation is more than usually questionable:

• The estimated decline of poverty in China to 17.2 per cent in 1998

depends almost entirely on a reported fall in incidence from 29.3 per
cent in 1993 to 17.1 per cent in 1996,14 as compared with a significant
rise in incidence in 1987–1990, a slight rise (if any at all) in 1996–1998
and a much slower fall in 1990–1993. Other sources (UNDP, 2001:
149; World Bank, 2000: 280) give a higher estimate of China’s 1998
dollar poverty incidence, at 18.5 per cent. This implies two things:
first, that dollar poverty incidence in China rose in 1996–1998 by
significantly more than our estimate suggests; and second, that if the
rate of change in 1990–1998 is extrapolated to 2015 using this new
evidence, this would mean slower poverty reduction than given in
Table 7.1 – dollar poverty incidence in China would fall to 5.9 per
cent in 2015 and global poverty would be estimated at 23.5 per cent in
2015; in other words, only 38 per cent of the poverty target would be
met.
• In Eastern Europe and Central Asia, poverty incidence more than
tripled, from 1.6 per cent in 1990 to 5.1 per cent in 1998 – this, too,
largely happened in three years, 1990–1993.

As in all other cases, we crudely extrapolate that the annual rates of

change in incidence in 1990–1998 will continue in 1998–2015. This adds
a projected 270 million dollar-poor to the world’s projected total in
1998–2015 from Eastern Europe and Central Asia, but deducts a projected
170 million for China. Of course, these numbers tell us only what will
happen if past trends continue. Things will go better in some countries
and worse in others, but the global picture of poverty reduction expecta-
tions, if 1998–2015 policies remain the same as those in 1990–1998, could
be about right. There is no obviously better alternative projection – and
no reason to believe the global rate of poverty change will ‘bend’ after
1998, for example that populations whose 1990–1998 poverty-reducing
performance improves in 1998–2015 will sharply outweigh those whose
performance deteriorates.
The first step is to estimate the 1990–1998 rate of reduction in poverty
incidence, for populations of developing and transitional economies,
from World Bank (2000: 23). If we do this for the total population, we
would derive a rate of change that is useless for estimating future global
rates of poverty change, even if the future is ‘like the past’. By assuming
156 Michael Lipton with Hugh Waddington
that the global 1990–1998 rate of poverty incidence reduction will prevail
globally after 1998, we imply that the whole world tends towards that rate.
Why is this implication unacceptable?

• First, suppose China’s reduction of poverty eventually takes incidence

in China to almost zero; thereafter, there is unfortunately nothing to
sustain the global rate of poverty reduction by causing China’s fast
(but now completed) rate of improvement to ‘infect’ very slowly
improving regions such as sub-Saharan Africa. Similarly, suppose
the sharp rise in poverty incidence in 1990–1998 in the transitional
economies continues in 1998–2015: applying a global rate of change
of poverty to all regions would assume that this somehow ‘infects’
other regions.
• Second, we need to allow for differential population growth among
regions.

Some optimistic forecasts of global dollar poverty incidence based on the

scenario that ‘past trends are continued’ may derive from erroneously
extrapolating the global poverty reduction rate, instead of (as is correct)
using each regional rate of poverty reduction to estimate that region’s
expected 2015 incidence, applying it to the region’s projected 2015 popu-
lation, and then calculating, as a weighted sum, 2015 global dollar poverty
incidence. Global dollar poverty incidence was 29.0 per cent in 1990 and
24.0 per cent in 1998 (World Bank, 2000: 23). Incorrect ‘global extrapola-
tion’ to 2015 would give an incidence of 16.1 per cent – close to the UN
target of halving incidence to 14.5 per cent (in fact, going 90 per cent of
the distance). Furthermore, incorporating the alternative estimates for
China as well as more recent ones for India (see below) gives an estimate
of 1998 global poverty incidence at 22.0 per cent, which extrapolates to
12.2 per cent by 2015 – implying that not only would the target of halving
incidence be met, it would be exceeded by 16 per cent.
Ideally, we would estimate the 1990–1998 rates of change of each
country’s poverty incidence, and project this to 2015. This would assume
not independence among countries, but that each country’s dollar poverty
incidence change affected each other country’s incidence similarly before
and after 1998. The information for this is not available except for South
Asia. We otherwise project changes for each main World Bank region, but
showing China and South Asian countries separately.
Having projected each region’s poverty incidence to 2015, we estimate
the 2015 population to which it applies from the UNDP Human Develop-
ment Report (2000: 223–226).15 We then estimate Indian, Chinese and
regional poverty incidences and finally global incidence for 2015, all
assuming that future incidence changes are at the same annual rate as
past ones for India, China and each region. These projections are pre-
sented in Table 7.1. World poverty, instead of falling from 29 per cent in
 Halving world poverty 157
1990 to 14.5 per cent in 2015, which is the target, falls to only 23.2 per
cent, which in fact is only 40 per cent of the target. In other words, if each
developing and transitional region (with China and each South Asian
country treated as a separate region) is projected to change poverty inci-
dence at the same annual rate in 1998–2015 as reported by the World
Bank for 1990–1998, poverty incidence in developing and transitional
countries in 1990–2015 will fall not by 14.5 per cent (from 29 per cent to
half that) but only by 5.8 per cent (from 29 per cent to 23.2 per cent) –
that is, achieving 40 per cent (or underachieving by 60 per cent) of the
target. Some countries will be able to do better by 2015 and some worse.
But we have no reason to believe that things are going to get so dramati-
cally better that we go to 14.5 per cent in 1990. In order to go past 23.2
per cent poverty in the world by 2015 and reach the 14.5 per cent target,
one of two things, or some combination of them, must be achieved: (1)
the rate of ‘world’ economic growth must be 2.5 times faster between 1998
and 2015 than it was in 1990–1998; or (2) the rate of transmission of eco-
nomic growth to poverty reduction must be 2.5 times better.
More recent provisional estimates for post-reform India (Datt and
Ravallion, 2002; see also Sundaram, 2001; Deaton and Tarozzi, 2000), if
confirmed, improve past performance, and hence the projection for 2015,
somewhat: ‘incidence of [dollar] poverty in India falls from 39.1 percent
in 1993–1994 to 34.3 percent in 1999–2000’ (Datt and Ravallion, 2002:
14). Using the same extrapolation techniques as above, this gives a trend
rate that if maintained until 2015 would cut Indian poverty incidence to
24.7 per cent. The implication of this is to reduce the above estimate of
2015 global poverty by about a further three percentage points, contribut-
ing to an achievement of 62 per cent of the ‘halving global poverty’ objec-
tive – better than 40 per cent, as implied in World Bank (2000), but still
far below target.

Getting back on track

Reviving progress on halving dollar poverty incidence by 2015 requires
changes in policy both in developing and in developed countries. Several
changes are needed to get back on track; they relate to process, to aid, to
agriculture, and to rural development.

Process
The process needs to ensure that underperformance is recognised where
it is happening. That does not just mean the countries, particularly in sub-
Saharan Africa, which have (at best) been reducing poverty slowly, but
also the groups within countries that have remained poor. Making moni-
toring serious at country level is a very important part of improvement of
the process. Each recipient country needs timely information not only
158 Michael Lipton with Hugh Waddington
about who are the poor, but about where the poor are getting poorer and
where they are getting less poor. This cannot be done with household
income-and-expenditure surveys that happen once every ten years; nor
with surveys, even if frequent, requiring several years of processing time;
nor with national samples too small to permit reliable disaggregation. A
few countries, including India, China and Indonesia, are in a position to
provide data of the frequency and quality needed for serious poverty
monitoring; most are not, and some of the poorest countries will not be
for many years. Participatory poverty assessment can help, especially in
such cases. What is needed is a comprehensive survey, for one year, of both
household expenditure and a range of other, more quickly obtainable
measures, known to be correlated with well-being, probably mostly gath-
ered in a participatory fashion. The latter measures can then be calibrated
into a ‘best estimator’ of measured narrow consumption poverty – now
mainly incidence below a currency poverty line; increasingly, one hopes,
intensity below a FEM poverty line – for that year.
In Ghana the Core Welfare Indicators Questionnaire (CWIQ) – a short
questionnaire (it does not collect income or consumption data) adminis-
tered to a sample large enough for reliable estimation at regional level –
provides ten ‘poverty predictors’ that have been validated against the
more comprehensive Ghana Living Standards Survey. Ravallion (1996)
tests the predictive powers of such ‘rapid appraisal’ welfare questions
against consumption data and concludes that, assuming a relationship
holds between the two types, rapid appraisal proxies for poverty can be
useful for monitoring of poverty when survey data are unavailable or inac-
curate. But, he cautions, credibility is undermined by the problems of
selecting weights for multiple indicators, particularly when actual con-
sumption data do not exist, of potential sampling biases introduced where
sampling methods are not rigorous, and, most importantly, of identifying
a (large enough) set of variables that can explain a significant proportion
of variation in consumption. However, until there is enough statistical
capacity for frequent household expenditure surveys, carefully selected,
quickly obtainable measures for various regions and groups should be
gathered every year or two and used to estimate consumption poverty.
The calibration between the measures and consumption poverty can
change, and needs to be rechecked every five to ten years. Further, care-
fully supervised participatory methods can be applied to understanding
whose poverty in many senses – not just measuring whose consumption
poverty – is changing, how fast and why.
Even with perfect information about poverty trends, published swiftly, it
does not follow that public action, by domestic or aid authorities, will be
taken to reallocate resources, or to change policies. Nor is appropriate
action always obvious. Suppose group or region A is cutting poverty at well
above the target rate, and group or region B well below it. Should anti-
poverty policy refocus on B (in greater need) or on A (where policy is
 Halving world poverty 159
better at reducing need)? Is it much costlier to reduce poverty by the same
amount for B than for A? Was B’s (or A’s) poverty reduced by public
action (or inaction) benefiting its poor people, or by their emigration
from region, or group, B (or A)? For some groups or regions, poverty may
be most cost-effectively reduced by easing migration controls, not by infra-
structure, institutions or even policy improvements in situ. But reasonably
reliable poverty data that are rapidly available and publicly discussed allow
these questions to be asked, and appropriate pressures on the polity
applied. Decent data can supplement ‘peer pressure’ from officials, acade-
mics or donors with poor pressure in civil society.

Aid
The value of aid in real terms (i.e. net aid disbursements as a proportion
of OECD gross domestic product) is now 5–10 per cent less than in the
late 1980s, and has fallen much faster in agriculture. The real value of aid
to agriculture was almost two-thirds less in 1998 than 1988 (IFAD, 2001:
41). The main reason for this collapse in aid to agriculture, over and
above the decline in total aid, is the fact that agriculture’s share of project
and sector aid has fallen sharply – from 20.2 per cent in 1987–1989 to 12.5
per cent in 1996–1998 (ibid.). Moreover, neither the public sector in the
recipient countries, nor private investors, have anything like fully replaced
the lost agricultural support that was formerly coming from aid.

Arguments for getting out of agriculture

Since aid to agriculture is falling so substantially, what are the reasons for
reducing support to agriculture? Two main arguments are advanced. The
first is fungibility: that aid supports government spending at the margin,
the part of spending that government least wants to do. So, the argument
goes, why aid agriculture if governments will only do less of it? This may
be reasonable in some cases and in an international cross-section, but it is
not a reasonable argument for some of the poorest countries of the world.
In sub-Saharan Africa, as well as Bangladesh, where aid is more than
public investment – indeed, a very large part of public expenditure, some-
times exceeding it – one cannot claim that aid to a sector simply displaces
public investment in it. Furthermore, fungibility can be dealt with by
sector agreements with the governments of the countries concerned, pro-
ceeding from the country’s own PRSP, which, as argued already, should
have sector policies in it. For two decades, some World Bank ‘hybrid
loans’ to a sector have been tied to agreed expansions in publicly financed
expenditure, providing infrastructure (including training and research)
for that sector.
The second main argument is that falling world prices damage rates of
return to some projects and sector spending in agriculture. This is less of a
160 Michael Lipton with Hugh Waddington
problem if the aid to this sector benefits principally smallholders and (via
extra work on labour-intensive small farms and activities) farm labourers:
if that is the case, there will be far less effect in glutting world agricultural
markets, since quite a large proportion, though by no means all, of that
extra income will be spent on food. Most of the evidence which has come
in recently through the World Bank comparisons does not suggest a
sharply falling rate of return on agriculture projects; the rate of return to
agricultural research, in particular, has not come down between the 1970s
and the 1990s (Alston et al., 2000). That of course may be in part because
there are fewer agriculture projects nowadays, and if old levels of spend-
ing were suddenly resumed, the rates would indeed fall. But these argu-
ments need to be seen as problems to be solved, not as excuses to allow
support to rural agriculture in aid programmes to collapse. It is hard to
see how such a collapse is consistent with anti-poverty priorities, since it is
on this sector that the bulk of the world’s poor still depend for their liveli-
hoods (IFAD, 2001).

Emphasis on rural development

We have said that rapid and significant reductions in global poverty of the
order of magnitude necessary to reach the halving world poverty target by
2015 can realistically be achieved by supporting the rural poor and rural
income-generating activities. The most important change is to reverse the
collapse of support for labour-intensive staples farming in smallholdings,
and to help revive the performance of smallholders in the developing
world, in particular in the area of food production for local consumption.
There is, however, a serious counter-argument against this view that agri-
culture, particularly local food production, is the way forward. This origi-
nates from the rationale of liberalisation – the outcome of which is that
each producer specialises in what they are relatively good at producing
and exchanges it with the ‘rest of the world’ – and the observation that
there is generally more rapid growth outside agriculture than inside.
However, in the early stages of transition out of extreme poverty, rapid
progress has almost always started with a breakthrough in food production
for local, or nearby, consumption. There is no reason why it should stop
with that – cash crops, the non-farm sector, industrialisation are the paths
to development that work – but they do not work usually before that initial
breakthrough in food production has happened.
The appropriate policies to support rural development will differ from
place to place. However, the Rural Poverty Report (IFAD, 2001) emphasised
several key strategies for rural development:

• Reverse urban bias in public service provision and support rural edu-
cation and health. Supporting education will help break the inherited
cycle of poverty and lack of education.
 Halving world poverty 161
• Increase access to physical assets such as improved farm technology,
water and land. This will reinforce the long-term poverty-reducing
aspects of ‘human asset’ policies such as health and education. Land
reform (which, incidentally, can be achieved consensually with a sup-
porting land fund that enables the poor to buy from large farmers)
may be the best way to provide protection for those without physical
assets against the risk of falling into extreme poverty.
• Improve communications and transport infrastructures and access to
marketing and extension services for rural communities, particularly
in sub-Saharan Africa. This will improve poor people’s access to
markets to sell goods and acquire production inputs (including credit),
and, through greater access to information, their bargaining power.
Better infrastructure will also improve access to public services.
• Complement such infrastructure policies (which can be regarded as
emphasising incentives to increase/diversify production) with policies
that enable yield increases and diversification, such as irrigation pro-
jects and agricultural research into staple foods. Agricultural research
must favour technologies that are labour-intensive if the impact on
poverty reduction is to be maximised.

Conclusion
This chapter has assessed the definition, process and feasibility of the UN
target of halving the incidence of dollar poverty in developing and transi-
tional countries in 1990–2015. If China, each country in South Asia, and
each other region changes incidence in 1998–2015 at the same rate as in
1990–1998, then overall dollar poverty incidence will fall from 29.0 per
cent in 1990 to 23.2 per cent in 2015. This is only 40 per cent of the UN
target reduction of halving world poverty to 14.5 per cent incidence. More
recent estimates for India and China provide a slightly brighter picture,
suggesting that we will reach around 60 per cent of the target, but this is
still well below what should and, with the right policies, could be achieved
globally. The halving of poverty does remain feasible in Asia and Latin
America over the 1990–2015 horizon, and, with suitable agricultural
technology and other policies, in sub-Saharan Africa too. However, we are
not at present on track; we are quite a long way behind it. That is not
because the process or the definitions are inherently flawed, but because
we have not fully recognised the effort needed in agriculture and the rural
sector. Re-emphasising employment-intensive agricultural and rural growth,
especially of staple foods, and asset distribution can revive faltering
progress towards the targeted halving of world poverty.
162 Michael Lipton with Hugh Waddington
Notes
1 Thanks to Howard White for useful comments. All remaining errors are ours.
2 About 95 per cent of people in developing and transitional economies are
covered by household income-and-expenditure surveys permitting periodic
estimation of the proportion in dollar poverty. The so-called ‘developed’ world
is not included; absolute dollar poverty hardly exists in those countries.
3 This oversimplifies: (1) there is a complex chain-linking procedure among
national, regional and global bundles; (2) national pricing is not (as would be
ideal) of the consumption bundle typical below each nation’s dollar-poverty
line, but of its mean gross domestic product (GDP); (3) rankings of countries
by dollar poverty, and its rate of change, are significantly affected by whether
the base date for PPP estimations is, say, 1985 or 1993. For an account of the
difficulties and possible routes to progress, see the May–August 2002 debate at
[Link] between Sanjay Reddy and Thomas Pogge (see the
summary of their ‘How Not to Count the Poor’), and Martin Ravallion of the
World Bank.
4 Most countries in Africa and South Asia, where there are a lot of poor people,
look much poorer than they are if we estimate dollar poverty as the propor-
tion with below US$1 a day at official exchange rates. That is because US$1,
exchanged into (say) rupees or renmimbi, buys more (four to six times in most
low-income countries) of the world consumption bundle than does US$1 in
the USA.
5 Other minor amendments could be suggested. In principle, private consump-
tion of public and subsidised goods and common property should be included;
this is seldom done in household surveys, from which measures of those in
dollar-a-day private consumption poverty are now derived.
6 See Reddy and Pogge (2002). Even the shift from 1985 to 1993 as the base-year
for selecting the PPP consumption bundle makes a big difference to the
numbers counted as dollar-poor in some countries, notably China.
7 There are three main objections to the food energy method (FEM)-based
poverty line:
(a) ‘People do not live by calories alone.’ FEM does not assume that they do,
but assumes that – at the level of consumption at which a person decides to ‘just
about meet’ calorie requirement – he or she also decides to allocate resources
so as to ‘just about meet’ other basic needs. The assumption is, in effect, that
consumers’ decisions normally reveal preferences, which are first directed towards meeting
‘basic needs’, balanced among types (food, shelter, etc.) appropriately for each particular
person’s survival and functioning. This ‘needs-oriented revealed preference’
assumption is likely to be close to reality for people near the margin of poverty.
To survive at all, they have to use resources reasonably rationally. There are of
course diversions due to addiction, debt, unpredicted change in needs or
income, waste or error; but for poor and near-poor people, such diversion is less
tolerable, since it constrains their already tight expendable resources. FEM,
unlike PPP, selects poverty lines that respond to a basic need for functioning
(the need for food, occupying some 65–75 per cent of the budget of the poor)
in the context of the need to spend on other basic needs for functioning. And
FEM, unlike the cost of basic needs (CBN) approach, avoids assumptions about
the ‘correct’ mix of consumption goods – assumptions that are paternalistic,
unlikely to be similarly applicable to all persons, and unresponsive to the fluctu-
ating needs of a family near the margin of survival.
(b) ‘FEM ignores within-household allocation of resources for food (and
other basic needs).’ In some countries, many households discriminate against
girls aged 2–4, providing them (relative to need) with less food than other
 Halving world poverty 163
household members. An ‘FEM poverty line’ could, in such countries, under-
state the consumption per equivalent adult at which all household members
achieve food adequacy. However, this objection (assumed fair division of con-
sumption among household members) applies to all consumption-based fixed
poverty lines. FEM at least allows the problem to be measured: in ‘gender-dis-
criminatory’ populations, caloric adequacy for all household members is
achieved at higher household consumption than elsewhere, so the FEM line
(properly measured) is also higher. This allows the costing of socially deter-
mined gender misallocation – as, indeed, that of addiction, or of interest on
debt.
(c) ‘FEM-based lines privilege consumer decisions, e.g. about food, that
reflect not need but choice, maybe luxury choice.’ Using FEM instead of CBN
recognises revealed preference, but the flip side is that the ‘FEM-poor’ in some
areas – but not others – include those habituated to costly diets and thus
calorie-underfed even at high consumption levels. Even poor people can be
pushed (e.g. by convenience of prepared snacks for male casual workers, pres-
sures on women’s cooking time, or peer pressure on children) to buy costly
calories. This has sometimes led to ‘necessary consumption levels for caloric
minima’ – FEM poverty lines – set absurdly higher (over 70 per cent at one
stage in Indonesia) in urban than in rural areas, leading to gross underestima-
tion of rural, relative to urban, poverty. However, such grossly food-inflated
FEM estimates are unusual, fall when urban survey techniques improve, and in
part reflect genuine needs (e.g. urban snacks at work, when work is far from
home).
8 See the discussion of China on p. 155.
9 In 1990–1998 the poverty gap index fell by 19.7 per cent for the total popu-
lation of developing and transitional countries (10.7 per cent excluding
China), as against 17.3 per cent (6.7 per cent) for incidence (Chen and Raval-
lion, 2000: tables 2, 4).
10 All this leaves open the issue of participatory versus top-down ways of assessing
poverty. However poverty is measured, the role of participatory methods is an
important but separate issue (see Chapter 5).
11 Successful farm growth first absorbs labour but eventually releases it, raising
townward migration. However, this is not the desperate pressure of rural work-
less on urban labour markets; it is the migration of hope, not of despair.
12 In fact, US$1.08 in constant 1993 purchasing power. Globally, this poverty line
corresponds roughly to US$1 a day in constant 1985 purchasing power, used in
many earlier estimates. However, for some countries, the change of PPP base
may significantly alter poverty estimates or even trends.
13 An apparently obvious point, appearing to justify or even to demand this pro-
cedure, is fallacious. It might be thought that growth of mean consumption
must, of itself, increase the responsiveness (elasticity) of PPP dollar poverty to
future growth of mean consumption – because earlier growth normally pulls
the poor nearer and nearer to the dollar poverty line, so that a given amount
of extra mean consumption pulls more people above the line (this would be
even more the case for each successive 1 per cent of steadily rising mean con-
sumption). However, this ignores the facts that a large proportion of the poor
become so (or become even poorer) in any given year, and that those left in
poverty by growth are likely to be those with most difficulty in escaping poverty
through growth. Certainly, though there is no systematic link between growth
and overall distribution, there is also no tendency for dollar poverty incidence
to become more elastic to mean growth of consumption or of GDP; for example,
these elasticities appear to have fallen somewhat in India as between 1975–1989
and 1992–1997.
164 Michael Lipton with Hugh Waddington
14 This would have reduced the number of Chinese poor by 138.4 million – 40
per cent of the 1993 number, and over 10 per cent of the world’s poor – in just
three years.
15 To estimate the populations of World Development Report (WDR) regions, we
adjust the populations of Human Development Report (HDR) regions where
they have different membership (UNDP, 2000: 285–286; World Bank, 2000:
334–335). The main differences are the following: ‘Developing East Asia and
Pacific’ (WDR) excludes Singapore and Hong Kong but includes Korea
(Democratic Republic); the opposite is true of the (otherwise almost identical)
HDR ‘developing countries in East Asia, South-East Asia and the Pacific’.
WDR’s ‘Developing South Asia’ includes Afghanistan but not Iran; HDR reverses
this. WDR’s ‘Developing Middle East and North Africa’ differs from HDR’s
‘developing Arab States’ by including Iran and Malta, and excluding Kuwait,
Qatar, the United Arab Emirates and Sudan (the latter counted in ‘developing
sub-Saharan Africa’ by WDR, whose sub-Saharan Africa definition otherwise
matches HDR’s). WDR’s ‘developing and transitional Europe and Central Asia’
include Bosnia and Herzegovina, the Yugoslav Federal Republic (Serbia and
Montenegro) and Turkey, but exclude Slovenia; HDR’s ‘developing Eastern
Europe and the Commonwealth of Independent States’ reverses this. In each
case we use the HDR projection for the 2015 population of the WDR region
(UNDP, 2000: 223–226), since WDR regions are used in poverty projections
(World Bank, 2000: 23).

References
Alston, J., Chan-Kang, C., Marra, M.C., Pardey, P.G. and Wyatt, T.J. (2000) A Meta-
analysis of Rates of Return to Agricultural R&D: Ex Pede Herculem?, International
Food Policy Research Institute (IFPRI) Research Report 113, Washington, DC:
IFPRI.
Chambers, R. (1997) Whose Reality Counts? Putting the Last First, London: Intermedi-
ate Technology Publications.
Chen, S. and Ravallion, M. (2000) ‘How Did the World’s Poorest Fare in the
1990s?’, Policy Research Working Paper 2409, Washington, DC: World Bank.
Collier, P. and Dollar, D. (2001) ‘Can the World Cut Poverty in Half? How Policy
Reform and Effective Aid Can Meet International Development Goals’, World
Development, 29(11): 1787–1802.
Datt, G. and Ravallion, M. (2002) ‘Is India’s Economic Growth Leaving the
Poorest Behind?’, Policy Research Working Paper 2846, Washington, DC: World
Bank, 11 February.
Deaton, A. and Tarozzi, A. (2000) ‘Prices and Poverty in India’, Research Program
in Development Studies, Princeton University, Version 4, mimeo, 29 July.
Foster, J., Greer, J. and Thorbecke, E. (1984) ‘A Class of Decomposable Poverty
Measures’, Econometrica, 52: 761–765.
Hanmer, L. and Naschold, F. (2000) ‘Attaining the International Development
Targets: Will Growth Be Enough?’, Development Policy Review, 18: 11–36.
International Fund for Agricultural Development (IFAD) (2001) Rural Poverty
Report 2001: The Challenge of Ending Rural Poverty, Oxford: Oxford University
Press.
Lipton, M. (1999) ‘Reviving Global Poverty Reduction: What Role for Genetically
Modified Plants?’ Sir John Crawford Memorial Lecture, Washington, DC:
Consultative Group on International Agricultural Research.
 Halving world poverty 165
Ravallion, M. (1996) ‘How Well Can Method Substitute for Data? Five Experi-
ments in Poverty Analysis’, World Bank Research Observer, 11(2): 199–221, August.
—— (2000) ‘On the Urbanisation of Poverty’, Université des Sciences Sociales
(Toulouse) and World Bank, mimeo, 6 February.
Reddy, S. and Pogge, T. (2002) ‘How Not to Count the Poor’, mimeo, Columbia
University, at [Link]
Sundaram, K. (2001) ‘Employment and Poverty in the 1990s: Further Results from
NSS 55th Round Employment–Unemployment Survey, 1999–2000’, Economic and
Political Weekly, 112 (11 August): 3046.
United Nations Development Programme (UNDP) (2000) Human Development
Report 2000, New York: Oxford University Press.
—— (2001) Human Development Report 2001, New York: Oxford University Press.
World Bank (2000) World Development Report 2000/2001, New York: Oxford Univer-
sity Press.
8 Towards universal primary
education1
Christopher Colclough

Introduction
The Universal Declaration of Human Rights, adopted by the United
Nations (UN) in 1948, stated that primary schooling should be free and
compulsory in all nations. More than fifty years later, the world still has far
to go in order to achieve this goal. The Millennium Development Goals’
reaffirmation of the aim of achieving schooling for all by 2015, and of
gender equality in schooling by 2005, again risks failure, particularly
in the poorest regions of the world. The challenge is greatest in the
countries of sub-Saharan Africa, where scarcely half the eligible children
attended primary school at the turn of the century. This chapter examines
some of the main causes of this disappointing progress, and indicates the
main policy changes needed if the development goals for education are to
be met.

Why should societies seek to educate everybody?

At the outset it is worth recalling the very many ways in which education
underpins the development process. Its impact is strong, particularly for
the lower levels of schooling. The following generalisations are broadly
supported by a large body of research:

• Education is productive. It more than repays its costs to society. Social

returns are high, particularly at primary and junior secondary levels,
and particularly if estimates for externalities are included. It provides
one necessary ingredient for economic growth.
• Primary education can directly help to alleviate poverty. Literacy and numer-
acy are key instruments with which to help the self-employed, includ-
ing those who are farmers, to increase their incomes. Providing the
poor with human capital can represent a very targeted form of spend-
ing: once created, the enhanced human capacity cannot be taken
away (except, perhaps, over the very long run, via obsolescence).
• Primary education also brings important gendered benefits. Better-educated
 Towards universal primary education 167
women incur larger costs if they have children, and have greater
incentives to have smaller families. They are also better able to control
their own fertility. World fertility surveys in thirty-eight countries show
a decreased number of children born, lower total fertility rates and
smaller desired family size for one to three years, four to six years and
seven-plus years of schooling compared with none. This relationship is
strong and widely documented at both macro and micro levels.
• There are other palpable social benefits primarily arising from educating girls
and women. The possession of more years of schooling by the female
population is associated with lower infant and child mortality. Child
nutrition is higher and school attendance itself is more strongly related
to the education of mothers than of fathers. Thus at the margin, the
education of girls can be said to have an even higher social priority
than that of boys.
• People also want education for its own sake. It increases their choices, both
in work and leisure, and is thus an important consumption good. Thus
notions of justice and equality underpin its status as a human right.

Accordingly, on grounds of consumption, production, distribution and

rights, the Millennium Development Goals in education – achieving uni-
versal primary enrolment by 2015 and gender equality in enrolments by
2005 – can be given a strong and well-defended endorsement as important
targets for all countries.2 The above generalisations imply that they are
also fundamental to the achievement of the other Millennium Develop-
ment Goals. Poverty alleviation, infant and child mortality, fertility and
sustainable development are all crucially influenced by the skills delivered
by schooling being widely held.3 Thus the education goals, in the long
term, are important to achieving all the others.

What do the targets mean – and what would count as

success?
The criteria for meeting the two education goals were set out, in some
detail, by the Organisation for Economic Cooperation and Development
(OECD) in 1996. As the following summary indicates, they are specific
and demanding.

Conditions for achieving universal primary education by 2015

• Net enrolment rates (i.e. the percentage of children of primary school
age actually enrolled in school) of 99 per cent by 2015. We should
note that even in industrialised countries there are reasons why not all
eligible children can go to school. Some are prevented from doing so
by illness, some are educated at home and some are absent for other
reasons. Accordingly, typical net ratios at primary level for the USA
168 Christopher Colclough
and European countries are 96–98 per cent. Thus a target of 99 per
cent for all countries looks ambitious.
• Completion rates (i.e. the percentage of Grade 1 pupils who subse-
quently complete four years’ schooling and who proceed to year 5) of
99 per cent by 2015. This is also highly ambitious, and, in combina-
tion with the first criterion, would seem to require universal Grade 1
enrolment by 2011.
• Literacy rates of 100 per cent amongst 15- to 24-year-olds by 2015.
This indicator is included as a measure of the quality of schooling.
Although, strictly speaking, literacy does not have to be delivered by
schooling, in practice primary education is the only viable instrument
for achieving widespread and permanent literacy in the younger age
groups. Yet if this is so, the criterion becomes even more demanding
than the first two, requiring universal Grade 1 enrolment and con-
tinuation by 2005 for it to be achieved.

Conditions for achieving gender equality

The following are the conditions for achieving gender equality:

• A ratio of girls to boys in primary and secondary schooling of 1:1 by

2005. Given the high levels of existing inequality between the enrol-
ment of males and females, this target will almost certainly not be
achieved in many countries by 2005.
• A ratio of literate females to males, over the ages of 15–24 years, of 1:1
by 2015. This requires both parity in Grade 1 enrolments and in sub-
sequent continuation rates by 2005, which, as indicated above, looks
impossibly ambitious.

What progress has been made so far?

All discussions of progress to date are hampered by a lack of data. The
data are not only weak, but often severely delayed in becoming available.
The United Nations Educational, Scientific and Cultural Organisation
(UNESCO) has the task of collecting and publishing data on education
systems world-wide. But many countries take several years to provide the
statistics; when they do, they are incomplete, and those that are provided
are often erroneous. These problems are greater in the poorer countries,
where the gaps between targets and outcomes are greatest. Thus in Africa
the most recent pupil enrolment data available in mid-2002 related to the
position in 1997. Even for that year, 15 per cent of countries were unable
to provide estimates for their gross enrolment ratios (GERs) for primary
schooling, and one-quarter could not provide net ratios (UNESCO, 1999).4
Many more could not provide information on intakes to school, on public
expenditures or on efficiency measures. No recent data of any reliable
 Towards universal primary education 169
kind exist for Nigeria, which accounts for 20 per cent of the population of
sub-Saharan Africa. So, the data published by UNESCO are based upon
samples of countries reporting, which are not necessarily representative of
those that do not. The picture, then, is almost certainly worse than it
seems. What does it tell us?
Table 8.1 shows trends in GERs by region. These data give an indication
of the capacity of school systems. Traditionally, universal primary enrol-
ment has been taken to be achieved when the GER has reached 100. But
because systems are so inefficient, late enrolment and high rates of repeti-
tion result in about 20 per cent of pupils in developing countries being
outside school age. So, GERs of 100 are consistent with 20 per cent of chil-
dren being out of school. There is a need, therefore, to move to using net
ratio data, but these are not yet reliably available on a trend basis.
The years 1965–1980 were a ‘golden age’ for the expansion of primary
enrolments. The table shows that GERs for all developing countries taken
together increased from 78 to 95 over those years, and it looked as though
universal primary education was within reach. In sub-Saharan Africa, expan-
sion was remarkable – almost doubling from 41 per cent to 77 per cent of
the age group, catching up with South Asia over those years. Things
changed, however, in the 1980s and 1990s. Latin America and East Asia
experienced steady enrolment increases from an already high base. South
Asian enrolments grew strongly. The Arab states, on the other hand, made
only small progress. But sub-Saharan Africa experienced a decline in GERs,
which re-attained their 1980 level only in 1997. For much of the last two
decades of the century, both the proportion and the number of African
children out of school increased, with their numbers being around 80 per
cent higher in 2000 than they had been two decades earlier.
As regards gender equality, Table 8.2 shows that this has not been
achieved in any developing region. The countries of Latin America, the
Caribbean and East Asia have a 10 per cent excess of boys enrolled. The
difference in other regions, however, is between 20 and 30 per cent. In

Table 8.1 Trends in primary gross enrolment ratio (GER) by region and gender,
1965–1997

1965 1980 1990 1997

Sub-Saharan Africa 41 77 73 77
Arab States n.a. 80 84 85
Latin America and Caribbean 98 105 107 114
East Asia and Oceania n.a. 110 118 118
South Asia 68 77 91 95
Developing countries 78 95 99 102

Source: UNESCO (1998, 2000a).

Note
n.a., not available.
170 Christopher Colclough
Table 8.2 Female enrolments at primary level in developing countries, relative to
male enrolments (per cent), 1980 and 1997

1980 1997

Sub-Saharan Africa 78 82
Arab States 73 80
Latin America and Caribbean 87 91
East Asia and Oceania 83 92
South Asia 66 73
All developing countries 79 84

Source: UNESCO (1998, 2000a).

South Asia, girls’ enrolments have been growing quickly. Although the
gender differential remains highest in this region, female GERs increased
rapidly after 1980. Male GERs in South Asia are over 100, and some nar-
rowing of gender differentials can be expected if policy reforms are con-
tinued. The countries of sub-Saharan Africa saw only a slight improvement
towards gender equality of enrolments over the period 1980–1997 – and
even then for the ‘wrong’ reasons: throughout the period the average
female GER in sub-Saharan Africa remained roughly unchanged at around
68 per cent of the age group. Thus the gender gap narrowed because boys
were withdrawn from school faster than girls over the two decades.
As a result, at the turn of the century sub-Saharan Africa had the lowest
ratios of both girls and boys in school of all the developing regions, and
there had been little or no progress in sub-Saharan Africa towards meeting
the education Millennium Development Goals over the previous twenty
years. It should be recalled that these data refer to gross enrolments. Net
ratios in Africa are scarcely higher than 50 per cent, so about half of the
primary-aged children in sub-Saharan Africa remained out of school. Pro-
jections of the trends implied by Table 8.1 suggest that, with no change in
current rates of population growth, by 2015 the total number of children
out of primary school by 2015 would fall from around 125 million to
around 80 million – that is, numbers would be absolutely reduced by one-
third, and schooling for all would not be achieved. Moreover, there would
be a tremendous change in the regional composition of underenrolment.
In particular, the number of children out of school in South Asia would be
dramatically reduced – from around 60 million to about 7 million chil-
dren. By contrast, those out of school in sub-Saharan Africa would increase
absolutely (by about 25 per cent) and they would account for almost three-
quarters of all children out of school. Thus on present trends, three out of
four out-of-school children will be African by 2015 (Watkins, 1999).
Accordingly, whether or not the Millennium Development Goals for
education will be achieved is likely to depend to a significant extent upon
what happens in Africa. Whether, in turn, Africa can achieve the targets
depends upon whether the causes of past trends can be removed or
 Towards universal primary education 171
changed. We therefore need to try to understand how it is that sub-
Saharan Africa has had such a dismal record in the recent past. What has
caused these outcomes, can they change and, if so, how?

Causes of enrolment outcomes in sub-Saharan Africa

Supply issues
We know that at a macro level, a number of economic variables jointly
determine both the number and the proportion of children attending
school. These comprise society’s expenditures on schooling, the average
cost per student and the size of the school-age population. The greater
the first of these and the smaller the last two, the greater the proportion
of children who will be enrolled in school. What has been happening to
these variables in sub-Saharan Africa?
Public spending on education rose to more than 5 per cent of gross
national product (GNP) during the 1990s. This was high in comparison
with other developing regions, and similar as a proportion of GNP to
levels in the countries of Europe and North America (UNESCO, 2000a:
118). Spending on primary schooling was also high, receiving 2 per cent
of GNP in sub-Saharan Africa, compared to 1–1.5 per cent in Latin
America and Asia during the 1990s.
However, although commitment to public spending on education, on
average, was strong, the region’s capacity to generate increased resources
for education was weakened by economic decline. Total education spend-
ing in current dollars increased by only 40 per cent in sub-Saharan Africa
from 1980 to 1997. By contrast, in India it increased by 150 per cent, in
Latin America it almost tripled and in East Asia it quadrupled over those
same years. Thus a reduced relative commitment to public spending on
education in South Asia still allowed increases in the value of real public
spending per pupil. In sub-Saharan Africa, on the other hand, economic
decline meant that it fell sharply and perilously – partly explaining the
dismal enrolment progress of the region in recent years.
Furthermore, although sub-Saharan African countries, on average, can
be said to have made greater efforts to achieve schooling for all children
in recent years, this is by no means true of all of them taken individually.
To demonstrate that this is so, Table 8.3 shows aspects of the costs of, and
expenditures upon, primary schooling for a group of sub-Saharan African
countries with GERs less than 100. All those having the relevant data are
shown in the table, for the most recent available year. As indicated,
according to UNESCO regional data, countries in sub-Saharan Africa were
allocating, on average, about 2.0 per cent of GNP to public expenditures
on primary schooling during the early 1990s. The (unweighted) average
value for this variable, for the sub-set of countries and years shown in
Table 8.3, was 1.9 per cent, indicating that they were not atypical of the
Table 8.3 Sub-Saharan African countries with gross enrolment ratios (GERs) less than 100: public expenditures on primary schooling
and related characteristics

Year of GER Female GNP per Expenditure Unit costs Population Expenditure
data GER as capita 1995 as percentage size required for
percentage (US$) GNP GER ⫽ 100
of male GER

(1) (2) (3) (4) (5) (6) (7) (8)

Low commitment – low cost

Eritrea 1996 54 83 180 0.6 8.5 13.6 1.2
Guinea 1995 48 54 540 0.7 8.9 16.6 1.5
Madagascar 1990 84 98 240 0.7 5.9 14.8 0.9
Ghana 1990 77 83 370 0.8 6.3 16.8 1.1
Chad 1996 65 52 210 0.9 8.3 16.1 1.3
Zambia 1995 89 99 350 0.9 4.7 20.9 1.0
Central African Republic 1990 66 64 350 1.2 11.0 16.0 1.8
Senegal 1996 69 82 550 1.2 10.4 16.2 1.7
Comoros 1995 74 n.a. 440 1.4 11.1 17.3 1.9
Mauritania 1996 83 90 450 1.5 10.9 16.1 1.8
Low commitment – high cost
Burkina Faso 1985 27 59 220 0.7 16.2 16.5 2.7
Mali 1995 34 66 250 1.0 16.9 16.5 2.8
Ethiopia 1996 37 57 110 1.2 20.8 16.2 3.4
Gambia 1990 64 68 350 1.3 14.2 14.6 2.1
Burundi 1990 73 84 150 1.5 13.4 15.8 2.1
Moderate commitment – high cost
Mozambique 1990 67 75 140 1.9 21.3 13.3 2.8
Djibouti 1990 38 71 2.0 32.2 16.1 5.2
Côte d’Ivoire 1996 71 74 670 2.1 17.7 16.7 3.0
Tanzania 1985 75 97 160 2.1 14.6 19.4 2.8
Benin 1995 72 57 350 2.5 19.5 18.0 3.5
Lesotho 1994 99 114 670 2.6 14.2 18.7 2.7
Kenya 1990 95 97 260 3.2 14.0 24.2 3.4
Average 66 77 334 1.5 13.7 16.8 2.3
Francophone average 62 71 368 1.3 14.0 16.4 2.3
Anglophone average 83 93 360 1.7 11.3 15.9 1.8
Sub-Saharan Africa Average 83 83 777 1.9 13.6 17.0 n.a.

Source: GNP per capita 1995, Atlas method World Development Indicators 2000 CD-ROM. All other data, UNESCO Statistical Yearbooks (1998, 1997, 1996,
1994), Paris.
Notes
n.a. not available.
Low Commitment ⫽ GER ⬍100 in years shown and where public spending on primary ⬍1.9 per cent of GNP.
Col. 5 ⫽ public spending on primary as a percentage of GNP.
Col. 6 ⫽ publicly funded costs per primary child expressed as a percentage of income per capita.
The calculated averages for sub-Saharan Africa, and for the francophone and anglophone groups, use all countries that have the relevant data, not just
those in the table.
174 Christopher Colclough
region as a whole. However, each of the first fifteen countries in the table
was allocating, in the early to mid-1990s, substantially less public spending
to its primary-school system than this sub-Saharan Africa average (column
5). Thus, as a group, they can be considered to have been demonstrating
only a low commitment to its provision.
Many of these ‘low-commitment’ countries also had extremely low
primary enrolments, with nine of them having GERs less than 70 (column
2). Furthermore, many of them were also spending very modest sums
per pupil (column 6). In six of them, this was equivalent to less than one-
tenth of income per capita (compared to an unweighted average value
of 13.6 per cent for sub-Saharan Africa as a whole). Such low average
per-pupil expenditures imply that for the first ‘low-commitment, low-cost’
group of ten countries shown in Table 8.3, universal provision of schooling
(as proxied by GERs of 100) would have been easily affordable. The last
column of the table indicates the amount of public spending on primary
schooling, expressed as a proportion of GNP, which would have been
required to achieve GERs of 100 in each country (assuming the mainte-
nance of the average schooling costs and population sizes prevailing in
the base year, and no demand constraints). It can be seen that the
‘required’ expenditure value for universal primary education in this first
group, under these assumptions, would in no case have exceeded 1.9 per
cent of GNP – the amount spent by the ‘average’ country in sub-Saharan
Africa on primary schooling. Furthermore, in many cases the required
financial allocation would have been much lower. It can be seen that the
governments of Eritrea, Madagascar, Ghana, Chad and Zambia would
have needed to allocate little more than 1 per cent of GNP for universal
provision of schooling to have been achieved. This was equivalent to only
about one-half of the average public expenditure allocations made by
other sub-Saharan Africa countries at that time.

Unit costs
These comparisons indicate that the level of primary enrolments attain-
able with a given aggregate expenditure depends upon the costs per
student at primary level. Although on average in sub-Saharan Africa,
public expenditures per student amounted to the equivalent of about 14
per cent of income per capita in the early 1990s, its variance was substan-
tial across countries – from about 5 per cent of income per capita in
Zambia to over 20 per cent in Ethiopia, Mozambique and Djibouti. In
some countries, high unit costs can act as a severe constraint, preventing
the achievement of universal school enrolment. For example, the middle
group of five countries shown in Table 8.3 would have been unable to
reach GERs of 100 even if public spending had increased to its regional
average value, because all of them had relatively high unit costs. These
‘low-commitment, high-cost’ countries mainly spent considerably more,
 Towards universal primary education 175
per primary pupil, than the average across sub-Saharan Africa. Although
universal primary education was still financially practicable for them, a
higher proportion of GNP would need to have been allocated for this
purpose than the average for sub-Saharan Africa. Burkina Faso, Mali and
Ethiopia were particularly affected by high costs: if per-pupil expenditures
were to have remained unchanged, between 2.7 and 3.4 per cent of GNP
would have been needed to secure universal provision. Accordingly, for
these countries, some combination of increased public spending on primary
schooling and reduced unit costs would probably be necessary if schooling
for all were to be achieved in the near future.
A further group of the low-enrolment countries were spending more
on primary schooling than the others, yet were still heavily constrained
by high costs of school provision. Enrolments in this third group of
‘moderate-commitment, high-cost’ countries were not mainly limited by
an unwillingness to commit public resources to primary schooling: each of
them was allocating 2–3 per cent of GNP for these purposes in the years
shown. Rather, enrolment growth was constrained by their primary
systems being relatively expensive, in comparison with their levels of eco-
nomic development. In a majority of these countries – Kenya, Benin, Côte
d’Ivoire and Djibouti – between 3 and 5 per cent of GNP would have been
needed in order to achieve universal provision of the system (column 8).
Here again, strategies to reduce unit costs would be critically important if
affordable expansion were to be secured.
The main determinants of the unit costs of primary schooling are
teachers’ salaries and pupil:teacher ratios. This is particularly so in Africa,
where only a small proportion of recurrent spending is typically available
for non-salary items. Accordingly, the substantial differences in costs
shown in Table 8.3 are the result of country diversity in the values for
these two elements. For example, where teacher earnings are high and
pupil:teacher ratios are relatively low (as in Ethiopia), both need adjust-
ment so as to allow costs to fall. By contrast, where teacher earnings are
low and pupil:teacher ratios are high (as in Zambia), unit expenditures
may need to increase so as to achieve a better quality of primary
provision.5 A wide range of values for each of these elements exists, requir-
ing different policy responses in each case.
We can conclude that a good number of sub-Saharan African countries
have not allocated sufficient resources to primary schooling: their commit-
ment has been insufficient to secure universal provision. More than half
of these countries could reach universal primary education by such spend-
ing changes (assuming that average and marginal costs did not differ
sharply). Others would need additional resources or reduced costs. A final
group has been constrained by both costs and resources, and would find
the shift to universal primary education very difficult to achieve in the
absence of much greater levels of external support.
176 Christopher Colclough
Demand issues
Much of the international attention given to achieving universal primary
schooling has been focused upon ways of increasing the supply of schools
and teachers so as to provide the opportunities for all to attend. However, it
is clear that this is often an insufficiently nuanced approach because it tends
to ignore the problems of demand. This has a number of dimensions.
First, the objective of securing universal primary education needs to
embrace the notion of school quality. Schools where classes are over-
crowded, and where provision for desks, teaching aids and textbooks are
each inadequate, are unlikely to be able to meet the learning needs of
most children. The achievement of primary schooling for all would
require both net enrolment ratios in the high 90s (implying GERs greater
than 100 over the medium term) and an acceptable and evenly distributed
level of school quality. This objective needs to replace the easy but over-
simplified emphasis on the attainment of GERs of 100, which has been
widely used in the literature to indicate the attainment of universal primary
education.6
Second, as suggested earlier, the effort to reduce the costs of schooling,
so as to make universal provision more affordable, can endanger school
quality. In that context, it is noteworthy that the first six countries shown
in Table 8.3, all of which were demonstrating low levels of commitment to
public spending on primary schooling in the early 1990s, also had very low
unit costs of school provision at that time. Whilst, as indicated earlier, this
means that the costs of universalising the system were modest, it also sug-
gests that the quality of primary school provision in those countries may
have been much too low.7
Under these circumstances it is likely that the present level of average
costs sharply underestimates those required to secure universal enrol-
ment. In other words, there may be a significant difference between
average and marginal unit costs, particularly where school quality is cur-
rently very low. There are, in any case, good reasons to suppose that the
costs of enrolling the final 10 per cent of an age group will be higher than
the average for the rest of the population. Often the people concerned
live in small or isolated settlements, in regions with difficult communica-
tions, or include minority or migrant populations who cannot be catered
for in the same way as other groups. But an additional, and often over-
looked, problem is that those who are out of school may prove difficult to
enrol because they do not believe that school attendance – at current
levels of school quality – would be in their interests. Although it is a
mistake to assume that public expenditures per pupil automatically
provide a proxy for school quality, in practice, and in the short run, it is
often very difficult to secure improvements in quality without some
increase in expenditures. This generates a dilemma. Although universal
provision may seem affordable at current cost levels, it may well actually
 Towards universal primary education 177
require a significant increase in both total and unit expenditures. In the
absence of the latter, a demand response may not appear. For example, in
Malawi there was initially a massive expansion of enrolments in the mid-
1990s associated with a shift to fee-free primary schooling. However, there
have been recent sharp falls in enrolment, which appear to be related to
reductions in unit expenditures by government and resulting low, and
declining, quality of provision (Kadzamira and Chibwana, 2000: 16ff.; Rose,
2002).
As the above discussion implies, demand issues, as well as those of
supply, are important in determining enrolment outcomes. In countries
where households face substantial direct and indirect costs in sending
children to school, policies that seek to reduce such costs can be expected
to have a positive impact on demand. Experience shows that at the macro
level the price elasticity of demand for schooling is quite high. Those
countries that have reduced or abolished fees at primary level have wit-
nessed significant increases in demand for places – particularly where
school quality has been maintained. However, it remains the case that
poverty can still be decisive: poor households need their children to
contribute to the household economy, and these opportunity costs can
often be significantly higher than the direct costs of school attendance. In
such circumstances the benefits of schooling to households need to be
palpable if the costs are to be overcome and the children are to be sent to
school.
Nevertheless, the opportunities for policy to affect outcomes are sub-
stantial. One way of considering its potential impact is to look at the vari-
ance of enrolment outcomes amongst countries at different levels of
income per capita. As we would expect, primary GERs tend to be lower for
poorer countries. However, even at rather low incomes, some countries
achieve high enrolments (Figure 8.1). Most of the sub-Saharan Africa
countries are concentrated in the US$300–500 per capita range, where
GERs range between 30 and well over 100 for countries at similar levels of
income. Thus countries with low incomes per capita need not necessarily
have to have low school enrolments: policy, or country history and circum-
stances appear to make a difference.
A further critical demand-side issue is that of gender. It is well estab-
lished that low primary-school enrolments are associated with large
gender gaps, and that gender inequality reduces as gross and net enrol-
ments grow (Colclough et al., 2000, forthcoming). In one sense they have
to do so: for net enrolment rates to be sustained at over 50 per cent,
female enrolment has to be higher than zero. But the range again is large.
Tanzania and Benin, for example, have had similar GERs (around 70), yet
the ratio between female and male enrolments in Tanzania has been close
to 1, compared with only 0.45 for Benin. The difference is not accounted
for by poverty, since income levels per capita in Tanzania have been con-
siderably lower than in Benin. Rather, Tanzania has had a long-standing
178 Christopher Colclough
Primary GER

Figure 8.1 Scatter plot of primary gross enrolment ratio (GER) against GNP per
capita, 1990: low- and middle-income countries.

social policy which has asserted the equality of men and women, advoc-
ated their equal need to have attended school, and established incentives
to encourage that result (Peasgood et al., 1997).
Evidence from nine African countries indicates the central importance
of gender-focused reforms if the demand for schooling is to be maintained
during the move towards schooling for all. These cannot be properly dis-
cussed in a chapter of this length. However, they would need to encompass
reforms focused on the household – where the gendered division of labour
is the major source of differential opportunity costs, and where parents
have less direct incentive to educate girls if future material support comes
mainly from sons. Reforms aimed at improving school conditions are also
required – where inadequate facilities, inadequate toilets, harassment
(both in school and en route), teacher male-bias, and the allocation of
school tasks all disadvantage girls. The labour market, too, is an important
site of discrimination – where formal or informal job reservation for males
is widespread, where wages are lower for girls and where there are few
female role models. Finally, reforms focused on changing broader tradi-
tions in society – where women’s roles are expected to be centred on the
family, and where early marriage robs many girls of incentives to stay in
school, or even the possibility of doing so – will also be required. It can be
shown that many of the relevant gender-focused, demand-side reforms cost
little. But they are as central to the success of policies to achieve schooling
for all as are those which focus on the supply side.8
 Towards universal primary education 179
The national reform process and the role of aid
The needs to increase public expenditures to finance the expansion and
qualitative improvement of primary school systems, to remove school fees
and to increase the incentives for the attendance of girls provide a demand-
ing agenda for many of the poorer developing nations. Many of them will
require the implementation of substantial reforms to improve the utilisa-
tion of existing resources, and to reduce costs in ways that do not compro-
mise quality. Most analyses suggest that the potential for reducing costs
and for increasing domestic resources is substantial. Efficiency savings
appear capable of reducing the expenditures required for primary school-
ing for all by between 10 and 20 per cent, whilst in some countries the
potential savings are very much greater (Colclough with Lewin, 1993:
205–241; World Bank, 2002). Equally, many countries would be capable of
increasing their expenditures on primary schooling substantially. Never-
theless, even if feasible policy reforms were fully implemented, and if
expenditure priorities were reassessed, financing gaps would remain,
particularly in much of sub-Saharan Africa.
A wide range of estimates of the costs of achieving universal enrolment
have been made. They vary from US$2.5 billion per year for forty-seven
‘most at risk’ countries (World Bank, 2002) to US$4.5 billion (Colclough
with Lewin, 1993), $9 billion (Delamonica et al., 2002) and as high as
$10–15 billion (World Bank, 2001) for all developing countries taken
together.9 However, these magnitudes are not strictly comparable, since
all estimates of additional costs have to make assumptions concerning the
amount of policy reform to be undertaken by each country. Some esti-
mates adopt a ‘no change’ scenario by calculating the additional resource
costs of enrolling all of the (growing) primary age group, over a determi-
nate number of years, at present levels of unit costs. Others build in sets of
policy reforms affecting both costs and expenditures, which substantially
change the implications for net additional resources required. Where gov-
ernments did not implement reforms affecting revenue sources, efficiency
gains and expenditure priorities, the requirements for external financing
would often be two or three times as large as those estimated.
This means that the search for a single estimate of the additional
resources required from external sources to achieve primary schooling for
all is probably misplaced. Recent work, for example, shows that annual aid
to education in eight African countries would need to increase by around
2.5 times recent levels, depending upon the assumptions made for
national revenue and expenditure growth over the next fifteen years (Col-
clough et al., forthcoming). Thus the amount of additional aid required to
achieve universal primary education will be substantial, but its precise
dimensions are very sensitive both to the particular countries included in
the calculation, and to the extent of policy reform each of them can be
expected to achieve. The search for agreement on one global estimate
180 Christopher Colclough
may be useful for purposes of advocacy – to raise expectations in the
North about the scale of desirable resource transfers needed to enable the
education targets to be achieved. But all estimates have to be premised
upon a particular amount of policy change occurring. How much will or
can happen for each country is impossible to predict, and we are left with
a financing range, from aid sources, which is uncomfortably large.
The first communiqué from the ‘High Level Group on Education for
All’ – a group of national and agency representatives convened by
UNESCO – reaffirmed that ‘no countries seriously committed to education
for all will be thwarted by a lack of resources’ (UNESCO, 2001). Yet the
experience of aid to education in the first decade since the 1999 Jomtien
Conference on Education, where the ‘Education for All’ initiative was
launched, leaves room for scepticism: the hopes raised by the promises of
increased support made at that time have not been realised. At the end of
the decade, educational aid from all sources amounted to about US$8
billion – roughly in the middle of existing estimates for the additional costs
of achieving primary schooling for all. Thus total aid to education may
need to double in real terms, to focus entirely upon the primary system, to
be refocused upon a sub-set of countries mainly in Africa, and to support
primarily recurrent costs. Notwithstanding recent commitments, these are
strong demands. They occur at the end of a decade when aid to education
from bilateral and multilateral sources actually fell by more than 30 per
cent from 1990 levels. Particularly hard hit was sub-Saharan Africa, where,
for example, World Bank International Development Agency (IDA) loans
to education in 1999 were running at half their real value at the start of the
decade. Suggested reasons include the presence of political turmoil in
some of the larger countries; the weak absorptive capacity of key institu-
tions; the reluctance of some governments to introduce policy reforms;
and too much aid focus having been placed on basic education (which, it
was implied, does not easily absorb large amounts of money) (World Bank,
2001). It is not clear that these factors are quickly disappearing in sub-
Saharan Africa – the region facing the greatest challenges to achieving uni-
versal participation in primary schooling.

Conclusion
The ways in which the Millennium Development Goals for education are
formulated, and the particular criteria established to indicate their suc-
cessful implementation, need to be adjusted. The most obvious problem is
that the gender target for achieving equality of enrolments by 2005 is
simply impossible to achieve. Because gender inequalities in schooling
have roots going well beyond the education sector, the possibilities of
achieving substantial reform, in the many countries affected, so as to
quickly facilitate this objective are remote. At best, most countries will
move towards gender equality only with their achievement of universal
 Towards universal primary education 181
primary education, and the likelihood that the gender parity targets will
be able to be achieved much before 2015 is low.
The quantitative targets for universal enrolment by 2015 are achievable,
but a 99 per cent net enrolment rate is higher than many rich countries
achieve. Targets for completion rates of 99 per cent and for universal liter-
acy amongst 15- to 24-year-olds actually imply the need for universal enrol-
ment at much earlier dates than 2015. These too need to be softened, to
enable realistic interim targets for progress towards primary schooling for
all to be employed.
The agenda for reform, however, remains substantial. National govern-
ments will need to implement new efficiency measures, together with
quality, demand-side and, particularly, gender-focused reforms. A substan-
tially increased spending commitment will also be needed on the part of
many governments. This will need supplementation by clearly targeted
and increased aid from the North. Broad sectoral support will be required,
flexibly delivered on a recurrent basis, and conditional not only upon
stated intention to reform, but also upon actual implementation. The
political economy of these changes is complex, and it is by no means clear
that their net implications will be supportive. Nevertheless, the dimen-
sions of the problem are clearer than in the past, and the mechanisms for
implementation are more articulated than in the past decade. These
factors, at least, present some opportunity for success.

