Global Monitoring Report 2004

GLOBALMONITORING REPORT 2004

Policies and Actions for Achieving the Millennium Development Goals and Related Outcomes

GLOBALMONITORING REPORT 2004

Policies and Actions for Achieving the Millennium Development Goals and Related Outcomes

© 2004 The International Bank for Reconstruction and Development / The World Bank1818 H Street, NWWashington, DC 20433Telephone: 202-473-1000Internet: www.worldbank.orgE-mail: [email protected]

All rights reserved.

1 2 3 4 07 06 05 04

This volume is a product of the staff of the World Bank and the International Monetary Fund. Thefindings, interpretations, and conclusions expressed herein do not necessarily reflect the views of theBoard of Executive Directors of the World Bank, the Board of Executive Directors of the InternationalMonetary Fund, or the governments they represent.

The World Bank and the International Monetary Fund do not guarantee the accuracy of the dataincluded in this work. The boundaries, colors, denominations, and other information shown on anymap in this work do not imply any judgment on the part of the World Bank or the InternationalMonetary Fund concerning the legal status of any territory or the endorsement or acceptance of suchboundaries.

Rights and PermissionsThe material in this work is copyrighted. Copying and/or transmitting portions or all of this workwithout permission may be a violation of applicable law. The World Bank encourages dissemination of its work and will normally grant permission promptly.

For permission to photocopy or reprint any part of this work, please send a request with completeinformation to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA,telephone 978-750-8400, fax 978-750-4470, www.copyright.com.

All other queries on rights and licenses, including subsidiary rights, should be addressed to the Officeof the Publisher, World Bank, 1818 H Street, NW, Washington, DC 20433, USA, fax 202-522-2422,e-mail [email protected].

ISBN 0-8213-5859-6

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 iii

Foreword ix

Acknowledgments xi

Abbreviations and Acronyms xiii

Executive Summary xvii

Millennium Development Goals (MDGs) xxii

Overview: From Vision to Action 1

Framework 19

1 Monitoring Framework 21

2 MDG Prospects: Reasons for Optimism, Grave Concerns 33

Developing-Country Policies 49

3 Overall Picture 51

4 Improving Enabling Climate for Growth: Economic and Financial Policies 57

5 Upgrading Public Sector Governance 81

6 Strengthening Infrastructure 93

7 Accelerating Human Development 107

8 Promoting Environmental Sustainability 125

Contents

IP A R T

IIP A R T

C O N T E N T S

iv G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

IIIP A R T

IVP A R T

Developed-Country Policies 131

9 Fostering Growth and Stability: Macro-financial Policies 133

10 Dismantling Barriers to Trade 143

11 Providing More and Better Aid 163

12 Fulfilling Responsibilities for Global Public Goods 181

Role of International Financial Institutions 189

13 Monitoring the IFIs’ Contribution 191

References 217

BoxesMillennium Development Goals xxii

1.1 An action plan for improving development statistics 241.2 Strengthening the links between PRSPs and the MDGs 302.1 East Asia and Pacific: Despite solid performance on MDGs,

challenges remain 385.1 Improving fiscal transparency through ROSCs 845.2 The African Peer Review Mechanism: self-assessing governance 895.3 Governance in Africa—progress on a difficult agenda 906.1 Water supply and sanitation in the MDGs 1047.1 Rwanda: HIV/AIDS and health expenditures 1127.2 The case of the missing money: monitoring public expenditure 1157.3 The Bangladesh Female Secondary School Assistance Program 1178.1 Multisectoral interventions to achieve the MDGs: lessons from child

mortality in rural India 1288.2 The United Nations Task Force on Environmental Sustainability 1289.1 Differences between remittances and capital flows 13910.1 The EU’s Common Agricultural Policy reform 14910.2 Lessons from Integrated Framework diagnostic trade

integration studies 15811.1 Estimating the cost of the MDGs 16711.2 Measuring aid selectivity 17111.3 Vietnam’s comprehensive government-led harmonization program 17612.1 International Task Force on Global Public Goods 182

12.2 Progress toward environmental sustainability through performance review and peer pressure 183

12.3 International environmental agreements: toward global cooperation,with some notable exceptions 184

13.1 Framework for assessing IFI contributions 19213.2 Results-based CAS 20013.3 Proposed IDA14 results-measurement system 20613.4 Sectorwide approach to primary education development 209

Figures1.1 Framework linking policies and actions with development outcomes 251.2 Monitoring: dimensions of developing-country policies 271.3 Monitoring: dimensions of developed-country policies 281.4 Monitoring: dimensions of development agency support 302.1 Growth prospects improve, but not enough 352.2 Most regions will reach the goal of halving poverty by 2015,

but Sub-Saharan Africa is seriously off track 372.3 Mortality at a given level of national income has been declining 402.4 A few regions are close to the target on primary education:

others are off track 412.5 Prospects for reaching the child mortality goal are dim 432.6 Reform combined with stronger partner support can substantially

boost prospects for achieving the MDGs 463.1 Developing countries’ policies have improved; governance and

institutions lag 523.2 Other ratings corroborate that developing-country policies

have improved but