Notes
1 This chapter has benefited from comments from Samer Al-Samarrai and
Howard White. The author is responsible for remaining errors.
2 A further, rather topical reason for ensuring the provision of schooling for all
has recently been suggested by a leading UN staff member: ‘Across the Muslim
World from West Africa to Asia, Islamic schools moved in as the state primary
education system failed. This set in chain a process which resulted in the funda-
mentalist groups replacing governments as the focus for popular loyalty, and
gave the terrorists their support structure’ (Malloch-Brown, 2001). This suggests
the importance of filling the educational space that may otherwise be captured
by fundamentalism.
3 The perceived usefulness of education to individuals is importantly influenced
by the opportunities available to them. Thus whereas in the 1960s the possession
of a primary education still influenced a person’s chances of getting a formal-
sector job – particularly in sub-Saharan Africa – this is no longer the case. It is
secondary and higher levels which now provide that access. Furthermore, many
people may be unaware of the other non-wage private benefits of primary
schooling, mentioned above. For these reasons, the possibility of influencing the
demand for primary education may turn out to be just as important a factor in
achieving the Millennium Development Goals as policies to influence its supply.
These topics are discussed separately later in this chapter.
4 Gross enrolment ratios express total enrolments as a percentage of the number
of children in the eligible age group. Net ratios exclude those pupils from the
numerator who are outside the official school age range.
182 Christopher Colclough
5 For further detail on these and other countries in sub-Saharan Africa, see Col-
clough et al. (forthcoming).
6 This notion of the importance of school quality, though absent from the wording
of the Millennium Development Goals, is present in the six ‘Dakar’ Goals for
achieving ‘Education for All’. There, the second goal explicitly states that ‘by
2015 all children . . . should have access to, and complete, free and compulsory
primary education of good quality’. Moreover, the sixth goal focuses upon
‘improving all aspects of the quality of education and ensuring excellence so
that recognized and measurable learning outcomes are achieved by all’
(UNESCO, 2000b: 8). To judge by these stated intentions, then, quality issues
will not be overlooked by the Education for All movement, notwithstanding
their more summary interpretation by the Millennium Development Goals.
7 Separate evidence demonstrates that this was so in Zambia, Ghana and Guinea
(Colclough et al., forthcoming).
8 Full discussion and evidence can be found in Colclough et al. (forthcoming).
9 Colclough with Lewin estimated the external requirements for primary school-
ing for all as US$2.5 billion in 1990 prices. This is roughly equivalent to US$4.5
billion at 2000 prices.

References
Colclough, C., Rose, P. and Tembon, M. (2000) ‘Gender Inequalities in Primary
Schooling: The Roles of Poverty and Adverse Cultural Practice’, International
Journal of Educational Development, 20: 5–27.
Colclough, C., Al-Samarrai, S., Rose, P. and Tembon, M. (forthcoming) Achieving
Schooling for All in Africa: Costs, Commitment and Gender, Aldershot: Ashgate.
Colclough, C. with Lewin, K. (1993) Educating All the Children: Strategies for Primary
Schooling in the South, Oxford: Clarendon Press.
Delamonica, E., Mehrotra, S. and Vandermoortele, J. (2002) ‘Is EFA Affordable?
Estimating the Minimum Cost of Education for All’, mimeo, New York:
UNICEF.
Kadzamira, E. and Chibwana, M. (2000) Gender and Primary Schooling in Malawi,
Brighton: Institute of Development Studies.
Malloch-Brown, M. (2001) Speech reported in The Independent on Sunday, 11
November.
Organisation for Economic Cooperation and Development (OECD) (1996)
Shaping the 21st Century: The Contribution of Development Cooperation, Paris: DAC.
Peasgood, T., Bendera, S., Abrahams, N. and Kisanga, M. (1997) Gender and
Primary Schooling in Tanzania, Brighton: Institute of Development Studies.
Rose, P. (2002) ‘Cost-Sharing in Malawian Primary Schooling: From the Washing-
ton to the Post-Washington consensus’, DPhil thesis, University of Sussex.
United Nations Educational, Scientific and Cultural Organisation (UNESCO)
(1998) Statistical Yearbook 1998, Paris.
—— (1999) Statistical Yearbook 1999, Paris.
—— (2000a) World Education Report 2000, Paris.
—— (2000b) The Dakar Framework for Action, World Education Forum (Dakar,
Senegal, 26–28 April 2000), Paris.
—— (2001) Communiqué from the first meeting of the High Level Group on
Education for All, Paris, mimeo, October.
Watkins, K. (1999) Education Now: Break the Cycle of Poverty, Oxford: Oxfam Inter-
national.
 Towards universal primary education 183
World Bank(2001) ‘A Chance to Learn: Knowledge and Finance for Education in
Sub-Saharan Africa’, Africa Region Human Development Working Paper Series,
Washington, DC.
—— (2002) ‘Achieving Education for All by 2015: Simulation Results for 47 Low-
Income Countries’, mimeo (processed), Washington, DC: World Bank, Human
Development Network.
9 Promoting gender equality
Ramya Subrahmanian

Introduction
What do the Millennium Development Goals represent for advocates of
gender equality in development? ‘Gender’ features explicitly in the Mil-
lennium Development Goals in two ways: in relation to improvements in
maternal health, and to the elimination of gender disparity in education.
Education is doubly emphasised, as it is also viewed as a vehicle for pro-
moting gender equality and empowerment. Improved educational status,
increased share of women in wage employment in the non-agricultural
sector, and the proportion of seats held by women in national parliaments
are represented as indicators of the increased gender equality and empow-
erment of women. The goals of poverty and hunger eradication and
reduced child mortality, however, are framed with no specific reference to
gender, although some agencies report gender-disaggregated indicators.
That gender equality and the empowerment of women figure as a spe-
cific goal in the Millennium Development Goals can be celebrated as sym-
bolic of the significant impact of feminist advocacy over many years in
making the case for gender-aware development. However, at the same
time, much cautionary evidence exists of the often rhetorical value of
statements on gender: language has often changed without significant
commensurate changes in institutional practices or financial commit-
ments, and policies targeted at redistributing opportunities to women
have not expanded to the extent necessary to address gender discrimina-
tion in development. The reduction of a form of differentiation and
inequality that cuts across sectors to one goal in a list is problematic for
the conceptual purist. It suggests a false and misleading separation
between ‘gender’, on the one hand, and development processes and out-
comes in a range of arenas on the other, when in fact the two intersect to
perpetuate patterns of inequality between women and men across the
range of development goals. To the pragmatist, however, the specification
of gender equality in any global statement represents a step forward,
further imprinting gender equality as a non-negotiable component of
development. It offers more, and perhaps new, spaces to continue
 Promoting gender equality 185
manoeuvring for greater financial and programme attention to gender
equitable outcomes, and serves as a lever for demanding greater account-
ability from governments and agencies.
This chapter seeks to locate an analysis of gender in the Millennium
Development Goals within a broader assessment of the use of such targets
in international development. Current literature on the Millennium
Development Goals indicates that there is much awareness of their limita-
tions, stemming partly from the fact that these goals crystallise the com-
mitments of donor agencies, and thus serve primarily as shorthand for
inter- and intra-donor communications. As a pithy summary of the chief
concerns facing the development agenda, the Millennium Development
Goals signal two significant features of the new consensus: the acceptance
of a multidimensional approach to poverty reduction (White, 2001), and
the importance of basic human well-being and good-quality basic social
services as a non-negotiable goal of development intervention.1
Some cynicism arises from the fact that targets have been set before,
but never met (White, 2001). Goals and targets flag destinations or the
finishing line, but not the route. They emphasise outcomes, but yield no
insight into the processes that are necessary to ensure that these outcomes
are achieved. The Millennium Development Goals are a selective prioriti-
sation of some goals over others. In some cases they specify inputs and
outputs; in others they tell us about desired outcomes (McGee, 2000).
They sketch assumed pathways and causalities between ends (desired out-
comes) and means (inputs and outputs) but are not helpful in illumin-
ating the complexities and multiple dimensions of lived poverty that often
obstruct efforts to reduce poverty. These limitations can be viewed as
benign, outweighed by the importance of having clearly defined goals gov-
erning international development co-operation, or they can be viewed as
significant, representing a weakly developed agenda with the associated
danger of deflecting attention from promoting the types of pro-poor
reforms that are really necessary to make a substantial dent in world
poverty.
However, the Millennium Development Goals may be somewhat differ-
ent from other ‘target’-based approaches. Although it is quite obvious
already that most of the Millennium Development Goals are unlikely to be
reached within their specified time-frames (Hanmer and Naschold, 2000;
Hanmer and Wilmshurst, 2000; McGee, 2000; White, 2001; and Chapter 6
of this volume), they provide a strategic talking-point for assessing what
the barriers to the achievement of goals are, and provide a tool with which
to hold both donor agencies and governments accountable (White, 2001,
2002).
Millennium Development Goals also need to be discussed within the
context of enhanced efforts towards increasing donor collaboration. As is
argued in Chapter 3 of this volume, the UK’s Department for Inter-
national Development (DFID) has internalised these goals to the extent
186 Ramya Subrahmanian
that organisational restructuring is taking place around the attainment of
the goals. New policy instruments such as Poverty Reduction Strategy
Papers (PRSPs) and sector-wide approaches also represent efforts to pool
donor resources for which a prior agreed common framework of goals is
essential to avoid the fragmentation of earlier project-based management
approaches (Mehrotra, 2001; White, 2002). Again, however, these approaches
and instruments have strengths and limitations that need to be assessed
carefully. While on the surface these priorities may be viewed as inter-
national commitments of both donors and governments, arising as they do
from a range of conferences and meetings, there is the danger that they
will be imposed as a fixed set of priorities on national governments as part
of new aid agreement processes and frameworks. If the Millennium Devel-
opment Goals dictate in a standardised fashion what aid is spent on,
without contextualised priorities negotiated at national and sub-national
level, then they provide an insufficient analytical map for enabling appro-
priate gender-aware means to be identified for poverty reduction. Thus
means and ends, and goals and institutional processes need to be looked
at as intersecting and hence inseparable for the purposes of assessing the
usefulness of the Millennium Development Goals in development co-
operation.
There are many levels on which the Millennium Development Goals
can be assessed from a gender perspective: the political value of such inter-
national statements; the substantive value of the ways in which goals and
targets have been set and identified; the institutional contexts within which
they are likely to be implemented; and the wider social, political and eco-
nomic contexts within which these identified global problems arise and are
perpetuated. This is a huge terrain for assessment, but this chapter will
touch on aspects of all these levels. The approach followed is to focus on
the current context within which these goals are being promoted within
development co-operation.
A central concern is whether the way in which gender equality is articu-
lated in the Millennium Development Goals is sufficient to ensure transla-
tion into concrete strategies that benefit women and address discriminatory
barriers to greater equality in the distribution of benefits and outcomes
across the range of goals. Gender awareness and gender equality are both
important means and ends in development. Specifying gender equality as
an end is important, but cannot be achieved without ensuring that the
means used in development are gender aware and address equity and
equality considerations at every step of the way. Thus the conceptual
frame of gender within the Millennium Development Goals provides an
entry point for raising questions about what kinds of interventions are best
placed to achieve these rather ambitious and broad goals. The next
section lays out the ways in which gender equality is articulated and
defined in the goals. The adequacy of these goals is reviewed in the light
of recent evidence of the challenges that remain in securing gender
 Promoting gender equality 187
equality. The analytical sufficiency of the target approach in addressing
the multidimensional characteristics of poverty and the social complexi-
ties of gender relations is assessed against the backdrop of broad trends in
achieving progress towards gender equality. In the subsequent section,
attention is focused on the changes in aid management approaches and
practices that are aimed at addressing the limitations of earlier project-
based funding. The prospects for integrating gender into these new
approaches need to be especially assessed in the light of the considerable
experience of what we may call first-generation approaches to gender
mainstreaming.

Gender in the Millennium Development Goals: an

assessment

The substance of gender

The presence of gender within the Millennium Development Goals reflects
the impact of long-standing efforts to mainstream gender into develop-
ment thinking and planning. Feminist advocacy efforts have long ‘ques-
tioned the presumed gender-neutrality of formal institutions’ (Razavi and
Miller, 1995: 2) and have fought for the inclusion of gender equality on
grounds of justice and merit, not just need (Kabeer, 1999b). The shift
towards justice as a primary rationale is yet to take root, but the arguments
against viewing women merely as instruments of development have been
made persistently, and have succeeded in some measure in transforming
the ways in which gender is now promoted by development agencies. Early
integrationist approaches had tended to ‘add’ women to development
projects in sex-specific ways,2 not questioning the division of labour and
resources that systematically keep women in a position of relative depen-
dence upon men in a wide range of aspects of life. The influence of
efforts to ‘mainstream’ gender into the mandates, structures, routine prac-
tices and procedures of development organisations is somewhat evident in
the Millennium Development Goals: clearly it is now acceptable for donor
agencies to commit themselves to gender equality and women’s well-being
as goals with intrinsic value for achieving development.3 This marks a
transition, in rhetoric at least, from the earlier focus on investing in
women solely in terms of the returns it would bring to development –
gender equality appears as both means and ends of development in the
Millennium Development Goals, albeit in very particular and rather limited
ways. As Kabeer notes, the persuasiveness of claims that investing in
women has policy pay-offs has made major agencies, most notably the
World Bank, unlikely champions of women’s empowerment (2001: 17).
However, as Jackson (1998) reminds us, gender has been assimilated into
development thinking in particular ways, and the assimilation has not
always reflected the richness and diversity of gender analysis.
188 Ramya Subrahmanian
‘Gender’ enters the Millennium Development Goals in two ways: goals
that are gender aware, and those that are gender specific (see Table 9.1).4
It is important to distinguish between these as ways of assessing how
gender is addressed, both to pinpoint the analysis that underpins gender
within the Millennium Development Goals, and to indicate the potential
for transforming unequal gender relations within efforts aimed at achiev-
ing the goals. The goal of universal primary education is gender aware to
the extent that it explicitly mentions both boys and girls, signalling that
closing the gender gap is a key means to achieving full universal primary
education. In addition, the Millennium Development Goals include goals
that are gender specific – that is, they specifically promote the interests of
women, in ways that transform unequal access to political and economic
opportunities. The choice of access to non-agricultural wage employment,
and representation in national parliaments, as indicators of women’s
empowerment and enhanced gender equality appears, at one level, a
relatively random selection out of the range of possible indicators for mea-
suring equality and empowerment. However, their inclusion in this list
merits closer attention. A positive interpretation suggests the recognition
of the importance of promoting non-traditional opportunities for women
that challenge ideologies that restrict women to traditional, often sex-
specific, roles in society. In the case of women’s economic participation,
this is indeed a crucial indicator signalling the extent to which women
have equal opportunities with men, and can move out of labour-intensive
and often unpaid forms of employment. The improved representation of
women in national parliaments, on the other hand, can be seen as a truly
international choice of indicator, as the representation of women in par-
liaments is relatively low world-wide, and does not follow any of the usual
divides: North/South; developed/developing.5 The choice of maternal
health is perhaps significant too; it suggests an intrinsic value placed on
women’s health (albeit the health of women in their maternal role), as
opposed to merely seeing them as the means for improved child welfare.6
However, a more realistic interpretation that suggests itself is also that the

Table 9.1 How gender enters the Millennium Development Goals

Gender-aware goals Gender-specific goals

Achieve universal primary education, Promote gender equality and empower

for boys and girls alike women (through education, increased
labour force participation of women in
non-agricultural sectors; improved
representation of women in national
parliament)
Improved maternal health

Source: Adapted from Kabeer (1999b).

 Promoting gender equality 189
emphasis on quantifiable indicators in the Millennium Development Goals
limits the choice of indicators to those that are measurable.
To translate goals into results, however, requires more than the adop-
tion of the right language. The content of agencies’ understanding of
development issues remains critical (Jackson, 1998). While the term
‘empowerment’ may embody the most noble results for women that can
be desired, debates on how to promote this outcome have emphasised the
inextricable connections between a wide range of resources and entitle-
ments and the importance of processes to enable women to exert agency
and exercise voice.7 Kabeer (1999a, 2001) has been influential in pointing
out the limitations of efforts to define empowerment in ways that are
amenable to measurement and quantification. While it has clearly become
desirable for agencies to claim that they are ‘empowering’ women, efforts
to show results in this regard have led to methodologically flawed and
reductionist approaches to both defining what empowerment means and
claiming success.
A notable, and disappointing, feature of the Millennium Development
Goals is the gender blindness of the goal of eradicating income poverty in
particular. The poverty goal aims at reducing the poverty and hunger of
populations at large without either specifying the target group or indicat-
ing that these phenomena may be differentially experienced by men and
women across ages. The goals and indicators are framed in aggregate
terms, neither necessitating nor reflecting gender-disaggregated analysis.
However, as much of the literature on gender and poverty has discussed,
poverty remains a highly gendered phenomenon, and in ways that are not
captured by income or head-count measures. As Cagatay (1998: 3) notes,
the concern with gender in poverty throws up three possible hypotheses,
not necessarily mutually exclusive: the suggestion that women have a higher
incidence of poverty than men; and/or that the incidence of poverty among
women is increasing compared to men; and/or that women’s poverty is
more severe than that of men. These indicate that poverty can be assessed
in terms of individual capabilities8 as well as outcomes that are fostered
through gender relations, and played out through processes and practices
of households and other institutions. In other words, outcomes of poverty
are embedded in processes and relations, and attempts to measure or
address them must do the same.
The additional value of separating out these three elements is to point
to the importance of evidence-based analysis in identifying whether women
are indeed poorer than men, as the term ‘feminisation of poverty’, used
widely as shorthand for discussion of gender and poverty issues, seems to
imply. As Jackson (1998) argues, development agencies have been keen to
collapse gender concerns within the wider category of poverty, as doing so
enables the use of a fairly depoliticised and needs-based discourse as
requiring focus on women within poor households, rather than gender dis-
advantage per se. However, analysis of the intersections between gender and
190 Ramya Subrahmanian
poverty remains challenging, as they do not conform to broad generalisa-
tions.9 At one extreme, measures of poverty focused on the household as a
unit have obscured the intensified and differential experience of poverty
by women within poor households (Kabeer, 1997); at the other extreme,
awareness of gender asymmetries in access to resources has been expanded
into the generalisation that all households headed by women must inevitably
be poorer than those headed by men (Jackson, 1998; Whitehead and Lock-
wood, 1999). Seeking simple conflations of women with poverty, either
through assuming they are affected as much as are their male kin by
processes of impoverishment, or through assuming they are always more
negatively affected, is unsatisfying and analytically potentially misleading.
The variations in the correlations between female headship and poverty
within and between countries indicate the need for nuanced, context-spe-
cific research.10 Feminist scholars argue that gender ideologies permeate
structures of economic and social reproduction and hence are an intrinsic
feature of processes of impoverishment; this in turn will differentially
impact the ways in which women and men experience poverty, and may or
may not increase the likelihood of individual women being poorer than
individual men. Thus while gender disadvantage may not determine eco-
nomic status, poverty often intensifies gender disadvantage (World Bank,
2001a). Further, generalised statements about gender and poverty serve to
demonstrate the rhetoric of commitment to gender, while at the same time
obscuring from policy attention the continued forms of disadvantage that
women in non-poor households face (Cagatay, 1998).
The above discussion seeks not to diminish the importance of focusing
on poverty reduction in the Millennium Development Goals, but to
emphasise the importance of including gender awareness within the
framing of this particular goal. Gender-aware poverty reduction entails
not just counting how many women are income-poor relative to men, but
taking account of the multidimensionality of poverty and the embedded-
ness of poverty in gender relations. Feminist advocates have significantly
influenced emphasis on the multiple dimensions of poverty and made
forceful arguments against relying purely on income and head-count
measures. They have long argued for approaches that recognise that
poverty is reproduced through inequitable distribution practices within
the family and other institutional domains, and have pointed out consis-
tently that social norms and customs underpin economic processes. Yet
approaches that are developed continue to fall short of the complexity
required to understand gender in its relational sense, rather than as a
characteristic of individuals. Despite appearing to build on a multidimen-
sional approach to poverty, the Millennium Development Goals fall short
in terms of mapping out a sufficiently coherent approach to poverty
reduction. Again here, the emphasis on measurable indicators is likely to
continue to limit the possibilities for the Goals to deal more comprehen-
sively with complexities. The danger is, however, that these limitations will
 Promoting gender equality 191
spill over into interventions that cite the Millennium Development Goals
as their guiding light.
Elson (1998) articulates this problem in relation to the introduction of
gender into the basic growth model, pointing to the tendency to conflate
gender analysis purely with gender disaggregation. While gender disaggre-
gation is important, for us to get a sense of relative deprivations and differ-
ences, analysis needs to go far beyond:

The basic problem with disaggregation is that it focuses on the separ-

ate characteristics of men or women (whether individuals or groups)
rather than the social institutions of gender as a power relation. . . .
The danger is that it does not draw sufficient attention to the recipro-
cal determination of the characteristics of women and men as eco-
nomic agents. . . . It would be useful to consider strategies which pose
the issue of economy itself as a gendered structure, rather than as a
gender-neutral structure within which men and women undertake dif-
ferent activities.
(Elson, 1998: 160)

Further, gender disaggregation may highlight divisions of roles, but remain

uninsightful in respect of the variations in configurations of power, author-
ity and entitlements that arise in different geographical and socio-cultural
contexts. The danger of standardisation of analysis of the factors that give
rise to gender inequalities must be avoided, and it can be done only if
management systems are responsive and open to learning and absorbing
lessons about what causes change. However, the development of guide-
lines, checklists and measures seems to indicate a tendency towards stan-
dardisation and reduction, rather than encouragement of analytical
diversity and expansion of categories.
Thus while gender equality and women’s empowerment find room in the
Millennium Development Goals, it is striking that the means to achieving
these are identified specifically in the area of maternal mortality and educa-
tion. This indicates that the key pathways that donors are pursuing towards
gender equality and empowerment of women are in relation to health and
education. In some senses this is a hangover from the past – a continuation
of the association of women with social sectors, and with issues of reproduc-
tion.11 This bias is not just a feature of the Millennium Development Goals;
Rodenberg (2002: 2) notes that the World Bank’s comparative evaluation
study of PRSPs discusses gender aspects only in relation to ‘the classic
sectors of education and health, while the study sees no relevance for these
aspects in central sectors like agriculture, the environment, transportation
and urban development’. The World Bank evaluation report (2002) notes
that the Bank’s assistance for poor women in economic development12 has
been minimal, restricted to one-quarter of projects outside health and edu-
cation. Social funds were also found to have lacked a gender strategy.
192 Ramya Subrahmanian
One measure of change, however, is the extent to which the shift from
instrumental rationales to recognition of the intrinsic worth of gender
equality as benefiting both men and women has taken place. The brief
overview of the way in which gender is framed within the Millennium
Development Goals indicates some movement away from instrumentalist
goals, but also points to a persistent fallacy or error that is made in rela-
tion to the rationales and methods through which gender is addressed in
development. The shift from ‘women in development’ to ‘gender and
development’ was fought for precisely on the grounds that being con-
cerned about ‘gender’ was not just about recognising the disadvantages
that women faced relative to men in gaining access to resources, benefit-
ing from investments and allocations of resources, and participating equi-
tably, but also about recognising the ways in which gendered ideologies
shaped the ways in which a range of institutions reproduce inequalities
and asymmetries. While the rhetoric of ‘engendering development’ has
spread widely within development institutions, whether the practice has
remains to be seen, and is an issue that will be taken up in greater detail
on p. 198.

Evaluating progress
Other chapters in this book provide assessments of the progress made in
achieving the specific targets relating to education, maternal mortality
and reproductive health (see Chapters 8 and 11). In this sub-section the
focus will be on providing the wider context of progress in achieving
gender equality. The selective focus of the Millennium Development
Goals contrasts sharply with the wider agenda elaborated at the Fourth
World Conference on Women in Beijing in 1995, which was influenced by
the advocacy of women’s movements around the world. The latter con-
tains twelve areas of action13 cutting across all aspects of development. It is
this broader agenda that merits reporting on, as an indicator of the areas
in which change has taken place, and where the status quo remains the
most entrenched.
The question of ‘progress’ depends very much on what it is that is being
valued. For example, the human capabilities approach, outlined in the
work of Amartya Sen, but most notably in relation to gender inequalities in
the work of Martha Nussbaum, sets out an approach towards identifying
the universal basic social minimum that allows for ‘a life that is worthy of
the dignity of the human being’, arrived at by focusing on central human
capabilities – that is, ‘what people are actually able to do and to be’ (Nuss-
baum, 2000: 5). Such an approach entails evaluating the opportunities and
liberties available to people in the widest sense of the terms.
A brief ‘report card’ is presented in Table 9.2, drawn from the Humana
World Human Rights Guides of 1986 and 1992, which construct indicators
of changes in gender equality in terms of three sets of interlinked arenas:
Table 9.2 Regional achievements in gender equality

Region Political and legal rights Social and economic rights Rights in marriage and divorce
proceedings

Ranking Relative change Ranking Relative change Ranking Relative change

1985 in 1990 over 1985 in 1990 over 1985 in 1990 over
1985 1985 1985

OECD 1 No change 2 No change 1 No change

Europe and Central Asia 2 Decline 1 Decline 2 Decline
South Asia 3 No change 7 Slight improvement 5 Slight improvement
Latin America and the Caribbean 4 Improvement 4 Slight improvement 4 Substantial
improvement
Middle East and North Africa 5 Slight decline 5 Slight decline 7 Improvement
Sub-Saharan Africa 6 Improvement 6 No change 6 Improvement
East Asia and the Pacific 7 Slight improvement 3 Slight improvement 3 No change

Source: Adapted from Humana Index as presented in World Bank (2001a).

194 Ramya Subrahmanian
political and legal equality; social and economic equality; and equality of
men and women in marriage and divorce proceedings.14 Although there
are many serious limitations associated with an index of this kind15 (see
Elson, forthcoming), the World Bank (2001a) report uses the data to
show changing patterns over time, thus providing a synoptic view of the
pressing challenges remaining in achieving gender equality. Here too it is
being used as a device to raise a series of issues about the correlations
between different kinds of freedoms and the types of achievement on
which the Millennium Development Goals are focused. Importantly, gender
equality is defined here in terms of the political and social, and in terms of
civil liberties: the right to choose to marry and the right to leave the insti-
tution of marriage. As the report highlights, asymmetries in rights and
privileges persist in many societies, fundamentally constraining choices
available to women and ‘often profoundly limiting the opportunities they
have in the economy and in society’ (World Bank, 2001a: 37).
In Table 9.2 I have assimilated the data from World Bank (2001a) to
demonstrate the relative performance of different countries in each of the
areas of concern. What the figures indicate broadly is that in general,
progress has been made in women’s rights in most regions of the world,
although the scores achieved by different regions in these different
dimensions shows that ‘in no region of the developing world do women
have equal rights with men in any of these dimensions’ (World Bank,
2001a: 37). Three points in particular from the analysis seem striking, with
implications for our discussion of the Millennium Development Goals,
and are elaborated on in the World Bank report. First, gains made in one
dimension are not necessarily correlated with gains in others. East Asia
and the Pacific demonstrate a higher achievement in gender equality
in socio-economic rights, and the provision of rights in marriage and
divorce, but not in political and legal rights. Second, gains are also
reversible. The case of Europe and Central Asia, where slight declines are
noted across all types of rights, is particularly striking. These declines are
noted in the context particularly of East European political and economic
transition, where many of the earlier securities provided for women have
been withdrawn. This cautions us against taking formal or de jure rights as
sufficient indicators of de facto enjoyment of rights. Macroeconomic envi-
ronments have a range of impacts on people’s ability to exercise choices
and claim rights, with potentially negative implications for impoverished
and socially excluded groups. Negative impacts under neo-liberal policy
environments on social development in particular have been well docu-
mented (see Molyneux and Razavi, 2002). Third, some indicators do not
necessarily reveal the quality of rights enjoyed: for instance, even though
women’s participation in the labour force has increased, particularly in
manufacturing and services and particularly in countries which have reori-
ented their economies for participation in the world market, wage gaps
between women and men persist. Further, such indicators can be mislead-
 Promoting gender equality 195
ing where the formal sector is only a small share of the economy, with the
bulk of employment opportunities, and women workers, in the informal
sector.

Pathways and causalities

Despite its level of generality, the data from the Humana Index show the
vulnerability of gains for women in the quest for gender equality, and how
changes in any one of the four dimensions – political, economic, social
and cultural – can have repercussions for the others. This illustrates the
difficulties entailed in attempting to chalk out very specific pathways as
measures of progress towards gender equality. It also illustrates the diffi-
culty of trying to view broader processes of social change in terms of
selected indicators and targets, even if these chosen markers are merely
intended as beacons along the route.
Education is a clear example of these difficulties. Female education is
widely believed to constitute an indicator of progress towards decreasing
gender inequality. Whilst this may well be the case, as any reversal of
restrictions placed on access to human resources can only bring positive
change, claims made for the multiple positive externalities associated with
education often have a hollow ring to them, especially when viewed from
the vantage point of the conditions under which women participate in the
economy, in society and in politics. The ‘silver bullet’ (Jeffery and Jeffery,
1998) approach to female education has been criticised for the burden of
expectations placed on this one resource in resolving a range of goals
relating to population and health, amongst others, but also because of the
lack of real interest that it suggests in testing the implicit assumption that
education translates into enhanced autonomy for women. The laziness of
this assumption is exposed by research which focuses on what Jeffery and
Jeffery (1998) refer to as the ‘education for what?’ question, which shows
that in some contexts education serves as a means to entrench, not chal-
lenge, women’s unequal position within conventional marriage and
labour markets, and indeed may or may not challenge structures of inequal-
ity (Subrahmanian, forthcoming). Further, as Whitehead and Lockwood
note, there is the danger of ‘spurious’ causality drawn between education
and household income, when in fact, ‘both . . . may be affected by under-
lying patterns of wealth organized through families’ (1999: 548).
The point is certainly not to suggest a counter-argument that investing
in women’s education is bad for women, but to highlight that getting
females into school may not necessarily improve the quality of their lives,
if the basis for evaluation is drawing on the notion of ‘capabilities’,
referred to earlier (Subrahmanian, 2002). Biases against girls within the
classroom and within the curriculum, and limited economic and political
opportunities outside the school, all conspire often to reduce the trans-
formative potential of education in women’s lives. Instrumentalist
196 Ramya Subrahmanian
arguments which suggest that reducing gender inequality in education
means enhanced autonomy for women often do a disservice to efforts to
challenge the gender biases within education systems, by making it all
appear far simpler to achieve than it really is. They also perpetuate
reliance on quantitative measures and statistical correlations to justify
funding (Baden and Green, 1994). This is dangerous, particularly given
the methodological flaws of many of the techniques used to assess returns
to education in general, and female education in particular (Bennell,
1996; see Subrahmanian, 2001, for a discussion).16
The other indicators of decreasing gender inequalities also need unpack-
ing to avoid making sweeping generalisations about their relationship with
women’s relative lack of participation and access to resources. Changes in
women’s representation in national parliaments may mean a range of
things, not all necessarily positively correlated with women’s empower-
ment. If the process by which women get to parliament does not reflect
any agency on the part of women to contest and represent their chosen
constituencies,17 then it will not tell us much. Alternatively, it may reflect
reductions in inequality between certain classes of women and men,
rather than representing a gain for all women. Similarly, women’s partici-
pation in non-agricultural employment will have limited meaning as an
indicator if applied to agrarian societies, which tend to be characteristic of
poor countries.
Statistical correlations often become the basis for expanded statements
of explanation about complex social decisions and processes. Thus corre-
lations between greater levels of education and smaller family sizes are
interpreted as indications of changes in the desires of women, as opposed
to exploring the possibility of other changes in the policy environment or
in incentives for changing family size. Policy efforts aimed at redressing
gender inequalities faced by women, and many theoretical approaches to
analysing gender inequality, are littered with assumptions about causality
that research tends to disprove (see Jeffery and Jeffery, 1998; Pearson,
1998). These include the assumptions that there is a direct causal connec-
tion between, inter alia, women’s wages and empowerment; and between
female education and increased autonomy. These in turn indicate that
social change is conceived of in these narrow policy discourses in terms of
changes made in the characteristics of individuals, rather than its being
viewed as combinations of change within individuals as well as the wider
societies in which they live.18 The ways in which individuals make trade-
offs between potential individual gain and the well-being and social coher-
ence of the wider societies to which they belong, however defined, suggest
that processes of social change are mutually constituted by changes in
individual circumstances and changes in the norms and ideologies of their
chosen or imposed groups of association. Importantly, they also suggest
that these processes of change may be read off shifts in a range of vari-
ables, and also at various moments of time.
 Promoting gender equality 197
The challenge of indicators: no pain, no gain
If social change in relation to eliminating inequality and empowering
people is a value-based process, subjective and defined differently by dif-
ferent development actors, then how can indicators be identified that
reflect this complexity, whilst at the same time setting benchmarks for
evaluating the extent to which desired change is taking place? The exer-
cise of finding indicators to match such broad yet important goals seems
doomed at the outset. Certainly, top-down indicators find no place within
such a conceptual terrain. Yet efforts to produce indicators have long con-
sumed efforts towards mainstreaming gender, part of the arsenal of methods
aimed at persuading gender-blind development actors to grasp the
importance of promoting change from a gender perspective.
Kabeer’s (1999a, 2001) schema for measuring interlocking dimensions
of change that transform women’s capabilities provides an extremely
useful way of thinking about indicators of change that are sensitive to this
complexity. An essential component of any attempt to measure
empowerment is a definition that frames the scope of change that is associ-
ated with the concept and the measures. Kabeer’s definition of empower-
ment unpacks the central notion of ‘power’ in terms of the ‘ability to make
choices: to be disempowered, therefore implies to be denied choice’ (2001:
18): ‘Empowerment thus refers to the expansion in people’s ability to make
strategic life choices in a context where this ability was previously denied
to them’ (ibid.: 19). Thus, measuring empowerment means measuring the
extent to which people are able to make new choices. Choice in turn has
three inter-related dimensions:

• resources or the conditions under which choices are made;

• agency or ‘the ability to define one’s goals and act upon them’; and
• achievements – or the outcomes of choices.

Thus Kabeer argues for measures of change that grasp the conditions, the
processes and the outcomes of choices made.
Such a schema, if applied to the case of education, discussed on p. 195,
would help locate education as one resource amongst many with the poten-
tial to transform women’s lives. Thus access to this resource would fulfil
only one condition of the indicator before claims could be made about its
empowering effects; the other conditions would be the extent to which
access helped strengthen women’s agency, and whether women’s agency
in addition to access to the resource would actually lead to beneficial out-
comes in terms of women’s capabilities. With such an approach, imputed
pathways are automatically subject to scrutiny to verify how access to edu-
cation enables women’s ‘empowerment’.
Resources and achievements are generally thought to be easier to
measure, but all depends on what achievements are being valued. It is not
the fact of change that is being sought to be measured, but the nature of
198 Ramya Subrahmanian
the change. Many definitions of empowerment are unclear or opaque in
their specification of what it is precisely that is being sought to be changed,
and whose perspective is being reflected. Further, as Kabeer notes, it is the
process by which resources translate into achievements that is central to
establish before claims can be made about ‘empowerment’. Any new
resource will filter through existing webs of gender relations, but the
potential to transform these existing gender relations will depend to a
great extent on whether an enabling environment is created for women to
exercise some agency, defined as their ability to determine what choices to
make.
But agencies need performance measures, and that essentially is what
the Millennium Development Goals represent to international develop-
ment agencies (White, 2001). A broad analytical approach as suggested by
Kabeer poses challenges both for methods that can be used and for attri-
bution, which are clearly at odds with the purposes for which quantifiable
measures are developed. Attribution remains a problem with all indic-
ators, particularly those that operate at the level of aggregation of the Mil-
lennium Development Goal indicators, as discussed in Chapter 3.
However, if agencies genuinely wish to occupy the high ground of women’s
empowerment, then they have no choice but to embrace complexity and
show conceptual sophistication in their discourses and approaches to
gender equality.

Mainstreaming gender in development: lessons learned

There is a vast literature documenting the difficulties development agen-
cies find in translating commitment to gender equality into reality (Goetz,
2003; Razavi and Miller, 1995). Many agencies have embraced the import-
ance of tackling gender inequality, but the record of efforts to institution-
alise such change remains uneven. The World Bank’s Policy Research
Report Engendering Development (2001a) is an example of a concerted attempt
to take a multidimensional approach to gender inequality. However, the
contrast with other reports on the extent to which gender issues are integ-
rated into guidelines, assessments, projects and other mechanisms of
policy design and delivery indicates the wide gulf that remains between
discourse and practice. Institutional issues still remain critical determin-
ants of the achievement of goals.
The Millennium Development Goals represent one attempt to change
donor approaches and practices in the disbursement and delivery of aid,
and need to be understood in the context of accompanying changes in
the instruments of aid and policy delivery. The shift to budgetary support
in the form of sector-wide approaches in key poverty-relevant sectors such
as agriculture, health and education, and to Poverty Reduction Strategies
for heavily indebted countries, provides the institutional context in which
resources for the achievement of goals are allocated. Thus efforts to assess
 Promoting gender equality 199
the achievability of goals with reference to gender in particular need to be
located within an assessment of the success of ‘mainstreaming’ efforts in
general, both within donor agencies and within national governments,
and also, crucially, at the intersection of the two, with particular reference
to the processes of building national ownership of donor-initiated policy
approaches.
Preliminary assessments of the success of mainstreaming gender into
the new policy instruments indicate that many of the old methods of
‘mainstreaming’ need to be recast to meet new challenges (Standing,
2001). The focus on budgetary support (rather than projects) has put
macroeconomic frameworks at centre stage of efforts to tackle poverty and
inequality. Thus skills in inserting gender into these budgetary and macro-
economic frameworks have become paramount,19 and gender advocacy
has now to refocus its efforts on discussion and negotiation with enor-
mously powerful and historically gender-blind ministries of finance and
planning (Rodenberg, 2002).
A review of what we may now call ‘first-generation’ approaches to main-
streaming gender within development institutions reveals a wide range of
lessons about the political and institutional challenges that need to be sur-
mounted for meaningful progress on gender equality to be achieved.
These first-generation mainstreaming approaches focused on the bureau-
cratic and technocratic dimensions of agenda management, and included the
establishment of special desks, officers and posts relating to gender exper-
tise which were to be the technical and political nerve centres within
development agencies, whether donor, government or non-government
organisations (NGOs). Assessments of the performance and achievement
of these forms of bureaucratised gender mainstreaming approaches throw
up mixed results – in some cases, they have succeeded in creating spaces
for women’s advocacy to infiltrate the state (Goetz, 2003), but in others,
they have served to limit the range of institutional actors involved in the
‘negotiating frame’ (Standing, 2001). The substance of the analysis of
these approaches has been focused, however, on the mismatch between
the resources provided to these bodies and the giant tasks of gender redis-
tribution and transformation that they were ostensibly created to promote.
In particular, Razavi and Miller (1995) note that understaffed and under-
resourced units of women in development/gender and development staff
were charged with tasks that actually cut across all departments of organi-
sations, including advocacy, technical support in the form of staff training
and developing guidelines and checklists, as well as oversight and moni-
toring functions, leaving little time for research, policy and strategy devel-
opment.20 The task of mainstreaming became displaced onto units that
did not have the resources or the power to influence agendas, and pre-
vented responsibility and accountability on gender issues from becoming
part of the mandate of every department and all personnel, as originally
intended.
200 Ramya Subrahmanian
What prospects, then, for the second generation of mainstreaming
approaches, which now focus on the more political and financial dimen-
sions of agenda-setting in development? A desk review of nineteen Interim
Poverty Reduction Strategy Papers (I-PRSPs) and four full PRSPs con-
ducted by the World Bank (2001b: 3) reveals that relatively little attention
is paid to gender issues in any of their core elements or sectors. The
report qualifies this further by noting that all the documents reviewed had
failed to meet the rather low standard that was being applied in the
review: ‘making passing reference to gender in diagnosis and indicating
vague intention to include men and women in their actions and indicators
in each of the eight sectoral areas’ (ibid.: 3). Although there are regional
variations within this overall poor performance on gender, the report
firmly concludes that ‘the low scores . . . do not reflect a lack of sophistica-
tion. Rather they reflect a failure to address gender issues, even with
simple approaches’ (ibid.: 5). Within the overall low attention to gender,
it was found that health and education sector discussions dealt with
gender issues with a greater degree of elaboration than the other ‘non-
traditional’ sectors such as labour markets, infrastructure, governance and
safety nets. It is widely noted that the addressing of gender concerns
within education, health, nutrition, population and social protection is far
easier to promote21 – in fact, female education is the one policy issue on
which actors at different levels of the World Bank agree (Whitehead and
Lockwood, 1999).
The low attention paid to gender in PRSPs is not just a matter of poor
government capacity or will. The World Bank (2001b) review examined
the Joint Bank and Fund Staff Assessments, which are meant to assess the
strengths and weaknesses of the PRSPs with a view both to strengthening
them and also to determining the levels of funding assistance that will
then flow. It found that fewer than one-quarter of Joint Staff Assessments
recommended further steps relating to gender, and when they did, the
nature of the comments was vague and general, and did not pick up on
the glaring omissions such as the failure to consult or consider consulting
women or women’s organisations (p. 8). However, there were also signific-
ant differences in capacity and commitment within the Bank. The review
found that the weakness of the Joint Staff Assessments focus on gender
contrasted sharply with the comments of the World Bank’s Executive
Board on the full PRSPs, which identified gender as an issue requiring
greater attention and provided specific comments on each of the four full
PRSPs.
In their review of six World Bank Poverty Assessments from four sub-
Saharan African countries (Ghana, Zambia, Tanzania and Uganda), White-
head and Lockwood (1999) note a range of limitations in gender analysis,
including variations in the way in which the language of ‘gender and
development’ is used and the different methodologies used to measure
and define poverty (see also World Bank, 2002). Their analysis and find-
 Promoting gender equality 201
ings are not dissimilar to the findings of the review of PRSPs cited above.
However, and importantly, they link the poverty of the gender analysis to
overall weaknesses in the commitment and capacities of the World Bank
with respect to gender issues.22 By perpetuating the long-institutionalised
association of women with ‘human resources’ alone, the Bank’s central
mandate and concern with economic growth issues have remained worry-
ingly ungendered. The diversity of methodologies and approaches used
could partly be explained by the weak operational guidelines prepared by
the Bank in helping national teams develop their assessments (Whitehead
and Lockwood, 1999).
A further element is the fit, or the perceived lack of it, between the
issues raised by gender analysis, on the one hand, and the policy mandates
within which these analyses are commissioned, on the other, revealing the
selectivity and partial nature of policy recommendations. This is a crucial
aspect explaining the analysis–policy gap. Gender analyses, where done
satisfactorily, are contained within reports as annexes, and are not integ-
rated into the overall analysis (World Bank, 2002). Whitehead and Lock-
wood’s review of the Poverty Assessments found that much of the analysis
was diluted or not taken into account in the drafting of the policy sec-
tions, particularly analysis drawn from the Participatory Poverty Assess-
ments (PPAs), which record the voices of poor women and men. They
note that despite the variety of methods and approaches employed in the
different Assessments,

there is a remarkable consistency of views expressed on how to reduce

poverty, with usually implicit, but occasionally explicit, implications
for the treatment of gender. These consistent views can be traced to
an orthodoxy in the World Bank regarding the nature of poverty and
policy on poverty reduction.
(1999: 545)

In particular, they note the probability of the standardising effects of peer

reviews on the policy sections on poverty reduction, and the implications
of the standardisation for the ‘filtering out’ of gender issues from the
assessments. Processes of dilution are evident across the different stages of
production of the Assessments – from the PPAs to the Poverty Assess-
ments; and from the empirically derived analysis sections to the policy sec-
tions (ibid.). This fits in well with Longwe’s famous discussion of ‘policy
evaporation’ (1997).
The lack of consensus within organisations such as the World Bank on
the operational definitions and perspectives, as well as approaches in
research and policy associated with gender analysis, is only likely to
increase as several agencies come together to combine their resources
towards the achievement of the Millennium Development Goals. It is not
the institutional capacities of the Bank alone that are at the root of the
202 Ramya Subrahmanian
uneven analyses of gender disadvantage. Whitehead and Lockwood (1999)
note the difficulties of co-ordinated approaches between national teams,
consultants, different donors involved and task managers who all bring
different understandings of gender to bear on their work, which only
serve to compound the weakness of the operational guidelines discussed
above. The struggle for consensus is likely to be the hardest challenge to
surmount. While the Millennium Development Goals represent one
attempt to fashion a consensus on priorities, managing co-ordination is
another challenge.23
New mechanisms for aid delivery have emerged out of a concern with
the limited effectiveness of project approaches and the belief in the
greater efficiency of resource pooling between donors and strengthened
processes of debate and dialogue about key areas of priorities with govern-
ments. Bringing the state back into the picture has meant that issues of
national ownership have been framed as central objectives of these new
policy tools. However, many commentators note that objectives of national
ownership, and – within national contexts – consultations with stakehold-
ers, particularly the traditionally voiceless, have scarcely been met, belying
the claim that these approaches constitute new participatory methods for
building sustainable development (Christian Aid, 2001; Rodenberg, 2002;
Standing, 2001). The World Bank (2001b) notes that consultation with
organisations within civil society, particularly women’s organisations, in
the I-PRSPs and PRSPs reviewed revealed low levels of consultation, with
only two of the nineteen I-PRSPs showing specific plans to consult poor
women or women’s organisations and only three in total showing any indi-
cation of incorporating better gender analysis in their plans. A Christian
Aid report comments that the involvement of poor people in PRSP-related
processes has been ‘minimal and superficial’ (2001: 2), noting particularly
the selectivity of issues on which consultation is invited and the lack of
support provided by international agencies to local groups whose views
contradict the policy directions of the PRSPs. Given the complexity of
gender issues, and the difficulties of organising and representing the
voices of women, it is clear that gender issues are likely to be excluded to a
greater extent from processes of negotiation and discussion.24 Even in
sectors such as health and education, which appear to engage more with
gender issues, stakeholder participation is traditionally very weak, given
that the nature of knowledge within those sectors is largely seen as techno-
cratic and expert-led (Standing, 2001).

Conclusion
For gender to make the transition from rhetoric to reality requires con-
certed attention at all levels of policy processes, from agenda-setting to
planning of all aspects, to accountability through transparent monitoring
and evaluation. Half-hearted or token measures result in uneven analytical
 Promoting gender equality 203
approaches and inadequate operational frameworks. Agencies appear to
be several steps behind the ever-increasing conceptual sophistication of
thinking on ‘gender’ in development scholarship, and continual processes
of learning are essential. There is a danger of ‘analysis fatigue’ concerning
gender, fuelled by reliance on particular kinds of checklists and tools
which demand little attention to nuance and contextual detail.
Further, ‘gender equity interests’ (Goetz, 2003) need to be addressed
within bureaucratic and political spaces to ensure that the content of
policy is informed by the voices of those most likely to be affected by pol-
icies. Given the constraints operating on women’s effective representation
and participation in policy processes, particular attention has to be paid to
ensuring that women are listened to, and enabled to influence policy
agendas. While a long wish-list has been established, and the conditions
under which gender interests are likely to flourish with broader develop-
mental benefits have been well documented, the implications for taking
the Millennium Development Goals forward in a meaningful way are
clear. Greater transparency around the means through which broadly
defined gender empowerment and equality goals is critical. This involves
greater transparency on the part of development agencies about the
processes by which they implement policies, the ways in which resources
are distributed and allocated, and the outcomes to which they give rise.
The World Bank evaluations cited in this chapter are a step in the right
direction, acknowledging as they do their failures to realise their claims
and objectives. The next steps entail meaningful action to address these
failures.

Notes
1 Mehrotra notes that almost all the International Development Targets relate to
basic social service provisioning (2001: 4).
2 Razavi and Miller (1995) note, for example, that the tendency has been to
focus on women as mothers only, especially in arenas linked to reproduction,
such as population services, nutrition education and child-related services,
whereas women have been absent figures in mainstream employment schemes
and land reform programmes.
3 In fact, it is argued that donor agencies have driven the gender and develop-
ment agenda, raising questions about the ownership of the agenda at the level
of national governments (Razavi and Miller, 1995). Razavi and Miller note that
financial resources for women-in-development (WID)/gender projects within
donor agencies have often come from particular donor governments through
the practice of ‘multi-bi’ funding, and thus the ‘feminist agenda has remained
excessively donor-driven’ (ibid.: 4).
4 The terms are used here as an adaptation from Kabeer (1999b). Kabeer
defines gender-aware policies as those that recognise that development actors
are both women and men, and that they are constrained in different and
unequal ways from participating and benefiting from development processes.
Gender-specific policies refer to those that are targeted to benefit a specific
sex, based on the recognition of differing gendered needs and constraints. If
204 Ramya Subrahmanian
they are designed to redress disadvantage and achieve gender justice, then these
policies can be considered to be transformative. Gender-specific approaches, as
Kabeer argues, can serve to integrate women in sex-specific ways that do not
challenge asymmetries in divisions of labour and resources; if, however, they
were based on an understanding of, and commitment to challenging, these
asymmetries, then they would have the potential to transform gender relations
in the direction of greater equality.
5 World-wide, women’s presence is less than 14 per cent in lower houses of par-
liament, a ‘share that is growing at slower than snail’s pace’ (UNDP, 2002: 70).
6 It is also striking to note, though, that this gender-specific goal is the one least
likely to be achieved of all the Millennium Development Goals, which points to
the difficulty of seeing results in areas where women need to be targeted for
themselves (Chapter 11 of this volume).
7 For instance, the World Bank’s recent policy research report discusses gender
equality in terms of ‘rights, resources and voice’ (2001a).
8 Capabilities refer to the potential that people have for achieving valued ways of
‘being and doing’ (Sen, 1985). The influence of Amartya Sen’s work on poverty
and gender has been critical for feminist analysis in development (Jackson and
Pearson, 1998), particularly for shaping the focus on the links between the
relations of gender and outcomes of gender inequality, and for allowing focus
on processes of negotiation and bargaining as part of understanding how
resources translate into outcomes.
9 Whitehead and Kabeer (2001) argue for the importance of locating under-
standing of intra-household gender inequalities within analysis of wider liveli-
hoods to obtain a fuller account of the nuances of the intersections between
gender and poverty.
10 See Cagatay (1998), Chant (1997), Jackson (1997, 1998) and Kabeer (1997) for
detailed discussions of concepts, approaches and debates relating to gender
and poverty in general, and on female headship in particular.
11 See note 4 above. The focus on education in recent times and the stress on
women’s education has similar roots in the concern with women’s role in
reproduction and child-rearing, emphasising as it does the role education,
particularly primary education, plays in bringing down fertility rates and improv-
ing child health.
12 It includes increased opportunities for income generation; increased access to
credit or other relevant economic services; and increased participation in train-
ing or skills upgrading activities.
13 The twelve areas of action highlighted in the Beijing Conference’s Platform for
Action include poverty; education and training; access to health care and
related services; violence against women; effects of armed or other kinds of
conflict, including effects on those living under foreign occupation; economic
structures and policies, in all forms of productive activities and in access to
resources; sharing of power and decision-making at all levels; mechanisms to
promote the advancement of women; human rights; women’s access to, and
participation in, all communication systems, especially in the media; the man-
agement of natural resources and the environment; and violation of the rights
of the girl child.
14 The results reported here are taken from World Bank (2001a). In this chapter,
the Humana scale is replaced with a 1–4 scale (whereas the Humana Index
used a 0–3 scale), with 1 representing consistent pattern of rights violations
and 4 representing unqualified respect for freedoms and rights (World Bank,
2001a: 37).
15 The level of regional aggregation is deeply problematic, as is the definition of
what each of these means. For instance, statutory rights are likely to vary depend-
 Promoting gender equality 205
ing on the legal system in place. Where, however, it is useful is that it looks at
relative equality for women vis-à-vis the rights enjoyed by men in each of those
dimensions.
16 Using education as an entry point to justify funding on gender raises the
further danger of obscuring from policy attention a range of inequalities and
disparities relating to the economic, political and social spheres. Participants
from East Asia at a Regional Gender Workshop organised by the World Bank
note that the Bank has ‘ignored gender issues’ in that region ‘because there is
no typical gender disparity in education enrolments’ (World Bank, 2002: 27).
17 For example, situations where parties may nominate wives of male politicians
as a way of strengthening the party presence, regardless of the mechanism in
place to promote women’s participation (party quotas; reservations of seats for
women; open competition). Political parties may co-opt spaces designed to
promote women’s participation, which would limit the potential for claiming
‘empowerment’ as well as dilute the terms of apparently reduced ‘inequality’.
However, as Goetz (2003) demonstrates, some mechanisms are more successful
in promoting women’s political participation and presence than others (see
also UNDP, 2002, for a discussion).
18 For instance, Croll (2000) documents evidence of the increasing scale of
daughter discrimination in East and South Asia, despite rising levels of eco-
nomic development and dramatic levels of fertility decline in South Asia and
China in particular. In fact, economic development has seen the entrenchment
of son preference in these regions, and ‘son-support has been incorporated
into most national strategies for economic development’ (p. 9).
19 ‘Gender budgets’ are fast emerging as a key approach to placing gender con-
cerns at the heart of development agenda-setting and resource allocation. See
Elson (1999) and Budlender (1999) for detailed discussions.
20 See Razavi and Miller (1995) and Goetz (1995) for studies of gender main-
streaming within multilateral, bilateral and government agencies.
21 See also World Bank (2002) for a report on the analysis of gender within the
Bank’s Sector Strategy Papers, Poverty Assessments and projects.
22 See Gender and Development Network (2000) for a brief introduction to the
ways in which gender is mainstreamed within the World Bank. While there are
high-powered gender bodies within the Bank, it is clear that capacity is unevenly
distributed. Recent publications of the Bank, like those cited in this chapter,
indicate, however, that there is an increasingly high level of awareness of the
importance of gender, but like other agencies, there are institutional dif-
ficulties in translating this into a more coherent strategy.
23 See Sedere (2000) for a more general discussion of the problems of donor co-
ordination in education sector funding in Bangladesh.
24 See Goetz (2003) for a more general discussion of issues relating to the representa-
tion of women’s interests through political processes. In addition, the complex-
ity of gender interests, given the heterogeneity of the category ‘women’,
conflicts with the need for civil society advocates to speak with one voice, which
often means ‘subordinating women’s gender interests to men’s’ (Geraldine
Terry, cited in Christian Aid, 2001: 14).

References
Baden, S. and Green, C. (1994) Gender and Education in Asia and the Pacific,
BRIDGE Report 25, Brighton: Institute of Development Studies.
Bennell, P. (1996) ‘Rates of Return to Education: Does the Conventional Pattern
Prevail in Sub-Saharan Africa?’, World Development, 24(1): 183–199.
206 Ramya Subrahmanian
Booth, D. and Lucas, H. (2001) Desk Study of Good Practice in the Development of PRSP
Indicators and Monitoring Systems: Initial Review of PRSP Documentation, London:
Overseas Development Institute.
Budlender, D. (1999) ‘The South African Women’s Budget Initiative’, Back-
ground Paper 2, Meeting on Women and Political Participation: 21st Century
Challenges, 24–26 March, New Delhi: UNDP.
Cagatay, N. (1998) ‘Gender and Poverty’, Working Paper 5, New York: UNDP,
Social Development and Poverty Elimination Division.
Chant, S. (1997) ‘Women-Headed Households: Poorest of the Poor? Perspectives
from Mexico, Costa Rica and the Philippines’, IDS Bulletin, (28)3: 26–48,
Brighton: Institute of Development Studies.
Christian Aid (2001) ‘Ignoring the Experts: Poor People’s Exclusion from Poverty
Reduction Strategies’, Policy Briefing, London: Christian Aid.
Croll, E. (2000) Endangered Daughters: Discrimination and Development in Asia,
London and New York: Routledge.
Elson, D. (1998) ‘Talking to the Boys: Gender and Economic Growth Models’, in
Jackson, C. and Pearson, R. (eds) Feminist Visions of Development: Gender Analysis
and Policy, London: Routledge.
—— (1999) ‘Gender-Neutral, Gender-Blind, or Gender-Sensitive Budgets? Chang-
ing the Conceptual Framework to Include Women’s Empowerment and the
Economy of Care’, Gender Budget Initiative Background Papers, London: Com-
monwealth Secretariat.
—— (2002) ‘Gender Justice, Human Rights and Neo-liberal Economic Policies’, in
Molyneux, M. and Razavi, S. (eds) Gender Justice, Development and Rights, Oxford:
Oxford University Press.
Gender and Development Network (2000) How to Challenge a Colossus: Engaging
with the World Bank and the International Monetary Fund, London: Womankind.
Goetz, A.M (1995) ‘The Politics of Integrating Gender to State Development
Processes: Trends, Opportunities, and Constraints in Bangladesh, Chile, Jamaica,
Mali, Morocco, and Uganda’, Occasional Paper 2, Geneva: UNRISD.
—— (2003) ‘Women’s Political Effectiveness: A Conceptual Paper’, in No Shortcuts
to Power: African Women in Politics and Policy-Making, London: Zed Press.
Hanmer, L. and Naschold, F. (2000) ‘Attaining the International Development
Targets: Will Growth Be Enough?’, Development Policy Review 18(1): 11–36.
Hanmer, L and Wilmshurst, J. (2000) ‘Are the International Development Targets
Attainable? An Overview, Development Policy Review, 18(1): 5–10.
Jackson, C. (1997) ‘Post Poverty, Gender and Development’, IDS Bulletin, 28(3):
145–155, Brighton: Institute of Development Studies.
—— (1998) ‘Rescuing Gender from the Poverty Trap’, in Jackson, C. and Pearson,
R. (eds) Feminist Visions of Development: Gender Analysis and Policy, London: Rout-
ledge.
Jackson, C. and Pearson, R. (eds) (1998) Feminist Visions of Development: Gender
Analysis and Policy, London: Routledge.
Jeffery, R. and Jeffery, P. (1998) ‘Silver Bullet or Passing Fancy? Girls’ Schooling
and Population Policy’, in Jackson, C. and Pearson, R. (eds) Feminist Visions of
Development: Gender Analysis and Policy, London: Routledge.
Kabeer, N. (1997) ‘Tactics and Trade-Offs: Revisiting the Links between Gender
and Poverty’, IDS Bulletin, 28(3): 1–13, Brighton: Institute of Development
Studies.
 Promoting gender equality 207
—— (1999a) ‘The Conditions and Consequences of Choice: Reflections on the Mea-
surement of Women’s Empowerment’, Discussion Paper 108, Geneva: UNRISD.
—— (1999b) ‘From Feminist Insights to an Analytical Framework: An Institutional
Perspective on Gender Inequality’, in Kabeer, N and Subrahmanian, R. (eds)
Institutions, Relations and Outcomes: A Framework and Tools for Gender-Aware Plan-
ning, New Delhi: Kali for Women.
—— (2001) ‘Resources, Agency, Achievements: Reflections on the Measurement
of Women’s Empowerment’, in Discussing Women’s Empowerment: Theory and Prac-
tice, Sida Studies 3, Stockholm: Swedish International Development Agency.
Longwe, S. (1997) ‘The Evaporation of Gender Policies in the Patriarchal Cooking
Pot’, Development in Practice, 7(2): 148–156.
McGee, R. (2000) ‘Meeting the International Poverty Targets in Uganda: Halving
Poverty and Achieving Universal Primary Education’, Development Policy Review,
18(1): 85–106.
Mehrotra, S. (2001) ‘The Rhetoric of International Development Targets and the
Reality of Official Development Assistance’, Innocenti Working Paper 85, Flo-
rence: UNICEF Innocenti Research Centre.
Molyneux, M. and Razavi, S. (eds) (2002) Gender Justice, Development and Rights,
Oxford: Oxford University Press.
Nussbaum, M. (2000) Women and Human Development: The Capabilities Approach,
Cambridge: Cambridge University Press.
Pearson, R. (1998) ‘ “Nimble Fingers” Revisited: Reflections on Women and Third
World Industrialisation in the Late Twentieth Century’, in Jackson, C. and
Pearson, R. (eds) Feminist Visions of Development: Gender Analysis and Policy,
London: Routledge.
Razavi, S. and Miller, D. (1995) ‘Gender Mainstreaming: A Study of Efforts by the
UNDP, the World Bank and the ILO to Institutionalise Gender Issues’, Occa-
sional Paper 4, Geneva: UNRISD.
Rodenberg, B. (2002) ‘Integrating Gender into Poverty Reduction Strategies:
From the Declaration of Intent to Development Policy in Practice?’ Briefing
Paper 2/2002, Bonn: German Development Institute.
Sedere, U. (2000) ‘Rethinking Educational Aid: Sector Support Approach to
Financing Basic Education: Lessons from Bangladesh’, Prospects: Quarterly Review
of Education, 116(4): 451–460, Paris: UNESCO.
Sen, A. (1985) Commodities and Capabilities, vol. 7 of Professor Dr. P. Hennipman Lec-
tures in Economics: Theory, Institutions, Policy, Amsterdam: North-Holland.
Standing, H. (2001) ‘Institutionalising Gender at a Sectoral Level: What Does It
Mean and Who Does It?’, paper given at the Dutch Aid Gender and Sector-Wide
Approaches Workshop, The Hague, 21–23 February, Brighton: Institute of
Development Studies.
Subrahmanian, R. (2001) Gender and Education: A Review of Directions for Social
Policy, report prepared for UNRISD.
—— (2002) ‘Engendering Education: Prospects for a Rights-Based Approach to
Female Education Deprivation in India’, in Molyneux, M. and Razavi, S. (eds)
Gender Justice: Development and Rights, Oxford: Oxford University Press.
UNDP (2002) ‘Deepening Democracy in a Fragmented World’, Human Develop-
ment Report 2002, New York: UNDP.
White, H. (2001) ‘Will the New Aid Agenda Help Promote Poverty Reduction?’,
Journal of International Development 13: 1057–1070.
208 Ramya Subrahmanian
—— (2002) ‘Using the Millennium Development Goals as a Basis for Agency-Level
Performance Measurement’ in Black, R. and White, H. (eds) Targeting Develop-
ment: Critical Perspectives on the Millenium Development Goals, London: Routledge.
Whitehead, A. and Kabeer, N. (2001) ‘Living with Uncertainty: Gender, Liveli-
hoods and Pro-Poor Growth in Rural Sub-Saharan Africa’, Working Paper 134,
Brighton: Institute of Development Studies.
Whitehead, A. and Lockwood, M. (1999) ‘Gendering Poverty: A Review of Six
World Bank African Poverty Assessments’, Development and Change, 30(3):
525–555.
World Bank (2001a) Engendering Development through Gender Equality in Rights,
Resources, and Voice, Policy Research Report, Washington, DC: World Bank and
Oxford: Oxford University Press.
—— (2001b) Gender in the PRSPs: A Stocktaking, Poverty Reduction and Economic
Management Network, Washington, DC: World Bank, Gender and Development
Group.
—— (2002) The Gender Dimension of Bank Assistance: An Evaluation of Results, Report
23119, Washington, DC: World Bank, Operations Evaluation Department.
10 Reducing infant and child death
Howard White

Ten thousand African children died today

Fifty years ago, close to two of every ten children born in East Asia died
during their first year of life, and another one of their sisters or brothers
did not reach their fifth birthday. The dramatic reductions in the rates of
infant and child death achieved in the second half of the last century are a
testament to the success of development efforts. Today, fewer than one in
twenty East Asian children die in the first five years of life. But there is still
far to go. Despite reduced mortality rates, a staggering number of chil-
dren still die. Two million African infants die each year, and as many
again before they reach 5. They would make a line of dead children that
would stretch the entire length of the United Kingdom. This figure has
not changed in over two decades. That is 10,000 avoidable deaths every
day of every week of every month for over twenty years. Every single child
death is a personal tragedy. In the words of a father who lost two children,
‘Can you tell me why this had to happen?’ (quoted in Howard and
Millard, 1997: 1). There is no answer to that question. It did not have to
happen. Yet, another 10,000 African children will die tomorrow, another
10,000 the day after and so on. What needs to change to stop the bodies
from piling up?
The Millennium Development Goal is for a two-thirds reduction in
infant and under-five mortality by 2015. The history of missed goals shows
that the existence of a goal is not enough to ensure that it is met. And, as
is argued in Chapter 3, the existence of a goal does not give any indication
of a strategy that might be employed to achieve that goal. Historical
trends, reviewed in the next section, show that progress in reducing mor-
tality has been very uneven. The current rate of reduction is insufficient to
meet the target. Yet in certain times and places, dramatic reductions in
mortality have taken place. What are the lessons from these different
experiences? In particular, what can be done to bring about more rapid
mortality decline amongst poor performers?
There has been a debate over the sources of mortality decline in the
now developed countries, most notably England (see p. 220), and this
210 Howard White
debate has been echoed in the literature on developing countries. The
opposing positions are held by those who attribute the decline more to
rising incomes and consequent improvements in nutrition, versus those
assigning a larger role to public health and other interventions. Straw
men are highly inflammable, and so should be avoided if we want sensible
policy advice rather than heated debate. No one denies that economic
growth matters for sustained improvements in social welfare. Few also
deny at least some role for government in health service provision. The
question is one of balance between the two.1 The position taken here,
developed in the section ‘Reasons for mortality decline’, p. 214, is that
health services and other public actions do play a key role. And substantial
improvements can be achieved independently of income growth, such as
through the promotion of breastfeeding, which is a vital aspect of child
survival chances. But the exhortation is not simply for more money; it is
also for better prioritisation of expenditure, including expenditure on
preventive measures (such as the promotion of breastfeeding), public
campaigns and an emphasis on female education. These policy implica-
tions are laid out in the concluding section.

The uneven record of mortality decline

Figure 10.1 shows trends in infant and child mortality, the former being
the probability of death in the first year of life and the latter that between
first and fifth birthdays. Under-five mortality, which is the subject of the
Millennium Development Goal, combines these two measures.2 The most
striking feature of the graphs is the secular decline in child and infant
mortality rates in all regions,3 with only two recent exceptions. The two
recent exceptions are child mortality in sub-Saharan Africa and infant
mortality in the transition economies of the former Soviet Union. But
these are exceptions, and should not detract from the very marked success
in reducing mortality, most notably in East Asia, but also the Middle East
and North Africa. Except in sub-Saharan Africa, where the rate is 65
deaths per 1,000 live births, child mortality is now at very low levels of 10
or fewer deaths per 1,000 live births in all other regions except South Asia.
In South Asia the rate is slightly higher at 15 deaths per 1,000 live births,
though this figure shows a suspicious kink in the trend and is probably
unreliable.
Lying behind these strong performances have been a range of social
measures, which are discussed in more detail for some specific cases
below. These measures, promoted by UNICEF and supported by a number
of international agencies, focused around GOBI: growth monitoring of
children under 2, spread of knowledge about oral rehydration therapy
(ORT), breastfeeding and improved early child feeding, and immunisa-
tion. The success of these initiatives can be seen both through intermedi-
ate indicators – such as raising average immunisation coverage in
Child mortality
Infant mortality

Figure 10.1 Trends in (a) child and (b) infant mortality by region (per 1,000 live
births).
Source: World Bank World Development Indicators CD-Rom, 2001.
212 Howard White
developing countries from 20 to 80 per cent during the course of the
1980s – and in the outcome of falling mortality. One estimate is that by
1995 these programmes had prevented some 25 million child deaths
(Jolly, 2001).
The second observation from Figure 10.1 is the difference in regional
performance, in terms of both levels and changes over time. The more
successful developing countries are reducing their mortality rates to approach
those in developed countries, although the rates in the latter countries
also continue to decline. The infant mortality rate in Cuba is the same as
that for the United States (8 per 1,000 live births). So a gap is opening up
between good and poor performers, with sub-Saharan African countries
mostly doing least well, particularly with respect to child mortality. Brief
reflection on the different regional experiences yields some initial insights
into the factors underlying mortality reductions:

• East Asia has been a region of strong economic growth. Whilst the East
Asian newly industrialising countries (NICs) have done well on social
indicators, over 80 per cent of the region’s population live in China,
which must therefore dominate the trend towards lower mortality.
Mortality reduction in China was strongest prior to the introduction of
market-based reforms in the mid-1970s, and evidence reviewed below
points to the key role of health services in this good performance.
• North Africa and the Middle East includes countries which have bene-
fited from oil wealth. There has been a tendency in the past to suggest
that this wealth was not turned into general improvements in the
population’s welfare. The data shown here suggest that this is not so,
though it can be argued that these improvements came about with a
lag.4 Table 10.2 reports data for Oman, which has reduced mortality
by nearly 90 per cent since 1950. Such a decline does not come about
automatically from the presence of oil but is the result of deliberate
government policy to utilise the income from oil wealth for the common
good. Oman’s National Health Programme (NHP) has had seven
separate components, including the Expanded Programme of Immu-
nisation (EPI),5 Mother and Child Health and Control of Tuberculo-
sis (Skeet, 1992: 115). The programme has been implemented with a
massive increase in health infrastructure: in the 1980s the number of
hospitals nearly doubled, health clinics increased in number by 75 per
cent and nursing staff fourfold (ibid: 112).
• Mortality started to rise in Eastern Europe and countries of the
former Soviet Union. This trend can be linked to income declines but
also to the associated collapse in health services as government
revenue dried up, resulting in falling immunisation coverage in some
countries (White et al., 2000). As documented elsewhere, the more
significant shift has been the rise in adult mortality (see Cornia and
Paniccià, 2000).
 Reducing infant and child death 213
• Infant and child mortality have declined in Africa despite deteriorat-
ing economic performance since the 1970s. The recent reversal of the
decline in child mortality in Africa is in part a consequence of HIV/
AIDS. But even before the effects of the epidemic were felt in the
1990s, there was a marked relative deterioration, indicating a failure
to get infectious diseases under control. However, mortality rates were
falling in countries in which income per capita was also falling (e.g.
Zambia), showing that there is not a necessary link between economic
growth and reducing mortality.

To pick up on this last point, previous studies have identified an ‘exoge-

nous component’ in mortality decline. That is, if we look at the relation-
ship between life expectancy and income (Preston, 1975) or infant mortality
and income (Hanmer et al., 2003), welfare is better (life expectancy
higher, infant mortality lower) in more recent periods than in earlier ones
for countries at the same level of real income.6 Hanmer et al. calculate that
this exogenous component has reduced mortality in low-income countries
by about 20 per 1,000 in the past thirty years. This exogenous decline can
be attributed to factors such as improved medical technology (e.g. the
development of heat-stable vaccines in the 1980s, and measures to
improve the monitoring of the cold chain)7 and the greater availability of
health services, as well as clean water in low-income countries. These
improvements can be attributed at least in part to flows of development aid.
The third observation from Figure 10.1 is that rates of mortality decline
are still generally insufficient to meet the target set by the Millennium
Development Goal. A two-thirds reduction in twenty-five years (from 1990
to 2015) requires an annual reduction of 4.3 per cent. As it happens, only
high-income countries have achieved this figure for infant mortality in the
past twenty-five years. On the other hand, all regions other than sub-
Saharan Africa have achieved sufficient progress in reducing child mortal-
ity to reach the target (Table 10.1). There may be some questions as to
Table 10.1 Annual mortality reductions over 25 years (1973–1998) by region (per
cent)

Infant Child Under-five

mortality mortality mortality

East Asia and Pacific ⫺2.6 ⫺7.1 ⫺3.8

Latin America and Caribbean ⫺3.7 ⫺5.9 ⫺4.2
Middle East and North Africa ⫺3.9 ⫺7.4 ⫺4.7
South Asia* ⫺2.3 ⫺4.5 ⫺3.2
Sub-Saharan Africa ⫺1.4 ⫺1.4 ⫺1.3
High income ⫺4.3 ⫺9.0 ⫺5.0

Source: Calculated from World Bank World Development Indicators CD-ROM, 2001.
Note
*Calculated over 24 years, and 1998 data point seems suspect.
214 Howard White
whether this rate of reduction can be sustained at lower mortality levels,
though it should be possible, since it has been done in developed coun-
tries. But the target is not for child mortality but for under-five mortality,
of which the largest part is now accounted for by infant deaths. Since
under-five mortality is dominated by infant mortality, only two regions, the
Middle East and North Africa, and Latin America and the Caribbean,
appear to be achieving sufficiently high rates of reduction.
Understanding why some regions are doing better than others is the
key to identifying the policies required to generalise rapid mortality
decline. But regional differences hide substantial variations in country
performance, so the next section begins with some country data.

Reasons for mortality decline

Table 10.2 shows infant mortality rates for selected countries since 1950
and the annual rate of change over the first twenty-five years and the next
twenty.8 Other than Oman, which was discussed above, in the countries
shown there have been high rates of mortality reduction in China, Cuba
and the Russian Federation (the data mostly referring to the latter region
whilst it was still part of the Soviet Union). What these three countries
have in common is that they all are or were centrally planned economies.
They have all done markedly better in reducing mortality than countries
that have followed a more market-oriented development strategy, such as
Brazil, India and Kenya. For example, prior to 1975, Vietnam achieved a
reasonable rate of reduction for a country at war, and since then it has
accelerated the rate of improvement after unification. Although moves
towards a market economy were made in the 1980s, the Vietnamese
government continues to exert strong control, including control over the
delivery of state-controlled health services. By contrast, the health service

Table 10.2 Infant mortality rates for selected countries (deaths per 1,000 live
births)

1950– 1960– 1975– 1995– Annual Annual

1955 1965 1980 2000 change, change,
1950–1975 1975–1995
(%) (%)

Brazil 135 109 79 42 ⫺2.1 ⫺3.1

China 195 212 52 41 ⫺5.1 ⫺1.2
Cuba 81 59 22 8 ⫺5.1 ⫺4.9
India 190 157 129 73 ⫺1.5 ⫺2.8
Kenya 155 127 93 65 ⫺2.0 ⫺1.8
Oman 231 207 95 27 ⫺3.5 ⫺6.1
Russian Federation 97 40 30 17 ⫺4.6 ⫺2.8
Vietnam 158 130 83 40 ⫺2.5 ⫺3.6

Source: UN World Population Prospects: The 2000 Revision, vol. 1.

 Reducing infant and child death 215
has deteriorated markedly in China in the past two decades, correspond-
ing to a considerable slowing in the rate of mortality reduction.
Contemporary wisdom has it that centrally planned economies perform
badly economically. If so, then improvements in economic well-being
cannot be the source of these countries’ success in lowering mortality.
Indeed, as just mentioned, the period of China’s rapid economic growth
has seen a slowing of mortality decline, hence the success of these coun-
tries seems more likely to rest with state provision of health services, as is
more thoroughly documented later in the chapter.
In this section the view that public health and other government inter-
ventions matter is supported by three arguments: (1) regression analysis
finds a robust effect of various measures of health service provision on
infant and child mortality; (2) case studies from both developed and
developing countries almost invariably report a significant role for a range
of policy measures; and (3) as mortality rates fall, the structure of death
changes, becoming focused on infants, whose survival is more dependent
on interventions than on general environmental conditions.

Evidence from robust regression analysis

There is a substantial literature analysing the determinants of infant mor-
tality using cross-country regressions. Infant or child mortality is taken as
the dependent variable, and is regressed on a range of factors such as
income per capita, maternal education and health expenditure. These
studies are reviewed by Hanmer and White (1999), who report that well
over thirty variables have been used as determinants of mortality. Under
such circumstances it is quite possible to come up with a regression in
which all coefficients are significant; but someone else could come up
with quite a different set of significant determinants. There are two
responses to this situation. The first is to abandon cross-country analysis in
favour of case studies, and that is a route I pursue on pp. 219–233. The
other is to adopt a robust regression approach.
Robust regression involves estimating all possible combinations of the pro-
posed explanatory variables. Hanmer et al. (2002) used twenty-four
independent variables to estimate 420,000 equations for infant and child
mortality. These found income to be a robust determinant of mortality, but
so were measures of health service provision, notably immunisation coverage
and the number of doctors per person. Measures of education and gender
inequality are also found to be robust. Claims, based on cross-country regres-
sion analysis, that health expenditure is unimportant in explaining country
differences in mortality (Filmer and Pritchett, 1999) are not borne out by the
robust regression approach. Money spent on health does indeed matter –
though of course so does how it is spent (see pp. 226–231).
With a variable selection justified by the robust regression analysis,
Table 10.3 shows some illustrative cross-country regression results for
Table 10.3 Regression results for infant mortality

(1) (2) (3) (4) (5) (6) (7) (8) (9)

GNP per capita (logged) ⫺0.52 – – ⫺0.49 ⫺0.51 ⫺0.45 ⫺0.46 ⫺0.10 ⫺0.09
(⫺46.4) (⫺36.9) (⫺39.0) (⫺29.4) (⫺30.0) (⫺1.45) (⫺1.23)
DPT immunisation – ⫺0.70 – ⫺0.26 – ⫺0.06 – 0.46 –
(logged) (⫺15.4) (⫺8.1) (⫺2.2) (4.51)
Measles immunisation – – ⫺0.67 – ⫺0.24 – ⫺0.06 0.51
(logged) (⫺12.3) (⫺7.3) (⫺1.9) (4.35)
Interactive term for income and – – – – – – – ⫺0.09
measles (⫺5.07)
Interactive term for income and DPT – – – – – – – ⫺0.08 –
(⫺5.44)
Sub-Saharan Africa – – – – – 0.67 0.66 0.68 0.68
(9.5) (9.3) (9.80) (9.77)
South Asia – – – – – 0.22 0.23 0.25 0.27
(2.1) (2.2) (2.52) (2.61)
East Asia and Pacific – – – – – 0.19 0.17 0.60 0.58
(2.9) (2.6) (9.48) (9.27)
Middle East and North Africa – – – – – 0.61 0.59 0.18 0.16
(9.5) (9.1) (2.83) (2.52)
Eastern Europe and Central Asia – – – – – ⫺0.25 ⫺0.25 ⫺0.26 ⫺0.24
(⫺3.5) (⫺3.5) (⫺3.76) (⫺3.49)
Latin America and Caribbean – – – – – 0.35 0.36 0.33 0.36
(6.2) (6.2) (6.02) (6.36)
1975–1979 – – – – – ⫺0.44 – ⫺0.52 –
(⫺1.6) (⫺1.89)
1980–1984 – – – – – ⫺0.55 ⫺0.35 ⫺0.55 ⫺0.32
(⫺2.0) (⫺2.6) (⫺2.09) (⫺2.38)
1985–1989 – – – – – ⫺0.69 ⫺0.48 ⫺0.68 ⫺0.41
(⫺2.5) (⫺3.5) (⫺2.54) (⫺3.03)
1990–1994 – – – – – ⫺0.79 ⫺0.57 ⫺0.77 ⫺0.50
(⫺2.9) (⫺4.2) (⫺2.89) (⫺3.67)
1995–1997 – – – – – ⫺0.94 ⫺0.73 ⫺0.93 ⫺0.65
(-3.5) (⫺5.3) (⫺3.48) (⫺4.81)
Constant – 6.37 6.25 8.29 8.34 7.61 7.44 5.38 4.99
(34.2) (28.0) (64.0) (59.7) (24.4) (35.4) (10.55) (9.49)
R squared 0.67 0.25 0.18 0.76 0.72 0.85 0.86 0.87 0.86
N 1,066 716 671 645 645 606 645 606 645
218 Howard White
infant mortality. The data are five-year period averages for 191 countries.
The variables have been logged so that the coefficients are elasticities,
showing that a 1 per cent rise in income per capita results in a reduction
in infant mortality of about 0.5 per cent. Two immunisation variables are
shown. The high correlation between the two immunisation variables (the
simple correlation coefficient is 0.8) means they are put into separate
equations, each having very similar coefficients (of ⫺0.7 in the simple
regression, reduced to ⫺0.25 when income is introduced and ⫺0.06 when
time and region dummies are used). A 1 per cent increase in immunisa-
tion coverage has a significant impact on infant mortality. The size of the
elasticity is much less than that for income – but increasing immunisation
coverage by 1 per cent costs much less than does increasing growth by
1 per cent.
These regressions make a simple point: infant mortality is strongly
related to income. Income is a means to many things which make lower
mortality possible – including immunisation (the correlation between
income and immunisation, both logged, is about 0.4). But the fit between
infant mortality and income is not perfect; equation (1) has an R squared
value of 0.67. Adding additional variables both weakens the significance of
income (as will be seen in a moment, it is possible to produce regressions
in which income is insignificant) as these additional variables pick up the
channels through which income affects mortality. Yet it is important that a
variable such as immunisation is significant when included with income,
since this shows that immunisation has an impact over and above that
from income – that is, health delivery makes a difference which operates
independently from income.
Equations (8) and (9) introduce an interactive term, this term being
the product of income and immunisation coverage. This term allows the
elasticity of mortality with respect to income to vary according to the level
of immunisation, or the elasticity for immunisation to vary according to
income. In these equations, income is no longer significant, which might
be interpreted as showing that higher income alone is no use in the
absence of immunisation. But not too much importance should be
attached to such an interpretation, and none should be attached to the
positive coefficient which now appears for immunisation. The correct
approach is to calculate the elasticities over the range of the data. That for
income varies between ⫺0.35 (immunisation coverage at 20 per cent) and
⫺0.48 (full coverage). The elasticity for immunisation varies from ⫺0.10
in the lowest-income countries to over ⫺0.4 in high-income ones. Allow-
ing for this interactive effect thus gives two important results: (1) the elas-
ticity of immunisation is much higher than appears from regressions that
ignore the interactive term, and (2) both elasticities are higher the higher
the value of the other variable, a finding which is easy to explain (see the
beginning of the next section on p. 219).
Equations (6)–(9) include time and region dummies. For the regional
 Reducing infant and child death 219
dummies, the reference group is developed countries, so the coefficient
shows how much higher or lower infant mortality is in that region once
the other variables are taken into account. Eastern Europe and Central
Asia have a negative coefficient, indicating that (despite the relative
decline in recent years) these countries enjoy low mortality relative to
their income (with their high immunisation rates also being allowed
for). For the time dummies, the reference period is the first five years,
1970–1975. The negative coefficients on these variables represent the
‘exogenous shift’ mentioned on p. 213. The size of the coefficient contin-
uously increases from one period to the next, indicating that the shift has
been present throughout the period.

Evidence from case studies

Much of the literature of the determinants of mortality is concerned with
attributing primary cause either to economic factors, including nutritional
status, or to public health and medical interventions. As suggested in the
introduction to this chapter, this debate should be seen as a false
dichotomy. More realistic is to utilise the Mosley–Chen framework, shown
as Figure 10.2. According to this framework, both sets of factors matter,
and neither one can operate independently of the other. Several channels
exist for this inter-relationship. Sustainable provision of health services
depends on their being financed, either directly by consumers or from
government revenues. The effectiveness of medical interventions also
depends on other risk factors. There is little point in immunising a child
who has no prospect of eating (hence the significance of the interactive
term in Table 10.3), or indeed one who faces too adverse a home environ-

Figure 10.2 The Mosley–Chen framework for analysing mortality (modified).

Source: After Mosley and Chen (1984).
220 Howard White
ment and a mother ignorant of good child-rearing practices such as the
importance of breastfeeding. For this reason, reducing mortality requires
a good domestic environment and non-hazardous working conditions.
Maternal education has been identified as a crucial mediating variable,
making it more likely that families will avail themselves of improved health
technologies and be open to preventive health practices. The implication
is that rising incomes will not be turned into reduced mortality unless the
mediating factors come into play. Moreover, there is scope to enhance the
‘productivity’ of a given level of income by policies aimed at mediating
factors (though in the end some improvement in socio-economic status is
necessary for further reductions).
Cross-country regressions tend to focus too much on the left-hand box
in Figure 10.2, ignoring the importance of mediating factors which may
be amenable to policy intervention, and which might also contribute to
lower mortality. On the other hand, much of the demographic literature
has focused unduly on the two right-hand boxes of Figure 10.2, document-
ing the immediate causes of death without tracing these back to their
proximate and underlying causes. The aim here is to look at the middle
ground, drawing on case-study evidence of the effectiveness of public
interventions in reducing mortality.
For developed countries the debate on the sources of mortality decline
has revolved around England’s experience in the late nineteenth and
early twentieth centuries and the contribution of McKeown. In England,
infant mortality began to decline around 1900 and that for older children
somewhat earlier. Having initially written of the importance of medical
advances, McKeown (1976) later reversed his position, arguing that better
nutrition was mainly responsible. This argument was hotly contested, and
one authority has concluded that whilst ‘specific therapeutic medical
treatments have played a minor role in mortality reductions in Western
countries . . . relatively little else of the McKeown thesis has survived’
(Preston, 1996: 532). Examination of more recent work bears this out.
Whilst McKeown is invariably mentioned in studies of European mortality,
his position is dismissed as simplistic. To take two recent examples, Woods
points to the role of the ‘health of towns’ movement, especially in ensur-
ing a supply of uncontaminated water, improved milk supply and food
quality, as well as the rise of female education (Woods, 2000: chapter 7).
Riley (2001) argues that success in reducing mortality cannot be reduced
to a single factor, but rather results from a multiplicity of factors which
have varied across time and between countries, and goes on to discuss
public health, medicine, income and nutrition.
Vögele’s (1998) analysis of towns in the UK and Germany provides a
good summary of the most accepted position that emerges from this liter-
ature.9 Mortality increased during the nineteenth century, and this
increase was largely associated with urbanisation. Vögele calls this the
‘urban penalty’ of higher mortality in urban areas on account of their
 Reducing infant and child death 221
squalid living conditions. At the beginning of the last century this penalty
began to be eliminated as conditions improved, most notably sanitation.
He is explicit that specific medical interventions played only a small role,
although others do point out that perinatal death declined with improved
delivery practices.
Mortality decline in Europe took place over a long period, with the
initial stages of industrialisation being associated with increased mortality
due to poor urban living conditions. The situation in developing countries
has been rather different. Where long-run data are available, they suggest
that mortality declines began somewhere around the 1930s, and there has
been no period of increasing mortality with urbanisation. This difference
in experience is due to the availability of immunisation, better access to
water and sanitation, and general availability of health services. Indeed, in
developing countries today there is arguably a ‘rural penalty’ which derives
partly from the worse access to health services of those in rural areas (see
p. 226 and p. 230).
Of particular interest is the situation in countries that have achieved
rapid mortality reduction in a very short time. Two cases are considered
here, China and the Soviet Union. Campbell presents a detailed analysis
of mortality in Beijing since the seventeenth century. The period of inter-
est here is immediately after the Communist Party came to power, from
1949 to 1955. In just six years, child mortality rate fell from 61 to 15 per
1,000 live births, and infant mortality fell from 115 to 56 per 1,000 live
births, with annual rates of reduction of 21 per cent and 11 per cent
respectively (Campbell, 2001: 231). Such rapid declines cannot plausibly
be attributed to economic growth, but rather must result from the atten-
tion paid to the development of health services:

[B]etween 1949 and 1955 the national government focused its atten-
tion on establishing an organisational infrastructure to implement
health policies and control certain infectious diseases. . . . Attention
was focused on pulmonary tuberculosis, parasites, sexually transmitted
diseases and certain other acute infectious diseases. To improve
maternal and child welfare, midwives were re-trained and new delivery
assistants were educated.
(Campbell, 2001: 233)

Measures included the greatly expanded production of vaccine, increased

numbers of clinics and free treatment for all those suffering infectious dis-
eases.
A similar story can be told about the Soviet Union after the Second
World War. Until the 1940s, infant mortality was still high at over 200 per
1,000 live births, with life expectancy at birth little more than 40. Within
twenty years, life expectancy had increased to close to 70 (65 for men and
74 for women), largely on account of reductions in under-five mortality.
222 Howard White
By 1965, infant mortality was 40 per 1,000 live births, an annual reduction
of close to 8 per cent. These improvements are mainly attributable to the
success of the health system in controlling infectious diseases (Shkolnikov
and Cornia, 2000: 254). China and the Soviet Union (and Cuba, not dis-
cussed here but shown in Table 10.2) are not isolated cases. For example,
infant mortality in Czechoslovakia was more than halved between 1950
and 1960, an annual reduction of close to 8 per cent (Blazek and Dzúrová,
2002: 309).
Cases of rapid mortality decline have taken place outside centrally
planned economies, and studies of these cases also highlight the role of
health services. The case of Oman was mentioned above. Another example
is Senegal, which had extremely high mortality rates of around 400 per
1,000 live births in the 1940s. By 1990 this had fallen to 130 per 1,000 live
births, with the decrease accelerating at the end of the 1970s. A study of
Senegalese mortality attributes this accelerated decline to

a new health policy, which emphasised primary health care and was
implemented during this period. . . . The proliferation of health infra-
structures (which had previously been highly concentrated in Dakar)
in the various regions and the implementation of the Expanded Pro-
gramme of Immunisation (EPI) probably contributed significantly. . . .
Continued mortality decreases in the coming years will depend sub-
stantially on the continuation of health programmes, in particular of
the sustained vaccination initiative, improved pregnancy monitoring,
and enhanced conditions of delivery.
(Pison et al., 1995: 155–157)

Studies of less dramatic mortality reductions also identify health services

as being of importance. For example, in their analysis of Ovamboland in
Namibia, Notkola and Siiskonen write that

the child mortality decline happened during the 1950s and this mor-
tality decline can be supposed to have been mainly caused by the
improved health care system. The health care system was based mainly
on the idea of preventing the disease epidemics. With the help of the
missionary health care system the Government was also able to carry
out the vaccination programme. . . . It is clear that the improved level
of education of mothers influenced infant mortality and the threshold
of using health services was lower due to improved education.
(2000: 110–111)

Finally, studies of specific interventions point to the effectiveness of immu-

nisation and other public health measures. For example, Kaseje (1992)
reports falls in infant mortality of as much as two-thirds as a result of anti-
malaria campaigns in Kenya and Nigeria (these falls being much greater
 Reducing infant and child death 223
than those in control communities). He also cites studies that find the
decline to be much less than would have been expected from the number
of malaria deaths prior to the intervention. He argues that these are cases
in which mortality would be high even in the absence of malaria, so that
infants die of something else once malaria is controlled. In such cases, the
elimination of malaria is a necessary step to reducing mortality, but is not
by itself sufficient. In the case of Sri Lanka, Preston says that the ‘best
estimate’ is that close to half the reduction in mortality from 1930 to 1960
can be attributed to the anti-malaria programme (1996: 533).
Case-study evidence thus almost overwhelmingly points to the import-
ance of public health measures, especially those against infectious diseases.
This is not to claim that these are all that matter. As shown by Figure 10.2,
there are a number of mediating factors, of which two are mentioned here.
Caldwell’s frequently cited 1986 paper ‘Routes to Low Mortality in Devel-
oping Countries’ argues strongly that women’s position in society plays a
critical role, and that female education is an important determining factor.
Educated women are more likely, and better able, to take advantage of
clean water, health services, and so on. They are also better placed to influ-
ence the allocation of household resources towards child health. Second,
there is a well-documented simultaneous (i.e. two-way) relationship between
fertility and infant mortality. Policies to reduce fertility, which certainly
include measures to promote economic growth and other means of redu-
cing insecurity, are undoubtedly important for reducing mortality.

The changing pattern of death

The left-hand panel of Figure 10.3 shows that there is a strong relationship
between the under-five mortality rate and the proportion of under-five
deaths occurring before the first birthday, using data from Demographic
and Health Surveys (DHSs) from the late 1980s to 2000. For the countries
Share of neo-natal mortality
Share of infant mortality

Under-five mortality rate Infant mortality rate

Figure 10.3 Relationship between structure and level of mortality.

Source: Based on data from [Link].
224 Howard White
with the highest mortality, over half of under-five deaths occur to those
aged 1 to 4. But for the countries with the lowest mortality rates, fewer than
one in ten of under-five deaths take place amongst older children. In
China, for example, the share of infant deaths in under-five mortality was
68 per cent in 1949, had risen to 80 per cent by 1955 and stood at 95 per
cent by 1990 (Campbell, 2001: 231). The right-hand panel shows infant
mortality and neo-natal deaths. Although the relationship is less strong, it is
the case that amongst countries with the lowest rates of infant death, the
majority of these deaths occur within the first month, whereas for countries
with higher mortality at least half occur between 1 and 11 months of age.
The same point can be made by looking at the experience of individual
countries. Figure 10.4 shows the neo-natal, post-neo-natal and child

Egypt, 1995 Ethiopia, 2000

Mortality rates

Mortality rates

India, 1999 Peru, 2000

Mortality rates

Mortality rates

Uzbekistan, 1996 Zambia, 1996

Mortality rates

Mortality rates

Figure 10.4 Trends in mortality rates for selected countries.

Source: Calculated from data from [Link].
 Reducing infant and child death 225
mortality rates for six countries for five time periods (at a five-year inter-
val). Reading the columns from right to left shows how rates have
changed over time, with the height of the column being approximately
equal to the under-five mortality rate.10 The first four countries – Egypt,
Ethiopia, India and Peru – all illustrate a clear trend of reduction in
under-five mortality. It is notable that the bulk of the reduction comes
from lower child mortality, with the least change in neo-natal mortality.
For example, in Egypt, which most clearly illustrates the pattern, child
mortality fell from 120 per 1,000 live births twenty to twenty-five years
before the survey, to 19 per 1,000 live births in the most recent five years.
Over the same period, neo-natal mortality fell from 63 to 30 per 1,000 live
births, and its share of under-five deaths rose from 27 to 38 per cent.
The exceptions really are exceptions. In Zambia, mortality has been
rising, in large part owing to HIV/AIDS (see Chapter 12 of this volume),
though the decline of health services has also been said to have played a
role (see Hanmer and White, 1999). Children born to mothers who are
HIV positive have an 80 per cent chance of being HIV positive themselves,
in which case they have virtually no chance of living to see their fifth birth-
day. This helps explain the rise in child mortality in that country. The situ-
ation of Uzbekistan exemplifies the social and economic collapse that
took place in the former Soviet Union in the 1990s. Indeed, Uzbekistan
fared better than most of its neighbours, experiencing the least fall in
income (White et al., 2000).11 Nonetheless, the data for the 1990s show a
reversal of the previous downward trend in mortality, with the largest
increase being for infant deaths.
The changing age at which death is likely to occur has implications for
the most appropriate policies to further reduce mortality, since the cause
of death varies by age group. Writing of Africa, Tarver (1996: 66–67) lays
this out as follows:

Most neo-natal deaths are due to congenital abnormalities and

injuries occurring during parturition. . . . In the post-neonatal period
of one to 11 months, malnutrition, respiratory, and diarrheal diseases
claim about 30 to 40 per cent of infant lives. After age one infectious
and parasitic diseases, such as measles and diphtheria, account for
about 25 per cent of all early childhood deaths.

So, child deaths are commonly linked to environmental factors and can
be brought down by campaigns to control diseases such as malaria and
by ensuring good immunisation coverage. Post-natal infants are also
amenable to improvements in environmental conditions, especially the
provision of clean water, though this can matter less if they are breast-fed,
and the other factors related to GOBI-type interventions. But neo-natal
deaths are very closely related to the conditions surrounding birth. In a
country such as India, child mortality is low, so most under-five deaths
226 Howard White
occur amongst infants. A growing percentage of these infant deaths are
neo-natal: between 1972 and 1995, this share grew from 51 to 65 per cent
(Claeson et al., 1999: 33). Antenatal care and attended births thus become
increasingly important in bringing about further reductions in mortality.
The high proportion of child deaths in high-mortality countries (most
of which are in Africa) reflects not only generally poor environmental
conditions but also weak delivery of health services, including immunisa-
tion. The result is a far higher level of communicable diseases than else-
where (Table 10.4). Infectious and parasitic diseases account for 40 per
cent of all deaths in sub-Saharan Africa, compared to just 3 per cent in
China, which is not much more than in developed countries. African
countries have the lowest levels of immunisation overall, with particularly
low rates amongst the income poor. Table 10.5 shows data on child mor-
tality and immunisation coverage for five countries. The population are
divided into asset quintiles, based on ownership of consumer durables and
housing quality. In four of the five cases the percentage of children having
no immunisation coverage is far greater in the poorest quintile than it is
in the top quintile, with a corresponding differential in mortality rates
(Chad is the exception to this). It does not have to be this way, as the case
of Zimbabwe, a country with a well-developed primary health-care system,
shows. However, the mortality differential is also evident for Zimbabwe,
showing that high immunisation coverage alone does not reduce these dif-
ferences – but overall mortality is much lower in Zimbabwe than in the
other countries.
These differences between income groups reflect in part rural–urban
differentials. In Ghana, for example, infant mortality in rural areas is 81
per 1,000 live births compared to 54 per 1,000 live births in urban areas.
For child mortality these figures are 69 and 34 respectively – that is, child
mortality is twice as high in rural areas as in urban ones. Such figures are
not atypical. Greater income poverty in rural areas is one factor here, but
so is worse access to health services. Nearly one-fifth of rural children are
not immunised, compared to only one-twentieth of urban children.
Seventy-three per cent of urban births in Ghana were attended by trained
medical personnel, compared to only 37 per cent in rural areas.

Policies required to achieve the infant and child mortality

Millennium Development Goal
Historical experience shows that it is possible to achieve the rate of mor-
tality reduction required to meet the Millennium Development Goals, but
that this will happen only with substantial public interventions for health
and related areas, notably water and sanitation, and good education cover-
age, especially for women. Will economic growth alone be sufficient to
bring this about? Higher incomes allow people to afford better health
care. But there are economic grounds for the public sector assuming at
Table 10.4 Causes of death for selected regions, 2000

Sub-Saharan Latin America China Developed

Africa and Caribbean countries

Male Female Male Female Male Female Male Female

Number of deaths (000s)

All causes 4,997 4,208 1,967 1,551 5,820 4,497 4,147 3,784
I. Communicable, maternal, perinatal 2,853 2,653 491 345 449 407 300 232
and nutritional conditions
A. Infectious and parasitic diseases 1,986 1,798 296 176 203 144 125 48
B. Respiratory infections 534 444 81 70 158 157 146 156
C. Maternal conditions – 111 – 11 – 9 – –
D. Perinatal conditions 262 234 83 57 69 71 20 14
E. Nutritional deficiencies 71 66 31 31 18 27 9 13
II. Non-communicable 1,200 1,152 1,138 1,086 4,665 3,584 3,547 3,395
III. Injuries 944 402 358 120 707 505 301 157
o/w War 210 154 12 9 1 – – –
Percentage of total deaths
All causes 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
I. Communicable, maternal, perinatal 57.1 63.0 25.0 22.2 7.7 9.1 7.2 6.1
and nutritional conditions
A. Infectious and parasitic diseases 39.7 42.7 15.0 11.3 3.5 3.2 3.0 1.3
B. Respiratory infections 10.7 10.6 4.1 4.5 2.7 3.5 3.5 4.1
C. Maternal conditions 0.0 2.6 0.0 0.7 0.0 0.2 0.0 0.0
D. Perinatal conditions 5.2 5.6 4.2 3.7 1.2 1.6 0.5 0.4
E. Nutritional deficiencies 1.4 1.6 1.6 2.0 0.3 0.6 0.2 0.3
II. Non-communicable 24.0 27.4 57.9 70.0 80.2 79.7 85.5 89.7
III. Injuries 18.9 9.6 18.2 7.7 12.1 11.2 7.3 4.1
o/w War 4.2 3.7 0.6 0.6 0.0 0.0 0.0 0.0
Memo item: total population (millions) 324 327 257 262 656 619 580 612

Sources: World Bank Web site and UN World Population Prospects: The 2000 Revision.
228 Howard White
Table 10.5 Child mortality and immunisation by asset quintile in selected African
countries

Quintile Chad Ghana Mozambique Niger Zimbabwe

Child mortality rate (per 1,000)

Poorest 99 86 111 173 34
Second 106 86 90 239 37
Third 119 62 84 227 16
Fourth 100 41 61 202 25
Fifth 91 30 55 107 15
Whole population 103 61 80 190 26
No immunisation (per cent)
Poorest 65 31 36 57 84
Second 51 18 27 45 85
Third 49 9 26 54 89
Fourth 33 14 5 37 85
Fifth 19 3 1 8 90
Whole population 43 15 19 40 86

Source: Data from World Bank Web site, calculated from DHS data.

least some of the responsibility. Arguing the case for public goods is not
currently fashionable, but there are good reasons to consider public health
investments as falling into this category.
Infectious diseases are, unsurprisingly, infectious, so their prevalence is
a clear case of an externality (something I do, or, in this case, have, that
affects you). Where externalities exist, markets will not bring about the
most socially desirable outcome as there will be under-investment in the
good, in this case actions required to reduce the risk of disease. Immunisa-
tion has a threshold effect, with about 80 per cent coverage required to
remove the possibility of an epidemic.12 There is a strong case for a
greater emphasis on preventive health care, and it makes little or no sense
to attempt cost recovery for such services. Similar arguments apply to
improved water and sanitation. The current fashion is for cost recovery in
this area, partly based on the failure of centralised government services to
maintain facilities. However, other participatory mechanisms, such as
those used in some social funds, have shown that it is possible to get
communities to take responsibility for facility maintenance without requir-
ing full cost recovery (Carvalho et al., 2002). And the economics suggest
that if there are increasing returns to scale, which is likely with infrastruc-
ture investments, a subsidy is required to support the optimal level of pro-
duction (based on marginal cost pricing – this is the starting point for
old-fashioned public-sector economics).
Despite considerable progress under the Expanded Programme of
Immunisation, coverage remains low in sub-Saharan Africa and South
Asia, having fallen in the former region in the 1990s. For sub-Saharan
Africa, coverage increased from 16 per cent in 1980 to 56 per cent in
 Reducing infant and child death 229
1990, but then fell ten points to just 46 per cent in 1999.13 Coverage in
South Asia rose steadily from 23 per cent in 1980 to 57 per cent in 1990,
remaining stable at around 60 per cent thereafter. Other regions have
faired better, achieving rates at or close to 90 per cent by the end of the
1990s. In contrast to other regions, coverage in Eastern Europe fell in the
1980s and early 1990s, but had recovered to 90 per cent by 1999.
Prospects for improving coverage now rest with the Global Alliance for
Vaccines and Immunisation (GAVI). Whilst the programme covers many
of the poorest countries, it is not intended as the sole source of finance.
Rather, it is seen as a catalyst for governments to commit their own expen-
diture, and a complement to bilateral aid funds. However, neither of these
sources has traditionally been prioritised towards the needs of the poor.
The Copenhagen Social Summit in 1995 agreed the 20:20 initiative in
which 20 per cent of government spending and 20 per cent of aid should
go to basic services for the poor, defined as basic health and education,
water supply and sanitation. This agreement was not amongst those used
to draw up the Millennium Development Goals. However, developing
countries ensured that the Millennium Development Goals included Goal
8 concerning global partnership, specifying some goals for their ‘part-
ners’. Indicator 33 is the ‘proportion of ODA [Official Development Assis-
tance] to basic social services (basic education, primary health care,
nutrition, safe water and sanitation)’, but there is no quantified target,
despite one being readily available from Copenhagen. This fact reflects
the unequal nature of the aid ‘partnership’, in which the side with the
power does not bind itself to actual targets. However, the good news is
that since data first became available, the share of Development Assistance
Committee (DAC) bilateral aid for basic health and education has increased
more than eightfold. The bad news is that it was initially less than 0.5 per
cent, and the most recent data (for 2000) put it at a paltry 4 per cent.14
Donor aid to basic health and education remains pitifully low. But
there is at least an upward trend, and the fact that attention is being paid
to the ‘use’ of debt relief funds and that sector programmes are of rising
importance does indicate that a genuine change of attitude has occurred
at least amongst some donors. But reliance on sector programmes means
that government expenditure must be going to the right things, and the
evidence here is not encouraging.
This evidence comes from two types of study: benefit incidence and
public expenditure tracking (PET) studies. Benefit incidence studies use
data on unit costs and use of facilities, taken from household surveys, to
estimate the subsidy (public expenditure, less any user payments) to dif-
ferent groups, most usually income quintiles. The limitation of this analy-
sis is that the data usually assume constant unit costs for a given facility
level (e.g. health clinic), which is almost certainly not the case, thus
imparting a bias which probably understates the skew of resources to the
better off. Table 10.6 shows results for selected countries. In all but one
230 Howard White
case, those in the top quintile are more likely to use all levels of service.
They are much more likely to use hospital facilities than are the poor, and
hospitals receive far more subsidy per patient than do primary facilities.
Thus in all cases the share of public spending on health which benefits
the top 20 per cent is greater than that going to the bottom 20 per cent.
The rich receive a subsidy twelve times that of the poor in the most
extreme case. These results are to a large extent driven by urban bias in
service provision.
PET studies analyse how much spending in a sector reaches the grass-
roots level. There is only one detailed study of health and education,
though several others are in the pipeline. The first such study, for Uganda,
has the most comprehensive results for education, finding that less than
30 per cent of intended non-salary spending actually reached schools in
the early 1990s (Ablo and Reinikka, 1998). The health sector lacked com-
parable expenditure data, but the authors cite another study which sug-
gested major leakage at the health unit level as staff sell drugs in order to
supplement their salaries. In the case of Ghana, Ye and Canagarajah
(2002) report that only 20 per cent of non-salary recurrent spending
reaches health facilities. Reorienting the spending of social sectors from
central ministries to service providers is recommended to address this situ-
ation (in some countries, education civil servants have been made into
teachers).
To close, it should be repeated that this chapter is not arguing that eco-
nomic growth is not important to reducing poverty in all its many dimen-
sions. But growth alone is not sufficient, nor is it always necessary.

Table 10.6 Benefit incidence of public spending on health in selected countries

Quintile shares of:

Primary Hospital Hospital All health

facilities outpatient in-patient a

Bottom Top Bottom Top Bottom Top Bottom Top

Côte d’Ivoire (1995) 14 22 8 39 n.a. n.a. 11 32

Ghana (1992) 10 31 13 35 11 32 12 33
Guinea (1994) 10 36 1 55 n.a. n.a. 4 48
Kenya (1992)b 22 14 13 26 n.a. n.a. 14 24
Madagascar (1993) 10 29 14 30 n.a. n.a. 12 30
Tanzania (1992/1993) 18 21 11 37 20 36 17 29
South Africa (1994) 18 10 15 17 n.a. n.a. 16 17

Source: Castro-Leal et al. (1999).

Notes
a Hospital subsidies combine in- and outpatient spending in Côte d’Ivoire, Guinea, Kenya,
Madagascar and South Africa.
b Rural only.
 Reducing infant and child death 231
Sustainable health services require a strong revenue base, but much can
be achieved in the meantime with external support. And some measures,
such as promoting breastfeeding, are relatively low cost. Government
action can facilitate an improved transmission from growth to poverty
reduction. It can do this by ensuring that growth is itself pro-poor. Target-
ing priority spending is a critical component of any anti-poverty strategy.
Of course, most governments claim to target priority spending, so it is
important to consider whether this simply has not happened because of
lack of capacity.
A government’s ability to carry out these actions is thus crucial. At the
extreme are collapsed states, the importance of which should not be
neglected in any discussion of the Millennium Development Goals.
During times of conflict, health services usually collapse. Immunisation
rates in refugee camps may be very low, and infant mortality rates as high
as 600 per 1,000 live births have been reported in acute emergencies
(Roberts et al., 2001: 12). Conflict prevention is thus a critical underpin-
ning for this, and all other, development goals. Other states are function-
ing but are often believed to lack the capacity to deliver on promises of
universal health and education. There are grounds for questioning such
judgements. Low-income countries, such as China, Cuba and Vietnam, have
put in place effective primary health care. To a lesser extent, Nicaragua and
Zimbabwe have also done so in the past. What is perceived as a lack of
capacity has often been a mixture of resource misallocation, which is
being rectified in response to both internal and external pressures, and
the consequent lack of resources at local level.
I would like to end on an optimistic note. Travelling in countries such
as Ghana, Malawi and Zambia, countries with supposed weak local capac-
ity, I have been struck by the idealism, commitment and ability of many
health and education workers, and of local officials. But unfortunately
these people can work only with the resources to hand, so my optimism
must be muted by the need to reorient aid and government budgets to
meet priority needs. Only if this is done can the Millennium Development
Goals on mortality be met.

Acknowledgement
I am indebted to Richard Jolly and Richard Black for useful comments on
an earlier draft of this chapter. The usual disclaimer applies.

Notes
1 This is the same debate as that between the World Bank and the United
Nations Development Programme (UNDP) during the 1990s (see, for
example, Ranis et al., 2000; Ravallion, 1997; White, 1999).
2 Infant, child and under-five mortality are the most commonly used categories.
232 Howard White
Any cut-off is of course arbitrary, though demographers argue for having two
under-five categories as the cause of death varies by age. The observed rates are
sensitive to the cut-off, especially up to about two years, which is one reason
why the United Nations Children’s Fund (UNICEF) has adopted the under-five
measure.
3 The linear nature of the decline is something of a statistical artefact as linear
interpolation has been used for some series.
4 Hanmer et al. (2002) show that Oman and Saudi Arabia used to be outliers in a
scatter plot of infant mortality against income per capita (logged), but are no
longer so.
5 Also known as universal coverage of immunisation (UCI), the name used by
UNICEF.
6 A similar result is found in the regression reported on pp. 216–217 of this
chapter.
7 The ‘cold chain’ refers to the logistical arrangements to maintain vaccines at
the required temperature from the time of production to use in health clinics.
8 Data reported by UNICEF (2002) show that around thirty developing countries
have achieved substantial reductions in mortality since the 1960s, leaving some
one hundred which have not. The discussion here draws general conclusions
from a selection of successful and less successful cases.
9 It is not possible to do justice to the substantial literature on the historical
experience of the now developed countries. For further examples, see the
papers in Bideau et al. (1997) and Schofield et al. (1991).
10 It is only approximate as the under-five rate is not the sum of infant and child
mortality, but this sum with child mortality adjusted to reflect those not surviv-
ing the first year.
11 However, life expectancy, though not under-five mortality, began to decline in
the Soviet Union in the 1960s (Cornia and Paniccià, 2000).
12 ‘Herd immunity’ at which the disease should not occur at all is achieved with
95 per cent coverage.
13 Data from the UNICEF Web site: [Link]
[Link].
14 Data from the DAC Web site: [Link]/dac/statistics (table 19).

References
Ablo, E. and Reinikka, R. (1998) ‘Do Budgets Really Matter? Evidence from Public
Spending on Education and Health in Uganda’, Policy Research Paper 1926,
Washington, DC: World Bank.
Bideau, A., Desjardins, B. and Brignoli, H.P. (eds) (1997) Infant and Child Mortality
in the Past, Oxford: Clarendon Press.
Blazek, J. and Dzúrová, D. (2000) ‘The Decline of Mortality in the Czech Republic
During the Transition’, in Cornia, G.A. and Paniccià, R. (eds) The Mortality Crisis
in Transitional Economies, Oxford: Oxford University Press.
Caldwell, J.C. (1986) ‘Routes to Low Mortality in Poor Countries’, Population and
Development Review, 12: 171–220.
Campbell, C. (2001) ‘Mortality Change and the Epidemiological Transition in
Beijing, 1644–1990,’ in Liu, T-j., Lee, J., Reher, D.S., Saito, O. and Feng, W.
(eds) Asian Population History, Oxford: Oxford University Press.
Carvalho, S., Perkins, G. and White, H. (2002) Social Funds: Assessing Effectiveness,
Washington, DC: World Bank, OED.
 Reducing infant and child death 233
Castro-Leal, F., Dayton, J., Demery, L. and Mehra, K. (1999) ‘Public Social Spend-
ing in Africa: Do the Poor Benefit?’, World Bank Research Observer, 14(1): 49–72,
February.
Claeson, M., Bos, E. and Pathmanathan, M. (1999) ‘Reducing Child Mortality in
India: Keeping Up the Pace’ HNP Working Paper, Washington, DC: World
Bank.
Cornia, G.A. and Paniccià, R. (eds) (2000) The Mortality Crisis in Transitional
Economies, Oxford: Oxford University Press.
Filmer, B. and Pritchett, L. (1999) ‘The Impact of Public Spending on Health:
Does Money Matter?’, Social Science and Medicine, 49: 1309–1323.
Hanmer, L. and White, H. (1999) Human Development in Sub Saharan Africa, The
Hague: Institute of Social Studies Advisory Service.
Hanmer, L., Lensink, R. and White, H. (2003 forthcoming) ‘Infant and Child
Mortality in Developing Countries: Analysing the Data for Robust Determin-
ants’, Journal of Development Studies, Brighton: IDS, University of Sussex.
Howard, M. and Millard, A. (1997) Hunger and Shame: Child Malnutrition and
Poverty on Mount Kilimanjaro, London: Routledge.
Jolly, R. (ed.) (2001) Jim Grant: UNICEF Visionary, Geneva: UNICEF.
Kaseje, D.C.O. (1992) ‘Malaria in Kenya: Prevention, Control and Impact on Mor-
tality’, in van de Walle, D., Pison, G. and Sala-Diakanda, M. (eds) Mortality and
Society in Sub-Saharan Africa, Oxford: Clarendon Press.
McKeown, T. (1976) The Modern Rise of Population, London: Edward Arnold.
Mosley, W.H. and Chen, L.C. (1984) ‘An Analytical Framework for the Study of
Child Survival in Developing Countries’, Population and Development Review, sup-
plement to vol. 10: 25–45.
Notkola, V. and Siiskonen, H. (2000) Fertility, Mortality and Migration in Sub-Saharan
Africa: The Case of Ovamboland in North Namibia, 1925–90, Basingstoke, UK:
Macmillan.
Pison, G., Hill, K., Cohen, B. and Foote, K. (1995) Population Dynamics of Senegal,
Washington, DC: National Academy Press.
Preston, S. (1975) ‘The Changing Relation Between Mortality and Level of Eco-
nomic Development’, Population Studies, 29: 231–248.
—— (1996) ‘Population Studies in Mortality’, Population Studies, 50: 525–536.
Ranis, G., Stewart, F. and Ramirez, A. (2000) ‘Economic Growth and Human
Development’, World Development, 28(2): 197–219.
Ravallion, M. (1997) ‘Good and Bad Growth: The Human Development Reports’,
World Development, 25(5): 631–638.
Riley, J. (2001) Rising Life Expectancy: A Global History, Cambridge: Cambridge Uni-
versity Press.
Roberts, L., Hale, C., Belyakdoumi, F. et al. (2001) Mortality in Eastern Democratic
Republic of Congo: Results from Eleven Mortality Surveys, New York: International
Rescue Committee.
Schofield, R., Reher, D. and Bideau, A. (eds) (1991) The Decline of Mortality in
Europe, Oxford: Clarendon Press.
Shkolnikov, V.M. and Cornia, G.A. (2000) ‘Population Crisis and Rising Mortality
in Transitional Russia’, in Cornia, G.A. and Paniccià, R. (eds) The Mortality Crisis
in Transitional Economies, Oxford: Oxford University Press.
Skeet, I. (1992) Oman: Politics and Development, New York: St Martin’s Press.
Tarver, J.D. (1996) The Demography of Africa, Westport, CT: Praeger.
234 Howard White
UNICEF (2002) State of the World’s Children 2002, Geneva: UNICEF.
Vögele, J. (1998) Urban Mortality Change in England and Germany, 1970–1913, Liver-
pool: Liverpool University Press.
White, H. (1999) ‘Global Poverty Reduction: Are We Heading in the Right Direc-
tion?’, Journal of International Development, 11(4): 503–519.
White, H. with Calandrino, M., Leavy, J., Litchfield, J., Edwards, C. and Amadi, J.
(2000) Poverty in Middle Income Countries, Report for DFID, Brighton: IDS, Uni-
versity of Sussex.
Woods, R. (2000) The Demography of Victorian England and Wales, Cambridge: Cam-
bridge University Press.
Ye, X. and Canagarajah, S. (2002) ‘Efficiency of Public Expenditure: Distribution
and Beyond’, Africa Region Working Paper 31, Washington, DC: World Bank.
11 Towards reproductive health for
all?
Hilary Standing

Setting international goals: the rise (and fall) of

reproductive health
This chapter differs from others in this part of the book in that it focuses
on the broad area of reproductive health, rather than the specific Millen-
nium Development Goal of reducing maternal mortality. This focus
reflects two concerns. First, as noted in Chapter 3, data on maternal mor-
tality are amongst the most difficult to pin down in terms of obtaining reli-
able information with which to analyse global or even country-specific
trends. This is reflected in Vandemoortele’s discussion of trends on mater-
nal mortality in Chapter 6, where he concludes on the basis of the limited
available data that the goal of a three-quarters reduction in maternal mor-
tality is unlikely to be met by 2015.
Second, it also reflects a particular chain of events in the setting of
targets for reproductive health. In general, international goals and targets
play an important role in focusing attention on key areas of development
need and practice. They are also important arenas within which different
political agendas are played out. This is particularly well illustrated in the
case of reproductive health. Yet the past decade has witnessed two major,
but in some ways contrasting, attempts to put the spotlight on reproduc-
tive health priorities.
The first of these was the International Conference on Population and
Development (ICPD), held in Cairo in 1994 (UNFPA, 1994 and 1999).
This produced international agreement on a vision and set of principles
for reproductive health and a costed Programme of Action which was
adopted by the United Nations (UN) General Assembly (and which is
popularly known as the Cairo Agenda). It departed radically from the old-
style vertically funded family planning and maternal and child health pro-
grammes in emphasising a comprehensive set of sexual and reproductive
health needs for both women and men through the life cycle.
The second has been the more focused attempt to identify priority
targets and indicators in the area of reproductive health through the
International Development Targets and the more recent Millennium
236 Hilary Standing
Development Goals. In the International Development Targets, reproduc-
tive health was represented both as part of a target on reducing mortality
rates, and separately as a restated commitment to the principles of the
Cairo Agenda:

• a reduction by two-thirds in the mortality rates for infants and chil-

dren under five and reduction by three-quarters in maternal mortality by
2015; and
• access through the primary health-care system to reproductive health ser-
vices for all individuals of appropriate ages, including safe and reliable
family planning methods, as soon as possible and no later than the
year 2015.

One of the most significant changes between the Targets and the Goals is
in reproductive health. In the Millennium Development Goals, we find
that while maternal mortality has become a target in its own right (Table
11.1), the goal of reproductive health for all has disappeared. Reducing
the spread of HIV/AIDS has been added as a separate goal, but family
planning appears only as an indicator for monitoring the HIV/AIDS
targets.1
Behind this change lies a tale of continuing political struggle over
women’s rights to sexual and reproductive health services which meet
their needs. Conservative opposition to the Cairo Agenda came from a
coalition led by the Vatican, with countries such as Sudan, Libya and Iran
as allies. They were joined by fundamentalist Christian groups in North
America, which have been particularly instrumental in constraining United
States Agency for International Development (USAID) funding for family
planning programmes, to ensure that they do not support abortion,
however indirectly. USAID is by far the largest bilateral funder of repro-
ductive health-related programmes. With the election of George W. Bush
in 2000, the US administration became the powerful new leader of this
alliance (Berer, 2001; Girard, 2001).
The writing out of reproductive health from the Millennium Develop-
ment Goals took place in the UN Secretary General’s office under pres-
sure from this US-led alliance and without any wider UN debate. Other
early targets of the current US administration have included references to
reproductive health in international agreements on commitments to ado-

Table 11.1 Millennium Development Goal for maternal health

Goal: Improve maternal health Indicators

Reduce by three-quarters, between Maternal mortality ratio

1990 and 2015, the maternal Proportion of births attended by skilled
mortality ratio. health personnel.
 Towards reproductive health for all? 237
lescent health, as well as American funding for UNFPA – supposedly on
the grounds that it supports coercive family planning practices in China
(Girard, 2001; Jacobson and Mallik, 2002).
There is, therefore, a political tension between these two major inter-
national initiatives, both of which many countries simultaneously subscribe
to. The dropping of the reproductive health target from the Millennium
Development Goals must be viewed with concern after so much inter-
national advocacy effort has been expended in pushing reproductive health
up the policy agenda.
This tension is also reflected in the structures for monitoring progress
in meeting goals and targets. The Cairo Programme of Action is moni-
tored through the UN and has strong formal and informal advocacy con-
stituencies keeping the broad vision alive in international and national
arenas.2 It remains to be seen what range of constituencies emerge in
monitoring progress towards the Millennium Development Goals.
The removal of reproductive health as an official Goal has not gone
unchallenged.3 In early 2002 the World Health Assembly passed a very
interesting resolution to guide the work of the World Health Organisation
(WHO) with regard to achieving the Millennium Development Goals.
This was passed with strong support from Latin America, Africa, the Euro-
pean Union and others, and against pressure from US delegates to remove
all mention of reproductive health. It urged member states

to strengthen and expand efforts to meet, in particular, international

development goals and targets related to reduction of maternal and
child mortality and malnutrition and to improve access to primary
health care services, including reproductive health, with special attention
to the needs of the poor and underserved populations.4

This effectively reaffirmed WHO’s commitment to the Cairo Agenda and

other international resolutions on reproductive health and human rights,
and mandated further action on these. It was also a recognition that the
targets laid out in the Millennium Development Goals cannot be met
unless reproductive health services are available and accessible, and coupled
with action on gender disadvantage. However, funding to turn these reso-
lutions into action will be seriously compromised without support from
the world’s richest nation. There is also the problem of the ‘balkanisation’
of reproductive health in the Millennium Development Goals which is
entailed in splitting it between different goals and targets.5
Reproductive health is thus a contested terrain in which are embedded
broader ideological struggles over different visions of human rights and
needs. It is also a terrain which is particularly susceptible to appropriation
by domestic political agendas and to ‘unholy’ alliances (Berer, 2001)
between otherwise unfriendly states.
In the rest of this chapter I will discuss some of the issues raised by both
238 Hilary Standing
the broad and the narrow commitments to reproductive health in the
Cairo Programme of Action and the Millennium Development Goals. I
will argue that there are some essential preconditions for making progress
in either direction. These relate not only to the willingness of developed
countries to implement their financial pledges, but also to the need for
systemic improvements in health-care delivery systems in poor countries in
particular. This in turn requires attention to the wider context of health
sector reforms and how they intersect with the substantive agendas of
maternal health, family planning and the range of services which together
constitute reproductive health care.
The next section outlines the vision for reproductive health, as effect-
ively set out by the Cairo Agenda. The section after that presents indic-
ators of the present situation in reproductive and maternal health. Two
further sections discuss health sector financing issues, and are followed by
a section that analyses performance of reproductive health in light of the
recent reforms in financing. The likely future of reproductive health, given
experience so far, is then examined.

The Cairo Agenda: a comprehensive vision for reproductive

health
The concept of reproductive health is now so ubiquitous that it is some-
times difficult to remember that it has quite a recent history in inter-
national parlance. It effectively came into being with the commitments
made at the 1994 International Conference on Population and Develop-
ment in Cairo (UNFPA, 1999). Cairo was the most recent of the decadal
world population conferences, which had been extremely influential in
setting the international tone of the debate about health, population and
development.6 Earlier conferences held in Bucharest and Mexico by and
large upheld a majority consensus that population policies meant control-
ling population growth in developing countries as a way of creating sus-
tainable development.
Whilst not necessarily entailed in the language and resolutions of the
earlier conferences, population policy in the 1970s and 1980s translated
into narrowly focused and aggressively targeted family planning pro-
grammes. The overriding concern was with meeting demographic targets,
and major donors funded large-scale population programmes in many
developing countries. These programmes were vertical in the sense that
they were dedicated to contraceptive service delivery and functionally
separate from other health programmes, such as maternal and child
health (MCH). They frequently received the lion’s share of external
funding for basic health services.
Programmes were heavily driven by numerical targets for family plan-
ning workers, coupled with incentives for women to use long-term or non-
reversible methods of contraception. In a number of programmes this led
 Towards reproductive health for all? 239
to a strong bias towards methods such as sterilisation at the expense of
reversible methods, concerns about coercion and lack of informed
consent, and encouragement to family planning workers to fiddle the
books for fear of having their salaries cut if they could not meet their
targets (Hartmann, 1995; Sen et al., 1994).
In the 1980s there was considerable disquiet about this approach, and a
coalition for change developed across advocacy groups, particularly women’s
health groups, NGOs and some international agencies (Sen, 1999). Two
key demands were articulated: an end to incentive-based family planning
targets and methods in favour of a menu approach and informed choice,
and a plea for integration of family planning programmes into the frame-
work of primary health care. Behind this lay much broader debates about
the causes and consequences of high population growth in poor coun-
tries, about human rights in the area of fertility decision-making, and
about the need for a stronger developmental approach to population
issues, including issues of gender equality.
Cairo was thus a watershed, in that it sealed a paradigm shift that had
been gradually taking shape in international and national thinking on
population issues (Box 11.1). A broad coalition for change was able to
push through at least a rhetorical commitment to a human rights approach
to health, social well-being and gender equality. This came together in a
new vision of reproductive health and rights.
It is important to note that agreements reached in Cairo with respect to
gender and reproductive rights were built upon earlier agreements on
women’s human rights that had been reached at the UN conference on
human rights in 1993 in Vienna. These in turn were reflected in the com-
mitments to gender equality and reproductive health that were reached at
the Fourth World Conference on Women, which became known as the
Beijing Platform for Action. Further reinforcement of the Cairo agenda
on reproductive health and rights came in the macroeconomic agenda of
the World Summit on Social Development. This produced the 20:20
agreement on the amount of funding which countries should allocate to
the social sector, to be matched by the same commitment from the inter-
national community.7

Box 11.1 Key moments in the development of the reproductive health approach

1974 World Population Conference, Bucharest

1984 World Population Conference, Mexico City
(1993 UN conference on human rights in Vienna)
1994 International Conference on Population and Development, Cairo
1995 Fourth World Conference on Women in Beijing
World Social Summit on Development, Copenhagen
1999 Cairo (ICPD⫹5) Review
240 Hilary Standing
These various conferences and summits are politically significant in
that their agreements become enshrined in UN conventions and resolu-
tions, countries can opt to sign up to them, and international mechanisms
are put in place to monitor them. For instance, in 1999 the Cairo +5
review looked at progress with the Cairo Agenda, including a detailed
examination of successes and failures in specific countries (Correa and
Sen, 1999).

What is the Cairo Agenda?

The Cairo Agenda – or, to give it its official title, the ICPD Programme for
Action – produced the following definition of reproductive health:

Reproductive health is the complete physical, mental and social well-

being in all matters related to the reproductive system. This implies
that people are able to have a satisfying and safe sex life and that they
have the capacity to have children and the freedom to decide if and
when to do so. Reproductive health care is defined as the constella-
tion of methods, techniques and services that contribute to reproduc-
tive health and well-being by preventing and solving reproductive
health problems.

We may note here the distancing from demographic objectives and targets.
The definition also in part follows closely the wording of the WHO decla-
ration on the concept of health.
It also carries a set of associated rights. Reproductive rights are defined
as the rights of couples and individuals to:

• decide freely and responsibly the number and spacing of their chil-
dren and to have the information, education and means to do so;
• attain the highest standards of sexual and reproductive health; and
• make decisions about reproduction free of discrimination, coercion
and violence.

These are matched by a similar set of definitions and rights for sexual
health.
Reproductive health may include the following services:

Family planning counselling, information, education, communication

and services, pre-natal care, education and services; safe delivery and post-
natal care, especially breastfeeding and women’s health care, preven-
tion and treatment for infertility; abortion care, including prevention
and management of consequences, treatment of reproductive tract
infections (RTIs), sexually transmitted infections (STIs) and other
reproductive health conditions, information and education on human
 Towards reproductive health for all? 241
sexuality, reproductive health and responsible parenthood; referral for
family planning, diagnosis, and treatment for pregnancy complica-
tions, delivery and abortion, infertility, RTIs, breast cancer and cancers
of the reproductive system, and STIs/HIV/AIDS.
(Cairo Programme of Action, para 7.6)

The Cairo reproductive health vision is based on some key concepts

and principles. First, it links sexual and reproductive health to broader
human rights for both women and men. This is particularly apparent in
the changed language around fertility and contraceptive choice. But it
also underpins a shift towards seeing health needs as implying associated
rights, such as the right not to die avoidably in childbirth.
Second, it signals a whole-lifespan approach to reproductive health.
Still, to a large extent, reproductive health is seen as the aggregate of
maternal and child health programmes and family planning for married
women of reproductive age only. The lifespan approach takes account of
the need to cater for the reproductive health needs of adolescents, unmar-
ried people and people in the post-childbearing years.
Third, there is a focus on both women’s and men’s roles and responsibil-
ities in securing women’s rights and health, particularly in respect of fertil-
ity decisions, protection against the risk of sexually transmitted illnesses,
and improving maternal health and nutrition. Increasingly, with the
HIV/AIDS epidemic and the associated role of other STIs as a co-factor in
its spread, the reproductive health agenda has widened further to empha-
sise men’s reproductive and sexual health needs.
Fourth, women’s empowerment is seen as an intrinsic aspect of securing
their health human rights. Improvements in reproductive health are thus
linked to a wider gender equality agenda, including addressing the gender
gap in education (see Chapter 9) and empowering women as citizens.
Finally, there is a strong statement of principle on the need for univer-
sal availability and accessibility of reproductive health services, including
services for underserved groups and minorities such as the poor and
adolescents. Related to this is a stress on the quality of services and,
particularly, respect for clients regardless of gender, class, ethnic or other
differences.
This Agenda represents a considerable achievement for coalition-
building at national and international levels. As noted above, it was negoti-
ated in the teeth of considerable resistance from an alliance of religious
conservatives from countries in both the North and South. Issues such as
abortion (which is a major cause of premature death where legally sanc-
tioned and accessible services are unavailable) and the rights of adoles-
cents to reproductive health services continue to be highly contentious
and contested. A number of countries, particularly in the Middle East,
have not endorsed the Cairo Agenda. Nevertheless, it moved the debate
on population and health decisively in a different direction.
242 Hilary Standing
Reproductive and maternal health: how are they faring?
In this section I will look at some of the main status indicators in repro-
ductive health at the start of a new millennium, focusing particularly on
maternal health. This is less easy than may be imagined. Probably the
most reliable figures are on contraceptive prevalence rate, due to the
major efforts of national family planning programmes over the past few
decades. Until recently, reproductive health status indicators have been
poorly defined compared to other areas of epidemiology. Few can be
described as accurate as there are huge problems in data collection and in
availability. What is most difficult to answer is how things have changed
over time. Reliable and appropriately disaggregated time series data are
rarely available. For instance, figures on maternal mortality are difficult to
interpret as much of the increase noted in the recent figures is thought to
be due to better and more accurate reporting.
The Cairo Programme of Action, and subsequently the Beijing Plat-
form of Action agreed at the 1995 World Conference on Women, precipit-
ated considerable efforts to develop appropriate reproductive health
indicators to monitor progress in fulfilling these goals. There has been
something of a tension between the view that countries should select
indicators most appropriate to their needs and capacities to collect data,
and the needs of international agencies to have information for inter-
national comparability and monitoring. Consensus on an accepted set of
indicators has thus been elusive, but WHO has used an expert panel
method for identifying a limited number of robust general indicators for
reproductive health in a given setting which can ‘stand for’ the broader
concept of reproductive health (WHO, 1997).8
However, many countries continue to lack the capacity and infrastructure
to collect such data through their existing health information systems.
Much of what we know about reproductive health status therefore relies
upon snapshot surveys carried out by different agencies, often using dif-
ferent methodologies and definitions. The most commonly used global
indicators relevant to reproductive health are the maternal mortality ratio
(MMR) and the contraceptive prevalence rate (CPR).
The maternal mortality ratio figures in both the International Develop-
ment Targets and the Millennium Development Goals.9 It is not difficult
to understand why. On average, one woman dies every minute from preg-
nancy or childbirth. Maternal mortality, like child mortality, provides a
telling proxy for the effects of poverty, inequality and lack of accessible
health services.10 More than 99 per cent of the estimated 585,000 deaths
annually from pregnancy and childbirth causes occur in developing coun-
tries. In developed countries, on the other hand, the risk of dying during
pregnancy or childbirth has declined by fiftyfold over the past seventy
years. Or, to put it another way, the lifetime risk for a woman of dying
from pregnancy- and childbirth-related causes is 1 in 4,000 in Western
 Towards reproductive health for all? 243
Europe and 1 in 48 in developing countries. In some sub-Saharan African
countries it is as high as 1 in 9 (Gelband et al., 2001; WHO, 2001a).11 This
reflects both the high level of risk from pregnancy and the high numbers
of pregnancies per woman.
The true picture may be even worse than this. Only seventy-eight coun-
tries (less than half) routinely record cause of death, and those that do
not tend to be the very poorest, where childbearing is the most risky. As
experts have noted, MMR estimations should be treated with caution.
Routine vital registration data are lacking in many developing countries,
and maternal deaths are generally thought to be under-reported for
various cultural, definitional and logistical reasons. Indeed, the first global
estimates of maternal mortality were not made until the late 1980s. Fur-
thermore, although the global figures are high, maternal deaths are relat-
ively rare events and accurate monitoring requires very large data sets,
which are expensive to collect (UNICEF, 1997).
But this is only part of the picture on maternal health. Beneath this
stark picture lies another layer of maternal morbidity – illness and disabil-
ity which does not result in death – and for which we have very little reli-
able information. However, we know that this affects many more women,
often for the rest of their lives (WHO, 1999). Maternal deaths and ill
health also have serious consequences for child survival. Maternal death
often results in the death of the baby, and surviving children have poorer
survival prospects than those with mothers who are alive.
Given the difficulties of getting reliable figures on maternal mortality,
trained assistance in childbirth is often used as a proxy for measuring
progress in reducing maternal mortality as it provides a measure of access to
essential maternal health services (WHO, 2001a). It is now one of the indic-
ators for the Millennium Development Goal on maternal mortality. Simms
et al. (2001) analysed data on this from Demographic and Health Surveys
(DHSs) in twenty-two developing countries. These showed national declines
in fifteen of these countries and improvements in only seven of them. The
picture is very mixed, with some countries and regions making progress and
others lagging or experiencing actual reverses, particularly in sub-Saharan
Africa. The impact of HIV/AIDS on health status and on health resources
in severely affected countries cannot be overstressed (see Chapter 12).
A similar mixed picture emerges in respect of contraceptive prevalence
rates. World-wide, it was estimated that in 1993, 40 per cent of reproductive-
age couples still did not have access to modern contraceptive methods.
But again, this disguises wide regional variation, as Table 11.2 indicates.
Africa again shows the greatest problems of lack of available, accessible
and utilised services. In West Africa the rate of contraceptive use is esti-
mated at only 8 per cent (UNFPA, 2001). This contrasts dramatically with
East Asia, where the very high rates of use of modern contraceptives partly
reflect historically high levels of compulsion, such as in the one-child
policy in China.
244 Hilary Standing
Table 11.2 Levels of current contraceptive use by major areas

Area (%) All methods (%)

World total 62 56
More developed regions 70 59
Least developed countries 60 55
Latin America/Caribbean 69 60
Africa 25 20
Eastern Asia 82 80
South Asia 48 29

Source: Compiled from UN Population Division (1998) and UNFPA (2001).

Prevalence rates are generally rising. In two-thirds of the countries for

which trend data are available, prevalence has increased by ten percent-
age points over the past decade (UN Population Division, 1998). However,
there is much that remains unsatisfactory about family planning delivery.
Method mix and choice continue to be very constrained in most develop-
ing countries, with female sterilisation accounting for nearly 40 per cent
of contraceptive use, as compared to 12 per cent in developed countries.
Service quality of reproductive health services generally is often of poor
technical quality and delivered rudely by inadequately trained, supervised
and remunerated staff (Hulton et al., 1998; Simmons and Elias, 1994).
Some of the key causes of this are the collapse of basic health services
as a consequence of economic crisis and government failure, endemic
conflict and war, and rising levels of HIV/AIDS infection placing intoler-
able burdens on formal health care. These factors all point to the need to
address much larger agendas on poverty, governance and international
responsibility in the context of global health crises such as HIV/AIDS.

Who pays for reproductive health?

In the past fifteen years, health sector financing has been a major area of
reform in most developing countries. The main shift has been towards a
far greater mix of financing sources and mechanisms for cost recovery
beyond public finance. There has been a move to much greater reliance
on private financing through encouraging greater official private-sector
involvement (Kutzin, 1995). The Cairo Programme of Action itself called
for a mixed financing strategy for mobilising resources nationally from the
private sector, along with the selective use of user fees and other forms of
cost recovery.
 Towards reproductive health for all? 245
What is happening internationally?
Cairo produced a set of projections about costs, together with commit-
ments from the international community and national governments. In
1995 the bilateral donor commitment to population assistance was US$1.4
billion, which was US$3.6 billion short of the total bilateral and multilat-
eral commitments for implementing the Cairo Programme of Action by
2000 (Forman and Ghosh, 1999). A total of 73 per cent of this came from
just four donors: the USA, the UK, Germany and Japan. It is clear on the
basis of this that the 2015 target of US$21.7 billion stands no chance of
being met.
Because reproductive health is not a sector but consists of different pro-
grammes embedded in a larger health system, it is difficult to plot pre-
cisely how much goes to it. But figures from 1990 estimate that 46 per cent
of external assistance to health and population sectors went to general
health services and 46 per cent to reproductive health. Of this, 42 per cent
went to family planning. Safe Motherhood programmes received just 0.2
per cent of total funds. As the Safe Motherhood initiative has been the
main international agency response to maternal mortality, this does not
give cause for optimism on funding commitments for the Millennium
Development Goal (Goodburn and Campbell, 2001).
Overall, grant aid appears to be declining, but international assistance
in the form of ‘soft’ loans from the World Bank and targeting for specific
diseases is increasing (Walt et al., 1999). Loan finance favours certain kinds
of reform policies in financing – including greater use of the private sector,
user charges at point of service and a move away from comprehensive
primary health care to basic service packages of selective interventions.
Disease targeting tends to favour a vertical approach to health problems.
Most recently, the Commission on Macroeconomics and Health, which
was a major international initiative to examine the health situation and
needs of poor countries in the context of global poverty reduction targets,
argued for an increased international investment in health of US$27
billion per year over the next five years (WHO, 2001b). Health, and
particularly the improvement of basic health services for the poor, has
come back onto the international agenda, owing to concerns about the
potentially destabilising effects of HIV/AIDS and the recognition of the
close links between poverty and poor health. It remains to be seen how far
those concerns are translated into firm commitments and how much pri-
ority is given to reproductive health.

What is happening nationally?

Again, the same difficulty of disaggregation applies, but the general
picture on the proportion of national public expenditure going to repro-
ductive health is almost certainly a mixed one. A recent series of country
246 Hilary Standing
studies on reproductive health financing finds that health and population
sector financing had increased in Bangladesh, Egypt and South Africa
post-1994 (Forman and Ghosh, 2000). However, it fell in Tanzania and
also in Mexico and Indonesia owing to economic crisis. In Tanzania the
bulk of earmarked funds are for contraceptive procurement, and in 1996,
98.5 per cent of the country’s population and reproductive health expen-
ditures were externally funded by four main donors, leaving little room
for national ownership of priorities. Some countries such as South Africa,
which are far less dependent on aid, have made efforts to shift resources
more towards primary care, which tends to benefit certain kinds of repro-
ductive health interventions.
Some of the very poorest countries are experiencing serious financing
problems. I noted the move towards basic service packages in a number of
countries. These are an attempt to provide a basic universal minimum of
cost-effective health interventions under conditions of financial resource
constraints. They are generally favourable to basic reproductive health care
as they tend to cover antenatal and obstetric care. However, a recent analy-
sis of health expenditures across forty of the poorest countries, mainly in
sub-Saharan Africa, showed average health expenditure per capita of
under US$10. This is up to 40 per cent below the level necessary to fund
the World Bank-recommended basic service package (Simms et al., 2001).
Such underfunding underlines the importance of the absolute lack of
financial resources in some regions. The spending base in most African
countries and in some Asian ones is still extremely low. In sub-Saharan
Africa, average incomes per capita are now lower than they were in the
late 1960s. This problem of low incomes is exacerbated by problems of
absorptive capacity for higher levels of external aid in countries where
governance and institutional structures are in a poor state.

So, who then is paying for reproductive health care?

In the health sector generally over the past decade there has been a shift
towards greater household expenditure on health care through the intro-
duction of user charges in the public sector, increased need to make infor-
mal payments to obtain treatment from public facilities, and increased use
of private-sector providers (Bloom and Standing, 2001; Kutzin, 1995).
This pattern is also present in relation to payment for reproductive health
care. Two examples illustrate this.
First, a study in Uganda looked at how households financed health care
in poor rural communities (Lucas and Nuwagaba, 1999). These researchers
found that many households face difficulties with the multiplicity of
demands for cash payments for services, notably in health and education.
Women faced particular difficulties as they rarely had access to cash, yet
mostly had to take responsibility for both their own and their children’s
health. At issue was not so much official user charges per se but the wide
 Towards reproductive health for all? 247
range of informal, illicit payments, which were effectively a user subsidy to
underpaid health workers and could be several times the official cost of
the service.
Another study looked at the hidden costs of maternity care in Dhaka,
Bangladesh (Nahar and Costello, 1998). Utilisation of public maternity
facilities is very low in Dhaka compared to other South Asian cities (less
than 15 per cent). This is despite the existence of a supposedly free
service. The authors examined the actual costs incurred by families in
using such facilities. These included informal payments to hospital staff,
drugs, and the costs of travel and food expenses. They found that the
mean cost for a normal delivery was a quarter of the average monthly
household income. The mean cost of a Caesarean operation was almost
equivalent to the average monthly income.

Meeting the costs – and improving access and quality

These kinds of findings raise serious issues about how to enable poor
people and, particularly, poor women to get access to affordable and com-
petent care. Doing so means looking at the need for cost exemptions for
essential interventions and improving the capacity of the very poorest coun-
tries to be able to finance these. It means looking at a range of other
mechanisms such as community-based and social insurance and medical
safety nets for the poorest. It also entails working on the supply side to
provide the right kinds of incentives to poorly paid health workers to
provide the services they are contracted for. Finally, it also means addressing
why people, including the poor, increasingly resort to private-sector services
(ranging from for-profit providers, through NGOs, to traditional healers).
One of the recurring themes here is that of quality and convenience.
It is important therefore to acknowledge that there is not a one-to-one
relationship between the amount of money spent on health care and out-
comes of better health. There is considerable variation across countries in
the quality of services provided at given levels of health expenditure per
capita (Filmer and Pritchett, 1997; Svennson, 1997).
First, provision of effective services is not just a question of resources.
Indeed, reproductive health advocates have also questioned whether the
funding targets for Cairo were realistic, arguing that better use of existing
resources is needed (DeJong, 1999; Petchesky, 2000). Health systems also
have to function with a reasonable level of efficiency and some minimum
probity. This points to the oft-stated need to improve institutional and
government accountability. Generally speaking, the poorer the governance,
the worse the services that are provided. Health service delivery on the
ground will not function well if drugs and equipment are siphoned off at
higher levels, if there is no effective oversight or supervision of health
workers, if salaries do not get paid and if officials responsible for these states
of affairs are beyond any bureaucratic or civil accountability. Reproductive
248 Hilary Standing
health services are no exception to this, and making them more effective
has also to be part of a wider agenda on governance and accountability.
Second, Cairo drew attention to another broader agenda – that of
gender equality and women’s empowerment. Analyses of the causes of
maternal mortality have drawn attention to the various delays which result
in obstetric emergencies turning into obstetric deaths. The first of these
delays is in decision-making related to reproductive health behaviour and
access to health care at household level. It is important to understand how
household and community power structures affect decisions, and how this
can lead to delays in seeking treatment, or to denial of care. There are
many examples of situations where pregnant women in emergency situ-
ations have not been able to get treatment owing to the absence of their
husband to give approval in accordance with law or social norms, with
tragic outcomes. This again suggests that we have to look to a broader
political and advocacy agenda around enabling women to take greater
control of decisions which affect their lives (Presser and Sen, 2000).

How has reproductive health fared in the context of health

sector reforms?
The Cairo Agenda inaugurated a hard-fought-for, comprehensive concept
of reproductive health which went beyond family planning to encompass
the lifespan health needs of women and men in relation to all aspects of
human reproduction. It was a major achievement in terms of getting inter-
national agreement among such a diverse collection of stakeholders. At
the same time, over the past decade most countries have been involved in
comprehensive reforms of their health sectors in response to the crises
and pressures already noted. Much of this process has been driven by
donors and international agencies and has entailed, in addition to financ-
ing, a wide range of institutional and regulatory reforms. What has hap-
pened to reproductive health in this process?
The Cairo Agenda owes its existence and vitality largely to the women’s
movement and particularly to women’s health advocacy groups in both
the South and North, which found ways of exploiting political spaces to
get progressive policies onto the international agenda. This entailed
working through relevant international conventions (e.g. the Convention
on the Elimination of All Forms of Discrimination against Women,
CEDAW) and pushing through resolutions in international forums such
as the Beijing Women’s Conference. It enabled women’s health groups
and NGOs to use the language and (albeit less successfully) the legal
apparatus of human rights in advocating for policies. For instance, an
issue such as maternal mortality rates in poor countries has been reframed
not as a health issue per se but as a rights violation.
The rights discourse has both strengths and limitations. It has proved
extremely powerful as an international advocacy tool. It has drawn atten-
 Towards reproductive health for all? 249
tion to the shocking neglect of even basic health entitlements for many
poor people. It opens up the possibility for a greater voice amongst those
lacking such entitlements (via demands from ‘below’ as opposed to need
identified and defined from ‘above’). Currently its limitations stem from
the enormous difficulty of any kind of enforcement. Conventions and
resolutions ratified by countries depend on self-regulation and on inter-
pretation according to local circumstances. Also, as we have seen, rich
countries have not exhibited the commitment to providing resources to
poor countries which would make compliance more feasible.
The Cairo Agenda embodies a vision of what should be achieved. It has
not proved so easy to translate into practice. From a sectoral point of view,
reproductive health has been concerned more with services and how they
are delivered. At a broader level, reproductive health advocacy has pio-
neered approaches to improving gender inequality as a prerequisite for
improved reproductive health. Neither of these has found space in health
sector reform programmes. For their part, reproductive health advocates
have only recently begun to understand the importance of engaging with
health sector reform processes (Standing, 2002).
A considerable gap remains between on-the-ground service delivery issues
with which reproductive health is associated, and system-level approaches to
health sector reform. There are, for example, a very large number of micro-
level initiatives in sexual and reproductive health. There is a lot of good
experience, particularly amongst NGOs, of delivering quality services, espe-
cially to female clients. But there is very limited experience of, or capacity
for, scaling up. Experience of delivering reproductive health services in the
comprehensive (i.e. Cairo) sense is very limited and has mostly been con-
fined to demonstration projects. Most programmes lack the capacity to
deliver integrated services (Mayhew et al., 2000; UNFPA, 1998).
The situation is changing quite slowly. In particular, management and
organisational capability needs to improve, and service provision needs to
be able to respond more effectively to local priorities. For instance, cervical
cancer is currently one of the fastest-growing health problems amongst poor
rural women in developing countries, but few if any such countries have
facilities and trained staff able to diagnose, refer or treat it. There has also
not been much progress on developing programmes on men’s reproductive
health consistent with the vision of reproductive health for all.
There are thus several reasons for the limited progress in realising the
Cairo Agenda. Reproductive health has been framed within a different lan-
guage from that of health sector reform. Reproductive health speaks within
human rights and women’s empowerment discourses, whereas health
sector reform is located firmly in the language of management and tech-
nical inputs. Whilst the women’s health movement has made very good
use of the tools of international advocacy, there has not been sufficient
dialogue with the national and international agencies driving health
sector reform. There is insufficient understanding on both sides as to how
250 Hilary Standing
good-quality reproductive health services can be expanded in the context
of health reforms as they are currently being framed and implemented.
Similarly, reproductive health has been focused on service delivery
issues, to the neglect of broader health systems thinking. As Fonn et al.
(1998) point out, a malfunctioning system cannot work for a woman in
labour when it does not work for a man with typhoid either. However,
systems issues are quite hard to address. Reproductive health tends to
need a visionary approach; it is not a technical area or sector with a
budget attached. One problem is where reproductive health, and exter-
nally driven programmes like Safe Motherhood, lie in health system
terms. Largely based in vertical programmes such as family planning and
maternal and child health, its components are often split between differ-
ent ministries and sectors, producing stakeholder conflicts between differ-
ent lines of management.
Another problem is the dominant focus in health reforms on the role
of the public sector and the neglect of increasing pluralism of health
providers. The private sector plays a very significant role in reproductive
health service delivery, often in areas where women find it most difficult
to access services, such as the provision of abortions. A systems approach
to reproductive health itself needs to take a broader view of the concept of
a system, going beyond the assumption that needs are wholly or largely
met by public-sector provision.
Progress continues to be restricted by problems of data availability. Few
data on reproductive health are available disaggregated by age, urban/
rural, class/income, religion/culture or ethnicity. Country studies of pro-
gramme implementation suggest little serious attention by policy-makers.
There are no agreed core indicators for monitoring a rights-based
approach to women’s health, as advocated, for instance, by the Beijing
Platform for Action. One advocacy organisation has suggested that wider
indicators of women’s health status should be developed, such as the
degree of gender-based violence (ARROW, 2001). The same organisation
also points to the need to develop a monitoring framework for financial
indicators which can both differentiate spending on specific services, and
monitor spending on comprehensive services from a reproductive health
point of view. Some countries, for example Bangladesh, are beginning to
develop the capacity to do this through the tool of the Public Expenditure
Review (PER) (Sen and Ensor, 2001).

Reproductive health since Cairo: progress and future steps

Despite these limitations, the recent Cairo ⫹5 review process, which exam-
ined progress in fifteen countries signed up to the Cairo Programme of
Action, noted some promising signs of progress (Correa and Sen, 1999).
First, there is the less-good news: reproductive health programmes
remain heavily dependent upon international assistance. In a number of
 Towards reproductive health for all? 251
countries, national programmes have simply been renamed as ‘reproduc-
tive health’ programmes but the bulk of resources are still going to family
planning. The review notes that reducing maternal mortality requires
investments in primary health programmes to be combined with the
improvement of referral systems and obstetric assistance. But currently
donors are reluctant to fund infrastructure, and structural adjustment
requirements have limited domestic investment in facilities and staff.
The findings also indicated that effective improvement of reproductive
health services has been very limited, especially in the case of urban poor
and rural populations. In many contexts, family planning and maternal
and child health programmes continue to be vertically organised and have
not established linkages with other programme areas, such as HIV/AIDS or
screening for reproductive cancers. Other recurrent obstacles to improving
the quality of services in either family planning or reproductive health
more broadly are described as the inadequate training, bureaucratic
mindset, and gender-insensitive attitude of health managers and providers.
More optimistically, in several countries efforts are being made to over-
come the lack of integration between the various components of reproduc-
tive health, and greater attention is being given to maternal mortality. For
instance, in India, policy-makers, providers and health activists are strug-
gling with the tremendous challenges of turning upside down a long-estab-
lished vertical and narrow family planning programme. In others, initiatives
are being developed to respond to adolescent reproductive health needs.
Encouragingly, in all countries examined, policy-makers, NGOs and the
media were found to be talking more openly about issues such as gender
equality and violence against women (Correa and Sen, 1999).
We have seen that the Cairo Agenda represents a very ambitious and
exciting attempt to rethink what is needed to produce good reproductive
health for all. We have seen equally that there is a very long way to go.
And this in itself has provoked a debate among reproductive health advo-
cates. This broad vision of reproductive health provides a compelling
vision to aspire to, but perhaps it is not realistic for all countries, certainly
in the time-frame of the Millennium Development Goals (DeJong, 1999;
Petchesky, 2000). Where resource and capacity constraints are severe,
there are strong arguments for focusing resources on priority populations
(e.g. poor rural women) and on key target areas such as maternal health.
Reproductive health is thus likely to remain a flexible construction for
the foreseeable future. Some middle-income countries can, and should be
expected to, take a more encompassing view and to expend resources
accordingly. Others will struggle to manage the most pressing reproduc-
tive health problems. An obvious point is the overwhelming impact of
HIV/AIDS on epidemiological profiles, health systems and coping strat-
egies in severely affected poor countries. This means that national plan-
ners and health advocates need to make their own context-specific
assessments of priority reproductive health needs.
252 Hilary Standing
Conclusion
In examining reproductive health, this chapter has noted the inherently
political nature of targets and goals in international development. The
Cairo Agenda and the reproductive health-related Millennium Develop-
ment Goals represent two different ways of setting international goals.
These differences relate to their respective histories. Whereas the Cairo
Agenda was advocacy led, the Millennium Development Goals were driven
by international bureaucracies. Careful coalition-building enabled the
Cairo advocates to drive through a progressive consensus against opposi-
tion from some national interests. Such advocacy was absent from the
setting of both International Development Targets and Millennium Devel-
opment Goals. This weakened the capacity of the bureaucracy to with-
stand pressure from a concerted group of countries and conservative
interests, resulting in the dropping of a long-agreed international health
goal. Despite this, reproductive health remains a powerful vision which
has recently been reaffirmed by the World Health Organisation in its
approach to the implementation of the Millennium Development Goals.
In general, this chapter has cautioned pessimism that the targets of
Cairo and the Millennium Development Goals can be met within their
respective timeframes. Financing of reproductive health is complex and
draws from a wide range of sources. International pledges for Cairo have
not been met and rich countries do not have a good record of meeting their
commitments to development aid. The extremely low levels of resources
allocated to the health sector in some of the poorest countries mean that
it is not possible to provide even basic universal coverage of essential
reproductive health services. Achievements over the past decade have
been mixed and not always easy to track, as reliable indicators are lacking
and data are missing or unreliable.
From the point of view of the user, national financing strategies
increasingly rely on a range of different mechanisms: essential services pack-
ages, user charges, insurances and safety nets. These different mechanisms
potentially have different impacts on women’s capacity in particular to
access reproductive health services, and it is important that policy-makers
are reminded of this. People are paying more for health services and they
are increasingly using the private sector, and this again needs more policy
attention than it is currently receiving.
At the same time, the link between greater resources and better health
outcomes is not necessarily direct and progressive. Funding needs to be
more judiciously targeted. Factors such as poor governance and account-
ability lead to low-quality, underutilised services. High levels of gender
inequality are also major determinants of poor reproductive health out-
comes and cannot be tackled by the health sector alone.
I have also argued that the different discourses of reproductive health
and health sector reform have contributed to slow progress in providing
 Towards reproductive health for all? 253
and improving reproductive health services. There has not been sufficient
engagement between the respective protagonists. Yet reforming health
systems is a prerequisite for improving reproductive health care and address-
ing the health-care delivery constraints which contribute to high levels of
maternal mortality.
Finally, I have suggested that we need to retain the broad vision of
reproductive health embodied in Cairo while attending to and respecting
priorities at different levels of developmental capacity. There continues to
be a key role for advocacy in keeping reproductive health on international
and national agendas. Reproductive health will always carry a political
charge as it concerns broad human rights and gender equality issues, yet
the narrower targets of the Millennium Development Goals in maternal
health, which place specific emphasis on the maternal mortality ratio, will
not be achieved without this broader commitment to these issues.

Notes
1 Condom use rate of the contraceptive prevalence rate (CPR).
2 For instance, the 1999 review of progress since the Cairo conference (ICPD⫹5),
which included inputs from international agencies, experts and technical and
civil society advocacy organisations.
3 Thanks to Françoise Girard, Senior Program Officer at the International
Women’s Health Coalition, for providing information on the background to
the dropping of the reproductive health target and institutional responses to it.
4 Fifty-Fifth World Health Assembly, agenda item 13.2: WHO’s contribution to
achievement of the development goals of the United Nations Millennium Dec-
laration, 18 May 2002, WHA55.19.
5 The most recent development is a plan to set up UN task forces to develop
strategies for achieving the Millennium Development Goals. Maternal and
child mortality will be combined in one and HIV/AIDS will be in another,
along with tuberculosis and malaria. Reproductive health advocates are con-
cerned that reproductive health will be lost in this reverticalisation of health
priorities.
6 It is estimated that 20,000 delegates from national governments, UN organisa-
tions, non-governmental organisations (NGOs) and the media attended this
nine-day conference.
7 Developed countries have, however, failed utterly to deliver on the 20:20 commit-
ment.
8 This list contains fifteen indicators, including ones on sexual and men’s health,
and on infertility.
9 The maternal mortality ratio records deaths per 100,000 live births. For clarifi-
cation of the difference between ratios and rates, see UNICEF (1997: section
3.1.1).
10 Maternal mortality is generally defined as death while pregnant or within 42
days of termination of pregnancy from any cause related to or aggravated by
the pregnancy or its management. However, there are expanded definitions
which lead to inconsistency across countries and data sets (Gelband et al., n.d.).
11 The lifetime risk is the cumulative risk of dying over the reproductive period of
a woman’s life, whereas the MMR is a measure of risk for a woman who is actu-
ally pregnant. For further discussion of these measures, see WHO (2001a).
254 Hilary Standing
References
Asian-Pacific Resource and Research Centre for Woman (ARROW) (2001) Advanc-
ing the Women’s Health Agenda in the Asia-Pacific Region: [Link]
org/arrow/[Link].
Berer, M. (2001) ‘Images, Reproductive Health and the Collateral Damage to
Women of Fundamentalism and War’, Reproductive Health Matters, 9(18): 68.
Bloom, G. and Standing, H. (2001) ‘Pluralism and Marketisation in the Health
Sector: Meeting Health Needs in Contexts of Social Change in Low and Middle
Income Countries’, IDS Working Paper 136, Brighton: Institute of Development
Studies.
Correa, S. and Sen, G. (1999) ‘Cairo ⫹5: Moving Forward in the Eye of the Storm’,
Social Watch No. 3, ITEM, Montevideo: [Link]
[Link].
DeJong, J. (1999) ‘The Role and Limitations of the Cairo International Confer-
ence on Population and Development’, Social Science and Medicine, 51: 941–953.
Filmer, D. and Pritchett, L. (1997) ‘Child Mortality and Public Spending on
Health: How Much Does Money Matter?’ unpublished mimeo, Washington, DC:
World Bank.
Fonn, S., Makhosazana, X., Tint, K.S. and Conco, S.V. (1998) ‘Reproductive
Health Services in South Africa: From Rhetoric to Implementation’, Reproductive
Health Matters, 6(11): 22–32.
Forman, S. and Ghosh, R. (1999) ‘The Reproductive Health Approach to Popu-
lation and Development’, Paying for Essentials Policy Paper, New York: Center
on International Cooperation, New York University.
—— (eds) (2000) Promoting Reproductive Health: Investing in Health for Development,
Center on International Cooperation, Boulder, CO: Lynne Rienner.
Gelband, H., Liljestrand, J., Nemer, L., Islam, M., Zupan, J. and Jha, P. (2001)
‘The Evidence Base for Interventions to Reduce Maternal and Neonatal Mortal-
ity in Low and Middle-Income Countries’, Commission on Macroeconomics and
Health Working Paper Series WG5.
Girard, F. (2001) ‘Reproductive Health under Attack at the United Nations’
(letter to the editor), Reproductive Health Matters, 9(18): 68.
Goodburn, E. and Campbell, O. (2001) ‘Reducing Maternal Mortality in the
Developing World: Sector Wide Approaches May Be the Key’, British Medical
Journal, 322: 917–920, 14 April.
Hartmann, B. (1995) Reproductive Rights and Wrongs: The Global Politics of Population
Control, Boston, MA: South End Press.
Hulton, L.A., Matthews, Z. and Stones, R.W. (2000) A Framework for the Evaluation
of Quality of Care in Maternity Services, Southampton: University of Southampton.
Jacobson, J. and Mallik, R. (2002) The Far Right, Reproductive Rights and U.S. Inter-
national Assistance: The Untold Story behind the Headlines, Washington, DC: Center
for Health and Gender Equity (CHANGE).
Kutzin, J. (1995) ‘Experience with Organizational and Financing Reform of the
Health Sector’, Current Concerns SHS Paper 8 (SHS/CC/94.3), Geneva: WHO.
Lucas, H. and Nuwagaba, A. (1999) ‘Household Coping Strategies in Response to
the Introduction of User Charges for Social Service: A Case Study on Health in
Uganda’, IDS Working Paper 86, Brighton: Institute of Development Studies.
Mayhew, S.H., Lush, L., Cleland, J. and Walt, G. (2000) ‘Implementing the
 Towards reproductive health for all? 255
integration of Component Services for Reproductive Health’, Studies in Family
Planning, 31(2): 151–162.
Nahar, S. and Costello, A. (1998) ‘The hidden cost of “free” maternity care in
Dhaka, Bangladesh’, Health Policy and Planning, 13(4): 417–422.
Petchesky, R. (2000) ‘Reproductive and Sexual Rights: Charting the Course of
Transnational Women’s NGOs’, UNRISD Occasional Paper 8, Geneva: UNRISD.
Presser, H.B. and Sen, G. (2000) Women’s Empowerment and Demographic Process:
Moving beyond Cairo, Oxford: Oxford University Press.
Sen, G.(1999) ‘Southern Feminist Perspectives on Population and Reproductive
Rights: Continuing Challenges’, Society for International Development 42(1): 25–28.
Sen, G. with Germain, A. and Chen, L. (eds) (1994) Population Policies Reconsidered:
Health, Empowerment and Rights, Cambridge, MA: Harvard University Press.
Sen, P.D. and Ensor, T. (2001) ‘Public Expenditure Review of the Health and
Population Sector Programme’, Research Paper 19, Health Economics Unit,
Policy Research Unit, Dhaka: Ministry of Health and Family Welfare.
Simmons, R. and Elias, C. (1994) ‘The Study of Client–Provider Interactions: A
Review of Methodological Issues’, Studies in Family Planning, 21(1): 1–17.
Simms, C., Rowson, M. and Peattie, S. (2001) The Bitterest Pill of All: The Collapse of
Africa’s Health Systems, London: Save the Children Fund UK.
Standing, H. (2002) ‘An Overview of Changing Agendas in Health Sector
Reforms’, Reproductive Health Matters, 10(20): 1–10.
Svennson, J. (1997) ‘The Control of Public Policy: Electoral Competition, Polariza-
tion and Endogenous Platforms’, unprocessed mimeo, Washington, DC: World
Bank, November.
United Nations Children’s Fund (UNICEF) (1997) Guidelines for Monitoring the
Availability and Use of Obstetric Services, New York: UNICEF/WHO/UNFPA.
United Nations Fund for Population Activities (UNFPA) (1998a) ‘Reorientation of
National Family Planning Programmes Towards a Broader Reproductive Health,
Including Family Planning and Sexual Health, Approach in East and South East
Asia’, Occasional Paper Series 5.
—— (1998b) Levels and Trends of Contraceptive Use as Assessed in 1998: Key
Findings, [Link]
[Link].
—— (1999) Report of the International Conference on Population and Development,
Cairo, 5–13 September 1994, and ICPD⫹5, 1999. [Link]
—— (2001) The State of World Population 2001: Monitoring ICPD Goals: Selected Indi-
cators, [Link]
Walt, G., Pavignani, E., Gilson, L. and Buse, K. (1999) ‘Health Sector Develop-
ment: From Aid Co-ordination to Resource Management’, Health Policy and Plan-
ning, 14(3): 207–218.
World Health Organisation (WHO) (1997) Monitoring Reproductive Health: Selecting
a Short List of National and Global Indicators, WHO/RHT/HRP/97.26, Geneva:
WHO.
—— (1999) ‘Reduction of Maternal Mortality: A Joint WHO/UNFPA/UNICEF/
World Bank Statement’, Geneva: WHO.
—— (2001a) Maternal Mortality in 1995: Estimates Developed by WHO, UNICEF,
UNFPA, WHO/RHR/01.9, Geneva: WHO.
—— (2001b) Macroeconomics and Health: Investing in Health for Development. Report of
the Commission on Macroeconomics and Health, Geneva: WHO.
12 The global challenge of
HIV/AIDS
Ronald Skeldon

The global context of HIV/AIDS

The Joint United Nations Programme on HIV/AIDS (UNAIDS) has esti-
mated that during 2001, 5 million people became newly infected with
HIV, the vast majority of them in developing countries. In one single
hour, over 500 people are infected with HIV and over 300 people die
from AIDS-related diseases. Globally, some 40 million people are esti-
mated to be living with HIV, 28.1 million of them, or 70 per cent of the
total, in sub-Saharan Africa. Just 3.8 per cent live in Western Europe and
North America combined (Table 12.1). The distribution of new infections
in 2001 reflected the total global distribution of HIV infections. Some 3.4

Table 12.1 Global situation of the HIV/AIDS epidemic, December 2001

Region Adults and Adults and Percentage Adult

children children distribution prevalence
newly living with of column rate
infected HIV/AIDS (2) (percentages)
with HIV (millions)
(000s)
(1) (2) (3) (4)

Sub-Saharan Africa 3,400 28.1 70.3 8.4

North Africa and Middle 80 0.4 1.1 0.2
East
South and South-East Asia 800 6.1 15.3 0.6
East Asia and Pacific 270 1.0 2.5 0.1
Latin America 130 1.4 3.5 0.5
Caribbean 60 0.4 1.1 2.2
Eastern Europe and Central 250 1.0 2.5 0.5
Asia
Western Europe 30 0.6 1.4 0.3
North America 45 0.9 2.4 0.6
Australia and New Zealand 0.5 0.2 0.0 0.1
Total 5,000 40.0 100 1.2

Source: UNAIDS, AIDS Epidemic Update, Geneva (2001: 3).

 The global challenge of HIV/AIDS 257
million new infections in the year 2001 were in sub-Saharan Africa, 68 per
cent of the total number of new infections. Only 1.5 per cent of those
newly infected were to be found in Western Europe and North America.
The prevalence rate is about 8.4 per cent in sub-Saharan Africa, compared
with 1.2 per cent globally and 0.6 per cent for North America.1
The importance of the target is discussed in the next section, and its
feasibility, in terms of whether it can be achieved and whether it is likely to
be achieved, is the subject of the two subsequent sections respectively.

Does the goal on HIV/AIDS matter?

This chapter presents a critical examination of the Millennium Develop-
ment Goal for HIV reduction. The Goal itself is to have halted and
reversed the spread of HIV/AIDS by 2015, with this Goal replacing an
earlier more specific target of a ‘25 per cent reduction in HIV infection
rates among 18–24 year olds in the worst affected countries by 2005 and
globally by 2010’. However, before we go on to examine this target, it is
important to consider briefly the implications of the current epidemic of
HIV/AIDS for the other key development goals dealt with in this volume.
Such a review helps to place into context why the target on HIV/AIDS
really does matter.
It is already clear that the scope and magnitude of the HIV/AIDS epi-
demic is putting at least some of these key targets under threat. For
example, and most obviously, the target of reducing infant and under-five
child mortality by two-thirds by 2015 looks particularly at risk, given the
significance of mother-to-child transmission of the HIV virus. According to
the UN (2000: 92), ‘between one fourth and one third of the children born
to HIV-positive women acquire the infection from their mothers’. Child
mortality declined in the three sub-Saharan countries for which data are
readily available, Cameroon, Kenya and Zambia, during the early 1980s but
saw an upturn from about 1986, reaching rates higher, at least in the cases
of Kenya and Zambia, than in 1981 (Figure 12.1). In Botswana, under-five
mortality in the absence of AIDS would have been 53 deaths per 1,000 live
births during the period 1995–2000, whereas it was actually more than
double that rate, at 107 child deaths per 1,000 live births (UN, 2000: 93).
The same source projects relative increases in under-five mortality for South
Africa, Namibia, Kenya, Malawi and Mozambique over the first five years of
the twenty-first century. The picture is clear: in the worst-infected areas,
child mortality is increasing rather than going down. Even though projected
rates will still be lower than those prevailing during the 1950s and 1960s, the
possibility of achieving a two-thirds reduction with reference to a 1990 base
year is clearly undermined.
Thus AIDS is responsible for the reversal of the long-term mortality decline
in several parts of Africa. The incidence of tuberculosis in rural Malawi
doubled between 1986 and 1994, largely because HIV-positive persons were
258 Ronald Skeldon

110
Relative under-5 mortality (1981=100) (%)

Zambia (HIV
prevalence
100 19.9%)

Kenya
(HIV
prevalence
14.1%)
90
Cameroon
(HIV
prevalence
7.7%)

80
1980 1985 1990 1995 2000

Figure 12.1 Trends in mortality from AIDS amongst children under 5 years old and
end-1999 adult prevalence rate, selected African countries, 1981–1996.
Source: Demographic and Health Surveys, Macro International, USA.

seven times more likely to develop the disease than those who were not so
infected (UN, 2000: 77–78). Despite the rising mortality and the post-
ponement of any progression along the epidemiological transition from
infectious diseases to degenerative diseases that has been seen elsewhere,
HIV-associated mortality is not yet causing populations to decline, even in
the worst-affected countries. It is a case of population growth rates slowing
down, which, perversely, can have some macro-level benefits for an
economy.
The impact of AIDS-related mortality on life expectancy can indeed be
profound, as can be seen in Table 12.2. Life expectancy at age 20 in
Botswana in the absence of AIDS would have been 54.6 years during the
period 2000–2005. However, with an adult prevalence rate of 22.1 per
cent, the estimated life expectancy at age 20 is only 27.6 years. For the ten
worst-affected countries, it is estimated that life expectancy at age 20 has
been reduced by at least one-third because of the epidemic. Figure 12.2
clearly shows the impact of excess AIDS-related mortality on the age struc-
ture of the most deeply affected country, Botswana, based on US Census
Bureau projections. In contrast, the worst-affected countries in Asia show
a much lower loss of life expectancy in the period 2000–2005 than almost
all countries in Africa, though the data from Asian countries nevertheless
suggest alarming trends.
As regards the Millennium Development Goals, it is clear that the
goal of universal primary education is also under threat as children are
 The global challenge of HIV/AIDS 259
Table 12.2 Impact of AIDS-related mortality on life expectancy at age 20, selected
countries, 2000–2005

Country Adult HIV Life expectancy at age 20 Absolute Percentage

prevalence difference difference
With AIDS Without AIDS

Botswana 22.1 27.6 54.6 27.0 49.5

Zimbabwe 21.5 28.2 52.0 23.8 45.8
Zambia 16.6 29.2 50.9 21.7 42.6
Namibia 16.1 30.7 49.0 18.3 37.4
Malawi 12.5 34.1 45.0 10.9 24.2
Mozambique 11.9 34.4 49.3 14.9 30.2
South Africa 11.8 33.4 48.4 15.0 31.0
Rwanda 11.2 35.9 53.0 17.1 32.3
Kenya 10.4 30.4 45.8 15.4 33.6
Central African 8.6 37.1 48.5 11.4 23.5
Republic
Côte d’Ivoire 8.5 37.2 48.9 11.7 23.9
United Republic 8.2 35.8 47.0 11.2 23.8
of Tanzania
Uganda 8.1 36.1 45.7 9.6 21.0
Ethiopia 7.7 36.9 45.4 8.5 18.7
Burundi 7.0 40.8 52.7 11.9 22.6
Togo 6.9 39.2 48.9 9.7 19.8
Lesotho 6.7 36.3 46.5 10.2 21.9
Congo 6.4 39.9 48.7 8.8 18.1
Burkina Faso 6.0 37.6 44.8 7.2 16.1
Haiti 4.1 42.6 48.0 5.4 11.3
Cameroon 4.0 42.8 47.1 4.3 9.1
Dem. Rep. of 3.5 42.3 49.7 7.4 14.9
Congo
Nigeria 3.4 41.9 46.5 4.6 9.9
Gabon 3.1 42.5 48.3 5.8 12.0
Liberia 3.0 43.1 47.4 4.3 9.1
Eritrea 2.6 38.4 41.5 3.1 7.5
Sierra Leone 2.6 42.2 45.3 3.1 6.8
Chad 2.2 49.8 53.2 3.4 6.4
Cambodia 2.0 44.4 47.7 3.3 6.9
Thailand 1.8 40.4 43.7 3.3 7.6
Benin 1.8 43.7 46.7 3.0 6.4
Guinea-Bissau 1.7 42.6 46.5 3.9 8.4
India 0.6 51.2 52.2 1.0 1.9
Brazil 0.5 51.9 52.5 0.6 1.1

Source: United Nations, World Population Prospects: The 1998 Revision, New York, Department
of Economic and Social Affairs, Population Division (2000, volume 3: 93).
260 Ronald Skeldon

80
75 Males Females
70
65
60
55
Age in years

50
45
40
35
30
25
20
15
10
5
0
140 120 100 80 60 40 20 0 20 40 60 80 100 120 140

Population (thousands)
Projected population Deficits due to AIDS
structure in 2020

Figure 12.2 Projected population of Botswana in 2020 showing deficits due to

AIDS-related deaths by cohort.
Source: United States Census Bureau, World Population Profile 2000, Washington, DC, 2001.

withdrawn from schools either to help with the family farm as parents
become sick or to care for sick relatives. In Swaziland, for example, school
enrolment is reported to have fallen by 36 per cent and girls may have
higher drop-out rates than boys, although both sexes are likely to have
been profoundly affected (UNAIDS, 2001a: 7).
The overall goal of reducing the proportion of people living in extreme
poverty by half between 1990 and 2015 is also under threat in the worst-
affected areas. HIV/AIDS is now generally seen as a ‘disease of poverty’, a
disease that afflicts the poor disproportionately. The third report of the
UK House of Commons Select Committee on International Development
states that ‘the fight against HIV/AIDS can be won only through progress
in the elimination of poverty’.2 These are fine words, but the path of the
epidemic threatens progress towards reducing poverty. Also, the relation-
ship between poverty and HIV is more complex than is often assumed.
Certainly, the epidemic is centred in the poorest region of the world, sub-
Saharan Africa, and clearly, too, the lack of physical and human capital
militates against the treatment of the disease. However, the HIV epidemic
is also intense in the wealthier economies of sub-Saharan Africa, such as
South Africa and Botswana, and it also profoundly affects members of
 The global challenge of HIV/AIDS 261
wealthier and better-educated groups, who are more able to afford mul-
tiple sexual partners. Cultural and societal factors as well as economic
deprivation are key to the spread of the epidemic.3
Although funeral costs, lost income, and savings and productivity losses
are devastating specific household economies and, most specifically, the
poorest households,4 it is much more difficult to calculate the macro-level
impact of the epidemic. Two economists, Bloom and Godwin (1997: 3),
have argued that there is a ‘resounding lack of evidence to support the
view that AIDS impedes economic growth’. Botswana, with the world’s
highest AIDS death rate, also experienced rapid rates of economic growth
between 1998 and 2001 averaging 9 per cent per annum, which was one of
the highest in the world at that time (IIASA, 2001). Thus high economic
growth can occur despite a high prevalence of HIV, although Botswana’s
economy is relatively small, is dependent on exports and is not reliant
upon a large labour supply. The World Bank has projected that in Africa,
income growth per capita is likely to be reduced by between 0.1 and 0.4
per cent per annum. GDP in South Africa in 2010 is likely to be 17 per
cent lower than it would have been in the absence of AIDS, with the epi-
demic costing that economy some US$22 billion (Barks-Ruggles, 2001).
However, development is not, or should not be, measured in macro-
level indicators alone, and unquestionably the disease is having a pro-
found impact on local economies and households. Thus the HIV/AIDS
epidemic is likely to slow, if not reverse, progress towards the attainment
of at least some of the targets contained in the Millennium Development
Goals themselves, particularly given the fact that many of these targets are
essentially goals of human development.
One other contextual issue requires discussion before the actual global
target for HIV itself is considered. This is the ‘illusion of numbers’. In
1991, projections were made that by the end of the decade 9 million
people would be infected in sub-Saharan Africa, with 5 million deaths
(UNAIDS, 2000: 7). The actual figures have proved to be three times
greater than that estimate. Thus a very big question mark hangs over the
quality of the available data and it must be stressed that these data are very
‘spongy’ indeed – full of holes, with a high degree of variance in the esti-
mates. For example, one of the principal sources of information that is
used for virtually every country in estimating infection levels is based upon
the number of women attending antenatal clinics. However, if fertility
among HIV-positive women is considerably lower than among all women,
then the estimates derived from this source may be too low.
Inadequacies in the available data should never be used as an excuse
for inaction, and it is necessary to learn to deal with the existing figures
with all their weaknesses. Although the quality of our information has
almost certainly improved over the past ten years, major areas of uncer-
tainty still remain and a degree of scepticism is always required in any
interpretation of the data presented.
262 Ronald Skeldon
Are the international targets for HIV/AIDS achievable?
The goal of halting and reversing the spread of HIV/AIDS by 2015 can be
achieved, but it does not go far enough. This statement can be amplified
as follows. The headline goal is realistic as, amidst all the gloom in the stat-
istics, there are signs of optimism, or if not of optimism – at least of hope.
The number of new infections in sub-Saharan Africa stabilised in 2000
and 2001, after rising inexorably up to 1999. In 2000, new infections in
that region were estimated to be 3.8 million, compared with 4 million in
1999 – a small decline, but perhaps a significant one. That decline con-
tinued into 2001, when the number of new infections was estimated at 3.4
million, representing a 15 per cent reduction in the number of new infec-
tions since 1999. It is useful to consider some of the ‘success’ stories of the
HIV/AIDS epidemic before returning to more sobering situations.

Success stories: Africa

In sub-Saharan Africa, Uganda is seen as one of the few countries that has
confronted the epidemic with significant results. Uganda was one of the
earliest countries to be affected by the epidemic in sub-Saharan Africa,
with infections dating from the early 1980s. It also has been one of the
most severely affected countries in the region, with overall prevalence
rates reaching 14 per cent in the early 1990s. By 2000, overall prevalence
is estimated to have declined to 8 per cent (UNAIDS, 2000: 9). In urban
areas, HIV infection among pregnant women attending antenatal clinics
has fallen for eight years in a row from 29.5 per cent in 1992 to 11.3 per
cent in 2000 (UNAIDS, 2001a: 17). Condom use rose from 7 per cent
nationwide since the early 1990s to 50 per cent in rural areas and 85 per
cent in urban areas by the late 1990s.
However, just how these figures were derived is not clear. In fact, it is
admitted that ‘due to weak monitoring, it is not possible to apportion the
observed decline between the three factors of abstaining, being faithful to
one’s partner, and condom use’ (UAC, 2000). Also, the prevalence rate
appears to have stabilised since 1996 at around 8–10 per cent, still very
high by global standards. Hence although Uganda may already have
reached the international target of a 25 per cent reduction in prevalence,
it has done so from a high base and is still in an unsatisfactory position. A
25 per cent reduction in HIV infection in the worst areas may still leave
countries in a very vulnerable position. Nevertheless, progress clearly has
been attained in Uganda, a country that ranked 141 out of 162 states in
the human development league tables of the United Nations Develop-
ment Programme (UNDP) in 2001.
In Lusaka, Zambia, HIV prevalence amongst girls aged 15 to 19 who
were attending antenatal clinics also declined, from 27 per cent in 1993 to
17 per cent in 1998. In Senegal, success of a different kind gives grounds
 The global challenge of HIV/AIDS 263
for hope. Senegal was never one of the sub-Saharan countries to be most
severely affected by the epidemic,5 but significantly, it appears to have
been able to take pre-emptive action in order to prevent the disease from
taking hold in the first place. The national adult prevalence rate has
remained low, at just under 1.8 per cent at the end of 1999.

Success stories: Asia

If the international goals are to be met and to have a significant impact on
development at a global level, this must occur in the most populous parts
of the world, and particularly China and India. The infection rates of
countries in Asia so far do not approach those of sub-Saharan Africa. At
the end of 1999 the Asian country with the highest rate of adult infection
was Cambodia at just over 4 per cent, followed by Thailand at just over 2
per cent (UNAIDS, 2000). As seen earlier, rates in excess of 10 per cent
are common in Africa, with Botswana and Zimbabwe over 20 per cent and
Zambia and Namibia over 15 per cent.
In Asia, Thailand is seen to be one of the success stories, with rates of
infection dropping through the 1990s. One of the most commonly cited
statistics is the marked decline of HIV prevalence amongst 21-year-old
military conscripts, which peaked at around 4 per cent in 1993 and has
currently fallen to below 1 per cent (Figure 12.3). Infection remains high,
however, amongst injecting drug users and commercial sex workers, and

14

12
Upper-
North
Percentage HIV-1 positive

10
Lower-
North
8
North-east

6
Central
4
South
2

0
1988 1990 1992 1994 1996 1998 2000 2002

Figure 12.3 HIV-1 seroprevalence among Royal Thai Army conscripts by region,
1989–2000.
Source: Royal Thai Army, Bangkok, 2000.
264 Ronald Skeldon
significant differences remain between urban and rural areas. Sex workers
in Bangkok, for example, are estimated to have prevalence rates of only 7
per cent, compared with 20 per cent for sex workers in rural areas.

Reasons for the success stories

These ‘successful’ countries have certain characteristics in common. They
all mounted intensive media campaigns; all raised awareness through edu-
cation programmes to disseminate information on the way the virus is
spread at all levels of society; all actively disseminated condoms through-
out the country; most important, however, the governments were willing
to confront the epidemic and raise the essentially sexual nature of its
transmission to a high profile. Such attitudes, unfortunately, have been all
too rare in other countries, and most publicly among senior policy-makers
in South Africa.
These success stories conclusively show that people can and do change
their behaviour. In Uganda the proportion of girls and boys aged 15 to
19 years who had never had sex rose from 20 per cent in 1990 to 50 per
cent in 1995. In Bangkok the number of men frequenting brothels is
reported to have halved between 1990 and 1993, with the number of sex-
based establishments dropping by about 60 per cent (UNAIDS, 1998).
The 100 per cent condom programme has been vigorously implemented
in the formal sex sector throughout the country. Using data from Africa,
the World Bank has demonstrated the effectiveness of condom use in a
population of 500 sex workers, four-fifths of whom were already infected
with HIV. Within a year, 10,000 infections could be averted, assuming
that condom use can be raised to 80 per cent (World Bank, 1999: box
2.6).
There is, however, no room for complacency in the successful coun-
tries. I have argued elsewhere (Skeldon, 2000) that there is evidence to
suggest that unsafe practices and high-risk behaviour have moved from
the formal sex sector in Thailand to a more informal level – that is,
amongst groups of friends or acquaintances at school, college or work-
place rather than with professional sex workers in brothels. Also, periodic
markets in rural and urban situations or temple fairs are emerging as key
nodes, or ‘hot spots’, in the diffusion of the epidemic in place of the
formal brothel or massage parlour environments. There is considerable
scope for a second stage in the development of the epidemic.

Are the targets on HIV/AIDS likely to be achieved?

Some countries have shown that infection rates can be reduced in line
with the target of a 25 per cent reduction globally by 2010. However, it
seems highly unlikely that the majority of countries will follow suit or that
the target will be achieved globally, with, as argued earlier, negative con-
 The global challenge of HIV/AIDS 265
sequences for the other Millennium Development Goals. Four often inter-
related sets of factors will probably prevent the target from being met: (1)
the fact that the epidemic is so recent in many parts of the world; (2) the
issue of political commitment; (3) the availability, or otherwise, of money;
and (4) geopolitical considerations. Each of these will be considered in
turn.

HIV/AIDS as a recent phenomenon

The first clinical evidence for the existence of AIDS dates from just twenty
years ago and the area most deeply affected since that time clearly has
been sub-Saharan Africa. Over 70 per cent of those living with HIV/AIDS
are currently found in that region, which in 2000 accounted for only 10
per cent of the world’s population. However, the population of East,
South-East and South Asia, at some 3.5 billion people, accounts for almost
58 per cent of the world population. For this reason, at the global level the
impact of HIV/AIDS must ultimately be determined by what happens in
the most populous region, Asia.
Thus far, the epicentre of the epidemic in Asia has been located in
Thailand, specifically in the region known as the ‘upper north’ in the area
north from Chiang Mai into the hill lands bordering Myanmar and Laos.
This is an area of relatively sparse population located strategically between
the densely populated regions of China to the north and east and South
Asia to the west. The fact that the epidemic should have developed in this
region is due to a complex combination of factors, of which the following
are important:

• These are areas of long-standing trade routes.

• The area has an extensive trade in narcotics, with relatively high local
consumption and a high proportion of injecting drug users.
• Cash is readily available from trade in border areas.
• The area has a sex culture in which men and women have been relat-
ively free in the selection of partners.
• Having multiple sequential partners is a much more accepted form of
behaviour than in other parts of Thailand.
• Women are sometimes incorporated at a very young age into sex
work.

The critical question revolves around what will happen if or when the epi-
demic takes hold in the densely populated parts of the Indian sub-conti-
nent and China. However, it is important to emphasise here that there is
no evidence to suggest, one way or the other, that the future path of the
epidemic in China or India will follow that of any of the sub-Saharan
African countries. However, even if the epidemic follows the path of
Thailand, with much lower levels of infection, the resultant human and
266 Ronald Skeldon
economic cost will be substantial. The case of Cambodia, one of the
poorest countries in its region, just might point to the future direction of
the epidemic in East Asia at least. Although Cambodia has the highest
adult prevalence rate in Asia, official government figures suggest that the
rate has declined from around 4 per cent in 1997 to 2.8 per cent in 2000
(Cambodia, 2001: 110). HIV prevalence amongst pregnant women, too,
appears to have declined from 1997 (UNAIDS, 2001a: 15), which again
suggests a real trend towards slower growth, even containment, and that
the path of the Asian epidemic may be different from that in Africa. Nev-
ertheless, even with lower overall prevalence rates, the human cost will be
devastating in the demographic giants of Asia.
In China, at the end of 1999 the number of adults and children living
with HIV/AIDS was recorded by UNAIDS at 500,000. This figure seems
unrealistically low for a population of 1.3 billion and it is generally accepted
that the incidence of the disease is being severely underestimated.
Chinese officials had admitted to a figure of 600,000 by June 2001 but by
then United Nations officials were talking of 1.5 million infected (Lange,
2001). The same source warns that there could be 20 million infected
Chinese by 2010. If that prediction proves to be accurate, it would severely
affect any global target for HIV reduction. Reported HIV infections rose
by two-thirds in the first six months of 2001 alone (UNAIDS, 2001a: 13).
Because of the country’s demographic weight, what goes on in China will
affect global trends, and if the incidence of infection increases amongst
the critical 15–24 years age group, then both the global goal of reversing
the spread of the disease, and the target of reducing prevalence rates, are
likely to be meaningless without a major pre-emptive programme within
China.
A scandal over blood-buying networks in Henan Province, in which vil-
lagers were paid US$5 for each blood donation, is estimated to have
infected some 100,000 people. Blood was collected to extract the plasma
to sell to drug manufacturers and then the pooled blood was re-injected
into the villagers so that they could donate more frequently. Needles were
shared and infection rates of 43–65 per cent have been reported among
blood donors as a result of needle sharing and re-injecting of pooled
blood (Lange, 2001). Despite this scandal, the diffusion of the disease is
not going to be the result of pooling blood products, sharing needles or
drug abuse but through heterosexual contact. The increasing mobility of
China’s population, with some 100 million undocumented migrants, and
an increasing commercial sex sector in the urban areas seem destined to
exact a grim toll and one which the government in 2002 finally appeared
to be addressing.
The whole issue of population migration and the spread of HIV is
extremely complex. There is not necessarily any relationship between the
movement of people and the spread of the disease as it is not the move-
ment as such that is the critical variable but the behaviour of the migrants.
 The global challenge of HIV/AIDS 267
Migrants who do not engage in high-risk practices will not spread the
disease. However, the movement away from tight-knit kinship and commun-
ity networks appears to place people in situations where there are higher
probabilities for them to engage in such practices. The fact that one of the
wealthiest sub-Saharan societies, Botswana, also came to have the highest
prevalence of HIV can be attributed greatly to the participation of its
population in the circular labour migration to South Africa’s mines. It is
well established that the disease is spread along truck and trade routes,
and the northern Thai epidemic was facilitated by the relatively high
mobility of its population. Development, irrespective of whether it is the
result of specific plans of action or of gradual evolution, tends to increase
the mobility of populations. The rapid rates of growth in Asian economies
and the drive for development on the part of most governments can but
stimulate further migrations, with significant implications for the spread
of HIV/AIDS, and particularly in China and India.6
At the end of 2000 it was estimated that in India some 3.86 million
people were infected with HIV/AIDS, an overall infection rate of less than
1 per cent. This prevalence rate is again low by African standards but, in
terms of absolute numbers, India has more people living with HIV than
any country in the world except South Africa (UNAIDS, 2001a: 13). The
state with the greatest number of victims is Maharashtra, which is also one
of the most prosperous states in India, containing, as it does, the major
industrial and commercial city of Mumbai. HIV prevalence among sex
workers in that city was estimated to have reached over 70 per cent in
1997 (UNAIDS, 2000: 13). The highest prevalence rates are to be found
among the more developed southern states, with HIV having made relat-
ively little inroad so far in the densely populated northern states such as
Bengal, or even within that state’s megalopolis, Calcutta. In sum, the
potential for a rapid spread of the epidemic in the two most populous
countries of the world, China and India, seems vast indeed. So, too, is the
potential for rapid spread in Indonesia, a country with some 218 million
people in 2002 but an adult prevalence rate of just 0.05 per cent at the
end of 1999.7

The issue of political commitment

Perhaps the most critical set of factors preventing the goal from being met
revolves around political commitment or the political recognition of HIV/
AIDS as a major policy issue. There are the ‘Three Ds’ of HIV/AIDS:
Denial, Delay and Do nothing.8 Many governments are reluctant to accept
that HIV/AIDS is a problem. Even if they do accept that it might be
a problem, they are reluctant to take action on an issue that touches
upon the most intimate aspects of the lives of their citizens. In all the cases
of successful intervention discussed earlier, the governments of the respec-
tive countries had responded in an open and vigorous way to the spreading
268 Ronald Skeldon
epidemic. It is only recently, if a headline in the Guardian (London) on 13
November 2001 can be believed, that China ‘finally wakes up to AIDS time
bomb’. A massive education programme of doctors and health workers
throughout the country is planned and anti-AIDS funding will become a
separate item in national and provincial budgets.
Other countries are, however, still in the ‘denial’ or the ‘do nothing’
stage. Myanmar does not recognise HIV infection as an issue, despite avail-
able evidence of a rapid spread in that country. The case of South Africa is
still notorious, with President Thabo Mbeki ordering a cut in the AIDS
budget as recently as September 2001 based on what a World Health
Organization official described as the ‘deliberate misinterpretation of old
statistics for political ends’. Even in March 2002, Mr Mbeki decried the
view that multiple sexual partners or rape might spread the disease as
‘demeaning’ and ‘insulting’ (The Economist, 11 May 2002). Only an even
more recent U-turn in official attitudes towards the disease, largely the
result of a court case ruling in favour of the distribution of anti-retrovirals,
offers some hope that the official face of South Africa may finally be
beginning to confront the enormity of its AIDS problem. As a result of the
constitutional court ruling, the government cannot appeal against the
high court decision ordering the supply of anti-retrovirals to hospitals,
which ultimately may lead to the saving of 30,000 babies a year from infec-
tion from their mothers.
Moreover, despite the fact that several of the ‘success stories’ discussed
above were from sub-Saharan Africa, the third report of the UK House of
Commons Select Committee on International Development observed that
‘The response to the HIV/AIDS epidemic, particularly in Africa, has been
a culpable and serious failure in political leadership and governance.’9 A
general consensus appears to be emerging that the only helpful response
to the epidemic is bold vision and strong and enlightened leadership.
Denying that the epidemic is a problem is counter-productive, and perhaps
this political dimension is the greatest global challenge in confronting the
diffusion of HIV/AIDS in the twenty-first century.

The availability, or otherwise, of money

Leadership may be the first priority but money and resources are an
important secondary need. Expenditure on HIV/AIDS has increased from
US$59 million in 1987 to US$293 million in 1998, almost a fivefold increase.
The numbers infected with HIV have, however, risen eight and a half
times, meaning that expenditure per capita has actually declined.
In April 2001 the Secretary-General of the United Nations (UN), Kofi
Annan, called for US$7–10 billion to be spent annually on the epidemic
and for a global health fund to be set up for HIV/AIDS. Up to May 2002,
just under US$2 billion had been collected and the scope of the fund was
broadened to cover both tuberculosis and malaria. The USA, the largest
 The global challenge of HIV/AIDS 269
donor, had given US$450 million and the UK some US$200 million. Bill and
Melinda Gates had donated US$100 million. To try to put the US$7–10
billion annual commitment into the context of other expenditure, it is
worth noting that the US Congress approved US$40 billion to aid recovery
efforts in New York City after the events of 11 September 2001, plus an
additional US$15 billion to assist the airline industry. Some US$1.2 trillion
was lost in stock market investments as a result of financial nervousness
following the 11 September attacks.
An annual spending on AIDS of US$7 billion annually would represent
more than a 10 per cent increase in total overseas development assistance
given by the Organisation for Economic Cooperation and Development
(OECD) countries, which seems a somewhat remote possibility. As noted
in Chapter 1, in 1970 the UN resolved that development aid should repre-
sent 0.7 per cent of gross national product (GNP). However, other than
the Scandinavian countries, only the Netherlands and Luxembourg have
met this target. The USA manages 0.1 per cent (although it is the second
largest donor after Japan in absolute amounts of development assistance).
The UK’s aid has at least risen in recent years, and in 1999/2000 was at its
highest level in real terms since 1976. Even so, the UK’s commitment to
development expenditure remains at around 0.35 per cent of GNP, or
half the UN target.

Geopolitical considerations
One critical issue is the direction that might be taken by the international
community after the tragic events of 11 September. As developed nations
become increasingly concerned about their own security, and as their
attention is turning towards Central Asia, a real danger exists that HIV/
AIDS will be relegated to the back burner and that the worst-affected
countries in sub-Saharan Africa will be still further marginalised from the
global community and economy. Long before these recent events, sub-
Saharan Africa was only tenuously linked to the global economy. Just 1 per
cent of foreign direct investment flowed to these countries throughout
the mid- to late 1990s. It is not so much a question of whether things can
get worse but whether there will be any serious effort to reverse this situ-
ation now that the eyes of the world community appear to have shifted to
concerns elsewhere.
The geopolitical issue can also be considered from another angle.
International peacekeepers have moved to Afghanistan. Parallels can
perhaps be sought in Cambodia, which has the highest HIV infection
rate in Asia. It is generally acknowledged that the presence of large
numbers of foreign troops based in a youthful population that had lost
enormous numbers of husbands and fathers was a major factor in the
spread of the epidemic in that country. Over 20,000 troops and civil ser-
vants of the United Nations Transitional Authority were at one time based
270 Ronald Skeldon
in Cambodia, where a significant commercial sex sector thrived on both
foreign and local clients (see Caouette, 1998: 15–16). It is estimated that
HIV sero-prevalence rates amongst commercial sex workers in 1998 ranged
from 25 to 64 per cent, depending upon the region (UNDP, 2000: 7), with
national group mean prevalence for direct sex workers in 1999 of 33 per
cent (Chantavanich, 2000: 22). Thus if history repeats itself, the introduc-
tion of an international peacekeeping force into Afghanistan may con-
tribute to a new AIDS problem. This would once again divert attention
from the situation in sub-Saharan Africa as the international community
seeks to prevent any increase in infection in a region that has already
seen rapidly rising infection rates among Central Asian republics such as
Uzbekistan.

Conclusion
Where does this leave the global goal and target for HIV/AIDS? At a
global level, the target of reducing HIV prevalence by 25 per cent by 2015
could and should be met. What is likely to act against the target is the
spread of the disease into previously little-affected populations, and pri-
marily into Asia. The Asian pattern, however, is unlikely to reach levels
currently seen in Africa. More realistically, regional rather than global
targets are required. Achieving the global target in some African countries
does not go nearly far enough, while it is irrelevant in areas that are only
now experiencing the first stages of an epidemic.
The mechanics of the spread of the epidemic are well known. In most
areas the people themselves know how the disease is spread. The best
practices to introduce in order to contain the disease are also known.
Most of all, it is understood that failing to achieve the targets for HIV/
AIDS is likely to undermine many of the other Millennium Development
Goals. The containment and reversal of the epidemic will be pivotal for
the achievement of other dimensions of development. The issue is now
primarily one of political will. Perhaps there should be annual targets for
the number of governments that seriously confront the epidemic. It can
be argued that the HIV/AIDS epidemic has moved firmly onto the polit-
ical stage, with demographic, economic and even medical aspects being
secondary. Targets are now needed in the political arena. Thus the main
global challenge of HIV/AIDS is to keep a clear focus on the existing pat-
terns of the epidemic. This might sound trite but the path ahead will not
be easy. The future course of the HIV/AIDS epidemic is firmly in the
hands of governments and policy-makers.
 The global challenge of HIV/AIDS 271
Notes
1 All figures from UNAIDS (2001a).
2 [Link]
cmintdev/354/[Link]#a1, Executive Summary, p. 1.
3 See, in particular, the work of Caldwell and his associates, with a synopsis in
Caldwell (2000).
4 See Knodel et al. (2002) for an analysis of the impact of the death through AIDS
of adult children on surviving parents in Thailand.
5 Prevalence is lower in West Africa than Eastern and Southern Africa.
6 For a discussion of the general issues of population mobility and HIV/AIDS, see
UNAIDS (2001b) and Skeldon (2000), also Chantavanich (2000).
7 See Hugo (2001) for a discussion of information gaps, current knowledge and
likely future directions of the epidemic in Indonesia.
8 I am indebted to Steve Kraus of UNAIDS, Bangkok, for this felicitous combina-
tion.
9 [Link]
cmintdev/354/[Link]#a1, Executive Summary, p. 9.

References
Barks-Ruggles, E. (2001) ‘Meeting the global challenge of HIV/AIDS’, Policy Brief
75, Brookings Resources, at [Link]
[Link].
Bloom, D.E. and Godwin, P. (eds) (1997) The Economics of HIV and AIDS: The Case
of South and Southeast Asia, New Delhi: Oxford University Press.
Caldwell, J.C. (2000) ‘Rethinking the African AIDS epidemic’, in Caldwell, J.C.,
Caldwell, P., Orubuloye, I.O. et al. (eds) Towards the Containment of the AIDS Epi-
demic: Social and Behavioural Research, Canberra: Australian National University,
Health Transition Centre.
Cambodia (2001) A Situation and Response Analysis of the HIV/AIDS Epidemic in Cam-
bodia, Phnom Penh: National AIDS Authority.
Caouette, T.M. (1998) Needs Assessment on Cross-Border Trafficking in Women and Chil-
dren: The Mekong Sub-Region, report prepared for the UN Working Group on
Trafficking in the Mekong Sub-region, Bangkok.
Chantavanich, S. (2000) Mobility and HIV/AIDS in the Greater Mekong Subregion,
Bangkok: Asian Development Bank and United Nations Development Pro-
gramme.
Hugo, G. (2001) Population Mobility and HIV/AIDS in Indonesia, Jakarta: Inter-
national Labour Office.
International Institute for Applied Systems Analysis (IIASA) (2001) ‘HIV/AIDS
and Development: Are Botswana, Namibia, and Mozambique on Sustainable
Paths?’, Options, Spring: 6–8, International Institute for Applied Systems Analy-
sis.
Knodel, J., Im-em, W., Saengtienchai, C., Vanlandingham, M. and Kespichayawat-
tana, J. (2002) The Impact of an Adult Child’s Death Due to AIDS on Older-Aged
Parents: Results from a Direct Interview Survey, Publication 266. Mahidol University,
Institute for Population and Social Research.
Lange, D. (2001) ‘A New Vein of Openness’, Far Eastern Economic Review, 4
October: 58–60.
272 Ronald Skeldon
Skeldon, R. (2000) Population Mobility and HIV Vulnerability in South East Asia: An
Assessment and Analysis, Bangkok: United Nations Development Programme.
Uganda AIDS Commission (UAC) (2000) The National Strategic Framework for
HIV/AIDS in Uganda, 2000/1 to 2005/6, Nairobi: UAC.
United Nations (UN) (2000) World Population Prospects: The 1998 Revision, New
York: Department of Economic and Social Affairs, Population Division.
United Nations Development Programme (UNDP) (2000) Cambodia HIV Vulnera-
bility Mapping: Highways One and Five, Bangkok: UNDP.
United Nations Programme on HIV/AIDS (UNAIDS) (1998) Relationships of HIV
and STD Declines in Thailand to Behavioural Change: A Synthesis of Existing Studies,
Geneva: UNAIDS.
—— (2000) Report on the Global HIV/AIDS Epidemic: June 2000, Geneva: UNAIDS.
—— (2001a) AIDS Epidemic Update, Geneva: UNAIDS.
—— (2001b) Population Mobility and AIDS, Geneva: UNAIDS.
World Bank (1999) Confronting AIDS: Public Priorities in a Global Epidemic, New York:
Oxford University Press.
13 Clean water for all
Richard Jolly

Introduction
Although the Millennium Development Goal for water does not refer to
sanitation and hygiene, these other two areas are as important for health
and poverty reduction as is safe water. In 2000, some 1.1 billion people
lacked access to safe water and 2.4 billion lacked access to adequate sanita-
tion. Allowing for population growth, this means that access to adequate
water is needed for an additional 1.5 billion people over the next fifteen
years and to adequate sanitation for an additional 2.2 billion. This would
represent about a doubling of the numbers who achieved access to safe
water over the 1990s and about a trebling of the numbers who got access
to adequate sanitation. No global data exist on those with inadequate
knowledge of hygiene.
This chapter argues that with political commitment, a clear policy focus
and a modest allocation or reallocation of resources, these goals are chal-
lenging but achievable. At the time of writing, some seventy developing
and transition countries, comprising about half of the population of the
developing world, seem to be on track. But another 100 are not: some
twenty-five are lagging or far behind while for another seventy-five no data
are available, a sign in most cases that the goal is not being taken seriously.
In almost all these 100 countries, including most of the poorer and least
developed countries, success will require a major acceleration of pro-
gramme action, much faster than that achieved during the 1990s. Success
will also require more focused development assistance, with more atten-
tion being given to low-cost water and sanitation activities in peri-urban
and rural areas.
Three myths about global goals need to be cast aside: that global goals
are never taken seriously; that they focus on the wrong issues; and that
global goals imply only top-down action by governments. Success during
the water decade of the 1980s and subsequently shows that global goals can
help focus and accelerate action, country by country. Moreover, the devel-
opment of a range of low-cost technologies and their increasing use show
what is possible. There are a number of success stories demonstrating
274 Richard Jolly
advance at a more rapid pace than the international goals require. What is
now needed is a co-ordinated international effort of mobilisation and
support, followed through at country level, with equal attention to all
three goals – of hygiene and sanitation as well as water – to ensure their
achievement by 2015.
The Goals for water and for sanitation and hygiene are elaborated in
the following two sections. Next, progress towards the targets is examined,
and evidence given on the present position of the developing world in
terms of access to safe water and sanitation, overall and broken down by
region. Then the increase in coverage required to fulfil the future world
population’s needs is discussed. The chapter then turns to what is needed
to achieve the targets: seven key areas of policy required are presented,
including social mobilisation, the role of public and private sectors, use of
appropriate technologies and pricing strategies, effective monitoring, and
the specific role of women in sanitation and hygiene. After this, action for
sanitation and hygiene is examined; then, more broadly, a strategy of ‘ten
commandments’ for turning formal goals into action is urged. Finally, the
prospects for achieving the targets for water, sanitation and hygiene are
examined.

The Millennium Development Goal for water

The seven Millennium Development Goals include one on water: halving
the proportion of the world’s population without sustainable access to
adequate quantities of safe drinking water by 2015. This water goal is
important, as a fulfilment of a basic human right, as a step to reducing
poverty more generally and as part of a sustainable strategy towards integ-
rated water resource management, globally and in every country.
Access to safe water has been recognised as a human right in a number
of international and national conventions – and explicitly in the Declara-
tion of 1977 in Mar del Plata11 and more recently in Article 24 of the Con-
vention on the Rights of the Child (UNICEF, 1990). Article 24 recognised
‘the right of the child to the enjoyment of the highest attainable standard
of health . . . through the provision of . . . clean drinking water, taking into
consideration the dangers and risks of environmental pollution’ (p. 57).
Though the Millennium Development Declaration and Goals refer only
to water, access to adequate sanitation and basic hygiene knowledge are
equally important. Indeed, in terms of their contribution to basic health,
sanitation and hygiene are arguably more important. Contrary to much
public opinion, recent research shows that hand-washing does more for
reducing child mortality and the incidence of diarrhoea than the provi-
sion of safe water or even basic latrines. For example, in a meta-analysis of
some fifteen micro-studies, nine from Asia, three from Africa, two from
Latin America and one from the USA over the years 1981–2000, Curtis
and Cairncross (2003) estimate that hand-washing on average is associated
 Clean water for all 275
with a 40 per cent reduction in the risk of infectious intestinal diseases.
This leads them to estimate that appropriate hand-washing with soap
could save between 0.4 to 1.5 million deaths from these diseases each year
– a mean estimate of avoiding almost 1 million deaths per year. Curtis and
Cairncross summarise the total number of deaths from infectious intesti-
nal diseases as 4.6 million in 1980, falling to 3.3 million in 1990 and 2.94
million in 1997, whilst they quote the latest World Health Organisation
(WHO) estimates as 2.2 million for 2000.
The conclusion to be drawn is not that safe water is unimportant, but
that people need access to adequate sanitation and basic hygiene know-
ledge as well as to safe water. All three contribute to health directly. Yet
indirectly, they also contribute to poverty reduction, since reducing the
incidence of ill health and disease of children and of adults frees the time
of women and men for more productive activities. Access to water and
adequate sanitation near the home can also save much of the time and
burden of collecting water far away, a cause of many girls being absent
from school as well as a heavy chore for children and women involved in
daily water collection. Less emphasised is access to improved sanitation
and hygiene as a step to ensuring dignity and privacy and, often, security,
especially for women and girls.

Goals for sanitation and hygiene

Though goals for sanitation and hygiene were not included in the Millen-
nium Declaration, a major effort was mounted to ensure the inclusion of a
sanitation goal in the declaration and plan of action of the World Summit
on Sustainable Development at Johannesburg in August and September
2002. Such a goal had already been adopted in several previous inter-
national meetings concerned with improving access to water, sanitation
and health.2
An important step in setting goals for hygiene and sanitation was taken
in the ‘Vision 21’ report presented to the Second World Water Forum at
The Hague in 2000. The goal proposed for sanitation was ‘by 2015 to
reduce by one-half the 1990 proportion of people without access to
hygienic sanitation facilities’ (WSSCC, 2000b). Meanwhile, for hygiene the
report set out several suggested targets for 2015:

• universal public awareness of hygiene;

• 80 per cent of primary schoolchildren educated about hygiene;
• all schools equipped with facilities for sanitation and hand-washing;
and
• a reduction in diarrhoeal disease by 50 per cent.

The report’s strategy to achieve water, sanitation and hygiene for all by
2025 was more challenging still, envisaging:
276 Richard Jolly
• good hygiene practices universally applied;
• all primary schoolchildren educated about hygiene; and
• a reduction in diarrhoeal disease by 80 per cent.

These targets are also summarised in other UN publications (WHO,

UNICEF and WSSCC, 2000: 2).

The present position

Table 13.1 summarises the numbers and proportions of the population in
developing countries lacking access to water and adequate sanitation in
2000. The figures presented in the table are the result of a world-wide
effort by WHO, the United Nations Children’s Fund (UNICEF) and the
Water Supply and Sanitation Collaboration Council (WSSCC) to assemble
data and estimate coverage, based on the latest evidence from as many
countries as possible in all regions of the world.
Though there are always problems of data quality and conceptual defi-
nition, this latest survey involved a number of improvements over previous
efforts to provide global estimates. More countries were involved than in
1990 and a wider range of sources were used, including both assessment
questionnaires and household surveys. These were distributed to country
offices through WHO country representatives with the support of local
UNICEF staff to relevant national agencies. Methodologically, the survey
was based on consumer response rather than data from service providers.
Even so, the quality of data is still often inadequate. For some countries,
including a number of industrial countries, data are totally missing. However,
by common consent, the quality and quantity of the data and the esti-
mates are much better than ever before.
Attempts were made in these surveys to apply clear definitions and

Table 13.1 Global water supply and sanitation coverage, 2000

World populationa (billions) Percentage of world population

Total With Without With Without

access access access (%) access (%)

Urban water supply 2.8 2.7 0.2 94 6

Rural water supply 3.2 2.3 0.9 71 29
Total water supply 6.1 5.0 1.1 82 18
Urban sanitation 2.8 2.4 0.2 86 14
Rural sanitation 3.2 1.2 2.0 38 62
Total sanitation 6.1 3.6 2.4 60 40

Source: WHO, UNICEF and WSSCC (2000: 8).

Note
a Figures have been rounded to the nearest 100 million.
 Clean water for all 277
common standards of access, based on the type of technology in use.
Table13.2 shows the classification of technologies, by those considered
‘improved’ and those ‘not improved’. The reasons for a technology to be
considered ‘not improved’ can include limitations on quantity as well as
quality, as illustrated by the inclusion of bottled water in the ‘not improved’
category.
Nonetheless, there are two other dimensions of the adequacy of water
and sanitation provision which could not be comprehensively included
through lack of data: access by distance to source of water (or sanitation
facility), and quality of water (or of upkeep of sanitation facility).
For water, access by distance (or by time taken from household to
source) should ideally consider separately differences such as having a
water supply inside the house or compound, within 100 metres or less, 100
metres up to 1 kilometre away, and more than 1 kilometre away. In densely
populated areas, even finer differences should ideally be taken into
account. For quality of water, maximum levels of iron and fluoride as well
as dangerous pollutants such as arsenic need to be considered.
For sanitation, some measures of distance and cleanliness also ought to
be included. But through lack of data, most of these dimensions have to
be ignored in global estimates – though of course they can and should be
brought into national and community surveys.

Access to safe water and sanitation by region

Tables 13.3 and 13.4 give the numbers and percentages with and without
adequate access to safe water and sanitation by region. In the case of Asia,
China and India alone comprised some 2,280 million people in the year

Table 13.2 Water supply and sanitation technologies considered to be ‘improved’

and those considered to be ‘not improved’

Water supply Sanitation

‘Improved technologies’
Household connection Connection to a public sewer
Public standpipe Connection to septic system
Borehole Pour-flush latrine
Protected dug well Simple pit latrine
Protected spring Ventilated improved pit latrine
Rainwater collection
‘Not improved technologies’
Unprotected well Service or bucket latrines
Unprotected spring (where excreta are manually removed)
Vendor-provided water Public latrines
Bottled water Open latrine
Tanker truck provision of water

Source: WHO, UNICEF and WSSCC (2000).

278 Richard Jolly
Table 13.3 Coverage of rural, urban and total water supply by region, 2000

Populationa (millions) Percentage

Total With Without With Without

access access access (%) access (%)

Rural
Africa 490 230 260 47 53
Asia 2,330 1,730 600 75 25
Latin America and 130 80 50 62 38
Caribbean
Urban
Africa 300 250 50 85 15
Asia 1,350 1,250 100 93 7
Latin America and 390 360 30 93 7
Caribbean
Total
Africa 780 480 300 62 38
Asia 3,680 2,990 690 81 19
Latin America and 520 440 80 85 15
Caribbean

Source: WHO, UNICEF and WSSCC (2000: 8).

Note
a Figures have been rounded to the nearest 100 million.

2000, or over 60 per cent of the region’s total population. Thus progress
in the Asian region is enormously influenced by performance in these two
countries. In this regard, one important caveat needs to be stressed:
because the Chinese authorities increased the standard by which they
assessed adequacy of sanitation in the 1990s, the gain in coverage by 2000
appears less than would otherwise have been the case.
This point is important in assessing progress in expanding water and
sanitation coverage (Table 13.5). Although the data presented in Table
13.5 for 1970 and 1980 are not strictly comparable to those of 1990 and
2000, they are included to give at least an order of magnitude of the
progress achieved. The most important conclusion to be drawn is that of
the enormous progress made over the past three decades. Over the 1980s,
during the years of the ‘Decade for Drinking Water and Sanitation’, access
to safe water more than doubled, whilst access to improved sanitation
nearly tripled.
Progress in the 1990s was quantitatively not so large but was still impres-
sive. For example, because of rapid increases in population, the propor-
tions of people with access to water and sanitation inched forward rather
than leaped ahead. Nonetheless, by 2000 probably five or six times more
people had access to safe water than thirty years before, and four or five
times more people had access to adequate sanitation. These are remark-
 Clean water for all 279
Table 13.4 Access to rural, urban and total sanitation by region, 2000

Populationa (millions) Percentage

Total With Without With Without

access access access (%) access (%)

Rural
Africa 490 220 270 45 55
Asia 2,330 710 1,620 31 69
Latin America and 130 60 570 49 51
Caribbean
Urban
Africa 300 250 50 84 16
Asia 1,350 1,050 300 78 22
Latin America and 390 340 50 87 13
Caribbean
Total
Africa 780 470 310 60 40
Asia 3,680 1,770 1,910 48 52
Latin America and 520 400 120 78 22
Caribbean

Source: WHO, UNICEF and WSSCC (2000: 8).

Note
a Figures have been rounded to the nearest 100 million.

Table 13.5 Estimated coverage of drinking water supply and sanitation in develop-
ing countries, 1970, 1980, 1990 and 2000

Population with access (millions) Percentage of total population with access

1970 1980 1990 2000 1970 1980 1990 2000

Water
Urban 320 530 1,100 1,480 67 75 95 94
Rural 180 570 1,190 1,480 14 29 66 71
Total 500 1,100 2,300 2,960 29 43 79 82
Sanitation
Urban 340 370 940 1,370 71 53 82 86
Rural 130 210 600 780 11 13 35 38
Total 470 580 1,540 2,150 27 25 55 60

Sources: 1990 and 2000 – as for Tables 13.1–13.4; 1970 and 1980 – WHO (1981).

able achievements, which cast a positive light on the possibility of expand-

ing coverage much further in the years up to 2015.
Many lessons have been learned nationally and internationally over the
past two decades about the actions required to ensure sustained access to
water and sanitation of adequate quality. Some of the most important
include:
280 Richard Jolly
• the importance of ensuring systems for self-sustaining maintenance as
opposed to simply installing hand pumps, tube wells and sanitation
facilities;
• the need to involve communities in the provision of facilities, their
maintenance and their financing;
• the need to ensure empowerment of women in decision-making and
action both about new programmes and in the running and mainte-
nance of existing ones;
• the multiplier effects of focusing on primary and secondary schools
and schoolchildren, to ensure that schools have adequate sanitation
facilities separately for girls and boys and adequate access to safe water.

Schoolchildren can also be agents of change: playing a vital and catalytic

role in initiating hygienic behaviour and in spreading new knowledge and
practice in their families and their communities.

Future needs
Table 13.6 shows the projections of population growth and their implica-
tion for halving the proportion of people without access to water and ade-
quate sanitation by 2015. The figures are large but they are not out of
proportion with past achievements except in the areas of urban sanitation
and rural water supply. Even in these respects, the major challenges are
those in Asia and Africa, not the Middle East or Latin America.
Achieving the goals for water, sanitation and health is neither extremely
difficult nor extremely costly. Examples exist in all regions of the world
that show how accelerated progress is possible and affordable within the
budgetary resources of the country and communities concerned. South
Africa, for instance, embarked on a major programme for the provision of
safe drinking water in 1994, when the democratic government came to
power. Within seven years, by 2001, the new government had halved the

Table 13.6 Additional population required to be covered to meet the 2015 targets
for water and sanitation

Total population (billions) Coverage (%) Additional

population
2000 2015 2000 2015 (billions)

Urban water supply 2.8 3.8 94 97 1.0

Rural water supply 3.2 3.3 71 85 0.6
Total water supply 6.1 7.2 82 91 1.6
Urban sanitation 2.8 3.8 86 92 1.1
Rural sanitation 3.2 3.3 38 69 1.1
Total sanitation 6.1 7.2 60 81 2.2

Source: WHO, UNICEF and WSSCC (2000: 32–33, table 5.1).

 Clean water for all 281
numbers without safe water, thus achieving the global goal fourteen years
ahead of 2015. South Africa now has the goal of achieving safe water for
all by 2008. Sanitation progress went more slowly – until the cholera out-
break of 2000; this seemed to act as a wake-up call. South Africa has now
increased expenditure and latrine programmes, in one year achieving
more than in the previous six years, with the goal of sanitation for all by
2010. South African experience demonstrates the importance of clear
political commitment.
Data are also available showing the different rates of progress in expand-
ing access to safe water made over the past few years against the rates
required to achieve the Millennium Development Goal for water. This
assessment is shown in Table 13.7, which highlights the number of coun-
tries that have already achieved the Goal, and how no country, at least of
those for which data exist, are moving in the wrong direction.
However, even with positive experience in many countries, barely half
of the population of developing countries live today in countries on track
to achieve the goal, whilst very few have already achieved it. About a third
of the population of the South live in countries lagging far behind.
Another 12 per cent live in countries without the data to assess the situ-
ation – a sign in most that achieving the goal is not in sight.

Table 13.7 Country progress on halving the proportion of people without sustain-
able access to safe drinking water by 2015

Number of countries Goal On track Lagging Far Slipping No data

achieved behind back

Sub-Saharan Africa 1 9 4 9 0 21
Arab States 0 8 0 3 0 6
East Asia and the Pacific 0 6 1 4 0 8
South Asia 3 4 0 0 0 1
Latin America and the 1 21 1 2 0 8
Caribbean
Central and Eastern 0 8 0 0 0 17
Europe and the
Commonwealth of
Independent States
Total number of 5 63 7 18 0 75
countriesa
Percentage of total world 4 39 5 27 0 10
population
Percentage of Third 5 45 6 32 0 12
World population

Source: UNDP (2002: 25).

Note
a Regions include only Human Development Index countries while the total includes all UN
member countries excluding high-income Organisation for Economic Cooperation and
Development (OECD) members.
282 Richard Jolly
Donor countries have a role to play in supporting action to achieve
these three vital goals. But the starting point has to be national action – to
recognise the goals as a national priority, to prepare action plans for
their achievement, to open opportunities for community action, and to
mobilise public awareness and support, especially for sanitation and
hygiene.

Seven key areas of policy

Experience over the past two decades underlines seven areas of policy and
action which are critical for accelerating country-by-country action towards
the achievement of the goals.
First, few developing countries will be able to achieve the goals for
hygiene, sanitation or water without widespread social and community
action, which in turn requires the mobilisation of populations. To rely on
government or the private sector alone in countries of limited resources will
require expenditures far beyond the revenues available or the capacity of
households and communities to pay. In contrast, approaches based on social
mobilisation, in which individual and community action is combined with
that of local or central government, can bring into play the additional labour
and additional finance to make the goals achievable. These approaches have
often been used in many of the countries and states that have seen rapid
expansion of water or sanitation facilities. Social mobilisation seems almost
the only way in which hygiene education can be expanded to the point
where behaviour is influenced on a major scale.
Second, involvement of the private sector will be needed. Several years
ago, privatisation in water was seen as the panacea and promoted widely by
institutions such as the World Bank. At the time of writing, this has given
way to a more pragmatic approach, in the light of experience, especially the
failure of some of the early private-sector schemes and the reluctance of the
large private-sector water companies to invest further in developing coun-
tries. Privatisation of the water system in Buenos Aires, hailed as a great
success three or four years ago, is now seen as a disaster.
But the private sector still has an important role to play. In matters of
sanitation, and in rural areas more generally, small-scale private entre-
preneurs have the skills needed and have often demonstrated a capacity to
contribute on an increasing scale. They are often ideal for making the
concrete ‘slabs’ required for simple latrines, for installing local-level water
taps and supplies, and occasionally for providing water through drums
and bowsers or tankers where these are needed.
Competition is important to ensure that these contributions are made
at low cost. In larger towns and cities, the private sector in the form of
national and international water companies, if willing, can also play an
important role in providing water and sanitation as part of management
of major schemes. Here, however, it is important to ensure that the con-
 Clean water for all 283
tracts for such companies require them to reach out to the peri-urban
areas, especially to poorer communities.
Third, more use of simple, low-cost technologies and approaches is
needed. Hand pumps, improved wells, rainwater harvesting, installations
using volunteer labour, community maintenance – all these are approaches
which are relevant and cost-effective in rural situations, as well as many
peri-urban areas. Vision 21 of the Water Supply and Sanitation Collabora-
tion Council (WSSCC) estimated that some US$9 billion per year would
be needed between 2000 and 2025 to meet the goals of water and sanita-
tion for all (WSSCC, 2000a: 28). These estimates were based on a cost of
US$15 per person for water in rural situations and US$50 in peri-urban
areas. For sanitation and hygiene promotion, the average costs were esti-
mated to be US$10 and US$25 per person respectively. These figures were
based on a small sample of actual projects and programmes. In contrast,
the World Water Commission (2000; see also Global Water Partnership,
2000) estimated that some US$180 billion would be needed each year for
ten years to ensure water and sanitation in line with the goals (Global
Water Partnership, 2000; World Water Commission, 2000). What explains
these enormous differences?
A review of the estimates shows that most of the difference is explained
by the much larger estimates in the World Water Commission calculations
for urban sanitation and waste disposal. For water, the Commission esti-
mated an additional US$17 billion per year or US$170 billion over ten
years compared with US$90 billion for water and sanitation and hygiene
in the WSSCC estimates. The difference between these estimates mostly
reflects the level of technology implied. The WSSCC estimates focused on
the needs of poor people in rural and peri-urban communities. The Com-
mission estimates focused on urban infrastructure, including large-scale
technologies.
A fourth key area of policy involves better monitoring, with the results
more publicly disseminated. Monitoring is required for efficient manage-
ment, but also for effective social mobilisation. Monitoring means the
development of a regular system of reporting, sufficient to demonstrate
progress in a way that can be reported publicly to sustain interest, enthusi-
asm and political support. Of course, monitoring is also important for
administration and efficient budgeting, although experience shows that
monitoring for social mobilisation is even more important for sustaining
action. Publicising progress in expanding water state by state in India
stimulated popular demand and support for water, just as it had in many
countries in areas such as child immunisation.
A fifth point is that it is important to set a price policy which is politic-
ally acceptable and equitable. Full-cost pricing has become the conven-
tional wisdom, pushed strongly by the World Bank, the International
Monetary Fund (IMF) and others working within ‘the Washington consen-
sus’. For example, the World Water Commission has stated that full-cost
284 Richard Jolly
pricing is the single most important policy required to deal with water
problems of the twenty-first century (WWC, 2000). In contrast, the Vision
21 report takes a very different position (WSSCC, 2000a). While recognis-
ing that issues of pricing are important, Vision 21’s approach is pragmatic.
It argues that prices must be set in relation to the capacity of different
groups to pay and in relation to the overriding commitment to ensure
access to water and sanitation for all, in line with international commit-
ments on human rights.
The Dublin Principles, agreed at a major international water meeting
in 1991, provided guidelines for integrated water resource management.
They emphasised that water needed to be recognised as an economic
resource. But the principles neither stated nor implied that full-cost
pricing was necessary. The UK government has given a balanced summary
of the issues, noting that, ‘it is important to continue to recognise water as
a social and ecological good as well as an economic good. Indeed, the
Dublin conference also recognised that access to clean water and sanita-
tion at an affordable price is a right for all human beings’ (DFID, 2001: 32).
In practice, price is an important means to ensure the economical use of
a scarce resource and an important means to raise the revenue required
for both installing and maintaining water supplies. The same can be said of
pricing for sanitation. But simply moving to full-cost charging is too simple.
It usually means charging full cost to those who are at present without
water or adequate sanitation facilities – and continuing subsidies to those
who have long had in-house connections at far below full-cost prices.
In respect of water, a pragmatic approach would start by considering
the relative benefits of different levels of charging for the main groups of
consumers – those already with access to supplies piped direct to their
houses and those relying on community standpipes and other such facili-
ties. Often, those with supplies direct to their own households are both
the main users and the ones receiving the greatest level of subsidy. Charg-
ing full-cost pricing to this group of consumers holds great potential for
generating more resources for the drinking water sector – though, of
course, it may involve political difficulties. But according to the WSSCC,
this should be the starting point.
For poorer communities, some form of cost recovery is also often desir-
able, both to generate resources and to encourage efficient use of sup-
plies. But given high levels of poverty and the need to ensure access to
water as a human right, systems of cost recovery and resource generation
should be judged in relation to capacity to contribute. Often this prag-
matic approach may lead to communities being encouraged to provide
labour in kind to ensure rapid installation of community supplies by hand
pumps and public facilities. Full cost recovery might be adopted for those
households wanting to establish connections to their own individual prop-
erties. A priority in all cases is to ensure that charging covers maintenance
costs, though again successful examples exist of communities providing
 Clean water for all 285
their own maintenance services. This has proved successful in the Swach
project in southern Rajasthan, as well as in Nigeria. Women often prove
more reliable in maintaining hand pumps, in part because they are
usually the main users, and also because when trained in maintenance
they are less likely to move from the community in search of a job else-
where. This was the logic behind many UNICEF programmes for training
women in pump maintenance in Sudan, India and Bangladesh.
The dangers of insisting on full-cost pricing in all situations cannot be
stressed too strongly. At its extreme, full-cost pricing is elevated to the
level of an unjustified piece of economic ideology, and can also be highly
misleading.
A sixth key area of policy concerns the need to ensure the full partici-
pation of women in the management and operation of water, sanitation
and hygiene programmes. Women are more affected by inadequacies of
arrangements in these critical areas and almost always are more motivated
to do something to improve the situation. Being the daily drawers and car-
riers of water, they are also usually the best informed about the inadequa-
cies of the present, and the clearest about what can and should be done.
However, very often their voices are not heard and household resources
are not theirs to control. Nonetheless, there are increasing examples of
how this can be changed if women are empowered to exercise more sani-
tation control and influence (WSSCC, 2000a, b). In Gujarat in India, in
Ethiopia even during the civil war, in Nigeria, Vision 21 gave many
examples and built on them. If the goals are to be reached, women’s
empowerment and control will need to become much more common in
water and sanitation than it is at present.
The seventh and final need is for development assistance to be more
directly focused on water and sanitation for the poorest. Given the strong
donor support for the Millennium Development Goals, one would imagine
that donor support for water and sanitation for the poorest would be
readily forthcoming. In practice, the bulk of international support goes to
urban schemes, at relatively high cost and benefiting the better-off urban
communities. Support for water and sanitation schemes in the rural and
peri-urban areas forms only a small fraction of the total – probably less
than 20 per cent.
This led to the call for ‘20:20’, an agreement first proposed in the
Human Development Report (UNDP, 1994: 77) with strong support from
UNICEF and subsequently endorsed as a recommendation of the World
Summit for Social Development in 1995. As discussed elsewhere in this
volume, under the 20:20 proposal donors would allocate 20 per cent of
their aid for key areas of basic needs – primary education, primary health
care, nutrition, family planning and reproductive health care – and low-
cost water and sanitation programmes. Developing country governments
would allocate a similar 20 per cent of their public expenditures to
the same priorities. Calculations showed that these two actions would be
286 Richard Jolly
sufficient on average to generate the financial resources required to
achieve the universal provision in the priority areas within a decade or so.
Some thirty developing countries indicated their willingness to support
this approach during the late 1990s – but only two or three donors. Now,
with the formal adoption of the Millennium Development Goals, the time
seems ripe to renew commitment to this practical measure. But whether
or not it is formally adopted, the underlying conclusion still remains. Both
donor and developing countries could generate the vast bulk of the
resources required to meet the Goals in general – and the goals for water,
sanitation and hygiene in particular – by a better allocation of resources
already going into these sectors. Donor contributions to water, sanitation
and hygiene need to be carefully reviewed in the light of the Millennium
Development Goal priorities, and reallocations or additional resources
provided to ensure the international support needed, especially for poorer
countries committed to achieving the goals.

Action for sanitation and hygiene

As regards sanitation, many low-cost schemes have relied on the sale of
concrete ‘slabs’, around which individual households can build a simple
latrine block to whatever standards they choose. One example is the venti-
lated improved pit latrine (VIP), which is often ideal. The ‘VIP’ can also
be relatively cheap, even though it provides a much higher and better
level of service, free of flies and smells. Communities also have developed
a range of latrines and simple shower and bathing blocks themselves.
Sometimes these are run by the community itself, or sometimes by an indi-
vidual entrepreneur or the local government. Such blocks can be made
available at a relatively low charge, sufficient to cover their maintenance.
In the past ten years an increasing number of successful examples have
been developed for all these approaches. In India the Sulabh movement
led by Dr Bindeshawar Patak has provided toilet facilities for over 10
million people. The twin-pit latrine has been especially successful.
However, sanitation means much more than the provision of latrines
with appropriate technologies. Recent years have witnessed the develop-
ment of many new approaches. Separate latrines for girls and boys in all
schools is one critical strategy, the starting point for any serious school
messages on hygiene but also a step in mobilising children as agents of
change in their homes and villages.
Strategies of ‘100 per cent sanitation’ have proved extraordinarily suc-
cessful in Bangladesh in motivating whole villages to improve their toilets
and waste disposal systems. The approach begins with a village-wide assess-
ment of the whole situation, involving the community and outsiders slowly
walking together along a transept across the whole village. Often it is
shame at the recognition of the situation that is revealed by villagers and
outsiders working together which triggers a collective determination to
 Clean water for all 287
make changes. It is then left to individual villagers to decide on what they
can and will do for their individual homes and plots. No resources are
spent in subsidising latrine construction, but rather funds are targeted at
supporting the whole process and promoting general awareness of the
need for better hygiene (Kar, 2003).
This appears to be a general rule for latrine construction and other
forms of household sanitation improvement. It is important to recognise
that individuals want to improve their houses, and will do so if low-cost
opportunities are available. There is often no need for subsidies. Rather, it
is possible to let the market work, with local craftspeople providing the
materials and often doing the work. Instead, resources can be channelled
into training and motivation, hygiene education and promotion.

The ten commandments for success with goals

Much of the current interest in setting global goals derives from the suc-
cessful experience of UNICEF in the 1980s and 1990s. In 1985, UNICEF
made a formal, major and public commitment to the global pursuit of the
goal of universal child immunisation as a critical step to reducing child
mortality. Universal child immunisation had already been adopted as a
key goal by WHO and UNICEF. However, it was only when UNICEF pub-
licly committed itself to the achievement of this goal by 1990, and mobilised
its own staff, resources and reputation to its achievement, that the goal
moved to high international profile.
In country after country, major efforts were undertaken to accelerate
action and increase coverage. These involved campaigns: mobilisation of
churches, mosques, teachers, women’s groups and political leaders, as well
as health workers, to the goal of reaching in each country at least 80 per
cent coverage of immunisation against tuberculosis, measles, polio, diph-
theria, tetanus and whooping cough. Contrary to the expectations of
many sceptics, by 1990 coverage of these six antigens in developing coun-
tries had reached 80 per cent or more on average. Some seventy-two devel-
oping countries had individually achieved the goals. These included many
countries in Latin America and sub-Saharan Africa, even though the 1980s
was economically a lost decade for economic development.
All this is a great example of the power of global goals, provided they
are made a priority for a UN agency and are used in a process of national
mobilisation at country level. In the case of UNICEF, the achievement of
the goals of immunisation in 1990 was used to set a range of other goals
for the next decade, which were formally adopted at the World Summit
for Children in September 1990. These goals covered ten major areas and
seventeen supporting sub-goals.
UNICEF’s experience in pursuit of goals was later summarised in a
tribute entitled ‘The Ten Commandments of Jim Grant’s Leadership for
Development’ (Box 13.1) (Gautam, 2001). These commandments showed
288 Richard Jolly
Box 13.1 Jim Grant’s ten commandments

1 Articulate your vision for development in terms of inspiring goals.

2 Break down goals into time-bound doable propositions.
3 Demystify techniques and technologies.
4 Generate and sustain political commitment.
5 Mobilise a grand alliance of all social forces.
6 Go to scale.
7 Select your priorities and stick to them.
8 Institute public monitoring and accountability.
9 Ensure relevance to the broad development agenda.
10 Unleash the full potential of the United Nations system.

Source: Jolly (2001).

how formal goals need to be turned into action nationally and globally in
ways that mobilise and sustain the interest of people. It requires a political
and people-centred process, rather than initiatives of a more technical or
administrative nature.
Implementing these ten commandments can help create the enabling
environment that in turn can set the stage for accelerating action country
by country. Top-down commitment must of course be matched by grass-
roots response. Action at both levels is required. When the two are com-
bined, achievement of the global goals becomes possible.

Prospects
Finally, we turn to the future. What are the prospects, at the time of
writing in 2002, that the goals will be achieved?
Without doubt, some countries will achieve them. Ten already have
achieved the goals for water and another fifty-eight are estimated to be on
track. But much could happen, positively or negatively, during the thirteen
years before the target date of 2015 is reached. All that one can do today is
to offer a conditional forecast of likely future achievements – and a highly
conditional forecast at that.
Recall first that there are three important goals, not one – for water,
sanitation and hygiene. Recall also that for each there are two quantitative
dimensions of achievement: halving the proportion of the total popu-
lation of developing countries without access and halving the proportion
without access within each developing country.
As regards the first, halving the proportion of the total population of
developing countries without access, the critical determinant is what
happens in China and India. These two countries alone comprise almost
half of the total population of developing countries. More important, they
comprise an estimated 56 per cent of the total additional population to be
 Clean water for all 289
reached with safe water and an estimated 64 per cent to be reached with
sanitation, if the 2015 goals are to be achieved. If one had to make a fore-
cast today, one would do well to bet that China and India will achieve the
water goals on the basis of past experience and achievements.
India is assessed as being on track for achieving the goals for safe water
and China, though assessed as lagging (UNDP, 2002), has made consider-
able advances and has demonstrated remarkable capacity in the past for
accelerated action in the social as well as economic sectors. Four of the
world’s five next most populous countries – Indonesia, Brazil, Pakistan and
Bangladesh, but not Nigeria – are also assessed to be ‘on track’, to judge
by their progress over the 1990s (UNDP, 2002: 46–49). If these six coun-
tries themselves achieve the water goal, their weight will do much to carry
the global average towards the global water goal.
Sanitation presents a less clear picture. China and India have each
expanded coverage over the 1990s to reach about an additional 10 per
cent of their population (WHO, UNICEF and WSSCC, 2000: 47–48). The
size of the backlog means, however, that in each case this is barely a
quarter of the distance to go to achieve the target by 2015. Both countries
are assessed to be ‘far behind’ and need about a doubling of their rate of
advance to reach the sanitation target. Without this, their weight in the
developing country average will make very difficult the achievement of the
global sanitation goal. The same four of the five next most populous coun-
tries are, however, all judged to be on track to achieve the sanitation goal.
But what about the prospects for goal achievement in other individual
countries? This brings us back to the key policy issues mentioned earlier.
Using low-cost technologies, the achievement of goals for safe water, sani-
tation and hygiene are all technically possible with only a modest alloca-
tion of additional resources or reallocation of existing ones. And, after all,
the goals are modest: only to halve the proportion without access, in coun-
tries where half to two-thirds already have access.
Whether the water goal will be achieved in any particular country will,
in my assessment, be mostly a matter of three factors:

• political commitment, to go to scale and to provide national leadership

to make the goal real and meaningful;
• willingness to adopt participatory approaches which enable and support
local participation and management, especially by women;
• willingness to ensure adequate resources for programmes in the rural
and peri-urban areas. In the poorest countries, resources will also be
required with sustained back-up from donors.

For sanitation and hygiene, the same three factors will be critical. But in
addition two others will be important. First, it will be necessary to ensure
that the education system gives proper attention to issues of hygiene and
sanitation with backing from the national media. Second, support will
290 Richard Jolly
need to be obtained from the private sector, especially from soap manu-
facturers. With such support, the goals for sanitation and hygiene can
readily be achieved in any country.
One final point needs to be made. In most of Asia and Latin America
the above factors cover the main issues. But in the least developed coun-
tries, and in most of sub-Saharan Africa, achieving the goals will also
depend on – and be part of – sustained long-run economic development,
of a sort that has not been seen for two or more decades. Civil conflict and
collapse of local administration hinder all aspects of development, includ-
ing pursuit of the goals for poverty reduction and the goals for safe water,
sanitation and hygiene. The priority problems of Africa are now widely
recognised. The presidents of South Africa, Algeria, Egypt, Nigeria and
Senegal have established the New Partnership for Africa’s Development
(NEPAD). The Millennium Development Goals for poverty reduction are
part of this agenda. Whether they are achieved in this continent of such
desperate need will depend on how seriously the goals are taken by the
individual countries and made part of the new partnerships with donor
governments and institutions of the international community.

Notes
1 United Nations (UN) Water Conference held at Mar del Plata, Argentina,
March 1977.
2 See the report of the International Conference on Freshwater held in Bonn,
December 2001 (Germany, 2000). See also WSSCC (2000a).

References
Curtis, V. and Cairncross, S. (2003) ‘Effect of Washing Hands With Soap on Diar-
rhoea Risk in the Community: a Systematic Review,’ The Lancet Infectious Diseases,
3(5): 275–281.
Department for International Development (DFID) (2001) Addressing the Water
Crisis: Healthier and More Productive Lives for Poor People, Strategy Paper, London:
DFID, March.
Gautam, K. (2001) ‘Ten Commandments of Jim Grant’s Leadership for Develop-
ment’, in Jolly, R. (ed.) Jim Grant: UNICEF Visionary, Florence: UNICEF, pp.
137–144.
Germany, ‘Federal Ministry for the Environment, Nature Conservation and Nuclear
Safety and the Federal Ministry for Economic Cooperation and Development
(2000) Water: A Key to Sustainable Development, report, Bonn: Courir-Druck GmbH.
Global Water Partnership (2000) Towards Water Security: A Framework for Action,
Stockholm: Global Water Partnership.
Jolly, R. (ed.) (2001) Jim Grant: UNICEF Visionary, New York: UNICEF.
Kar, K. (2003) ‘Subsidy or Self Respect? Participatory Total Community Sanitation
in Bangladesh’, IDS working paper, WP184, Brighton, Sussex: IDS.
United Nations Children’s Fund (UNICEF) (1990) First Call for Children, Conven-
tion on the Rights of the Child, New York: UNICEF.
 Clean water for all 291
United Nations Development Programme (UNDP) (1994) Human Development
Report 1994, New York: Oxford University Press.
—— (2002) Human Development Report 2002, New York: Oxford University Press.
Water Supply and Sanitation Collaboration Council (WSSCC) (2000a) Vision 21: A
Shared Vision for Hygiene, Sanitation and Water Supply and A Framework for Action,
Final Report, Geneva: WSSCC.
—— (2000b) Vision 21: A Shared Vision for Hygiene, Sanitation and Water Supply and a
Framework for Action, Proceedings of the Second World Water Forum, The
Hague, 17–22 March 2000, Geneva: WSSCC.
World Health Organisation (WHO) (1981) Drinking Water and Sanitation 1981–1990:
A Way to Health, Geneva: WHO.
World Health Organisation (WHO), United Nations Children’s Fund (UNICEF)
and Water Supply and Sanitation Collaboration Council (WSSCC) (2000) Global
Water Supply and Sanitation Assessment 2000, Report, Geneva: WSSCC.
World Water Commission (2000) Commission Report: A Water Secure World, Mar-
seilles: World Water Council.
14 Achieving sustainability in Africa
James Fairhead

Introduction
This chapter reflects on the form that ‘sustainability’ takes within the
seventh Millennium Development Goal, which is broadly to ‘ensure environ-
mental sustainability’. In particular, it considers the linked target, to
‘integrate the principles of sustainable development into country policies
and programmes and reverse the loss of environmental resources’. This
Goal and target are themselves adapted from the International Develop-
ment Target, which envisaged ‘the implementation of national strategies for
sustainable development in all countries by 2005, so as to ensure that
current trends in the loss of environmental resources are effectively
reversed at both global and national levels by 2015’.
A huge literature has emerged concerning sustainability and develop-
ment, prompted first by the World Summit at Rio and again in the build-up
to the 2002 World Summit on Sustainable Development at Johannesburg,
and its outcomes. I do not attempt a review here. Rather, I am concerned
to explore how the concept of sustainability is being elaborated around
the Millennium Development Goal, and the targets and strategies and
programmes to achieve them, considering this particularly in relation to
Africa. The main argument is that its elaboration in documentation and
initiatives minimises or obscures the way local issues are linked to global
forces, whether economic or environmental. Crucial aspects concerning
the transnational corporate and global dimensions to current resource
uses and their sustainability are rendered marginal, not centre stage, in
the focus on country policies and programmes of the Millennium Develop-
ment Goals. In particular, I argue that the impact both of the ‘new scram-
ble for Africa’ by transnational corporations, and of impending global
climate change on African environmental resources, threatens to under-
mine or override country policies for sustainability.
The Millennium Development Goals do have a Goal concerning global
partnership which considers transnational issues, but this neither men-
tions environmental issues nor balances the potential benefits that corpo-
rations can bring for technology-sharing with critical evaluation of their
 Achieving sustainability in Africa 293
terms of engagement – and there is no evidence that the rhetoric of partner-
ship has translated into more equal international relations (see Chapter 16).
The significance of global economic and environmental forces was
highlighted in deliberations and negotiations during the 2002 World
Summit on Sustainable Development. It has, for example, been politically
difficult for European and American negotiators to consider lifting pro-
tectionist trade barriers, and to regulate the problematic conduct of trans-
national corporations. It has been politically impossible for American
negotiators to deliberate global climate agreements further. It has been
easier for the same negotiators to focus on support to local and national
conservation and sanitation initiatives in poorer countries, assisted in this
by a rhetoric which places the poor centre stage. The focus on ‘country
programmes’ in the International Development Target and the Millen-
nium Development Goal reflects the ease with which donor nations of the
Organisation for Economic Cooperation and Development (OECD) can
involve themselves with sustainability in poorer countries through local
and national initiatives there. It also reflects the political difficulty that
they face when addressing what I shall argue are the far more important
questions of international trade and global climate change.
It should be said, however, that at the World Summit there were
attempts among Northern politicians and pressure groups (amongst them
the UK leadership) to address environmental sustainability by concentrat-
ing on the very problems of global political economy and global climate
change. The UK Prime Minister, Tony Blair, for example, made a speech
proposing codes to ensure transparency in payments made by trans-
national corporations to African governments; payments which have been
at the heart of unaccountable government and conflicts in the region. He
re-emphasised the importance of global climate change. It is to be hoped
that such speeches, and the debate at Johannesburg that signalled their
pertinence, might help mark a change in emphasis in approaches to
poverty and sustainability, and in the formulation of the Millennium
Development Goal.
As it is, however, these international political-economic dimensions to
resource use and ‘sustainable development’ are conspicuously absent
from most documentation and strategies linked to the Millennium Devel-
opment Goal. Instead, the ways that sustainability is discussed parochialise
it. Let us consider the UK Department for International Development’s
(DFID’s) Strategy Paper for achieving the sustainability International
Development Target, called Achieving Sustainability: Poverty Elimination and
the Environment. This document certainly acknowledges that it is the
wealthy who cause most environmental degradation. The solutions it pro-
poses, however, still seem to lie with the poor: thus to quote the executive
summary, ‘The main causes of environmental degradation are unsustain-
able consumption, particularly of the rich, both in developed and devel-
oping countries’ (DFID, 2000a: 8) and ‘globally, most environmental
294 James Fairhead
degradation is caused by the non-poor as the consumption levels of the
poor are still low relative to the rich’ (ibid.: 12). Yet when actions are dis-
cussed to meet International Development Targets, they ‘are looked at
primarily from a poor country perspective. . . . Emphasis is placed on
working with the poor and on improving often weak and ineffective
systems of governance.’
The poor are highlighted because they suffer from degradation, not
because they cause it. ‘Pro-poor’ policy, however, fades into ‘working with
the poor’ rather than with the conditions that lead to poverty. It responds
to the supposed cycle of ‘poverty and degradation’, rather than with the
far worse cycle of wealth and the degradation of poor people’s resources.
The problem is not that these issues are absent, but that they are mar-
ginal. Thus the DFID strategy paper acknowledged that a pro-poor focus
will involve ‘controls on environmentally damaging activities of the non-
poor’ (ibid.: 26). Yet this is hardly pursued. There is, after all, much less
experience in this field to draw on in order to identify best practice. It has
long been easier to help the poor than hinder the rich. More significantly,
the extent to which this is the local or the international wealthy is not
addressed, and neither is the capacity of the wealthy to circumvent such
control.
The second pillar of the Millennium Development Goals, or at least of
the strategies promoted to date by official development agencies to achieve
them, is trade liberalisation. Again, to take the DFID strategy paper as
exemplar, trans-boundary concerns are raised in relation to globalisation.
Yet any potential negative environmental impacts of trade liberalisation
are to be addressed by national strategies and environmental policies:
‘Without the implementation of appropriate environmental policies in all
countries, trade liberalisation could lead to environmental damage.’ So, it
is

a mistake to argue against further liberalisation on environmental

grounds. Marginalising poor countries from the benefits of global
trade and investment will not prevent environmental degradation and
will alienate developing countries from negotiations on global environ-
mental issues. Rather the challenge of globalisation is that it intensi-
fies the need for better enforcement of appropriate environmental
policies at the national level.
(ibid.: 20)

Within this logic, the national strategies of the Millennium Development

Goal are proposed as the ‘main vehicle for integrating pro-poor economic
growth with social improvement and a responsible approach to environ-
mental management’ (ibid.: 11; see also DFID, 1997). In the DFID strat-
egy paper there is considerable deliberation over the criteria that could be
used to assess whether a country has successfully incorporated considera-
 Achieving sustainability in Africa 295
tions of sustainability in its development policies and programmes. Indeed,
the UK/DFID input into the 2002 Summit was envisaged as focusing on
‘how far countries have effective sustainable development processes in
place.’
The emphasis is, however, on poor countries to enforce regulation, rather
than on the international financial system. This intensifies the need for
‘better enforcement’ at a time when there are fewer national financial
resources, and, as will become clear, at a time when there is intensified inter-
national/multinational interest in African natural resources, and increased
financial dependency among African governments on these. The call is for
the indebted to regulate the enriched.
Exemplars of success in such regulation should not mislead us. Espe-
cially in instances where donors have been willing to support the creation
and implementation of regulatory policy, there have been moves towards
such regulation. The DFID strategy documents highlight Uganda and
Ghana as cases. Yet it can be questioned whether these are replicable
exemplars of the way forward. Are they not better understood as ‘policy
oases’ lush on exterior support, or even ‘policy mirages’ stronger on docu-
mentation than implementation, which would be misleading as a basis for
sustainable planning?
Moreover, pitting ‘national programmes’ against the ‘forces of globali-
sation’ misses the mutuality between governments of low-income coun-
tries (and their elite), transnational corporate finance, and international
geopolitical alliances. It is to this issue that I now turn, with a particular
focus on sub-Saharan Africa. The next section examines the issue of sus-
tainability in Africa in the context of the international political economy
and is followed by a section that discusses global and local causes of
climate and environmental damage.

Sustainability and international political economy in Africa

The very same month that the UK published its White Paper Eliminating
World Poverty: Making Globalisation Work for the Poor (DFID, 2000b), the US
National Intelligence Council (NIC) – part of the US Central Intelligence
Agency (CIA) – published its own view of global futures in its Global Trends
2015 (NIC, 2000). This acknowledged straightforwardly that by 2015 – the
maturity date for the Millennium Development Goals – ‘most African
states will miss out on the economic growth engendered elsewhere by
globalisation and by scientific and technological advances’ (NIC, 2000:
71). This dovetails with the view of many African negotiators expressed at
the Johannesburg World Summit when they argued that freeing restric-
tions on transnational corporate engagement with Africa is leading less to
African development through inward investment, than to a more parasitic
and exploitative engagement.
The NIC vision for Africa in 2015 is of a catastrophe, reproducing the
296 James Fairhead
image of ‘hopeless continent’ beloved of contemporary journalists (e.g.
The Economist, 2000). In some contrast to the image of despair, however,
the NIC report goes on to predict that Africa will be supplying 25 per cent
of North American oil imports by 2015. New oil discoveries in the Demo-
cratic Republic of Congo, Uganda, Sierra Leone and Côte d’Ivoire will be
adding to the huge reserves to be found in Nigeria, Sudan, Chad, Nigeria,
Angola, Gabon, Congo, Equatorial Guinea and others. Despair, however,
is maintained, as despite these resources, the Council argues, ‘patterns of
oil wealth fostering corruption rather than economic development will
continue’ (NIC, 2000: 73). This is an extraordinary statement. The NIC is
happy to affirm this as if African corruption had nothing to do with the
strategic interests of the Great Powers, and the commercial interests of
their corporations. Nothing could be further from the truth.
Through geological fate, African countries are now known to hold
major oil reserves. Several also monopolise global supplies of several stra-
tegic minerals such as cobalt and tantalum, on which modern industrial
powers have come to depend. The Democratic Republic of Congo (DRC),
for example, has 60 per cent of global cobalt reserves (the rest are in
Canada and Cuba), and 80 per cent of global tantalum. The heat and
electrical properties of tantalum make it a crucial ingredient in capacitors
for third-generation mobile phones and other advanced electronics, as
well as for jet engines, night vision goggles and fibre optics. Cobalt is a
critical ingredient in the superalloys used in air- and land-based turbine
engines as well as for rechargeable batteries for cell phones and comput-
ers. Moreover, environmental pressures for zero-emission vehicles will
massively increase demand for this ‘oil of the future’.
As with oil, industrial nations need to secure their supplies, and corpora-
tions controlling supplies can profit hugely, so their governments and cor-
porations inevitably work together. African natural resources have long
been of strategic significance (Hveem, 1986). This continues. Over the past
decade, international security officials have been paying even greater atten-
tion to intensified competition over strategic materials, and this has regained
centrality in US security planning. Indeed, Klare, writing in Foreign Affairs,
goes further, arguing that there is now a ‘new geography of conflict, a
reconfigured cartography in which resource flows rather than political and
ideological divisions constitute the major fault lines’ (2001: 52).
Strategic resources have now acquired further importance following
the attacks on the USA on 11 September 2001. First, Africa’s ‘non-Islamic
oil’ has become central to the geopolitics of energy. Second, as former
Assistant Secretary of State Susan Rice argued to the US House of Repre-
sentatives’ deliberation on Africa and the war on global terrorism,

We must recognise that regimes lacking legitimacy and failed states

are convenient safe havens as well as breeding grounds for terrorists.
If we are serious about our anti-terrorism commitment . . . the US
 Achieving sustainability in Africa 297
must become more rather than less engaged in the difficult task of
peacemaking, peacekeeping and national reconstruction – from the
Great Lakes to Sierra Leone, from Liberia to Sudan and Somalia. We
must also find effective ways to secure Africa’s vast natural resources –
its diamonds, cobalt, uranium, oil, timber, coltan, its gold – so they do
not provide currency for the world’s terrorists.
(Rice, 2001: 12)

New demand for African resources in this second scramble for Africa has
not been helping most Africans. For example, Oxfam (2001) reported that
poverty and health problems are worse in developing countries that are
dependent on oil and mining. They are associated with civil war, military
expenditure and with governments that, being financially independent of
their populations, cease to become accountable to them (see also Moore,
2000). Many researchers have reiterated the centrality of environmental
resources, and more particularly the political-economic conditions of their
extraction, to poor governance and conflict (e.g. Auty, 1998; Collier and
Hoeffler, 2000, 2001; Nafziger et al., 2000; Reno, 1999). This association has
come to be dubbed the ‘resource curse’ (World Bank, 2002). Geopolitical
and commercial rivalry play into national and regional political rivalry, and
patterns of social differentiation, so it would be foolish to reduce causality in
contemporary conflicts to this (e.g. Mamdani, 1996; Mkandawire, 2002).
Nevertheless, it would be equally foolish to ignore it.
The 2003 World Development Report pursues this analysis. Speaking of the
resource curse, it says:

The adverse effects are magnified when the natural resource endow-
ment, like petro-chemicals and minerals, are ‘lootable’ or ‘point
sourced’, meaning that their production and revenue patterns are
concentrated [as opposed to ‘diffuse’]. Coffee and cocoa take on
point source characteristics when shipped.
(World Bank, 2002: 149)

Timber, and even ‘conservation’, can take on such characteristics too

when control over resources is granted through state timber concessions
or conservation concessions. Of the 45 countries whose growth was not
sustained, ‘all but 6 are point-source economies. The majority of the coun-
tries with point-source natural resources also suffered violent conflict in
the 1990s’ (ibid.: 149). In cases where this ‘curse’ has fuelled conflict, the
ensuing international response has lobbied for trade embargoes linked to
the vocabulary of conflict diamonds, conflict oil, conflict timber, conflict
tantalum, and so on. There have been some successes. Conflicts, however,
are the more extreme and high-profile manifestations of a corporate and
political blight that currently undermines governance and resource-use
sustainability in many other countries.
298 James Fairhead
The most documented case of corporate culpability in transforming
governance and fuelling conflict is to be found in the DRC. Here, North
American, European and Asian corporations have struck deals with govern-
ments and rebels; rebel movements themselves have become self-financing
through their mining and resource extraction; and neighbouring countries
(especially Rwanda and Uganda), allied with rebel movements, have
enriched themselves through illegal looting. This has been well docu-
mented both by the United Nations (UN) and by assorted non-govern-
mental organisations (NGOs) and academics (e.g. UN, 2001).1
Assessments suggest that a staggering 2.5 million people have died of this
curse in the DRC alone since 1998 (All Party Parliamentary Group, 2001;
IRC, 2001).
At the same time, the world’s leading mining corporations have sup-
ported protagonists in the conflict and profited both from their interests
there and indeed, from the increased value of their reserves elsewhere
when conflict disrupts Congolese production.
Many of the corporations operating in the DRC and its neighbours
have strong political connections. These are hard to research in detail, but
some indication of the links can be gauged from the ‘revolving door’ that
sees senior figures in government moving into and out of these corpora-
tions. In recent years, for example, Barrick Gold, which has large interests
in DRC, has counted three former CIA directors and two former North
American presidents in its management (George Bush Snr, Richard
Helms, Robert Gates and Brian Mulroney). The Bechtel corporation,
which also worked with Congolese rebels, has counted former US Secre-
tary of State George Schultz, alongside Philip Habib and Casper Wein-
berger, as well as CIA directors William Casey and Richard Helms working
for it. The oil company Halliburton, which – together with its subsidiary
Brown and Root – has major oil and other interests in the region, had
current US Vice President Richard Cheney as its Chief Executive Officer
(CEO) until his nomination. Another oil company, Chevron, which is now
involved in a US$2bn investment in the DRC, until recently counted Con-
doleezza Rice as a member of the board. Prior to that, she was Special
Assistant to George Bush. Now she has become US National Security
Advisor. Corporate expansion in conflict zones has coincided with an
explosion in private security firms. The largest private security transnation-
als not only work closely with the US government, but also are owned by
former members of government and the military. The Vinnell Corpora-
tion (based in Fairfax, Virginia), for example, is owned by BDM Inter-
national Inc. and controlled by the Carlyle Group, which counts former
CIA deputy director Frank Carlucci and US Secretary of State James Baker
among the shareholders. Carlucci was Secretary of Defense under Ronald
Reagan and George Bush and, it might be noted, was also former second
secretary at the US embassy in the DRC. The nature of corporate and
political links can perhaps be gauged well in the work of Cohen and
 Achieving sustainability in Africa 299
Woods, a Washington-based lobbying firm. This firm has worked both for
mining and security firms seeking contracts and concessions in Africa
(such as MPRI in Angola), and for African leaders seeking US political
support (such as the late Laurent Kabila). Cohen and Woods is managed
by Herman Cohen (former Assistant Secretary of State for Foreign Affairs)
and Jim Woods (former Assistant Secretary of State of US Defense).
Liberia provides a contrasting case in which transnational corporations
and conservation organisations appear to operate more independently of
the political–military–industrial complex. Thus at a time when there has
been a UN arms and trade embargo on Liberia, and at a time when the
US has been hostile to Liberia, there have have been large external
corporate investments in the different resources of the country. United
States tele-evangelist Pat Robertson, for example, and his ‘Freedom Gold’
brokered a deal with the president, Charles Taylor, to prospect in south-
eastern Liberia. A friend of the current administration, he has been lobby-
ing for a shift in US policy towards Taylor. Timber concessions have been
sold to the ‘Oriental Timber Company’ and others. Conservation Inter-
national signed a deal to establish and manage protected area networks
covering 1.5 million hectares with a view to establishing ‘conservation con-
cessions’, inaugurating, perhaps, an era of ‘conflict conservation’.2 Insta-
bility is not necessarily fuelled simply by buying arms and paying soldiers,
although this is asserted too, and is a powerful image in international lob-
bying (Global Witness, 2001). It is also fanned by the undermining of citi-
zens’ rights, government accountability and the rule of law. Thus the
logging operation, for example, drew public criticism concerning the des-
tination of payments, the granting of work permits to foreign workers, dis-
respect for indigenous rights and resources, non-compensation for damage
to property, the exclusion of national regulatory authorities, and the viola-
tion of laws and regulations.
After reviewing the role of Western corporations in the unfolding of
the current Congo holocaust, Montague argues that ‘In order for invest-
ment to be used as an effective tool for development, multinationals must
understand that massive corporate investments have a negative impact on
society in the absence of state stability. Indeed, their financial leverage
only exacerbates state instability’ (2002: 115).
Environmental resources are fundamental to poor and indebted modern
African economies. As I have argued, we see less patterns of governance able
to influence management of environmental resources, than the inverse: the
transnational dimensions to the management of environmental resources
influencing patterns of governance. It is this that modern policy must
address. Viewed in this way, it does not become sensible to differentiate
between ‘renewable’ and ‘non-renewable’ resources.
Indeed, making a strong distinction between renewables and non-
renewables has long enabled the transnational dimensions to resource use
and sustainability to be downplayed. The focus on renewables, and on
300 James Fairhead
their scarcity, leads analysts to explore the causes of conflict in a localised
way linked to local use of resources. It is ‘population increase’ on location
that is considered to increase demand for renewable resources and their
value, or poverty or social transformation on location that undermines
long-term sustainable management. The way international and global eco-
nomic forces affect governance and resource governance, and thus local
environments and resource use, is eclipsed. We lose sight of the way that
timber and agribusiness (including cocaine and opium), and the conser-
vation business (including land, rights and resource alienation), affect
governance, conflicts and, through this, environmental use and its sustain-
ability. The distinction between renewables and non-renewables has long
been ambiguous (encapsulated in analysis of soil or forest ‘mining’, for
example), and from the vantage point of their commercialisation and
their relations of commercialisation, their distinction becomes irrelevant.
Ignoring the distinction enables us to speak about the way transnational
interests in African environments – their diamonds, oil, tantalum, cobalt,
timber, cocaine, opium – and biodiversity play into located political
economies. It enables other, locally relevant distinctions to get noticed.
Modes of transnational engagement shape, for example, how people in
different places understand their ‘environment’ and ‘environmental quality’.
For example, it influences whether or not people consider deforestation
to be ‘degradation’, and how views differ. In several locations in West
Africa, the return of lands to high forest is considered by many farmers to
be degradation. The vast majority of Sierra Leonian, Liberian, Ivorian and
Ghanaian forests, after all, have grown over once prosperous places,
where cycles of bush fallow have been disrupted following depopulation
during the slavery era, and reservation during the colonial one (Fairhead
and Leach, 1998). Much of today’s Liberian forest did not exist twenty
years ago, having grown over war zones. Is the subsequent loss of this
forest again ‘degradation’? For whom? How does the alienation of rights
over heavy timber to state authorities influence these perspectives? How
does the alienation of land to reserves, protected areas or conservation
concessions influence this? How does the alienation of diamond resources
in these areas influence patterns of governance and modes of resource
use? How does this alter local engagement with the state, timber com-
panies and conservation organisations?
Whatever the failings of the National Intelligence Council vision for
2015, it does provide a counterpoint with which to escape the utopian and
parochial language common to development strategising around sustain-
ability. Yet predictably, despite this contrast, even the NIC, like the Millen-
nium Development Goal and associated strategies, plays down the current
problematic international engagement of transnational corporations with
African globalisation. International engagement in 2015, the NIC sug-
gests, will increasingly be through ‘international organizations and non-
state actors of all types: transnational religious institutions, international
 Achieving sustainability in Africa 301
nonprofit organizations, international crime syndicates and drug traf-
fickers; foreign mercenaries; and international terrorists seeking safe
havens’ (NIC, 2000: 73). The operation of transnational corporations is
omitted.
Deniability concerning the negative impacts of corporate involvement
in environmental resources has been maintained in many ways. First, it is
perfectly clear that many of the financial transfers, political support and
indeed military interventions are covert in nature, and firm analysis of
them does not emerge. Second, accountability is clouded by the complex-
ity of modern corporate structures. Third, major corporations have
learned to speak the language of corporate social and environmental
responsibility. Belief in the latter will certainly have been shaken by recent
corporate and accounting scandals. It remains to be seen whether corpor-
ate scandals will reach the African dealings of major corporations. Fourth,
blame is easily transferred to the African partners.
In short, approaching issues of sustainable development simply
through national programmes appears enormously naive. It is in recogni-
tion of this corporate co-responsibility for failures in African governance
that the philanthropist financier George Soros has recently linked with
‘Global Witness’ and dozens of other international organisations to
lobby for enforced transparency in international corporate financial
transfers. Their ‘Publish what you pay’ initiative would require trans-
national resource extraction companies to ‘publish net taxes, fees, royal-
ties and other payments as a condition for being listed on international
stock exchanges and financial markets’.3 This would facilitate both
corporate and government accountability. This approach to sustainabil-
ity and poverty alleviation is the inverse of that envisaged in the Millen-
nium Development Goal. Action is not to be through national pro-
grammes, but through calling on the G-7 nations – the world’s wealthiest
– ‘to take leadership and promote transparency over resource revenues
worldwide’. It was this (albeit in voluntary rather than regulated guise)
that the UK’s prime minister, Tony Blair, supported in speeches made
around the Johannesburg Summit on Sustainable Development.
However, this approach runs up against both the military–industrial
complex of contemporary geopolitics as much as the nether world of
independent transational entrepreneurs and national ‘rent-seekers’.
Certainly there would be immense problems in implementation, but it is
nevertheless the only way forward. The only caveat that I would add is
that it should apply as much to financial transfers for those seeking to
control resources as for conservation and transfers relating to extrac-
tion.
302 James Fairhead
Concerning ‘global and local causes’
A second arena that is marginal to policy documentation on sustainability,
poverty and the environment is the impact of Northern industrialisation
on African climate and environment. International Development Target
strategies make much of the ‘many opportunities to meet local environ-
mental priorities while also contributing to global concerns, such as the
build up of greenhouse gas emissions in the atmosphere’ (DFID, 2000a:
8). Missing, however, is an appreciation of the way global climate changes,
with roots in Northern industrialisation, impact on African localities.
Whilst it has been easy to dismiss as fanciful, rhetoric arguments that
African droughts and famines result from climate changes linked to
Northern industrialisation, evidence is now building that the rhetoric lies
with the complacent dismissal. Evidence is accumulating rapidly that the
wide-scale impact of climate changes on African vegetation and agro-
ecology threatens to override many local conservation and development
initiatives. Future historians may find such initiatives, whether rooted in
participation or exclusion, to have been ‘fiddling while Africa burns’. This
argument is premised on alarming evidence that has emerged during the
past fifteen years that African climates have experienced major fluctua-
tions (at times catastrophic deteriorations) in historical times. African
forests appear to be particularly susceptible to global climate fluctuations
of the sort associated with global warming.
The sensitivity of African climate and vegetation to global change is
indicated by evidence from forest and climate history. For example,
archaeological, climate historical and oral evidence combine to suggest
that forests in West and Central Africa have responded massively to major
climate fluctuations during recent centuries and millennia. In a review,
Maley (2002) has concluded that 2,500 years ago a climatic deterioration
lasting several centuries led to a catastrophic destruction of central-West
African forests, almost halving the current forest range. In recent cen-
turies, forest vegetation has been recovering from this period, and from
other less extreme periods of deterioration around 800–1,200 years ago.
In Cameroon, more than 1 million hectares of savanna has become forest
in this region alone since 1952, whether in the presence or absence of cul-
tivation (Maley, 1999). For equatorial Africa, Vincens et al. (1999) suggest
that the earlier dry phase appears to have lasted for a very long time, with
the start of the current humid phase and recolonisation of forest dating
back only to 600–900 years ago.
Preliminary analysis of work conducted by Overpeck and colleagues
based on sediment cores and tree remains from Ghana’s Lake Bosumtwi
indicates that ‘Ghana was gripped by a “megadrought” for much of the
little Ice Age’ (Overpeck pers. comm.). As in Central Africa, West African
forests appear to have recovered from this period of climatic deteriora-
tion, colonising savannas in recent centuries. The quality of the forest is
 Achieving sustainability in Africa 303
thus not as ‘pristine’ or ‘biodiverse’ as one might expect. As Hawthorne
has noted on the basis of tree species distribution, the vast majority of
Ghana’s forest is effectively a ‘scar tissue, a recently assembled group of
mainly widespread, well-dispersed species, covering up after some immense
disruption of this area and barely infiltrated by rarer species which could
occur there’ (1996: 138). My own research in Guinea, and more widely in
West Africa, indicates that much of what is today classified as the ‘forest
region’ has recently been savanna land (Fairhead and Leach, 1998).
It is thus beginning to appear that African forests have been very sensi-
tive to global climate changes in historical times. The alarming problem is
that climatic conditions that led to these African droughts – at least which
led to the catastrophic destruction of the Congo forest – appear to be the
very conditions predicted for the region by modern climate models linked
to human-induced global warming. As Maley (2002) makes clear, if current
predictions concerning anthropogenic global warming are accurate, this
is likely to have a catastrophic effect on the region’s forests. If, as is the
case, the key indicator of whether the Millennium Development Goals are
met is the ‘proportion of land covered by forest’, and ‘land area protected
to maintain biological diversity’, these overriding global dimensions to sus-
tainability need to be addressed – indeed, need to have been addressed.
Other recent studies indicate that industrial emissions from North
America and Europe may have caused the recent severe droughts that
have afflicted the Sahel region of Africa. Industry and power-generation
emissions have been creating aerosols that affect cloud formation, altering
the temperature of the Earth’s surface. In climate modelling, temperature
change causes the Earth’s surface in the north to cool relative to the
south, appearing to drive the tropical rain belt southwards and causing
droughts in the Sahel.
As we gain more knowledge about global climate relations and history,
we shall be made increasingly aware of these underlying dimensions to
sustainability. Future phases of desiccation associated with forest fires and
future droughts might well be attributable more to greenhouse-gas and
other emissions from industrialised countries than to land users in West
Africa. It will appear, however, from attention to proximate causes that
‘Africans are responsible’. This should not, however, deflect attention to
underlying causality and responsibility. Attention to the impact of global
climate change on African environments, and especially its threats on the
Congo forest, puts in its place the fanfare that accompanied the recent
announcement at Johannesburg by an interagency partnership to ‘save
the Congo forest’ through located interventions.
The other lesson from African climate history is that it undermines the
idea of stability and predictability that has infused many interpretations of
‘sustainability’. This poses some problem for the expansion of ‘results’- or
‘target’-based planning of international development and the Millennium
Development Goals into the field of environmental sustainability, when
304 James Fairhead
the results of social interventions are measured simply in relation to
environmental qualities. The focus should be less on getting institutions
right, as if there were a single solution, than on questions of power, polit-
ical economy and the processes shaping the evolution of institutions of
resource governance. Equally, the focus should be less on ‘stabilising the
forest’ than on flexible adaptation in response to non-equilibrial ecolo-
gical and economic environments (Leach et al., 2002).

Conclusions
The aim of this chapter has been to indicate a central deficiency in the
Millennium Development Goal concerning sustainability. The ‘country
programmes and policies’ which are its linchpin may well be necessary,
but they will be by no means sufficient. Clearly there is a huge importance
to national policies which I have not explored in this chapter. Yet national
policies themselves must be contextualised within international strategies
that respond realistically to the way the global political economy currently
shapes resource use and its governance, and to strategies that address
global climate change. Otherwise it will become increasingly apparent that
the focus on national strategies is a comfortable ideological device pro-
mulgated by OECD countries and donors which has the effect of transferring
responsibility for current patterns of unsustainability to African countries.
Several international organisations, including the OECD, have already
developed linked initiatives to address this. The OECD countries have, for
example, ratified a ‘Convention on combating bribery of foreign public
officials in international business transactions’. This entered into force in
1999 but awaits effective implementation (OECD, 2002). Many donors
have also allied themselves in anti-corruption initiatives, such as the Euro-
pean Union anti-corruption resource centre and the World Bank anti-
corruption initiative, having ‘identified corruption as the single greatest
obstacle to economic and social development’.4 These initiatives could
have provided the foundation for a Millennium Development Goal and
more concerted action on corruption. Yet this has not happened, and this
chapter has raised issues which help to explain why. The ‘publish what
you pay’ initiative is the single most important initiative to have emerged
in the field of development and sustainability in the past fifty years.
Attempts to craft governance for sustainability without it may be doomed
to failure.
Debates concerning sustainability at the Johannesburg conference also
went well beyond the narrow remit of the Millennium Development Goal
– and indeed beyond its narrow indicators. This was exemplified in the
speech of the British prime minister, Tony Blair, highlighting the central
issues of transparency of financial transfers and global climate change.
As it is, the prospect, at present, is that Africa’s oil (and other environ-
mental resources) will undermine sustainability twice over: their extraction
 Achieving sustainability in Africa 305
leading to problems in governance and accountability so fundamental to
any policy of sustainability, and their use leading to major problems of
global climate change.

Notes
1 See also the addenda to this UN report, and NGO reports posted at [Link]
[Link]/security/issues/[Link].
2 See the official agreement, [Link]
3 [Link]
4 See [Link] and [Link] respec-
tively.

References
All Party Parliamentary Group (2001) All Party Parliamentary Group on the Great
Lakes and Genocide Prevention, at [Link]/downloads/
report_2001.pdf.
Auty, R.M. (1998) Resource Abundance and Economic Development: Improving the
Performance of Resource-Rich Countries, Research for Action 44, UNU/WIDER,
Studies in Development Economics, New York: Oxford University Press.
Collier, P. and Hoeffler, A. (2000) ‘On the Incidence of Civil War in Africa’,
Journal of Conflict Resolution, 46(1): 13–28.
—— (2001) Greed and Grievance in Civil War, World Bank.
Department for International Development (DFID) (1997) Eliminating World
Poverty, a Challenge for the 21st Century, London: DFID.
—— (2000a) Achieving Sustainability: Poverty Elimination and the Environment,
London: DFID.
—— (2000b) Eliminating World Poverty: Making Globalisation Work for the Poor, White
Paper on International Development, London: DFID.
Economist, The (2000) ‘The Hopeless Continent and The Heart of the Matter’, 13
May.
Fairhead, J. and Leach, M. (1998) Reframing Deforestation, London: Routledge.
Global Witness (2001) Taylor-Made: The Pivotal Role of Liberia’s Forests in Regional
Conflict, 2nd edition report, September, at [Link]
campaigns/forests/liberia/[Link].
Hawthorne, W. (1996) ‘Holes and the Sums of Parts in Ghanaian Forest: Regener-
ation, Scale and Sustainable Use’, Proceedings of the Royal Society of Edinburgh,
104B: 75–176.
Hveem, H. (1986) ‘Minerals as a Factor in Strategic Policy and Action’, in Westing,
A. (ed.) Global Resources and International Conflict: Environmental Factors in Strategic
Policy and Action, Oxford: Oxford University Press, pp. 55–84.
IRC (2001) Mortality Study Eastern D.R. Congo, at [Link]
[Link]?fa=show&topicID=86&wwwID=440.
Klare, M. (2001) ‘The New Geography of Conflict’, Foreign Affairs, June 2001.
Leach, M., Fairhead, J. and Amanor, K. (eds) (2002) ‘Science and the Policy
Process: Perspectives from the Forest’, IDS Bulletin, 33(1).
Maley, J. (1999) ‘L’Expansion du palmier à huile (Elaeis guineensis) en Afrique
306 James Fairhead
Centrale au cours des trois dernier millénaires: nouvelles données et interpréta-
tions’, in Bahuchet, S., Bley, D. and Pagezy, H. (eds) L’Homme et la forêt tropicale,
Paris: Bergier, pp. 237–254.
—— (2002) ‘A Catastrophic Destruction of African Forests about 2500 Years Ago
Still Exerts a Major Influence on Present Vegetation Formations’, IDS Bulletin,
33(1): 1–12.
Mamdani, M. (2001) When Victims Become Killers: Colonialism, Nativism, and the Geno-
cide in Rwanda, Princeton, NJ: Princeton University Press.
Mkandawire, T. (2002) ‘Post-colonial Rebel Movements in Africa’, Journal of
Modern African Studies, 40(2), 181–216.
Montague, D. (2002) ‘Stolen Goods: Coltan and Conflict in the Democratic
Republic of Congo’, SAIS Review, 22(1) (Winter/Spring): 103–118.
Moore, M. (2000) ‘Political Underdevelopment’, paper presented at the Tenth
Anniversary Conference of the Development Studies Institute, London School
of Economics, London, 7–8 September.
Nafziger, W., Stewart, F. and Vayrynen, R. (eds) (2000) The Economic Causes of
Complex Humanitarian Emergencies, Oxford: Oxford University Press.
National Intelligence Council (NIC) (2000) Global Trends 2015: A Dialogue about the
Future with Nongovernmental Experts, NIC 2000, at [Link]
2015_files/2015htm.
Organisation for Economic Cooperation and Development (OECD) (2002) Report
by the Committee on International Investment and Multinational Enterprises:
Implementation of the Convention on Combating Bribery of Foreign Public Officials in
International Business Transactions and the 1997 Recommendation, OECD 2002 (date
declassified), Paris: Directorate for financial, fiscal and enterprise affairs, at
[Link]
Oxfam (2001) Extractive Sectors and the Poor, Boston: Oxfam America, at
[Link]
Reno, W. (1999) Warlord Politics and African States, Boulder, CO: Lynne Rienner.
Rice, S. (2001) Submission to ‘Africa and the war on global terrorism’, Hearing
before US House of Representatives, Committee on Social Relations (Subcom-
mittee on Africa), 107th Congress, 1st Session (no. 107–46), 15 November.
UN (2001) UN Report of the Panel of Experts on the Illegal Exploitation of Natural
Resources and Other Forms of Wealth of the Democratic Republic of the Congo, 12 April,
at [Link]
Vincens, A., Elenga, H., Reynaud-Farrera, I. et al. (1999) ‘Forest Response to
Climate Changes in Atlantic Equatorial Africa During the last 4000 Years BP and
Inheritance on the Modern Landscapes’, Journal of Biogeography, 26: 879–885.
World Bank (2002) World Development Report 2003, Washington, DC: World Bank.
15 Building a global partnership for
development?
Peter Clarke

Introduction
The history of development programmes in recent decades has been char-
acterised by steadily increasing use of conditioned aid to influence recipi-
ent government policy. Concerns about project and programme failure
have served to justify conditionality in ever-wider policy areas. At the same
time, concern has been growing that the sustainability of policy reform
depends on the degree of local political will behind reform, and that con-
ditionality tends to have a negative impact on so-called ‘ownership’. This
has led to widespread demands for a change in North–South aid relation-
ships away from coercive conditionality and towards more equal ‘partner-
ship’ to such an extent that the policy statements of most major multilateral
and bilateral donors now centre around this concept. The Millennium
Development Goal of building a ‘global partnership for development’ rep-
resents one example of this increased emphasis on partnership, with
targets encompassing a more open, rule-based, predictable and non-dis-
criminatory trading and financial system, a commitment to good gover-
nance, increased levels of development assistance, measures that address
the problems of debt, provision of access to essential drugs and new tech-
nologies, and specific attention to the needs of young people, in the least
developed countries as well as landlocked and small island states.
The rhetoric of partnership includes a shift away from the traditional
government-to-government power structure in the aid relationship, expand-
ing policy and aid management processes to encompass the private sector
and civil society in both North and South. That there are tensions and
ambiguities about the interests served by such partnership is evident from
questions surrounding the legitimacy of non-governmental organisations
(NGOs) as representatives of ‘the poor’ and, for some, by the mere fact
that the private sector is also seen as a partner striving for the common
goal of ‘development’. However, the discussion here focuses on the prob-
lems surrounding relationships between donor and recipient govern-
ments. Does the discourse of ‘partnership’ indicate a fundamental move
to more equal relationships between North and South, or instead mask
308 Peter Clarke
ever wider and deeper infiltration of development power relations in the
South? Moreover, what forms of resistance to these interventions can be
detected, and to what extent can development agencies or individual
actors challenge existing power relations?
In this chapter, I address these questions through an examination of
the partnership discourse of the UK government’s Department for Inter-
national Development (DFID). The chapter first seeks to locate DFID
policy in the broader context of the partnership discourse in develop-
ment, before examining the discourse of partnership in DFID policy docu-
ments from 1997 to 2001. A key critique of the discourse of partnership is
that it represents a ‘political technology’ (Foucault, 1978: 86; Shore and
Wright, 1997: 4) that functions to break down boundaries limiting the
global penetration of power, rather than an attainable relationship between
North and South. Yet there remains scope for Southern actors to contest
and work within this discourse, through complex negotiations over the
course and meaning of development at a local level.

Partnership discourses in development

The use of the term ‘partnership’ to prescribe the appropriate relation-
ship between international aid donors and recipients has spread remark-
ably in the past few years and achieved a central place in the policy
documents of all major multilateral and bilateral donors, as well as in the
Millennium Development Goals themselves (German and Randel, 2000).
The notion of a partnership for development dates back at least to the
Pearson Commission, set up in 1968 by the World Bank in the context of
concern about the limited impact of ‘twenty years of development assis-
tance’ (Pearson, 1969: vii), and its report was entitled Partners in Develop-
ment. The objective of development assistance1 at that time was defined in
a limited and straightforward way as ‘self-sustaining economic growth’
(ibid.: 130). Partnership was represented as a new kind of development
relationship that would avoid ‘friction, waste of energy, and mutual irrita-
tion’, while maintaining ‘clear and accepted channels’ for ‘advice, consul-
tation and persuasion’ (ibid.: 127). The precise objective of development
made it possible to satisfy the ‘natural’ interest of aid providers in the use
of their resources, while avoiding donor monitoring of the whole of social
policy in developing countries, by defining two indicators of progress:
‘adequate and sustained increases in the ratio of domestic savings to
national income and in the ratio of exports to imports’ (ibid.: 132),
simultaneously taking appropriate account of factors outside the country’s
control. Since partnership implies mutuality, relations between donors
and recipients were to be ‘based on an informal understanding expressing
the reciprocal rights and obligations’ (ibid.: 127). From the beginning,
partnership relationships were to be supported by improved donor co-
ordination, and the ‘donor community’ would accept its responsibility to
 Building a global partnership for development? 309
guarantee predictable and long-term performance-related aid flows. But
also from the start, doubts were expressed in the South about the credibil-
ity of the proposals:

Unfortunately, the concept of a genuine partnership in development

somehow lacks credibility. There has never been any real sense of
equality between donors and recipients even when they attend the
same consortium meetings and sit around the same table in many
other forums. . . . The donors have parliaments and public opinion
which reign so supreme that a mere reference to them should silence
all criticism, whereas the recipients should obviously be able to manip-
ulate at will their parliaments and public opinion in the interest of
appropriate development policies. A mere equality of opportunity in
engaging in dialogue cannot establish parity in decision-making. Nor
can the platonic world of knowledge as a sufficient basis for right
conduct be easily summoned into existence.
(Patel, 1971: 305)

A significant early application of the partnership concept was the Lomé

Convention, signed in 1975 between the European Economic Community
and the African, Caribbean and Pacific (ACP) countries. The approach
differed from the Pearson proposals by replacing the informal agreement
between partners with a formal contract. Maxwell and Riddell (1998: 260)
argue that this formal definition of mutual commitments and procedures
for redress is essential for symmetrical accountability in partnership, but
consider that the initial European Union (EU) model was too inflexible
and failed to permit response to political and policy changes.
The partnership discourse began to dominate mainstream policy in the
1990s following discussions in the Development Assistance Committee
(DAC) of the Organisation for Economic Cooperation and Development
(OECD). The Committee’s 1995 statement Development Partnerships in the
New Global Context was followed by the more detailed 1996 report which set
out the International Development Targets. Comparison of these reports
with the partnership proposals of the Pearson Commission report shows
important similarities, but also substantial differences. There is continuing
emphasis on donor co-ordination and the clear definition of mutual
responsibilities (OECD, 1996: 2). On the other hand, the OECD gives
increased attention to ‘coherence between aid policies and other policies
which impact on developing countries’. Whereas the earlier report defines
a clear and narrow aim for development assistance and closely circum-
scribes the recipient policy areas for donor concern, the 1996 document
explicitly refers to a change of conception to encompass a ‘much broader
range of aims’ (ibid.: 13). Thus in addition to proposing the adoption of
the International Development Targets, there is a simultaneous emphasis
on ‘locally owned development strategies’.
310 Peter Clarke
Since 1996 the discourse of partnership has achieved a place at the
centre of development policy, and in the presentation of a wide range of
new policy instruments (German and Randel, 2000: 19). In the UK, a new
Labour government was elected in 1997, and introduced a new develop-
ment policy centred around ‘partnership’. The newly established Depart-
ment for International Development published the first White Paper on
development since 1975, and this, alongside documentation of the new
policy in other official publications and the writings of leading officials,
offers an excellent opportunity to analyse the discourse of partnership as
it is conceived by a major aid donor.

Partnership in DFID policy discourse

The first DFID White Paper dedicated twenty-eight of its eighty pages to a
section entitled ‘Building partnerships’ (DFID, 1997: 22–49), and almost
every imaginable institutional divide is to be crossed by this type of relation-
ship: public–private, civil society–state, North–South, and so on. However,
the argument here concentrates on relationships between DFID as a
Northern government donor agency, and Southern recipients, which are
mainly governments.
How is the new type of relationship justified, and how is it char-
acterised? Partnerships are represented as an alternative to ‘old condition-
alities’ (DFID, 1997: 37) and are described as ‘putting developing
countries in the lead, devising and taking forward their own develop-
ment strategies’ (DFID, 2000a: 91). The justification for a less coercive
relation is fundamentally instrumental (ibid.: 92), as ‘over-prescriptive
aid conditionality has a poor track record in persuading governments
to reform their policies’. A subsequent account goes on to state that
‘recipient governments have learnt to play the game. They may have
no real commitment to the conditions which they sign up to’ (DFID,
2001: 28). According to the then Chief Economist of DFID, Andrew
Goudie,

Most fundamental to the partnership approach is the perception that,

without the full political support of developing country partner gov-
ernments, any efforts to eliminate poverty in those countries will be
seriously constrained. . . . Strategies and policies to eliminate poverty
need fundamentally to be designed and initiated by the government
itself. . . . Partnership seeks to build on existing political will within
partner countries and provide support to expand and generate a
broader constituency.
(Goudie, 1998a: 171)

This political will is considered necessary for sustainable policy reform and
is argued to be dependent on creating policy ‘ownership’ (e.g. DFID,
 Building a global partnership for development? 311
2001: 10) in the Southern government. This concept is especially prob-
lematic and will be further discussed below.
Since conditionality has proved incapable of achieving policy owner-
ship, this is instead to be achieved primarily through partner selection.
DFID’s partnership policy is based on more focused aid to priority coun-
tries, selected on the basis of their need and their commitment to poverty
elimination, as expressed in ‘sensible policies’ (DFID, 1997: 38). This
selectivity is not limited to defining a threshold for aid, but is also applied
to determine the depth and duration of partnerships. Goudie argues that
‘partnership . . . necessarily carries with it the implications of a degree of
selectivity in the manner in which we work with partner governments’
(1998a: 171). The problems of applying selectivity in practice have been
discussed by White (2001: 1063).
The partnership approach is associated in DFID policy with a broad
range of policy changes. In the following summary, the most characteristic
changes are isolated for the purposes of discussion, although there are
important links between them.

Comprehensive interventions
The partnership approach is associated with ‘reducing support for stand-
alone projects’ (DFID, 2000a: 93), which are characterised as ‘isolated and
poorly integrated actions’ (Goudie, 1998a: 170). Instead, development
interventions are to become increasingly comprehensive and long term,
with ‘increasing support for sector-wide reforms’ (DFID, 2000a: 93), or
even ‘the economy as a whole’ (DFID, 1997: 38).

Donor co-ordination
Coherence is also to be served by donor co-ordination. According to the
DFID 2000 White Paper, ‘we have worked hard to promote greater har-
monisation among development agencies’ (2000a: 93). The reduction of
the administrative burden on developing country governments is argued to
justify ‘working with the international community to strengthen joint
working including providing resources through common funding mechan-
isms’ (DFID, 2000a: 44). Goudie suggests additional benefits in terms of
more coherent policy influence:

[We] should look for a consistency between our bilateral response

and the multilateral response – it clearly makes little sense for our
programme and, for example, a Bank/Fund ESAF [enhanced struc-
tural adjustment facility]/structural adjustment programme to be
moving along different tracks and sending conflicting and confusing
signals to government about our multi-sided partnerships.
(Goudie, 1998b: 9)
312 Peter Clarke
Comprehensive policy concern
The comprehensive nature of donor policy interventions is matched by
similarly comprehensive policy concern, as shown in the selection criteria
for prospective partners, spelt out in the 1997 White Paper:

We would expect partner governments to: have a commitment to the

principles of the agreed international development targets and be
pursuing policies designed to achieve these and other UN targets
which they have agreed; be committed to pro-poor economic growth
and conservation of the environment, and be pursuing appropriate
policies; wish to engage with us and with the donor community to this
end; pursue policies which promote responsive and accountable
government, recognising that governments have obligations to all
their people; promote the enjoyment of civil, cultural, economic,
political and social rights; and which encourage transparency and
bear down on corruption in the conduct of both the public service
and the business sector.
(DFID, 1997: 39)

In this way, relationships move away from conditionality focused on

‘certain very specific measures and actions’ (Goudie, 1998a: 173), to part-
nerships in which ‘we have both a necessary and legitimate interest in a
potentially vast range of partner country affairs’ (Goudie, 1998b: 3). This
is consistent with a substantial broadening in the development agenda
since the Pearson Commission report – it is argued to be no longer pos-
sible to limit intervention to a single objective, nor monitoring to a few
indicators.

Means of legitimising policy prescriptions

The broadening development agenda creates an equally growing chal-
lenge to justify the legitimacy of a vast range of policy prescriptions. As the
above quotation suggests, DFID policy documents consistently construct
this array as a taken-for-granted international policy consensus. On a few
occasions the attribution is more explicit – for example, when referring to
‘the almost universal consensus that has developed since the end of the
Cold War in favour of democracy and on creating an enabling environ-
ment for a liberalised economy’ (DFID, 2001: 9). Descriptions of the
policy consensus present two kinds of problems. Some reduce a series of
highly contestable and complex issues to repeated keywords (Shore and
Wright, 1997: 18) such as ‘democracy’, ‘participation’ and ‘sustainability’.
Others refer to highly specific and fashionable solutions through expres-
sions such as ‘effective and efficient public sector management’. In both
cases the specifications are subject to widely different interpretations. The
 Building a global partnership for development? 313
assumption of consensus also ignores the experience that even among
donors, policy prescriptions have shown a tendency for drastic change
every decade.

Means of legitimising intervention

Closely related to the issue of legitimising policy prescriptions is the need
to legitimise external intervention in recipient government policy, espe-
cially at the comprehensive level proposed. This is constructed in various
ways – first, with reference to the responsibility of DFID to taxpayers and
Parliament, and the justification given to them for international aid.2 As
Goudie explains,

The Secretary of State for DFID is . . . bound to account to Parliament

in a formal sense . . . for the effectiveness and efficiency of the manner
in which she deploys her scarce resources . . . to address legitimate
concerns of her own UK constituencies. . . . We should, therefore, be
under no illusions that the developmental concerns of the UK con-
stituency play a key role in driving departmental concerns and . . . we
have both a necessary and legitimate interest in a potentially vast
range of partner country affairs.
(1998b: 3)

A further contribution to legitimation is made by reference to the Inter-

national Development Targets (DFID, 1997: 39). A similar and more com-
prehensive basis is provided by reference to ‘universal human rights’.
Goudie explicitly addresses the issue of ‘the legitimacy of the interest of
the external community in the governance of any other country’ (1998b:
2), and starts from the foundation of the Universal Declaration of Human
Rights, which, he argues, is ‘perhaps, the best example of an international
acceptance of this legitimacy’. Despite concern about the lack of universal
agreement about the boundaries of acceptable interference, he believes
that ‘certainly in the governance field, and particularly the human rights
field, external interest is broadly accepted’.3 A final contribution to legiti-
mise intervention is based on the contract implicit in the aid partnership:

Once a partner government has explicitly decided to enter into a

form of partnership with an external interest, and where the external
party – as is the case with the Department for International Develop-
ment – has established a programme of development assistance that
entails the transfer of resources in some form, then the legitimacy of
the external interest will take on a greater intensity.
(Goudie, 1998b: 2–3)
314 Peter Clarke
Mechanisms for recipient compliance
Even within the partnership framework, some mechanism is still required
to achieve recipient compliance with policy prescriptions. The documents
sometimes slip back into a language of direct conditionality: ‘international
support is conditional on economic, social and environmental policies
which will systematically reduce poverty’ (DFID, 2000a: 91), but there are
other formulations more characteristic of the partnership approach. A
first requirement is some mechanism of policy monitoring, and this is
facilitated by donor coordination, associated with International Monetary
Fund (IMF) surveillance (DFID, 1997: 71) and such mechanisms as donor
consultative group meetings (German and Randel, 2000: 20). Second,
rewards and sanctions are required, and a well-differentiated range of
these is proposed within the framework of partnership. The rewards are
articulated through partner selectivity, since this is not limited to entry
conditions but also promises a graded scale of aid intensity and duration,
according to policy compliance. The access to resources and flexibility in
their use are to depend on ‘the confidence we have in [partner govern-
ments’] policies and actions’ (DFID, 1997: 40). Sanctions are also main-
tained within partnership, with the threat of partial withdrawal when
policies begin to deviate. Thus Goudie argues that ‘we shall need to recon-
sider the scale and nature of . . . partnerships if the commitment of
partner governments continues to cause concern’ (1998b: 8). The 1997
White Paper defines appropriate measures where the government policies
of a poor country fail to win approval:

Where poor countries are ruled by governments with no commitment

to helping the poor realise their human rights, we will help – where
we can do so – through alternative channels. These will include insti-
tutions of civil society, voluntary agencies and local government.
(DFID, 1997: 39–40)

This alternative may be viewed with concern by Southern governments,

since in many countries civil society organisations are platforms for
opposition politicians. Goudie acknowledges difficulties, but reluctantly
lays them aside:

We should be deeply uneasy about working with civil society if it is

somehow to the exclusion of government or as a substitute for a
government or it acts to displace government. At times, we may need
to live with this unease while we work to build government partner-
ships.
(1998b: 9)

More extreme policy deviations are seen to justify more extreme

sanctions:
 Building a global partnership for development? 315
Where there are large-scale violations of international humanitarian
law and crimes against humanity, and where the government in ques-
tion is unable or unwilling to halt the atrocities, the UK believes that
the international community should take action. . . . Once all non-
violent measures have been exhausted, it may, in exceptional circum-
stances, be necessary and appropriate to use force to achieve the
humanitarian purpose.
(DFID, 2000a: 102)

Consistency of UK government policy

DFID responds to the criticism of development and aid discourses
that they construct a boundary around certain relations with developing
countries defined as ‘development interventions’, or certain transfers
defined as ‘aid’, leaving other areas of Northern government policy
‘free’ to pursue self-interest in a way that causes substantial prejudice to
developing-country interests and to net transfers. Its 1997 White Paper dedi-
cates a substantial section to these issues, under the title ‘Consistency of
policies’ (DFID, 1997: 50–76).4 The argument is made as follows:

[T]here is a complex web of environmental, trade, investment, agri-

cultural, political, defence, security and financial issues which affect
relations with developing countries. . . . To have a real impact on
poverty we must ensure the maximum consistency between all these
different policies as they affect the developing world. Otherwise, there
is a risk that they will undermine development, and development
assistance will only partly make up for the damage done.
(ibid.: 50)

The White Paper commits DFID to ‘ensure that the full range of Govern-
ment policies affecting developing countries . . . takes account of our sus-
tainable development objective’.

Partnership as political technology

How do the characteristics of the partnership approach described in the
previous section interact with the expressed aspiration for more equal
relations between donor and recipient? Renaming them as partners does
not change the condition of donor and recipient, and arguably the depen-
dence of the latter on the flow of resources from the former will tend to
dominate the relationship. Attempts to ignore this power relation are,
with the best will in the world, likely to run up against the expressed
‘legitimate’ donor needs to demonstrate timely results of their spending
316 Peter Clarke
to their parliaments and taxpayers. As Helleiner (2000) argues, agencies
are pushed to undermine local control by varied and entrenched forces:
their own procedural and accounting requirements, cross-country prior-
ities, demands to spend budgets, as well as career interests. The tendency
seems to be that rather than diminishing, these pressures are currently
increasing, with a growing emphasis on measurable aid impact and rigor-
ous programme auditing.
In this context, the move from limited project interventions to much
less bounded actions supporting substantial policy areas over the long
term in whole national territories similarly opens up this much larger
space/time to donor personnel and interference, in parallel with (legitim-
ate) comprehensive donor concern with virtually the whole of recipient
government policy.5
Donor co-ordination has the effect of closing Southern governments’
options to diversify their foreign co-operation portfolio, further tightening
their policy straitjacket. This is exacerbated by the increasing costs of non-
compliance due to the potential impact of sanctions by a co-ordinated
international community, right up to the extreme of the ‘just war’.
The policy prescriptions themselves are also problematic. Influences on
international consensus are unlikely to be equal, even in forums such as
the United Nations (UN) where each member has a relatively equal voice.
It could be argued that if Southern governments had played a greater role
in the development of international targets and the Millennium Develop-
ment Goals, other issues – such as Northern trade protectionism and debt
– might well have been included earlier, giving them more weight and
clearer objectives.
Similarly, the supposed benefits of a coherent UK government policy in
relation to developing countries6 – breaking down the artificial bound-
aries between policy areas – disappear when faced with the realities of
power.7 DFID does not set the agenda for other government departments;
the objectives of the Department of Trade and Industry, the Foreign
Office or the Ministry of Defence might still be expected to predominate,
despite significant increases in DFID funding.8
This review of the impact of the policy reforms associated with
‘partnership’ shows how the pursuit of coherence, comprehensiveness, co-
ordination and consistency itself consistently promotes a breaking down of
boundaries: between limited project intervention or conditionalities and
broader policy areas, between different donor programmes, between
Northern development policy and other policy areas. As the boundaries
are opened, power may flow more freely (and at lower cost) into every last
corner.
How can the discourse of partnership and ownership be understood if
its practice is really so contradictory? The extent of the contradictions is
shown by an example of World Bank pressure on Mozambique to remove
protection of a key industry. According to a Mozambique official,
 Building a global partnership for development? 317
the World Bank told us we must say this is our policy and stop saying it
is imposed by the World Bank . . . now we must lie to get World Bank
approval. And we will. But we remain totally opposed to a policy that
will destroy our cashew industry.
(Hanlon, cited in White and Dijkstra, 2002)

Helleiner (2000: 85) quotes a donor representative during a recent study

of aid relationships as remarking that ‘ownership exists when they do what
we want them to do but they do so voluntarily’. In this sense, the concept
of ‘ownership’ can be seen as part of an attempt to represent a political
problem as a technicality,9 and its use seems to refer less to a reality than
to an illusion that must be constructed precisely because true policy
ownership is absent. When recipient governments respond to aid con-
ditions with façade compliance, donors resolve to seek out governments
that share their policy perspectives, and apply a test for ownership. In this
way, ownership adds a further criterion to the burden of (effective) condi-
tionality, to the extent that it seeks to appropriate the agency of aid recipi-
ents. Recipients, as well as complying with policy conditions, are even
compelled to be convincing about wanting to comply.10 The costs of domi-
nation through coercion are reduced inasmuch as the recipient govern-
ment internalises the agenda of power. Nevertheless, the stick of IMF
disapproval (or, ultimately, military intervention) and the carrot of flexi-
ble aid transfers remain in place. In this view, the discourses of partner-
ship and ownership are revealed as examples of policies as ‘political
technologies’ concealing an ever wider and deeper penetration of power
(Foucault, 1978: 86; Shore and Wright, 1997).
The development policy narrative (Apthorpe, 1996: 8; Roe, 1991)
described thus far, with its emphasis on coherence, comprehensiveness,
co-ordination and consistency, suggests the construction of a monolith,
but are there no cracks in the edifice? Must any counter-narrative be so
monolithic? The remainder of this chapter explores this question, and
whether these cracks allow a place for human agency and resistance to
dominant conceptions of ‘partnership’, in an attempt to move beyond the
definition of human actors as instruments of development and/or its
victims. As Leach and Fairhead (2000: 36) warn, ‘subsuming bureaucratic
practice into discourse absolves the actors involved of consciousness,
intentionality and responsibility’. How can agency be given a place? An
alternative view of the same issues as are addressed by this chapter is pre-
sented by Chambers et al. (2001: 3). Their analysis of partnership also
points to the contradiction between its rhetoric and its practice, but
focuses on the responsibility of development professionals, arguing that
‘personal behaviour and attitudes are pivotal in helping or hindering
change, and directly influence wider norms and relationships’. The argu-
ment follows Chambers’s earlier publications (e.g. 1997), in which he
argues for a personal transformation of development professionals, who
318 Peter Clarke
are charged with a reversal of power relationships, to release the agency of
the poor. However, by focusing on the role of the ‘developers’ as the
problem, they paradoxically become the basis for the solution, and their
agency assumes an exaggerated importance. If neither analysis achieves a
balanced view of agency, how is the dilemma to be approached?
If attention is turned away from the abstraction of policy documents
and towards specific situated arenas of development practice, the picture
changes substantially, suggesting that the wood may have obscured our
view of the trees. Leach and Fairhead (2000: 36) point to the value of
actor-oriented sociology as an approach to these issues. As Arce and Long
argue,

[W]hereas discourse scholars . . . give priority to understanding how

Western science and development models enrol, discipline and
transform forms of knowledge rooted in other cultural traditions,
actor-oriented research focuses upon the ‘diverse and discontinuous
configurations of knowledge’ . . . that we encounter in specific devel-
opment arenas.
(Arce and Long, 2000: 24)

This approach shows how external interventions ‘enter the existing life-
worlds of the individuals and social groups affected, and in this way are
mediated and transformed by these same actors and structures’ in the
contexts of their own projects and strategies (Long, 1992: 20). The impli-
cation is that interpretations of the diversity of actor behaviour by develop-
ment professionals as compliance or resistance are based on their own
narrow and self-important perspective: actors are simply getting on with
their own particular projects, and development may or may not create
opportunities or obstacles for them (cf. Crewe and Harrison, 1998: 1,
159).
Viewed in this way, aid relationships take on a more differentiated char-
acter. The division between aid donors and recipients becomes hard to
sustain in specific development arenas – aid flows through many hands so
that even DFID is a recipient of funds (from the UK government budget),
while a Southern community group may also be a donor in relation to
households or individual members. Agency staff may be dependent on
recipients through their need to disburse funds on time. As Foucault
argues, the pervasiveness of power is matched by the resistance it encoun-
ters everywhere (1980: 142). The UK Secretary of State for Development,
in campaigning to defend a boundary between commercial and develop-
ment interests and to abolish the tying of aid, may be viewed as having
been resisting, as may a development bureaucrat who subverts established
procedures in order to permit local institutions to work at their own
pace, or an NGO which refuses project funding rather than accept donor
conditions.
 Building a global partnership for development? 319
Conclusion
A discourse of ‘partnership’ forms the core of the last Millennium Devel-
opment Goal, and has taken centre stage in the development policy of
DFID and other major multilateral and bilateral donors. The analysis of
key statements of DFID associated with this policy, in the context of
unequal power relations between donor and recipient, shows that far from
moving towards more equal relations, these measures can be viewed as
constituting a ‘political technology’, breaking down boundaries that limit
the global infiltration of power. The discourses of partnership and owner-
ship can in this sense be seen as attempts to mask an ever wider and
deeper penetration of development power relations in the South.
This ‘political technology’ and its associated discursive structure should
not be viewed as monolithic, converting human actors into victims or
instruments of development. But care should equally be taken to avoid
exaggerating agency and grounding solutions exclusively in the personal
transformation of development professionals. Actor-centred approaches
to the analysis of local development arenas suggest a more differentiated
picture of development relationships, in which clear divisions between
donors and recipients break down, and both power and resistance can be
detected in complex dynamic interactions.
The analysis of this chapter gives only a partial view of the meaning of a
‘Global Partnership for Development’, as it is limited to policy documents
and does not engage with the relations of partnership in everyday prac-
tice. As Sivaramakrishnan and Agrawal (1999: 21) argue, ‘development is
most fruitfully studied at the several loci of its practice, and in the mul-
tiple genres of its enactment’. Just as a ‘macro’ analysis, such as this, tends
to give an oversimplified picture, a ‘micro’ study may see only complexity
and differentiation. Further analysis of these issues would therefore
benefit from an endeavour to combine the two levels. How would a more
balanced analysis of development relationships look, one which might
contribute to a clearer understanding of relationships between the power
of discursive structure and individual agency?
Harrison’s study of the complexity of power relations in a Zambian
aquaculture project suggests how this may be done. Her multi-sited
ethnography (Food and Agriculture Organisation (FAO) Rome, Harare
and Luapula Province, Zambia) looks at ‘the ways in which . . . “local” con-
cepts of development and modernity are adopted and internalised’ and
analyses ‘the extent to which the supposed cohesion of the development
bureaucracy is in fact real or illusory’ (Harrison, 1995: 6). She argues that
it is necessary to go beyond a sharp and simple division between ‘us’ and
‘them’ towards ‘a fuller exploration of how the boundaries between one
apparent category of social actors and another are bridged, transformed
and shifted’. The study rejects the assumption that ‘developers always
develop while local people resist’ (ibid.: 266). A picture emerges of complex
320 Peter Clarke
negotiations of interests and a multiplicity of competing interpretations of
development. A fuller assessment of whether a ‘Global Partnership for
Development’ might ever be possible would surely depend on similarly
nuanced analysis of how development targets are reinterpreted and
mobilised at a local level.

Notes
1 The gradual move from development ‘assistance’ to ‘cooperation’ implies an
increasingly active role for donors.
2 The DFID 2000 White Paper states this in terms of both justice and self-inter-
est: ‘The policies set out in this White Paper . . . will increase social justice.
They are also in the UK’s self-interest because they will contribute to a more
stable and prosperous world at a time when no country can be isolated from
global developments’ (DFID, 2000a: 104).
3 This application of human rights instruments is consistent with the consider-
able importance given by DFID to ‘a rights-based approach to development’
(DFID, 2000b).
4 See also the Development Assistance Committee statement ‘Development Part-
nerships in the New Global Context’ (OECD, 1995: 2): ‘Other policies need to
be coherent with development goals.’
5 Helleiner demonstrates the absurd level of detail of policy intervention by
citing the example of World Bank insistence on the privatisation of the dog-
sniffing service at the airport in Jamaica (2000: 83).
6 The supposed benefits of policy ‘coherence’ are also questioned by Duffield
(2002), who concentrates on its impact on the work of humanitarian agencies.
7 In a parallel way, NGOs are increasingly entering into broader ‘partnerships’
with donors, which give them improved access to resources, but become an
opportunity for increasing donor control of their development policies.
8 If DFID manages sometimes to defend ‘developmental’ criteria in UK govern-
ment policy, this may perhaps be to the credit of the Secretary of State – which
raises issues of the role of agency and resistance, discussed later in this section.
9 cf. Shore and Wright (1997: 8): ‘[P]olicies appear to be mere instruments for
promoting efficiency and effectiveness. This masking of the political under the
cloak of neutrality is a key feature of modern power.’
10 Interestingly, for the purposes of the donor it is sufficient that the recipient
government believes itself in control, while for the recipient it is sufficient if
the donor believes in its shared policy conviction.

References
Apthorpe, R. (1996) ‘Reading Development Policy and Policy Analysis: On
Framing, Naming, Numbering and Coding’, in Apthorpe, R. and Gasper, D.
(eds) Arguing Development Policy: Frames and Discourses, London: Frank Cass.
Arce, A. and Long, N. (2000) ‘Reconfiguring Modernity and Development from
an Anthropological Perspective’, in Arce, A. and Long, N. (eds) Anthropology,
Development and Modernities: Exploring Discourses, Counter-tendencies and Violence,
London: Routledge.
Chambers, R. (1997) Whose Reality Counts? Putting the First Last, London: Intermedi-
ate Technology Publications.
Chambers, R., Pettit, J. and Scott-Villiers, P. (2001) ‘The New Dynamics of Aid:
 Building a global partnership for development? 321
Power, Procedures and Relationships’, IDS Policy Briefing 15, Brighton: Insti-
tute of Development Studies.
Crewe, E. and Harrison, E. (1998) Whose Development? An Ethnography of Aid,
London: Zed Books.
Department for International Development (DFID) (1997) Eliminating World
Poverty: A Challenge for the 21st Century, White Paper on International Develop-
ment, London: The Stationery Office and at www.dfi[Link]/Pubs/files/
[Link].
—— (2000a) Eliminating World Poverty: Making Globalisation Work for the Poor, White
Paper on International Development, London: The Stationery Office, and at
www.dfi[Link]/Pubs/files/[Link].
—— (2000b) Realising Human Rights for Poor People: Strategies for Achieving the Inter-
national Development Targets, London: DFID, at www.dfi[Link]/Pubs/files/
tsp_human.pdf.
—— (2001) Making Government Work for Poor People: Building State Capability,
London: DFID, www.dfi[Link]/Pubs/files/tsp_government.pdf.
Duffield, M. (2002) ‘Politics vs. Aid?’, Insights: Development Research, 39: 1–2, ID21,
Brighton: Institute of Development Studies.
Foucault, M. (1978) The History of Sexuality, Harmondsworth, UK: Penguin.
—— (1980) Power/Knowledge: Selected Interviews and Other Writings, Brighton: Har-
vester Wheatsheaf.
German, T. and Randel, J. (2000) Some Notes on Conditionality and Ownership,
Reality of Aid, at [Link]
Goudie, A. (1998a) ‘Eliminating World Poverty: A Challenge for the 21st Century.
An Overview of the 1997 White Paper on International Development’, Journal of
International Development, 10: 167–183.
——(1998b) ‘Is a Good Government Agenda Practical? An Approach to Gover-
nance’, talk given at Overseas Development Institute, 25 March, at [Link].
[Link]/speeches/[Link].
Harrison, E. (1995) ‘Big Fish and Small Ponds: Aquaculture Development from
the FAO, Rome, to Luapula Province, Zambia’, [Link]. thesis, Brighton: Univer-
sity of Sussex.
Helleiner, G. (2000) ‘External Conditionality, Local Ownership, and Develop-
ment’, in Freedman, J. (ed.) Transforming Development: Foreign Aid for a Changing
World, Toronto: University of Toronto Press.
Leach, M. and Fairhead, J. (2000) ‘Fashioned Forest Paths, Occluded Histories?
International Environmental Analysis in West African Locales’, Development and
Change, 31: 35–59.
Long, N. (1992) ‘From Paradigm Lost to Paradigm Regained? The Case for an
Actor-Oriented Sociology of Development’, in Long, N. and Long, A. (eds)
Battlefields of Knowledge: The Interlocking of Theory and Practice in Social Research and
Development, London: Routledge, pp. 16–43.
Maxwell, S. and Riddell, R. (1998) ‘Conditionality or Contract: Perspectives on
Partnership for Development’, Journal of International Development, 10: 257–268.
Organisation for Economic Cooperation and Development (OECD) (1995) Devel-
opment Partnerships in the New Global Context, Paris: Development Assistance Com-
mittee, at [Link]/dac/pdf/[Link].
—— (1996) Shaping the 21st Century: The Contribution of Development Co-operation,
Paris: Development Assistance Committee, at [Link]/dac/pdf/[Link].
322 Peter Clarke
Patel, I.G. (1971) ‘Aid Relationship for the Seventies’, in Ward, B., Runnalls, J.D.
and D’Anjou, L. (eds) The Widening Gap: Development in the 70s, New York:
Columbia University Press, pp. 295–311.
Pearson, L.B. (1969) Partners in Development: Report of the Commission on International
Development (IBRD), London: Pall Mall Press.
Roe, E. (1991) ‘Development Narratives’, World Development, 19(4): 287–300.
Shore, C. and Wright, S. (1997) ‘Policy: A New Field of Anthropology’, in Shore,
C. and Wright, S. (eds) Anthropology of Policy: Critical Perspectives on Governance and
Power, London: Routledge, pp. 3–39.
Sivaramakrishnan, K. and Agrawal, A. (1999) ‘Regional Modernities in Stories and
Practices of Development’, paper presented at South Asia Anthropology Group
Conference, School of Oriental and African Studies, University of London.
White, H. (2001) ‘Will the New Aid Agenda Help Promote Poverty Reduction?’,
Journal of International Development, 13: 1057–1070.
White, H. and Dijkstra, G. (2002) Beyond Conditionality: Programme Aid and Develop-
ment, London: Routledge.
16 Aid, trade and debt
How equal is the global
partnership?
Kevin Watkins and Juliana Amadi

Introduction
Sixty years ago, the Marshall Plan laid the foundations for the social and
economic recovery of Europe after the Second World War. It was motiv-
ated by the view that prosperity and security in one part of the world could
not be protected if mass poverty and hunger reigned elsewhere. Political
leaders of the day also had the vision to act accordingly.
The contrast with today is striking. While governments in the developed
world seldom miss an opportunity to offer rhetorical commitments on
poverty alleviation, they have collectively cut aid budgets to their lowest-
ever levels in real terms, failed to dismantle the trade barriers facing devel-
oping countries, and implemented ‘debt relief’ programmes which
leave poor countries with levels of external debt that are inconsistent with
the financing requirements for achieving the Millennium Development
Goals.
There are other important differences between the role of develop-
ment assistance under the Marshall Plan and international co-operation
today (White, 1999). The Marshall Plan was administered by the Organisa-
tion for European Economic Co-operation (OEEC), the forerunner of
today’s Organisation for Economic Cooperation and Development
(OECD). Receipt of Marshall Aid was based on national plans subject to
scrutiny by OEEC, with no special status accorded to the United States by
virtue of its position as donor. By contrast, developed countries have
retained the upper hand in their dealings with developing countries. Aid
management functions fall under the OECD body the Development Assis-
tance Committee (DAC), despite attempts of the more representative UN
body, the United Nations Conference on Trade and Development
(UNCTAD), to play this role. Attempts to create a large grant-giving aid
body within the UN in the early 1960s (the Special United Nations Fund
for Economic Development, SUNFED) failed, donor countries instead
favouring the creation of the International Development Association
(IDA) within the World Bank – an agency whose voting structure is based
on economic strength rather than the one country, one vote principle of
324 Kevin Watkins and Juliana Amadi
the UN. Today the World Bank and the International Monetary Fund
(IMF) act as ‘gatekeepers’ to bilateral aid, with donors insisting on com-
pliance with their loan conditions as a requirement for transferring aid. In
short, developing countries are very much the junior partners in the
development assistance contract.
The language of Goal 8 of the Millennium Development Goals, which is
‘to develop a global partnership for development’, sounds like a break with
the recent past. But does the reality of actions in the main areas identified
under Goal 8 – notably aid, trade and debt – match up to the rhetoric?
This chapter argues that the global partnership is far from equal. Developed
countries retain the upper hand, and are taking inadequate steps toward
fulfilling the commitments implied by their adoption of the Millennium
Development Goals. The next section begins by arguing that Goal 8 is not
taken as seriously by developed countries and development agencies as are
the other goals. Then, in the three following sections, attention is turned
to the three areas in which targets have been set in relation to Goal 8 – on
aid, trade and debt. Finally, we outline actions that are necessary for the
Millennium Development Goals to be achieved – actions broadly related to
aid, debt relief and international trade.

Last amongst equals: Goal 8 and the Millennium

Development Goals
The International Development Goals proposed in the 1996 OECD docu-
ment Shaping the 21st Century set out targets to be achieved by developing
countries, with no mention of the supporting actions required by developed
countries to help these goals be fulfilled. As Fairhead argues in this volume
(Chapter 14), it is inappropriate to focus on strategies by poor countries
when the goal is to attain environmental sustainability. Yet feasible strat-
egies for poverty reduction more generally also depend on international
conditions – hence the importance of a set of goals for actions to ensure
that poor countries, and the poor people within them, benefit from glob-
alisation through increased access to financial resources, markets and
technology. Strategies to close the huge gaps in health, education and
living standards between rich and poor will otherwise fail.
Whilst the first seven Millennium Development Goals are mutually rein-
forcing and are directed at reducing poverty in all its forms, Goal 8 is
about the mobilisation of the financial resources and other actions required
to attain the first seven Goals: specifically, the elimination of trade barriers
to developing country exports, debt relief for heavily indebted poor coun-
tries (HIPCs), additional financial assistance for the poorest developing
countries, and ensuring access to technology, including pharmaceutical
products. In a broad sense, Goal 8 is about creating an international
enabling environment in which national development strategies can
succeed.
 Aid, trade and debt 325
The term ‘global partnership for development’ implies a partnership
in the fight against poverty: partnership in the aid relationship, trade
relationship and debt-relief relationship between donor governments
and developing countries. Simply put, the bargain on offer can be
summarised as follows: commit to poverty reduction and good gover-
nance, and in return expect more development finance and greater flex-
ibility in the use of resources (Maxwell and Riddell, 1998). Though the
philosophy is nothing new, it marks an important shift in approaches to
development. It is based on the proposition that development is as much
about autonomy, self-determination and self-respect as it is about
income growth. Yet admirable as the principle of a new partnership with
developing country governments committed to poverty alleviation and
good governance might be, translating principle into action is difficult.
Genuine participation implies joint ownership, with mutual rights and
obligations. It involves a greater degree of formal reciprocity than cur-
rently appears to be on offer from the industrialised world. It also
implies a contractual relationship, with procedures for redress in the
case of default.
Behind the rhetoric, Goal 8 has an inferior status to that of the other
Development Goals. Moreover, this is an area where the gap between com-
mitments and action is exceptionally large, even by the normal standards
of development discourse.
First, it is notable that the indicators selected for Goal 8, unlike those
for the other seven Goals, do not include any time-bound commitments.
This is the case even though there are long-standing targets that could be
used. For example, there is no target for aid listed amongst the targets, let
alone, say, ‘achieve aid as 0.7 per cent of gross national product (GNP) by
2010’ or to meet the promise made at the 1995 Copenhagen Social
Summit under the ‘20:20’ initiative that 20 per cent of aid should be for
basic services. Second, the same effort is not put into producing reports
on progress for these indicators as for the others. Neither the World Bank
nor the OECD Web sites for the Millennium Development Goals report
these indicators, whereas they do report indicators for the other seven
Goals. This is the case despite the fact that some of the data, such as aid as
a percentage of GNP and the debt service ratio, are readily available from
World Bank and OECD sources.
Moreover, industrialised countries are already back-tracking on their
commitments. Aid commitments are but one illustration. At Monterrey in
2002, governments reaffirmed their commitments to achieving the Millen-
nium Development Goals. They also reaffirmed that developing countries
would need to supplement domestic resources in order to achieve the
Goals since most have limited sources of capital available to them and are
unable to attract substantial private-sector investment. The Monterrey
meeting also provided what many see as a last chance to mobilise the neces-
sary financial resources needed. Yet despite the international commitment to
326 Kevin Watkins and Juliana Amadi
achieving the Millennium Development Goals, donors refused to pledge
the necessary additional aid resources.
Since the terrorist attack of 11 September 2001, many governments
have spoken of the link between a peaceful global order, inequality and
poverty. Governments from the industrialised world have embarked on a
war against the evils of terrorism. But they have yet to commit themselves
seriously to the war against the evils of mass poverty, disease and illiteracy.
This failure to act will reinforce inequalities between rich and poor coun-
tries. It also calls into question the willingness of industrialised countries
to support inclusive forms of globalisation, or partnership for that matter.
While flows of private capital to poor countries are increasing, those
countries with the most entrenched poverty are being bypassed. Without
increased development finance, they face an increasingly marginalised
future.
On current trends, the 2015 Millennium Development Goals will not
be achieved in many parts of the world. Numerous countries are off track
and the gap between the required rate of progress and actual outcomes is
in some cases widening. According to the United Nations Development
Programme (UNDP), thirty-three countries, accounting for more than
one-quarter of the world’s people, will not achieve even half of the Millen-
nium Development Goals. Sub-Saharan Africa faces particularly acute
problems. Twenty-three of the countries that are off track are located in
the region, and another eleven lack sufficient data to make an assessment.
The share of people in the region living on less than US$1 a day was the
same at the end of the 1990s as at the start, at around 47 per cent (UNDP,
2002). Nothing more powerfully illustrates the human costs of failing to
meet the Millennium Development Goals than child mortality trends. The
number of additional child deaths that will occur as a result of the gap
between the target rate for achieving the Millennium Development Goals
and current trends represents a cumulative total of around 56 million
deaths. By 2015, sub-Saharan Africa will account for around 55 per cent of
total child deaths, compared with around 30 per cent in 2000 (Oxfam cal-
culations based on UNICEF, 2002).
Fortunately, trend is not destiny. All these outcomes and the loss of
potential and suffering associated with them are avoidable. According to
Oxfam, an additional US$100 billion a year in aid would be enough to
realise the Millennium Development Goals and honour the commitment
to the world’s poor (Oxfam, 2002b). For many countries, debt relief could
play a critical role in filling financing gaps. At the same time, measures to
strengthen the links between trade and poverty reduction could provide a
powerful catalyst for human development. But without the political will
that is needed to make these changes possible, there is a real danger that
the Millennium Development Goals will be missed by a wide margin.
Northern governments face a choice. They can continue their current
policy of using UN summits to deliver a large volume of rhetoric on
 Aid, trade and debt 327
poverty reduction, devoid of any financing commitments. Or they can
commit themselves to the investments in poverty reduction, health and
education that could transform the lives of poor people, creating the
foundation for shared prosperity.

Global trends in aid

A first target as part of the Millennium Development Goal to develop a
global partnership is to address the special needs of the least developed
countries, in part by more generous Official Development Assistance
(ODA) for countries committed to poverty reduction. There has been a
restatement of the pledge to raise aid to a target level of 0.7 per cent of
GNP, last reiterated at the Rio de Janeiro Earth Summit in Agenda 21, the
programme for action that set out policies for combating poverty and
improving living standards. Since Rio, aid as a percentage of GNP has
declined substantially, reaching its lowest level of 0.22 per cent in 1997
(OECD, 2001). Having increased in the last two years of the 1990s, ODA
as a percentage of GNP again fell back to 0.22 per cent in 2000. Only five
donor countries – Denmark, the Netherlands, Sweden, Norway and Lux-
embourg – have managed to meet the UN target of 0.7 per cent or
increase their aid as a percentage of GNP. Sixteen Development Assis-
tance Committee (DAC) donors – including Italy, France and Japan –
have been cutting aid, while most G-7 donors have allowed their aid as a
percentage of GNP to decline over the past ten years. The USA, Canada,
Germany and Italy are allowing aid contributions to stagnate at
exceptionally low levels. Even as the Financing for Development Summit
approached, G-7 donors allowed their aid to fall by 3 per cent in real terms
between 1999 and 2000. On average, the G-7 countries – Canada, France,
Germany, Italy, Japan, the UK and the USA – in 2000 gave just 0.19 per
cent of GNP in aid – even lower than their 0.21 per cent figure for 1999.
Several DAC members – Canada, Greece, Ireland, Switzerland and the
UK – have targets for increasing aid. But these commitments must be seen
in their proper context. The UK, the largest donor committed to increases,
has been making substantial progress. But even so, if current targets are
achieved, UK aid as a percentage of GNP will still be significantly below
the level achieved when a Labour government last left office in 1979.
Canadian aid is unlikely to rise above 0.33 per cent – far below the level
maintained from 1970 to the mid-1990s. Sweden plans to reach 0.81 per
cent of GNP in 2003, but it does not have a timetabled commitment to
return to the previous level of 1 per cent of GNP achieved in both 1982
and 1992. Even the weak proposals tabled for Monterrey by the European
Commission (EC), which called for a target of 0.33 per cent by 2006, were
accepted by member states only after much prevarication. So while
planned increases are welcome, there is a clear record of DAC
governments failing to deliver on commitments, and many of the
328 Kevin Watkins and Juliana Amadi
commitments are in any case to reach levels lower than those achieved at
earlier dates.
Perhaps the bleakest picture is the fact that five of the G-7 donors – the
USA, Italy, Germany, France and Japan – show no real signs of reversing
the declines that have occurred in their aid. More generally, the record of
even some of the strongest performers, including the UK, falls short of
what is required to achieve the Millennium Development Goals. In the
case of the USA, the Bush administration did announce on the eve of the
Finance for Development conference in Monterrey a 50 per cent increase
in aid by 2006 – representing resources of around US$5 billion a year.
However, this still leaves the USA rooted at the foot of the donor league
table. Moreover, the Bush administration has chosen to administer the
funds on a unilateral basis through a Millennium Challenge Account, with
aid eligibility governed by a wide range of economic reform criteria. This
is making it more, rather than less, difficult to allocate resources to where
they are most needed.
If we look at the long-term trend in aid, the 0.7 per cent target is now far
removed from actual trends. Through the 1980s, donors maintained aid at
around half the UN 0.7 per cent GNP target; even a decade ago, aid as a
percentage of GNP was stable at 0.33 per cent. But any optimism that the
end of the Cold War would result in a new world order in which the fight
against poverty was prioritised quickly evaporated, with aid declining sharply
to an all-time low of just 0.22 per cent of the combined GNP of DAC coun-
tries. Optimism that a post-11 September world would result in a stronger
commitment to reduce poverty must be seen against this background.
More worrying is the fact that the countries most dependent on aid
have suffered major losses. In 2000 only half of all aid went to low-income
economies with an income per capita of less than US$700. The other half
went to middle-income economies, where income per capita ranged from
US$700 to US$9,000. Real aid per capita fell from US$34 to US$20 in sub-
Saharan Africa in the second half of the 1990s, and halved in South Asia
over the same period (White, 2002).
Some of the deepest cuts have fallen in areas that have the most poten-
tial to reduce poverty. For instance, at the end of the 1990s aid flows
directed towards agriculture were running at one-third of their level in the
late 1980s (IFAD, 2001). This is despite the fact that rural communities
account for the overwhelming bulk of global poverty, and despite the
pressing need for public investment in infrastructure, marketing and
extension services.

Trade
A second general goal on partnership focuses on trade, calling for further
development of an open, rule-based, predictable, non-discriminatory trading
and financial system. This includes a commitment to good governance,
 Aid, trade and debt 329
development, and poverty reduction – both nationally and internationally.
This is important, since international trade as one of the motors for glob-
alisation is far more important than aid in defining poverty reduction. As
a source of economic growth, it has never been more important for global
prosperity, yet the benefits of trade are distributed unequally. While some
developing countries – notably in East Asia – have tapped into the benefits
of globalisation, others are being left behind.
At an international level, a variety of forces are at play, systematically
skewing the benefits of trade towards rich countries, perpetuating a highly
unequal pattern of globalisation in the process. Protectionist trade pol-
icies targeted at Southern exports, the disposal by rich countries of heavily
subsidised agricultural surpluses on world markets and poorly designed
IMF–World Bank trade liberalisation programmes all play a part. More
broadly, the ‘rules-based’ system enshrined in the World Trade Organisa-
tion (WTO) is facing a crisis of legitimacy. There is a gathering perception
in the developing world that many of the rules reflect blatant hypocrisy
and double standards on the part of industrialised countries. Viewed from
the developing world, the WTO is reinforcing a system that leaves coun-
tries representing four-fifths of the world’s population with less than one-
fifth of world exports.
Since the Uruguay Round of world trade talks concluded in 1996,
promises have been in steady supply. Industrialised countries pledged to
phase out protection against imports of textiles and garments, to scale
down agricultural subsidies and to remove trade barriers against the poorest
of countries. They made commitments to ensure that WTO rules on intel-
lectual property and investment do not undermine development
prospects. They also promised technical assistance to enhance the capacity
of developing countries to participate in the WTO and trade. In practice,
however, the attempts of industrialised countries to open up the markets
of developing countries through reduced tariffs, allowing trade in services
and protecting intellectual property, have not been matched by reciprocal
behaviour where the actions would adversely affect pressure groups in
developed countries.

Market access
The trend in the past thirty years has been towards increased openness in
international trade. With the advent of structural adjustment in the early
1980s, developing countries have been strongly encouraged to open their
markets. In addition to ensuring that developing-country markets are
open to exports from developed countries,1 the Uruguay Round intro-
duced new concerns of particular interest to industrialised countries, in
particular trade in services and protecting intellectual property rights. At
the same time, little progress has been made on areas of most interest to
developing countries: dismantling state support for agricultural production
330 Kevin Watkins and Juliana Amadi
in developed countries and removing distortions in the global footwear
and garments market.
Improved access to industrialised-country markets would help create
employment opportunities in developing countries and achieve a fairer
distribution of global wealth. In this respect, trade is far more important
than aid. According to Oxfam (2002a), every 0.7 per cent increase in
exports by developing countries generates as much income as they receive
each year in aid. But increased market shares require increased access to
markets.
Since the mid-1980s, South Asia, Latin America, East Asia and sub-
Saharan Africa have all halved average tariffs. Industrialised countries
have responded by maintaining exceptionally high trade barriers. These
protectionist barriers are costing developing countries approximately
US$100 billion per annum – twice the amount they receive in aid (Oxfam,
2002a). Far from supporting poor countries, industrialised countries are
actively discriminating against them. Again, according to Oxfam (ibid.):

• For manufactured goods, tariffs on developing country exports to

industrialised countries are on average four times higher than those
facing the exports of industrialised countries.
• High tariff and non-tariff barriers are concentrated in areas of special
interest to developing countries, such as agriculture and labour-
intensive goods.
• Between them, the USA and the European Union (EU) have launched
234 anti-dumping cases against developing countries since the end of
the Uruguay Round trade talks in 1994.

In the 1994 Uruguay Round, the Agreement on Textiles and Clothing

(ATC) was seen as a step in the right direction. It provided a commitment
by industrialised countries to phase out quotas on textiles and garments in
four stages by 2005. Developing countries account respectively for 50 per
cent and 70 per cent of these export categories. Despite this agreement,
the vast majority of quotas are still in place. Industrialised countries
have found ways to comply with the letter of the ATC while comprehen-
sively violating its spirit, for example by back-loading liberalisation and
‘liberalising’ goods not previously subject to quotas. Consequently, devel-
oping countries continue to face excessive trade barriers in textiles and
garments:

• The EU and the USA should have phased out over 70 per cent of the
quota restrictions. In reality, the EU has removed one-third of the
quotas on goods subject to restriction, and the USA one-tenth (Inter-
national Textile and Clothing Bureau, 2002).
• The average industrialised country tariff on textiles and clothing
imports from developing countries is 11 per cent – three times higher
 Aid, trade and debt 331
than the average tariff on imports from industrialised countries.
Tariffs will remain in excess of 10 per cent even after the Multi-Fibre
Agreement (MFA) phase-out.

These measures are having devastating effects on developing countries.

Textiles and clothing account for 10 per cent of total developing-country
exports; South Asia alone is estimated to lose around US$2 billion per
year as a result of trade barriers erected by industrialised countries. More
broadly, the World Bank estimates that industrial country restrictions on
trade in textiles and garments have prevented the creation of well over 20
million jobs in developing countries (IMF and World Bank, 2002). In
many countries, these jobs would have been taken by women workers.
Notwithstanding important problems relating to labour rights, health and
safety conditions and wage discrimination, many of these jobs would
have created an escape route from rural poverty, while at the same time
providing vulnerable populations with income for health and education
spending.
The failure of industrialised countries to match trade reforms has
played a determining role in the failure of trade reform in developing
countries to deliver anticipated benefits. This unbalanced liberalisation is
denying poor countries the opportunity to share in the benefits of globali-
sation.
Set against the limited trade liberalisation undertaken by Northern gov-
ernments, many developing countries have dramatically lowered tariff and
non-tariff barriers. This has been encouraged under IMF–World Bank
loan conditions, which frequently require rapid trade liberalisation. One
effect of the intervention of the Bretton Woods agencies in this area has
been to reinforce the asymmetric pattern of liberalisation outlined above:
in contrast to commitments undertaken at the WTO, Northern govern-
ments do not have to reciprocate unilateral liberalisation carried out by
developing countries under IMF–World Bank programmes. There is also
evidence that poorly designed and weakly sequenced liberalisation pro-
grammes have undermined the livelihoods of the poor, reducing the
contribution of trade to poverty reduction. In the case of Haiti, rapid lib-
eralisation of the rice sector resulted in the country being flooded with
heavily subsidised US rice, with damaging consequences for rural poverty
(Oxfam, 2002a: chapter 5).
The IMF and World Bank attempt to justify their role in promoting
trade liberalisation by citing econometric evidence purporting to show
that more open economies achieve more rapid growth and poverty reduc-
tion. However, the evidence itself is at best contentious – and at worst a
sophisticated irrelevance. By using trade:GDP ratios to measure openness,
the World Bank has diverted attention from more important indicators of
trade policy, including the speed and depth of liberalisation. Many of the
most successful developing countries in terms of reaping the benefits of
332 Kevin Watkins and Juliana Amadi
integration into global markets – such as China, Vietnam and Mauritius –
have combined rising trade:GDP ratios with relatively high levels of pro-
tection and a slow pace of liberalisation. There are no blueprints for
success, but this raises questions over the liberalisation blueprint advo-
cated by the IMF and World Bank.

Agricultural policy
Agricultural trade has a major bearing on poverty reduction efforts.
Approximately three-quarters of the poorest people in developing coun-
tries live in rural areas. Their livelihoods are affected both by export
opportunities and by competition from imports. Subsidies in industri-
alised countries exclude poor countries from world markets. They also
result in unfair competition in local markets, since smallholder farmers
cannot compete on price with subsidised exports.
Declared to be the start of a new era in which the withdrawal of subsi-
dies in industrialised countries would open new opportunities for poor
countries, the Uruguay Round agreement on agriculture was heralded as
a triumph for resolve and political will. In fact, the new era has yet to start.
The agreement committed industrialised countries to subsidy and tariff
reductions of 36 per cent. The action, however, has been minimal. By
choosing a reference period (1986–1988) of very low world prices and
high subsidisation as a yardstick for cuts, industrialised countries have
been able to avoid meaningful reductions. As a result, there has been no
real decline in agricultural protection.
While the headline figures point to subsidy cuts, these have been
achieved through a reclassification exercise. Income transfers have con-
tinued but are classified as ‘support payments’ rather than subsidies.
Annual emergency payments to US farmers, permissible under the WTO,
have grown rapidly. The net effect has been to create the appearance of
subsidy cuts while allowing past practices to continue.

• At the end of the 1990s, subsidies accounted for almost 40 per cent of
the value of OECD agricultural output – the same as in 1986–1988.
• The average tariff imposed by industrialised countries on agricultural
goods from developing countries is close to 20 per cent, almost five
times higher than the average tariff on all goods.
• Tariff peaks for commodities such as groundnuts in the USA, and
meat and dairy products in the EU, exceed 100 per cent.
• Processed food products attract tariffs at least as high as those on
unprocessed products, and usually higher.
• In the cotton sector, US farmers produced a level of output valued
at US$3 billion in world price terms, but received US$3.9 billion in
subsidies.
 Aid, trade and debt 333
These barriers represent a major obstacle to trade for developing coun-
tries seeking to break into export markets, and are estimated to cost them
approximately US$20 billion per year. Developing countries also lose from
the price-depressing effects of rich-country exports in third markets. For
example, West African cotton farmers are estimated to have sustained
foreign exchange losses in excess of US$200 million in 2001 as a direct
result of US cotton subsidies.2
The continuation of export subsidisation has been equally damaging.
Agriculture is the only area in the WTO where the practice of dumping,
or the sale of exports at prices below the cost of production, is institution-
alised as an acceptable practice. Of the twenty-five countries that reserved
the right to use export subsidies under the Agreement on Agriculture,
twenty-three were industrialised countries. Between them, these countries
account for 93 per cent of the US$21 billion of export subsidies in the
base period (Konandreas, 2002).
Oxfam’s verdict on the Agreement on Agriculture is that it was designed
to let industrialised countries continue with essentially the same policies.
The Agreement has introduced minimal restraints, notably by creating a
so-called Green Box arrangement under which certain forms of direct pay-
ments to farmers provided by the EU and the USA are not counted as
trade-distorting subsidies. Imbalances in the agreement highlight the way
in which the WTO framework has been subordinated by rich countries to
the vested interests of large farmers and powerful lobbies in the agribusi-
ness sector.

A better deal for the least-developed developing countries

At the 1996 Singapore Ministerial Conference, governments agreed to a
‘Plan of Action, including provisions for . . . duty free access aimed at
improving the overall capacity of Least Developed Countries to respond to
opportunities provided by the international trading system’ (WTO, 1997).
During the 2001 third UN developing country conference, industrialised
nations failed to act on this commitment. As in other areas, there has
been no real progress towards policies that might help the poorest coun-
tries capture larger shares of the benefits from trade.
The forty-nine countries classified by the UN as least developed coun-
tries (LDCs) are the poorest in the world. Around half of their population
– some 300 million people – live below the poverty line. Collectively they
account for less than 1 per cent of world trade. Yet several years after the
Uruguay Round, their exports continue to face stringent protectionist bar-
riers in industrialised countries:

• In the USA and Canada, only around one-tenth of all tariffs are above
5 per cent. Yet in both countries, approximately half of all LDC
exports face tariffs higher than this.
334 Kevin Watkins and Juliana Amadi
• Imports into industrialised countries from LDCs are twice as likely to
face tariffs in excess of 15 per cent as imports from other industri-
alised countries.
• Trade barriers are highest in sectors where LDCs have a potential
comparative advantage. Restrictions are particularly high for sugar
exports to the EU and clothing and footwear exports to the USA and
Canada.

According to the World Bank, eliminating duties and quotas on LDC

exports would generate US$2.5 billion in additional export earnings
(Hoekman et al., 2001). These are very large, static losses for countries
facing extreme foreign exchange constraints. Even so, they understate the
dynamic losses associated with lost opportunities for investment and growth.
These losses translate into large-scale losses of employment opportunities
and income for vulnerable populations, denying them a stake in the
potential benefits of trade. In some countries, such as Canada and the
USA, trade barriers against LDC imports cost more than is given in aid,
demonstrating how bad trade policies can seriously undermine develop-
ment assistance.
To its credit, the EU attempted to act on the commitment to improve
market access for LDCs. Its ‘Everything but Arms’ (EBA) proposal called
for the removal of all tariffs and quotas on LDC imports. However, the
proposal adopted by governments was substantially watered down. Follow-
ing intensive lobbying by farmers and agribusiness, liberalisation of
trade in key products such as rice and sugar – the very products which
offered the largest potential foreign exchange gains for LDCs – was
postponed.

Special action for Africa

The special problems facing sub-Saharan Africa have consistently been
acknowledged by developed nations. At the Lyon G-7 summit in 1996 they
launched a New Global Partnership for Development, with a special focus
on Africa. At the Okinawa summit, developed countries agreed to give
HIPCs and other low-income developing countries a stake in world trade
and to improve access for these countries to international markets. Similar
pledges were made at the end of the Uruguay Round of world trade talks
and in the Doha Development Agenda. Once again their performance has
been less impressive than the rhetoric.
The challenge facing Africa is immense. It has 12 per cent of the
world’s population, but accounts for less than 1 per cent of exports – one-
quarter of the share it enjoyed in the 1970s. Africa is the only region in
which the incidence of poverty has increased during the 1990s. More than
citizens in any other part of the developing world, Africans are being
bypassed by the benefits of globalisation.
 Aid, trade and debt 335
Industrialised countries have failed to act on their pledge to improve
market access. Unrestricted access to industrialised countries would gener-
ate an additional US$2.5 billion in non-oil export earning – a rise of 14
per cent. Several initiatives have been launched ostensibly aimed at improv-
ing Africa’s trade prospects. The US Africa Growth and Opportunity Act
provides in theory for duty-free and quota-free access in textiles and gar-
ments. The EU’s ‘Everything but Arms’ proposal provides similar conces-
sions for a far wider group of products. However, generosity in each case
is highly circumscribed. In the case of the USA, unrestricted duty- and
quota-free access is open only to products using American yarn. Similarly,
the EU modified its original ‘Everything but Arms’ proposal in the light of
lobbying by agro-industries. Both sugar and rice are now subject to a far
slower pace of liberalisation. On one estimate, Mozambique is losing
around US$80 million a year because of lost export opportunities in sugar
(Oxfam, 2002d).
But the problems facing Africa go beyond market access. The slump in
global commodity prices has had a devastating impact on the region,
causing economic collapse and large increases in poverty. Oxfam inter-
views with coffee farmers in Tanzania indicate that many are taking their
children out of school in the face of a 50 per cent decline over three years
in the price they receive for coffee. Primary commodities account for
three-quarters of Africa’s exports. Without concerted international effort
to address the causes of low commodity prices, which are rooted in
chronic over-supply, there is little prospect of Africa reversing its decline
in world export shares (Oxfam, 2002a: chapter 6).
The proposal to establish an International Trade Organisation (ITO)
in 1948, alongside the IMF and World Bank, included an objective to
secure ‘commodity prices as are fair to consumers and provide a reason-
able return to producers’. Fifty-three years later, this promise has yet to be
fulfilled. The issue of commodities has been kept off the WTO agenda, in
stark contrast to the issues of concern to industrialised countries, such as
investment or intellectual property.

Global patent rules that safeguard public health in poor countries

The application of the Trade-Related Aspects of Intellectual Property
Rights (TRIPS) agreement to pharmaceutical products was one of the
most controversial parts of the Uruguay Round agreement. The agree-
ment established for the first time a global intellectual property regime
enforceable through trade sanctions. That regime included a minimum
twenty-year period of patent protection (Lanjouw and Cockburn, 2001;
Oxfam, 2001a).
Developing-country governments raised concerns about the potential
effect of more stringent patent protection on the affordability of vital medi-
cines to the poor, and on development more generally. These concerns
336 Kevin Watkins and Juliana Amadi
were partially reflected in provisions (Article 31) allowing for compulsory
licences to override patent claims on public health grounds by authorising
local production. The agreement also included a provision allowing coun-
tries to import a patented drug from another country if the patent holder
was charging a higher price domestically – the so-called parallel importing
provision. However, neither of these loopholes addressed the fundamen-
tal tensions in the agreement. In particular, they failed to address the
rights of countries lacking a strong generic industry and therefore capable
of producing cheaper versions of patented drugs. They also opened the
door to extensive litigation on the part of companies seeking to restrict
parallel importing rights. The weakness, from a public health standpoint,
of the original agreement rapidly became apparent.
In South Africa, thirty-nine drugs companies began a court action to
prevent the South African government from importing cheap generic
copies of patented HIV/AIDS drugs. This case was followed by the US
decision to take Brazil to a WTO dispute panel (see Oxfam, 2001b;
Watkins, 2001). Once again, the aim of the complaint, subsequently with-
drawn in the face of public protest, was to prevent Brazil from producing
generic copies of vital drugs. The potential inflation in drugs prices
related to more stringent patent protection is part of a broader problem.
In a given year, 14 million people in developing countries will die from
infectious diseases. Many factors contribute to this distressing figure,
including poverty, weak health infrastructure, inadequate access to water
and sanitation, and poor policies. But many of the deaths could be pre-
vented if people could afford basic medicines. The TRIPS agreement
poses an acute threat because it will raise the cost of medication. Patented
medicines frequently cost more than ten times the price of generic
equivalents. And for poor people, price differences of this scale can be a
matter of life and death. The counter-case, argued by bodies representing
the pharmaceuticals industry, is that patents are vital to create incentives
for research into diseases affecting the poor. That such research is desper-
ately needed is not in doubt. In its second White Paper on international
development, the UK government pointed out that 90 per cent of drugs
research is on diseases affecting 10 per cent of the world’s population
(DFID, 2000). It proposed to support research into drugs of use to the
poor by guaranteeing the purchase of these products. The problem with
seeking to create market incentives through the patent system is that the
very basis of the incentive – namely, higher prices – has the effect of
excluding the poor from markets.
Despite acknowledging the difficult public health issues raised by
patenting, industrialised governments have supported the efforts of trans-
national pharmaceutical corporations to enforce the most stringent inter-
pretation of the TRIPS agreement. The USA in particular has used the
threat of bilateral trade sanctions to demand that the patent claims of US
companies be enforced. Countries such as India, Argentina, the Domin-
 Aid, trade and debt 337
ican Republic, Brazil, Vietnam and Thailand have all been threatened
under the ‘Special 301’ provision of US trade law. This contrasts in stark
fashion with actions in industrialised countries.
Industrialised countries are guilty not just of threatening the health of
vulnerable people in developing countries, but of extreme double stand-
ards. The US and Canadian governments have shown themselves willing to
threaten to override patents at home when faced with bio-terrorist threats
to their own citizens. Although no compulsory licences for patented
antibiotics were eventually issued, the threat of purchasing low-cost gener-
ics was successfully used to bargain down prices. The application of one
set of rules when North American public health is threatened, and
another for the health crisis in poor countries, is unacceptable. Whatever
the future threat posed by anthrax, the number of casualties that
prompted the change in approach to patents pales into insignificance
against the deaths associated with HIV/AIDS, which claims 2 million lives
each year in Africa alone. Moreover, the budget constraints of Northern
governments are far less severe than those of developing countries.
When trade ministers met at the WTO summit in Doha, in November
2001, they finally adopted a ‘public health declaration’ specifying that
patents would not be allowed to take precedence over public health
claims. Since then, there has been a protracted deadlock over the most
appropriate measures for translating this commitment into action, not
least because of the requirement that it reflects the provisions of the ori-
ginal agreement. Several developing countries have argued that in coun-
tries lacking a strong generic drugs industry, governments should be
allowed to issue a ‘compulsory licence’ to override patent claims, and that
generic manufacturers elsewhere should be automatically entitled to
export to that country. This approach has been resisted by the USA, which
wants each case to be considered separately on its merits, and by the
global pharmaceuticals industry.

Aid and technical assistance to developing countries

At the end of the Uruguay Round, industrialised countries promised tech-
nical assistance to developing countries to help them meet the costs of
implementing the Uruguay Round agreements, and to enhance their
ability to participate in the WTO. Less developed countries were promised
special treatment, but there is a huge gap between this promise and the
actual disbursement of funds.
For many developing countries, the cost of implementing the Uruguay
Round agreements is prohibitive, and places a huge burden on limited
human resource capacities. It will cost Tanzania US$10 million to meet
WTO customs evaluation standards, for example. The cost of drafting and
enforcing new laws on intellectual property in Bangladesh is estimated at
more than US$l million per annum. Despite this, at the end of the 1990s
338 Kevin Watkins and Juliana Amadi
the WTO budget for technical assistance was only US$500,000, sufficient
to meet less than one-fifth of the requests made for technical assistance.
The Integrated Framework to provide technical assistance to develop-
ing countries, launched in 1996, has an even more abysmal record. By the
end of the 1990s it had failed, and was relaunched in 2002. To date, indus-
trialised countries have provided US$7 million to undertake a ‘needs
assessment’ in a small group of pilot countries. There are no concrete
funding commitments for the future to address the priorities that emerge.
Failure to provide adequate technical assistance is reflected in the huge
imbalances in negotiating strength and institutional capacity at the WTO:

• The average developing-country trade mission at the WTO has three

people, compared with seven for developed countries. Even a large
country like Bangladesh has only one representative.
• Of the thirty-eight African countries in the WTO, fifteen have no resi-
dent delegate; four maintain only one-person offices.
• On average, there are forty-six delegate meetings per week in the
WTO. There are complex negotiations across large areas of industrial,
agricultural, investment and services policy that have profound implica-
tions for human development. Yet many of the world’s poorest coun-
tries lack the capacity to monitor, let alone influence, the direction of
these negotiations.

Industrialised countries also promised action to help developing countries

acquire a greater share of the benefits of international trade. Financial
and technical assistance is crucial to help them take advantage of new
market opportunities. In particular, developing countries need support to
address the constraints in producing goods for export, such as inadequate
infrastructure and limited technical facilities and skills to add value to
domestic produce and ensure that goods meet quality and other export
standards.

Will the WTO help create the conditions for sustained growth and
poverty reduction in developing countries?
The Doha Round of trade talks has been dubbed the ‘development round’
– and there is much encouraging rhetoric for developing countries. The
document adopted by trade ministers at Doha acknowledged that ‘There is
need for positive efforts designed to ensure that developing countries . . .
secure a share in the growth in international trade commensurate with the
needs of their economic development’ (WTO, 2002). Industrialised coun-
tries claim that nothing in the WTO will hamper the ability of developing
countries to achieve this goal. This is untrue: WTO agreements restrict gov-
ernments from introducing policies that might enable their countries to
reap the benefits of integration into the global economy.
 Aid, trade and debt 339
Under the Uruguay Round agreement, developing countries lost the
right to implement many of the policies that had been central to East
Asia’s success. These included the selective protection of domestic indus-
tries, targeted subsidies for domestic firms, restrictions on foreign investors,
the copying of patented technologies, and requirements on foreign
investors to link with the local economy.
The TRIPS agreement is a particular concern. Meanwhile, the WTO
Trade-Related Investment Measures (TRIMs) agreement, concluded in
1994, poses similar problems. It severely restricts the right of governments
to impose ‘local content requirements’ – an obligation to source inputs
from local industry – on foreign investors. Countries such as South Korea
and Taiwan used this local content rule extensively to build dynamic link-
ages between the export sector and domestic firms. Today, these two
countries account for over one-third of medium- and high-technology
exports from developing countries. Unlike countries such as Mexico, they
have succeeded in entering dynamic new markets on the basis of domestic
innovation and enterprise – and they capture a larger share of the value of
their exports as a result. Yet the policies behind their success have been
outlawed through the WTO.
The General Agreement on Trade in Services (GATS) is another area
of concern. Negotiations in this area cover not just financial and technical
services, but also utilities such as electricity, water and education. To date,
the effects have been minimal. However, powerful corporate lobbies, led
by the Coalition of Service Industries, and strongly supported by both the
EU and the USA, are seeking to advance an agenda for radical liberalisa-
tion. In principle, developing countries could gain from some aspects of
service market liberalisation, especially in areas such as software and con-
struction. However, firms in developing countries are ill-equipped to
compete with transnational corporation service providers in areas such as
finance and insurance.
There is a real danger that these WTO agreements will lock developing
countries into a subordinate position in the global trading system, leaving
them unable to upgrade their exports. Instead of supporting the develop-
ment of East Asian-style dynamic export growth, the WTO is promoting
Mexican-style dependency on investment by transnational corporations,
weak linkages between the export sector and the domestic economy, and
low wages.

Debt relief
A third broad Goal on global partnership concerns debt relief. Unsustain-
able debt presents a huge barrier to progress in the fight against poverty
(Oxfam, 2002e). Debt repayments by some of the poorest countries in the
world are diverting the necessary resources required for development
financing. The Heavily Indebted Poor Country Initiative, announced in
340 Kevin Watkins and Juliana Amadi
1996, was to eliminate debt as an obstacle to poverty reduction. In 1999
the HIPC Initiative was reformed. The new Enhanced HIPC initiative pro-
vided for earlier and deeper debt relief, and sought to establish a close
link between debt relief and poverty reduction. However, the level of debt
repayment after the Enhanced HIPC initiative debt relief remains far too
high, undermining the necessary investment needed to accelerate poverty
reduction.
The HIPC Initiative has already freed up resources from debt servicing
for twenty-six low-income countries, enabling pro-poor expenditure and
some progress towards the Millennium Development Goals (IMF and IDA,
2002). Preliminary analysis of the HIPC Initiative’s achievements shows
that in some countries debt relief has resulted in demonstrable social and
economic gains (World Bank, 2001a). For 2001–2003, the HIPC Initiative
reduces the average debt service paid by HIPC graduates by about one-
third. Among these countries, social expenditures are expected to increase
in 2000–2003 from the levels in 1998–1999. Where countries have had
resources freed up from debt, the proceeds have resulted in some new
development programmes and economic progress, such as a free immuni-
sation programme for children in Mozambique; the abolition of user fees
for primary education in Uganda, Malawi and Tanzania, and in rural areas
of Benin; and planned increases in spending on HIV/AIDS prevention in
Mali, Mozambique and Senegal. Moreover, the requirement to engage in
a consultation process in designing Poverty Reduction Strategy Papers
(PRSPs) has helped to increase the potential for people to influence
national resource allocation processes.
But the HIPC Initiative stops short of what is needed. Although the
examples demonstrate that debt relief can generate additional resources
that contribute to furthering human development, the socio-economic
gains are by no means universal and, where they exist, they are limited and
precarious. Worse, the development gains made with the small additional
resources provided by the Enhanced HIPC Initiative may be swept away
without additional financing. HIPC countries, like all low-income countries,
continue to face development challenges such as the spread of HIV/AIDS,
low literacy levels and poor nutrition, and they face them equipped with
scarce and highly vulnerable domestic resources. Given the fragile social
and economic conditions prevalent in HIPCs, the benefits derived from
limited amounts of debt relief are likely to be small or easily reversed.
For example, in almost all HIPCs, private-sector flows will not make up
for chronic resource deficits. The marginalisation of the African continent
from global trade is equivalent to a loss of 21 per cent of regional GDP, or
US$68 billion per annum. For Africa in 2001, after adjusting for inflation,
non-fuel commodity prices are at one-half the annual average value for
the period 1970–1981. The World Bank and IMF estimate that eight to ten
of the HIPCs most affected by the slump in commodity prices will have
higher debt:export ratios by completion point than the 150 per cent
 Aid, trade and debt 341
target set by the HIPC Initiative itself. Instead, the HIPCs continue to rely
on external official assistance, particularly in the form of grants, to fund
their domestic spending and balance of payments gaps. Despite optimistic
projections in decision-point documents, new HIPCs are not receiving the
levels of external finance anticipated that will in turn help them to achieve
the Millennium Development Goals.

Levels of debt relief

Under the Enhanced HIPC Initiative, debtors receive debt relief under a
two-phase process. After complying with an IMF programme and demon-
strating progress towards the development of a PRSP, they reach a decision
point. At this stage, calculations are made of the level of debt reduction
needed for them to reach sustainability, defined in terms of a range of
threshold indicators. The (net present) value of debt stock is measured
against exports of goods and services. If the ratio of debt to exports is
greater than 150 per cent after the full application of the traditional debt-
relief mechanism, the country’s debt is considered unsustainable. It then
qualifies for interim debt-service relief to reduce the level below the
threshold. Provided that it continues to comply with the IMF programme
and finalises a comprehensive PRSP, it can then graduate to completion
point and receive debt-stock relief. The arrangement is intended to
provide a permanent exit from unsustainable debt.
Four countries – Bolivia, Uganda, Tanzania and Mozambique – have so
far reached completion point. Another twenty-two are receiving interim
debt relief, having reached decision point. IMF and World Bank staff
reports invariably express the level of debt relief provided in terms of
long-term changes in debt stock, or debt:service rates. For example, the
April 2001 review noted that nominal debt relief for the twenty-six coun-
tries covered amount to US$40 billion (and US$25 billion in net present
value terms) (IMF and World Bank, 2001a). For the same group of coun-
tries, the average annual debt servicing as a percentage of exports for
2001–2005 was projected to fall by almost one-third from the 1998–1999
level. In a similar vein, debt service relative to government revenue is pro-
jected to fall from an average of 24 per cent a year to 13 per cent over the
same reference period. Viewed in the context of strategies for mobilising
financial resources for poverty reduction, this represents a huge drain on
public finances:

• Out of twenty HIPCs at decision point, Mali, Niger, Sierra Leone and
Zambia will have annual debt payments due in 2003–2005 which will
be higher than their annual debt service paid in 1998–2000.
• Five countries (Ethiopia, Guinea-Bissau, Honduras, Nicaragua and
Uganda) will be paying as much in debt service payments as before
the HIPC Initiative.
342 Kevin Watkins and Juliana Amadi
• In six countries, annual debt service will be reduced by a modest
US$15 million in 2003–2005.
• The medium- to long-term projections on debt servicing are also
alarming: Senegal’s debt service jumps by 61 per cent in 2004 and
Honduras’s by 93 per cent.

Savings measured in terms of the gap between projected debt servicing

pre- and post-Enhanced HIPC Initiative (much of which would not have
been paid in any case) have limited relevance when it comes to real
current spending capacity. Similarly, nominal debt stock figures provide
some insights into financial sustainability, but offer little information
about budget resources. Since it is these resources that dictate what gov-
ernments are able to finance, the sustainability of debt should be assessed
against budget criteria, as well as foreign exchange ratios.

Debt and government revenue

The current system of debt relief, the Enhanced HIPC Initiative, is not
working effectively. When it is measured against domestic resource mobil-
isation, its shortcomings are painfully apparent. The problem is that debt
repayments continue to absorb a large share of the limited revenue base
available to governments. If sustainability is measured against the criteria
of financing for human development, the debt burdens of many countries
– especially those at the centre of HIV/AIDS crisis – are unsustainable.
Consequently, some countries will soon be left with unsustainable debts
once again. Not only are some HIPCs spending more on debt relief, they
are overshooting the World Bank and IMF’s own definitions of debt sus-
tainability.
Of the twenty-six countries receiving Enhanced HIPC debt relief in
2002, half of them are still spending an average of 15 per cent or more of
government revenue on debt repayments. Only two of the twenty-six
HIPCs currently receiving debt relief have debt-service repayments equival-
ent to less than one-half of total spending on health. While several HIPCs
now spend less than 10 per cent of revenue on debt servicing, the repay-
ments are crowding out vital public investment in health, education and
other areas. Thirteen of the twenty-six countries receiving debt relief are
still spending more on debt than on public health. These are some of the
worst cases:

• Zambia and Malawi have amongst the highest HIV/AIDS prevalence

rates in the world. But while Zambia has almost 1 million people
affected, the country is spending 30 per cent more on debt than on
health. Malawi’s health budget is equivalent to its debt servicing.
These are countries in which HIV/AIDS is driving an increase in child
mortality.
 Aid, trade and debt 343
• In Cameroon, HIV prevalence rates have passed 5 per cent. Debt
repayments amount to three and a half times the spending on health.
• Zambia, Mali, Niger and The Gambia, amongst other countries, all
spend more on debt than on education.
• Several countries – including Cameroon, Sierra Leone and Maurita-
nia – spend more than twice as much on debt as on education.

Unfortunately, the tensions between debt servicing and financing for basic
services are not untypical. Even HIPC countries that have benefited from
significant debt relief face an acute mismatch between the claims of credi-
tors and the resources allocated to investment in basic public services:

• For every $1 that Mali spends on health, $1.60 is transferred to credi-

tors.
• Niger, with the highest child mortality in the world, continues to
spend more on debt servicing than public health even after debt
relief.
• Sierra Leone, with one of the world’s highest maternal mortality rates,
will spend 2.5 times more on debt servicing than on health in 2002 on
current projections.

Debt repayment obligations inevitably clash with efforts to help finance

development strategies called for under UN targets. For instance, the
national strategic plan for HIV/AIDS developed by the government of
Malawi plans to allocate around US$24 million annually of domestic
resources. Yet its capacity to undertake these investments is being compro-
mised not just by a disastrous famine, but also by a debt-service profile that
resulted in transfers to creditors of US$57 million in 2002.
Heavily indebted countries in West Africa face acute problems.
National adult HIV/AIDS prevalence rates have already passed 5 per cent
in several countries, reaching 7 per cent in Burkina Faso. Even countries
with relatively low debt-service:revenue ratios face debt-related financing
constraints. In Burkina Faso the financing provisions for the national
AIDS strategy amount to approximately one-half of the amount that every
woman, child and man in Burkina Faso currently transfers to external
creditors.
The finance and service delivery challenge facing HIPC governments is
of daunting proportions. When the human and financial implications of
HIV/AIDS are considered, the inadequacy of the Enhanced HIPC Initi-
ative is as apparent as the gap between rhetoric on development and
action (or more accurately, their inaction) on development financing.
According to the Commission of Macroeconomics and Health, govern-
ments of low-income countries need to increase spending on health by 1.6
per cent of GNP a year to 2015 (based on 2002 costs) in order to move
towards universal coverage (WHO, 2001). Current spending on debt
344 Kevin Watkins and Juliana Amadi
servicing after Enhanced HIPC Initiative debt relief amounts on average
to 3 per cent of GDP. In other words, the additional health spending
could be financed to a significant degree by a transfer of resources from
external creditors to domestic service providers.

The mirage of debt sustainability

There is a misconception over the time at which countries achieve debt
sustainability, defined by the World Bank and the IMF as a 150 per cent
debt:exports ratio. Board papers assert that debt sustainability will be
achieved on reaching completion point, and explicitly act on the assump-
tion that debt relief will be ‘delivered unconditionally’. But rather than
occurring immediately, the debt relief is implemented over as long as
thirty- to forty-year periods, depending on the relief method chosen.
As a result of the slide in commodity prices, some countries are already
overshooting the World Bank- and IMF-defined threshold of debt sustain-
ability. The international financial institutions’ response of supplementing
the HIPC Initiative with a ‘one-off’ post-completion point ‘topping-up’
facility is little guarantee that the Initiative’s graduates will be in a position
to sustain their debt-servicing liabilities in the short term, let alone meet
the promise of a ‘robust exit from unsustainable debts’. Uganda, the first
HIPC graduate, currently has debts of over 200 per cent of the debt:
exports ratio. This will be the third time Uganda has exceeded its debt sus-
tainability after reaching completion points. The March 2002 Completion
Point Board paper for Burkina Faso concludes with the expectation that
after receiving its ‘topping up’, the country may achieve debt sustainability
by the year 2016! However, contrary to most preceding definitions (net
present value to exports) of debt sustainability produced by the World
Bank and IMF, it asserts that ‘debt sustainability is not endangered’
because there is sufficient liquidity to cover debt servicing. This is the first
time, and against the agreed rules, that the World Bank and IMF have
changed definitions of debt sustainability to include liquidity as the opera-
tive criterion.
World Bank and IMF projections and estimates for future growth,
investment rates and financial inflows have been systematically over-
optimistic and bear no relation to rates achieved in the past. The use of
wholly unrealistic assumptions about the future financial and economic
performance of HIPCs is bound to lead to unrealistic debt sustainability
analysis and countries overshooting their sustainability thresholds.
Amongst the multiple difficulties faced by HIPCs is their extreme vul-
nerability to external shocks. In particular, their high concentration of
exports on few commodities leaves them acutely sensitive to external
shocks in commodity prices and climatic conditions. The current criterion
of the net present value of debt to exports for debt sustainability analysis
therefore has a limited use. Because of its reliance on the narrow and
 Aid, trade and debt 345
highly volatile variable of export earnings as a means of calculating future
debt sustainability, it is the key failing of the HIPC Initiative. Also, exports
alone do not reflect the resources available to HIPC governments for
poverty reduction expenditures. It would be quite possible, under the
current criteria, for a country’s debts to be considered sustainable from
the point of view of external viability while that country has insufficient
resources to meet even the most basic poverty reduction expenditures.
For most HIPCs, exports are therefore an unreliable predictor of medium-
term and, for some, even short-term debt sustainability.
Effective debt relief could help by releasing the resources needed for a
concerted assault on poverty. The financial implications of HIV/AIDS for
the attainment of the Millennium Development Goals point to a wider
need to reassess the adequacy of the HIPC Initiative. Changing that
picture for the better in the face of the challenge posed by the epidemic
will require unprecedented commitment to resource mobilisation. If gov-
ernments are serious about achieving agreed human development targets,
they need to assess – or reassess – the costs of doing so in the light of the
HIV/AIDS epidemic. And creditors need to balance their claims against
the financing needs of debtor countries.
Notwithstanding a US$1 billion ‘top-up’ at the Kananaskis G8 summit,
developed countries have consistently failed to respond to the inadequa-
cies of the Enhanced HIPC Initiative. No attempt has been made to revise
debt sustainability indicators in the light of the financing requirements for
addressing the HIV/AIDS crisis and achieving the Millennium Develop-
ment Goals. Meanwhile, many countries are being forced back into acute
debt unsustainability by a protracted decline in commodity prices, weak
aid flows and wildly over-optimistic export growth projections by IMF and
World Bank staff.

Actions required to achieve the Millennium Development

Goals
More than additional financing will be required if the Millennium Devel-
opment Goals are to be met and a genuinely equal global partnership
achieved. Improvements in the delivering and quality of service are a pre-
condition for progress. Money is not the only input that matters – or even
the most important one. However, many of the poorest countries lack the
domestic financial resources needed to achieve the 2015 targets. More-
over, there is now a credibility gap between the rhetoric of the industri-
alised countries, the World Bank and IMF, and the reality of development
financing. Action needs to be taken in relation to aid, trade and debt for
the Millennium Development Goals to be reached.
346 Kevin Watkins and Juliana Amadi
Financing the gap
Various estimates have been made of the costs of achieving the Millen-
nium Development Goals. The World Bank suggests an indicative range of
US$40–60 billion in additional aid for the next fifteen years. Whilst it is
difficult to calculate exactly how much money is needed, the estimates
made are, in Oxfam’s view, significant understatements of the resources
needed. This is because the World Bank’s estimates understate the cost of
achieving the Millennium Development Goal for health, and associated
investments in water and sanitation. According to Oxfam, the real cost of
achieving the Millennium Development Goals by 2015 will be approxi-
mately US$100 billion extra per year (Oxfam, 2002b).
The headline figure is large, but affordable. Had donors met their
pledge to spend 0.7 per cent of GNP on aid, they would now be spending
an extra US$114 billion. The financing requirements for achieving the 0.7
per cent target are modest in relation to government expenditure. The
average increase in government spending required for the G-7 countries
would be around 1.4 per cent. Instead, they have cut their aid budgets.
The cost of this investment in human development has to be assessed
against the potential benefits, both human and economic. According to
the Commission on Macroeconomics and Health, aid investment equival-
ent to 0.1 per cent of the GNP of industrialised countries could avert 8
million deaths a year by 2015 (WHO, 2001). Using extremely conservative
estimates, the Commission suggests that the increased wealth generated
by improved health would represent three times the costs of increased
health spending by rich and poor countries.
Just as the different aspects of deprivation are mutually reinforcing, so
are the benefits to human development. For instance, each additional
year of education is associated with an increase in output of around 9 per
cent among smallholder farmers, with the adoption of new technologies.
Improvements in girls’ education are intimately related to better health,
especially for children. Comparative research across countries has found
that each additional year of maternal education reduces child mortality by
around 8 per cent. Improved health is also one of the requirements for
improved learning.
The uncertainty that continues to surround debates on financing for
Millennium Development Goals highlights the need for all developing
countries to work with donors in developing reliable national estimates.
This exercise should be a central part of the process for preparing Poverty
Reduction Strategy Papers (PRSPs). The financing requirements should
be reflected both in government financial frameworks and in IMF/World
Bank programmes. More immediately, and notwithstanding current donor
reluctance to increase aid, it is important that the UN and World Bank
avoid tailoring figures to suit the preferences of industrialised countries.
Whatever the precise costs of achieving the Millennium Development
 Aid, trade and debt 347
Goals, two things are clear. First, while there is scope for additional financ-
ing through resource mobilisation in least developed countries (LDCs),
through improved efficiency, greater equality and enhanced revenue col-
lection, poverty imposes financial constraints. The Millennium Develop-
ment Goals will not be achieved without a major and sustained increase in
aid spending. Second, increased aid will not be enough. Past experience
shows that political commitment on the part of aid recipients is vital.
Building health and education systems that are responsive to the needs of
the poor, and reprioritising public spending to emphasise poverty reduc-
tion, are the keys to success.
Developed countries should set a five-year time-frame for achieving the
0.7 per cent aid target. This would generate US$130 billion a year in addi-
tional financing by 2007 – sufficient not just to achieve the Millennium
Development Goals, but also to sustain a broader campaign against
poverty. The costs of that campaign would amount to an annual increase
in government expenditure of 0.2 per cent over five years for countries
such as the UK and Germany, and 0.3 per cent a year for the USA. The
cost of success in the war against poverty is modest when compared with
other priorities adopted by governments:

• The US$11 billion annual increase in spending required for the USA
represents around one-quarter of the increase in military spending
scheduled for 2003, and one-seventh of the tax cuts for the period
2002–2014.
• The EU could reach the 0.7 per cent target if it were to increase aid by
an amount equivalent to the subsidies provided under the Common
Agricultural Policy (US$35 billion).

Enhancing the Enhanced HIPC Initiative

The HIPC Initiative marked an important step forward in addressing the
debt problems of low-income countries. It provided for the first time an
integrated framework for dealing with all categories of debt, and it set
limits on creditor demands linked to a notion of sustainability. The
Enhanced HIPC Initiative provided for earlier, deeper and broader debt
reduction. But it has not gone far enough. The Initiative is teetering on
the brink of failure in its central objective, namely, the provision of a cred-
ible guarantee that countries entering it will be provided with a once-and-
for-all exit from unsustainable debt.
As a debt-relief strategy, the benefits of the Initiative are being eroded
by wider pressures and mismanagement. This is doubly unfortunate, since
there is evidence that it is contributing in a powerful way to poverty reduc-
tion efforts across a wide range of countries. Five interacting problems are
contributing to this outcome: (1) failure to prioritise the government
revenue: debt service ratio in assessing sustainability; (2) failure to factor
348 Kevin Watkins and Juliana Amadi
in low and unstable commodity prices; (3) persistent over-estimation of
export growth prospects by the IMF; (4) insufficient and uncertain debt-
relief provision prior to completion; and (5) inadequate flows of aid.
From the outset, levels of debt relief have been dictated by what credi-
tors deem affordable, rather than by the needs of debtors. Moreover, the
criteria for determining levels of debt relief reflect a narrowly defined
financial perspective, focused on foreign exchange indicators. No attempt
has been made to develop debt sustainability indicators that reflect
the financing requirements for achieving poverty reduction goals. The
inadequate weight attached to the proportion of government revenue
allocated to debt servicing reflects this bias. Another problem has been
the central role of the IMF in managing the Enhanced HIPC Initiative.
Assessments of country performance by IMF staff that are often at best
weakly related to poverty reduction considerations have resulted in delayed
disbursement of debt relief, introducing high levels of uncertainty into
the framework.
Current responses to the problems posed by the Enhanced HIPC Initi-
ative betray a short-sighted piecemeal approach. Successive meetings of
G7 finance ministers have witnessed repeated wrangles over financing
‘top-ups’ to compensate for adverse commodity price trends – and for the
reckless projections of export performance developed by IMF staff. No
attempt has been made to confront the more fundamental challenge of
integrating debt relief into a coherent resource mobilisation strategy for
realising the Millennium Development Goals.
What is needed is a bold new strategy, which should include the follow-
ing elements:

Assessing the financing implications of HIV/AIDS

The financial assessments of the implications of HIV/AIDS in HIPCs
should inform evaluations of debt sustainability and levels of debt relief
provided. As part of the PRSP process, all countries should cost their
national AIDS plans, identify the financing gaps that could be filled
through debt relief and other measures, and ensure that financing provi-
sions are reflected in national budgets and medium-term expenditure
frameworks.

Deepening debt relief

Current debt sustainability indicators suffer from two problems: they are
tangentially related to resource mobilisation for poverty reduction goals,
and they are not sufficiently generous. An upper ceiling of 5 per cent
should be set on the proportion of government revenue allocated to debt
servicing. Such a limit would have mobilised an additional US$1.6 billion
in the 26 countries currently receiving Enhanced HIPC debt relief. While
 Aid, trade and debt 349
this implies real costs for creditors, these costs represent less than 3 per
cent of existing aid flows. The human costs of continuing with business as
usual are beyond estimation.

Broadening debt relief

There is a strong case for broadening debt relief in response to the threat
posed by HIV/AIDS and wider poverty reduction challenges. For instance,
both Kenya and Angola have been deemed ‘sustainable debt’ cases under
the existing framework, and Nigeria is not covered. The limitations of the
HIPC Initiative have been further exposed by crises in the private capital
market. Indonesia – the world’s fourth most populous country – has been
allocating more than one-fifth of government revenue to debt servicing
since the 1997 financial crisis, rising to over one-third in some years. No
effective debt-relief mechanisms exist (despite the fact that Indonesia has
a lower income per capita than Honduras, which is eligible for HIPC debt
relief). This has hampered social and economic recovery. It has also
undermined efforts to curtail HIV/AIDS. After more than a decade with
negligible rates of HIV, the country is now seeing infection rates increase
rapidly. There is an urgent need for the international community to look
beyond the confines of the HIPC Initiative to provide effective protection
from the claims of private capital market creditors, including negotiated
debt write-offs and standstill agreements.

Recruiting new gatekeepers

The IMF and the World Bank, with the former first among equals, remain
the gatekeepers to entry into the HIPC Initiative, and the arbiters of com-
pliance with conditions during the interim debt-relief period. There is
little evidence that the IMF in particular prioritises approaches to public
spending aimed at achieving poverty reduction goals. This suggests a
strong case for the involvement of other agencies, including specialised
UN bodies.

Strengthening and democratising Poverty Reduction Strategy Papers

Most PRSPs provide little more than cursory treatment of the links
between HIV/AIDS and poverty, and few set out clear estimates of the
resource implications of responding to the crisis. Such estimates would
help to clarify debt relief and aid needs. More generally, there is a need
for a renewed emphasis on costing poverty reduction goals, and on creat-
ing effective public expenditure management systems. Civil society could
– and should – make an important contribution to debates on the use of
debt relief to support HIV/AIDS strategies and wider poverty reduction
efforts.
350 Kevin Watkins and Juliana Amadi
Implementing an ‘Education for All’ action plan
Increased public investment in education is urgently needed in HIPC
countries, both to absorb the costs associated with HIV/AIDS, and to
support effective preventive work. In Apri1 2002, finance and develop-
ment ministers from rich countries finally agreed to support an action
plan aimed at getting all children into school by 2015. Implementation
will require US$4 billion a year, but no financing deal has yet been
agreed. This should be seen as a key element in the HIV/AIDS strategy.

Trade
In the area of trade, hypocrisy and double standards have characterised
the behaviour of industrialised countries towards the poor. Such behavi-
our has not only undermined the efforts of developing countries to reap
the rewards of trade for development, but has also undermined the credi-
bility of the multilateral trading system. To ensure a genuine partnership
in development in addition to the attainment of the Millennium Develop-
ment Goals, the cycle of broken promises must end. Otherwise, the Mil-
lennium Development Goals will join a growing list of failed development
strategies.
In the area of developing countries’ access to open markets, industri-
alised nations need to ensure that average tariffs for imports from devel-
oping countries are no higher than tariffs on imports from developed
countries. Tariffs in areas like agriculture and labour-intensive manufac-
tured goods that are of special interest to developing countries must be
scaled down. A moratorium on anti-dumping actions against developing
countries must be declared. In terms of market access for textiles and gar-
ments, industrialised countries should catch up within a year with the
schedule for phasing out the MFA quotas. They should eliminate tariffs
and quotas on all textile and garment exports from developing countries
by January 2004.
In the area of reduced agricultural protectionism, a comprehensive ban
on all agricultural export subsidies must be introduced in all industrialised
countries. The rights of developing countries to protect their agricultural
systems for food security reasons have to be acknowledged. The structure of
agricultural support to promote social and environmental objectives, includ-
ing a transition to less intensive agriculture, must be re-geared. Developed
countries must substantially reduce tariffs against developing-country agri-
cultural exports, including processed food products.
A better deal for the least-developed developing countries requires
industrialised countries to agree to provide duty-free and quota-free access
to all products exported from the least-developed developing countries.
In terms of special action for Africa, the New Partnership for Africa’s
Development (NEPAD) provides an important opportunity to strengthen
 Aid, trade and debt 351
Africa’s capacity to benefit from trade. Industrialised countries should
support it by substantially improving market access for all products exported
from sub-Saharan Africa. They should also convene an international con-
ference to explore strategies for tackling the crisis in commodity markets.
The system of global patent rules that safeguards public health in poor
countries – the TRIPS agreement – needs to be reviewed. The social and
development objectives of TRIPS should be paramount. Each provision of
TRIPS should be interpreted in this light. Health obligations should take
precedence over intellectual property rights. Nothing in the TRIPS agree-
ment should prevent countries from adopting measures to protect public
health. Governments should have an absolute right to introduce compul-
sory licences in order to meet pressing public health needs, and to import
patented drugs from the cheapest source. They should agree to an in-
depth review of the agreement from a health and development perspect-
ive, with a view to amending it in this light. The length and scope of
pharmaceutical patents should also be reduced.
In the area of aid and technical assistance, industrialised countries
must increase substantially the funding of trade-related technical assis-
tance and capacity-building to improve the participation of developing
countries in the WTO, and their ability to take advantage of new market
opportunities.
For the WTO to help create conditions for sustained growth and
poverty reduction in developing countries, future meetings need to focus
on areas of priority concerns to developing countries. The WTO can do
this by providing meaningful special and differential treatment for devel-
oping countries by extending transition periods for developing and least
developed countries to comply with TRIPS, in line with their achievement
of health and development milestones. Industrialised countries have to
give developing nations more flexibility in areas such as protection for
infant industries and regulation for foreign investment. There also needs
to be a review of the implications of TRIPS for access to technology by
developing countries.

Notes
1 However, developing country tariff barriers are on average higher than those of
developed countries. Developing country exports generally face higher barriers
in entering other developing countries than they do developed ones (World
Bank, 2001b).
2 This is based on data provided by the International Cotton Advisory Committee.
For a detailed analysis of the structure of US protectionism in cotton, see Oxfam
(2002c).
352 Kevin Watkins and Juliana Amadi
References
DFID (2000) Realising Human Rights for Poor People: Strategies for Achieving the Inter-
national Development Targets, London: Department for International Develop-
ment.
Hoekman, B., Ng, F. and Olarreaga, M. (2001) Tariff Peaks in the Quad and Least
Developed Country Exports, Development Research Group, World Bank, February,
at [Link]
International Fund for Agricultural Development (IFAD) (2001) Rural Poverty
Report 2001, Rome: IFAD.
International Monetary Fund (IMF) and International Development Agency
(IDA) (2002) Financial Impact of the HIPC Initiative: First 26 Country Cases, July
2002.
International Monetary Fund (IMF) and World Bank (2001) ‘HIPC Progress
Report’, mimeo, April, Washington, DC: World Bank.
International Monetary Fund (IMF) and World Bank (2002) ‘Market Access for
Developing Countries – Selected Issues’, paper prepared by staff of the
IMF/World Bank, 26 September.
International Textile and Clothing Bureau (2002) ‘Agreement on Textiles and
Clothing: Where It Stands’, mimeo, April, Geneva: ITCB.
Konandreas P. (2002) ‘Agriculture and agricultural policies in the developing
countries ten years after the Uruguay Round Agreement on Agriculture’,
mimeo. Paper presented to conference on agricultural policy reform and the
WTO: where are we heading? Capri (Italy), June 24–26.
Lanjouw, J. and Cockburn, I. (2001) ‘New Pills for Poor People? Empirical Evid-
ence after the GATT, World Development, 29(2), 265–289.
Maxwell, S. and Riddell, R. (1998) ‘Conditionality or Contract: Perspectives on
Partnership for Development’, Journal of International Development, 10(2):
257–268.
Organisation for Economic Cooperation and Development (OECD) (1996)
Shaping the Twenty-First Century, Paris: OECD.
—— (2001) Development Co-operation Report, Paris: OECD.
Oxfam (2001a) ‘Patent Injustice: How World Trade Rules Threaten the Health of
Poor People’, Oxfam Briefing Paper, 02/01, Oxford: Oxfam.
——(2001b) ‘Drugs Companies versus Brazil: the Threat to Public Health’, Oxfam
Briefing Paper, Oxford: Oxfam.
—— (2002a) ‘Rigged Rules and Double Standards: Trade, Globalisation and the
Fight Against Poverty’, Oxfam Make Trade Fair Briefing Paper, Oxford: Oxfam.
—— (2002b) ‘Last Chance in Monterrey: Meeting the Challenge of Poverty
Reduction’, Oxfam Briefing Paper 17, Oxford: Oxfam.
—— (2002c) ‘Cultivating Poverty – The Impact of US Cotton Subsidies on Africa’,
Make Trade Fair Briefing Paper 30, Oxford: Oxfam.
—— (2002d) ‘The Great EU Sugar Scam: How Europe’s Sugar Regime is Devastat-
ing Livelihoods in the Developing World’, Oxfam Briefing Paper 27, Oxford:
Oxfam.
—— (2002e) ‘Debt Relief and the HIV/AIDS Crisis in Africa: Does the Heavily
Indebted Poor Countries (HIPC) Initiative Go Far Enough?’, Briefing Paper 25,
Oxford: Oxfam.
 Aid, trade and debt 353
United Nations Children’s Fund (UNICEF) (2002) State of the World’s Children, New
York: UNICEF.
United Nations Development Programme (UNDP) (2002) Human Development
Report, New York: UNDP.
Watkins, K. (2001) ‘Pharmaceutical Patents’, in Bircham, E. and Charlton, J. (eds)
Anti-capitalism, Sydney: Bookmarks.
White, H. (1999) ‘Foreign Aid’, in Encyclopaedia of Political Economy, London: Rout-
ledge.
—— (2002) ‘Long Run Trends and Recent Developments in Official Assistance
from Donor Countries’, paper prepared for WIDER conference on external
finance, August 2002, Helsinki.
World Bank (2001a) Financial Impact of the HIPC Initiative: First 23 Country Cases,
June 2001, [Link] [Link].
—— (2001b) Globalization, Growth, and Poverty: Building an Inclusive World Economy,
New York: OUP for the World Bank.
World Health Organisation (WHO) (2001) Macroeconomics and Health: Investing in
Health for Economic Development. Report of the Commission on Macroeconomics
and Health, Geneva.
World Trade Organisation (WTO) (1997) Comprehensive and integrated WTO
plan of action for the Least-Developed countries, adopted on 13 December
1996, at [Link]
—— (2002) The Road to Doha and Beyond: A Road Map for Successfully Concluding
the Doha Development Round, at [Link]
roadtodoha_e.pdf.
Index

abortion 241 Barrick Gold 298

accountability 3, 12–13, 17, 48, 301; BDM International 298
asymmetric 40; reciprocal 40, 43 Bechtel 298
achievability/feasibility of targets 48, 53–5, Beijing Platform for Action 239, 242
124–35 benchmarks 101
Achieving Sustainability 293–5 benefit incidence studies 229–30
administrative systems 114–19; incentive Benin 177
issues 115–16 Black, R. xiv, 1–23
Afghanistan 269, 270 Blair, Tony 293, 301
Africa: and climate change 302–4; cobalt blood-buying networks 266
reserves 296; conflicts in 297–8; economic Booth, D. xiv, 77–95, 96–121
growth prospects 295; education in Botswana: economy 261; HIV/AIDS 267; life
sub-Saharan Africa 170–8; HIV/AIDS expectancy 258
262–3; infant and child mortality 209, 210, bottom-up systems 42–3, 56, 133
213, 226; market access 334–5; natural Brazil 336
resources 296–7; oil reserves 296; poverty bribery 304
reduction 26, 31; safe water and sanitation British Aid Statistics 63
290; sustainable development 292–305; budgetary spending 12, 98–9, 135–8
and trade 334–5 Burkina Faso 99, 101, 105, 343
agriculture 332–3, 350; cotton farming 333; business climate surveys 118
and poverty reduction 36, 37, 146, 153;
subsidies 332; technologies 153–4; value of Cairo Agenda 235, 236, 237, 238–41, 248–9,
aid 159–60 252–3
aid programmes: for agriculture 159–60; Cameroon 302
conditionality of aid 96–7; for education Cambodia 116, 266, 269, 270
179–80, 229; for HIV/AIDS 268–9; Canadian International Development
recipient compliance 314–15; technical Agency (CIDA) 11
assistance 337–8; trends in 327–8, see also Carlyle Group 298
donor agency performance measurement centrally planned economies 214–15, 222
AIDS see HIV/AIDS cervical cancer 249
Amadi, J. xiv, 323–51 Chambers, R. 147
Annual Review of Development Effectiveness Chevron 298
56–7 child malnutrition see nutrition
Asia: HIV/AIDS 263–4, 265; infant and child child mortality see infant and child mortality
mortality 209, 212; poverty reduction 26, 31 China 89, 90, 92; blood-buying networks
Assessing Aid (World Bank) 86 266; and debt relief 150; decline of
asymmetric accountability 40 poverty 155, 156; DFID programme focus
attribution 50, 55–6, 57 62; health services 215, 231; HIV/AIDS
averaging process 133–5 265–6; income poverty 126–7; infant and
child mortality 212, 221; land
balance of measures 47–8, 50–1 redistribution 153; safe water and
Bangladesh 89, 90, 92; education 133; risk of sanitation 289
poverty 148; sanitation strategies 286–7 citizen-based organisations 151
 Index 355
citizenship 33–5 Agreements 11, 58–62, 64–5, 68–73;
Clarke, P. xiv, 307–20 Service Delivery Agreements 58, 59–61,
climate change 293, 302–4 64–5, 68–73; and sustainability 69–71,
cobalt reserves 296 293–5, see also strategy papers
Colclough, C. xv, 166–81 Dhaka 247
Collier, P. 54 Djibouti 174
Colombia 134 Doha Round 338–9
community participation 106, 116–17, 282 Dollar, D. 54
conditionality of aid 96–7 dollar-a-day poverty 145, 146–7, see also
conflicts 71, 231, 297–8 poverty measures
Congo 296, 298 Dominican Republic 134
consistency of policy 315 donor agency performance measurement
contraceptive use 3, 243–4 47–74; achievability of targets 48, 53–5;
contracts 116 attribution 50, 55–6, 57; balance of
Copenhagen Social Summit 51, 145, 325 measures 47–8, 50–1; data
Core Welfare Indicators Questionnaires availability/quality 52–3, 63–4; definition
(CWIQs) 118, 158 of targets 51–3; desirable features of 47–8;
corruption 3, 304 International Development Targets 50–8;
cost of basic needs (CBN) indicator 147 logic models 49–50; management
costs of Millennium Development Goals information systems 56–7; Millennium
135–8, 346–7 Development Goals 50–8; and
Cotonou Convention 40 organisational culture 49; relevance of
cotton farming 333 measures 47–8, 50–1; using outcome
Country Strategy Papers (CSPs) 58–9, 62–3, measures 48–50, see also progress
77–88; consultations 79–80; debt relief 86; monitoring
‘missing middle’ 82–3; outcome targets donor coordination 311
83; partnerships and process 78–80, 85; drinking water see water access
poverty analysis 80–2; poverty definition droughts 303
80–1; programme design 86; sections 78 drugs companies 336
Cuba: health services 215, 231; infant and Dublin Principles 284
child mortality 212
Eastern Europe 212
data collection: commissioned studies 118; economic well-being 4
death rates 108; education 168–9; final economies: of Africa 295; Botswana 261;
outcome data 108–9; HIV/AIDS 261; centrally planned 214–15, 222; and
household data 101, 109, 112–13, 158; HIV/AIDS 261; macroeconomic policy
and performance monitoring 52–3, 63–4, 36–8; pro-poor growth 86–7
100; and progress monitoring 104–5, education 4, 5, 7, 124, 133–4, 166–81; aid
105–6, 111; reproductive health 242, 250; programmes 179–80; in Bangladesh 133;
water access 276, see also surveys completion targets 168; data availability
death rates: data collection 108; from 168–9; demand-side issues 176–8; DFID
malnutrition 130; maternal mortality 53, targets 72; and fertility rates 167; and
114, 132, 242–3; registration of deaths gender equality 168, 169–70, 184, 188,
243, see also infant and child mortality 195–6; gendered benefits of 166–7; gross
debt relief 86, 137–8, 149–50, 339–45, enrolment ratios (GERs) 167–8, 169–70,
347–50; and government revenue 342–4; 174; and HIV/AIDS 131; literacy targets
and HIV/AIDS 342–3, 349; levels of 341–2 168; in Peru 133–4; price elasticity of
debt servicing 342 demand 177; progress monitoring 127–8,
debt sustainability 9, 344–5 139; public spending on 171, 174; school
Democratic Republic of Congo (DRC) 296, quality 176; social returns 166; in
298 sub-Saharan Africa 170–8; supply-side
Demographic and Health Surveys (DHSs) issues 171–4; in Swaziland 260; teacher
107, 129, 133–4 salaries 115, 175; in Uganda 230; unit
denominator problem 106 costs of schooling 174–5, 176, 179; and
Department for International Development women 166–7, 178
(DFID): Achieving Sustainability 293–5; Egypt, infant and child mortality 224, 225
education targets 72; health targets 72–3; Elson, D. 191
partnership discourse 310–18; poverty empowerment of women 28, 184, 189,
reduction targets 68–9; Public Service 197–8, 241
356 Index
Engendering Development (World Bank) 198 72–3; in Ghana 215, 230, 231; in Malawi
England 220 215, 231; maternal health 188–9, 242–4; in
Enhanced Heavily Indebted Poor Country Nicaragua 215, 231; patented medicines
Initiative 339–45, 347–50 336–6; progress monitoring 115;
equality see gender equality reproductive health services 4, 5; in
Ethiopia 174 Uganda 246–7; in Vietnam 215, 231; in
Europe 220, 221 Zambia 215, 231; in Zimbabwe 215, 231,
exit polls 100, 118 see also reproductive health
expenditure surveys 101, 109, 112–13, 158 Heavily Indebted Poor Country (HIPC)
Initiative 149–50, 339–45, 347–50
Fairhead, J. xv, 292–305 HIV/AIDS 3, 5, 7–8, 54, 124–5, 129, 256–70,
family planning programmes 236, 238–9; 348, 349–50; achievability of targets
contraceptive use 3, 243–4 262–70; in Africa 262–3; aid programmes
feasibility/achievability of targets 48, 53–5, 268–9; in Asia 263–4, 265; in Botswana
124–35 267; in Cambodia 266, 269, 270; in China
feedback mechanisms 100–1 265–6; data availability 261; and debt
fertility rates and education 167 relief 342–3, 349; and economic growth
focus groups 100, 118 261; and education 131; geopolitical issues
Food Economy Zones 114 269–70; and household economies 261; in
food energy method (FEM) 147–8, 149 India 265–6, 267; and infant and child
forests 300, 302–3 mortality 257; and international
Foster, M. 39 peacekeeping forces 269–70; and life
expectancy 258, 259; in Malawi 343;
Gambia 101, 112, 115 Millennium Development Goals 256–70;
GATS (General Agreement on Trade in and political commitment 267–8; and
Services) 339 population migration 266–7; and poverty
gender disaggregation 191 reduction 260–1; progress monitoring
gender discrimination 128 130–1, 139; scope and prevalence of
gender equality 4, 5, 7, 184–203; economic 256–7; in South Africa 268; and teachers0
participation of women 188; and 131; in Thailand 263–4, 265; in Uganda
education 168, 169–70, 184, 188, 195–6; 131, 262, 264; in Zambia 262–3
mainstreaming gender 198–202; and Household Responsibility System 153
poverty 189–90; progress in achieving household survey data 101, 109, 112–13, 158
192–5; regional achievements 193, 194; Human Development Reports (HDRs) 28, 107
reversibility of gains 194 human rights 313
genetic modification (GM) 154 hygiene 273, 274, 286–7; hand-washing
geographical targeting 113–14 274–5; Millennium Development Goals
Ghana 89, 90, 92; forests 302–3; health 275–6; prospects for 288–90, see also
services 215, 230, 231; infant and child sanitation
mortality 134, 226; maternal mortality 53;
poverty predictors 158 IMF (International Monetary Fund) 2
Gini coefficient target 35 immunisation coverage 129, 210, 212, 218,
global partnerships see partnerships 226, 228–9
Global Trends (NIC) 295–6 implementation monitoring 100
Goudie, A. 310, 311, 313 incentive issues 115–16
governance indicators 3, 63 incomes: income poverty 107–8, 125–7; and
Government Performance and Results Act infant and child mortality 213, 218, 220;
49 nurses salaries 115; in sub-Saharan Africa
Grant, J. 287–8 177; teacher salaries 115, 175
gross enrolment ratios (GERs) 167–8, India 89, 90, 92; debt relief 150; HIV/AIDS
169–70, 174 265–6, 267; infant and child mortality
225–6; poverty incidence 148, 157; safe
Halliburton 298 water and sanitation 289
hand pump maintenance 285 indicator choice 103–7
hand-washing 274–5 Indonesia 134
Hanmer, L. 215 infant and child mortality 134–5, 209–31; in
Hauge, A. 110 Africa 209, 210, 213, 226; causes of death
health services: in Cambodia 116; in China 129, 227; and centrally planned
215, 231; in Cuba 215, 231; DFID targets economies 214–15, 222; in China 212,
 Index 357
221; cross-country regressions 215, 220; in Living Standards Survey 158
Cuba 212; definitions 51–2; in East Asia local content requirements 339
209, 212; in Eastern Europe 212; in Egypt logic models 49–50
224, 225; in England 220; in Europe 220, Lomé Convention 309
221; in Ghana 134, 226; and hand-washing Lucas, H. xvi, 96–121
274–5; and HIV/AIDS 257; and
immunisation coverage 129, 210, 212, 218, McKeown, T. 220
226, 228–9; and incomes 213, 218, 220; in macroeconomic policy 36–8
India 225–6; indicators 5; and infectious mainstreaming gender 198–202
diseases 228; International Development malaria deaths 222–3
Targets 4; and intestinal diseases 275; Malawi 83, 84, 89, 90, 93; debt servicing 342;
malaria deaths 222–3; and maternal health services 215, 231; HIV/AIDS
education 220; Millennium Development strategic plan 343; school enrolments 177;
Goals 7, 209, 229; Mosley-Chen analysis tuberculosis 257–8
219; neo-natal deaths 225–6; in Oman malnutrition see nutrition
212; in Ovamboland 222; patterns of management information systems 56–7
deaths 223–6; policy recommendations Mar del Plata Declaration 274
226–31; post-natal infants 225; progress market access 9, 329–32, 334–5
monitoring 129, 134–5, 139; reasons for Marshall Plan 323
decline in 209–10, 214–26; record of maternal education 220
mortality decline 210–10; regional maternal health 188–9, 242–4, see also
performance 212–13; robust regression reproductive health
analysis 215–19; in Senegal 222; in the maternal mortality 53, 114, 132, 242–3
Soviet Union 221–2; under-five children Mauritania 99
51–2, 124, 129, 134–5, 210, 214, 223–4; in Maxwell, S. xvi, 25–44
Uzbekistan 224, 225; in Vietnam 214; in measles 129
Zambia 115, 224, 225; in Zimbabwe 129, Medium Term Expenditure Frameworks
226, see also death rates (MTEF) 29, 99
infectious diseases 228 Millennium Summit xviii, 1, 6
input monitoring 98–9, 110–11 mining corporations 298
integrity surveys 118 Monterrey 325
intellectual property rights 335–7, 339 mortality see death rates
intensity of poverty 148 Moser, C. 35
International Development Statistics 63 Mosley-Chen analysis 219
International Development Targets 1–2, Mozambique 89, 91, 93, 174, 335
3–6, 50–8, 293–4 multidimensionality 107–8, 148–9
international peacekeeping forces 269–70
International Strategy Papers (ISPs) 58–9 National Intelligence Council (NIC) 295–6
intestinal diseases 275 natural resources 296–7
neo-natal deaths 225–6
Jolly, R. xv, 273–90 Nepal 89, 91, 93
New Plant Type 153
Kabeer, N. 197–8 Nicaraguan health services 215, 231
Kenya 49, 79, 89, 90, 92, 100 Nigeria 115
Norton, A. 35
land redistribution 153 nurses salaries 115
land reform 145 Nussbaum, Martha 192
latrines 286 nutrition 39, 51; cost of basic needs (CBN)
league tables 11, 14, 15 indicator 147; food energy method (FEM)
least developed countries (LDCs) 333–4 147–8, 149; malnutrition progress
legitimation 312–13 monitoring 130
liberalisation of trade 294, 329, 330, 331
Liberia 299 oil reserves 296
life expectancy 258, 259, see also death rates Oman 212
lifespan approach to reproductive health organisational culture 49
241 outcome measures 48–50
Lipton, M. xvi, 145–61 outcome oriented approach 97–8, 99, 107–9,
literacy targets 168 112–14
livelihood strategies 116 outcome targets 83
358 Index
Ovamboland 222 Poverty Reduction Strategy Credits (PRSC)
ownership of policies 310–11 29
Oxfam International 38 Poverty Reduction Strategy Paper (PRSP)
xviii, 29, 36, 78–9, 96–111, 145–6, 149–51,
parliamentary participation of women 196 349–50; budget system 98–9; and gender
Participatory Poverty Assessments (PPAs) equality 200; indicator choice 103–7;
35–6, 114 information access 101–2; input
partnerships 39–40, 41, 78–80, 85, 307–20, monitoring 98–9; Interim-PRSP 99, 100;
324–5; donor coordination 311 ‘missing middle’ problem 103; policy
patent protection 335–7, 351 process 96–9; sector policies 150
Pearson Commission 308 poverty targeting 113–14
peer reviews 40 price elasticity of education demand 177
performance measurement see donor agency pricing policy for water access 283–5
performance measurement primary schooling see education
performance targets 12–16 PRISM 62, 65
Pergau Dam 10 private sector and water access 282–3
Peru 133–4 privatisation 282
Philippines 134 pro-poor growth 86–7
policy implementation monitoring 97, 100 process conditionality 97
policy ownership 310–11 process monitoring 114–19
population migration 266–7 progress monitoring 51–3, 96–121, 138–40;
post-natal infant deaths 225 child malnutrition 130; child mortality
Poverty Aim Marker (PAM) 82 129, 134–5, 139; and community
poverty funds 86 involvement 106; denominator problem
poverty gap index 148 106; education 127–8, 139; feedback
poverty measures 80–1, 89–94, 146–9; mechanisms 100–1; geographical
dollar-a-day poverty 145, 146–7; intensity targeting 113–14; health monitoring 115;
of poverty 148; multidimensional poverty HIV/AIDS 130–1, 139; implementation
148–9; purchasing power parity (PPP) monitoring 100; income poverty 125–7;
127, 146–8, 154; severity of poverty 148 indicator choice 103–7; input monitoring
poverty profiles 109 98–9, 110–11; maternal health 242–4;
poverty reduction 10, 25–44, 145–61; and maternal mortality 132; outcome oriented
agricultural growth 36, 37, 146, 153, approach 97–8, 99, 107–9, 112–14; process
159–60; analysis of poverty 80–2; Asia and monitoring 114–19; Public Expenditure
the Pacific 26, 31; bottom-up approach Tracking (PET) Studies 99–100; water
42–3; and citizenship 33–5; DFID targets access 132–3, 139, 281, see also donor
68–9; feasibility of targets 151–7; and agency performance measurement
gender equality 189–90; Gini coefficient project monitoring 11
target 35; and HIV/AIDS 260–1; Human protectionism 293, 330
Development Reports (HDRs) 28; public expenditure tracking (PET) studies
identification and use of targets 12, 30–3; 99–100, 229, 230
International Development Targets 3, 4, 5; Public Service Agreements (PSA) 11, 58,
and macroeconomic policy 36–8; Medium 59–62, 64–5, 68–73
Term Expenditure Frameworks (MTEF) purchasing power parity (PPP) 127, 146–8,
29; Millennium Development Goals 7, 154
26–7; Participatory Poverty Assessments
(PPAs) 35–6, 114; partnerships for 39–40, recipient compliance 314–15
41; and politics 43; Poverty Reduction reciprocal accountability 40, 43
Strategy Credits (PRSC) 29; process refugee camps 231
conditionality 97; results-based registration of deaths 243
management 30; rights-based approaches relevance of measures 47–8, 50–1
33–5; risks of new construction approaches reproductive health 4, 5, 235–53; access to
30–40; rural poverty 151–3, 160–1; services 247–8; Cairo Agenda 235, 236,
sector-wide approaches (SWAPs) 29, 36, 237, 238–41, 248–9, 252–3; cervical cancer
38–9, 87–8; sub-Saharan Africa 26, 31; and 249; contraceptive use 4, 243–4; data
trade policy 37–8; United Nations targets collection/availability 242, 250; family
145–61; urban poverty 152; value of aid planning programmes 236, 238–9;
159–60; World Development Report strategy financing sources 244–7; future of 250–1;
27–9 International Development Targets 235–6;
 Index 359
lifespan approach 241; Millennium school 111; service delivery 117–18, see also
Development Goals 235–6, 237, 252; data collection
progress on 242–4; quality of services sustainable development 3, 4, 5, 8, 292–305;
247–8; Safe Motherhood programmes and climate change 302–4; DFID targets
245; services 240–1; in Tanzania 246 69–71; and political economy 295–301
reproductive rights 239, 240 Swaziland 260
results-based management 30
results-based management (RBM) 10–11, 56 Tanzania 89, 91, 93; participatory
Rice, Susan 296–7 monitoring 101; Public Expenditure
Rights of the Child (UNICEF) 274 Reviews (PER) 99; reproductive health
rights-based systems 14, 16–17, 33–5 246; school enrolments 117–8
rural development 160–1, 285 target setting 12–16
rural poverty 151–3, 160–1 Target Strategy Papers (TSPs) 58–9
Rural Poverty Report (IFAD) 145, 160–1 tariffs 330, 350
Russia 54 teachers: and HIV/AIDS 131; salaries 115,
175
Safe Motherhood programmes 245 technical assistance 337–8
safe water see water access ten commandments 287–8
salaries 115–16, 175 textile and clothing industries 330–1
sanitation 273, 274–5, 286–7; in Bangladesh Thailand, HIV/AIDS 263–4, 265
286–7; latrines 286; Millennium timber concessions 299, 300
Development Goals 275–6; prospects for top-down systems 56, 133
288–90; regional statistics 277–80; in trade 328–37, 350–1; and Africa 334–5; and
South Africa 281, see also hygiene least developed countries (LDCs) 333–4;
schools see education liberalisation policies 294, 329, 330, 331;
sector-wide approaches (SWAPs) 29, 36, market access 9, 329–32, 334–5; patent
38–9, 79, 87–8 protection 335–7, 351; and poverty
Sen, Amartya 192 reduction 37–8; protectionism 293, 330;
Senegal 101, 222 tariffs 330, 350; textile and clothing
Service Delivery Agreements (SDA) 58, industries 330–1; Uruguay Round 329,
59–61, 64–5, 68–73 330, 337, see also agriculture
service delivery surveys 117–18 TRIMS (Trade-Related Investment
severity of poverty 148 Measures) 339
Shaping the Twenty-First Century (OECD) 1 TRIPS (Trade-Related Aspects of
Short, Clare 58 Intellectual Property Rights) 335, 339, 351
Skeldon, R. xvi–xvii, 256–70 tuberculosis 257–8
social development 4
social returns from education 166 Uganda 89, 91, 94, 99, 102, 109–10;
Social Summit 51, 145, 325 education spending 230; health services
South Africa 81, 89, 91, 93; HIV/AIDS 268; 246–7; HIV/AIDS 131, 262, 264; poverty
safe drinking water programme 280–1; reduction 6, 99, 108; school surveys 111;
and sanitation 281 surveys 118
Soviet Union 221–2 under-five mortality 51–2, 124, 129, 134–5,
Standing, H. xvii, 235–53 210, 214, 223–4, see also infant and child
sterilisation 244 mortality
strategy papers: Achieving Sustainability UNICEF 274, 287
293–5; Country Strategy Papers (CSPs) United Nations (UN): poverty reduction
58–9, 62–3, 77–88; International Strategy targets 145–61; Social Summit 51, 145, 325
Papers (ISPs) 58–9; Target Strategy Papers Universal Declaration of Human Rights 166
(TSPs) 58–9, see also Poverty Reduction urban poverty 152
Strategy Paper (PRSP) Uruguay Round 329, 330, 337
sub-Saharan Africa see Africa USAID programmes 55–6, 64
Subrahmanian, R. xvii, 184–203 Uzbekistan 224, 225
subsidiarity 42
subsidies 332 value for money 73
surveys 117–19; business climate 118; Vandemoortele, J. xvii, 124–40
demographic and health 107, 129, 133–4; ventilated improved pit latrines (VIP) 286
household survey data 101, 109, 112–13, Vietnam 89, 91, 94; health services 215, 231;
158; integrity 118; living standards 158; infant and child mortality 214
360 Index
Vinnell Corporation 298 education 166–7, 178, 220; empowerment
of 28, 184, 189, 197–8, 241; and hand
Waddington, H. xvii, 145–61 pump maintenance 285; involvement in
wages 115–16, 175 water access programmes 285; maternal
wars see conflicts education 220; maternal health 188–9,
water access 124, 273–90; in Africa 290; in 242–4; maternal mortality 53, 114, 132,
China 289; community action approaches 242–3; parliamentary participation 196;
282; cost recovery systems 284–5; costs reproductive rights 239, 240; Safe
283; data quality 276; distance from Motherhood programmes 245;
supplies 277; Dublin Principles 284; in sterilisation 244, see also gender equality;
India 289; Millennium Development reproductive health
Goals 274–5; monitoring results 283; and World Bank 2, 86; Poverty Assessments
population growth 280; present position 200–1; rating system 56–7
276–7; pricing policy 283–5; private sector World Development Indicators (World Bank) 2
involvement 282–3; progress monitoring World Development Report strategy 27–9
132–3, 139, 281; prospects for 288–90; World Trade Organisation (WTO) 329,
regional statistics 277–80; rural area 338–9
programmes 285; social mobilisation
approaches 282; in South Africa 280–1; Zambia 62, 118, 174, 319; debt servicing 342;
technology used 283; women involvement and education 174; health services 215,
in programmes 285 231; HIV/AIDS 262–3; infant and child
Watkins, K. 323–51 mortality 224, 225; infant mortality 115
White, H. xvii, 1–23, 47–74, 77–95, 209–31 Zimbabwe 89, 91, 94; child mortality 129,
women: Beijing Platform for Action 239, 134; health services 215, 231; infant and
242; economic participation 188; and child mortality 226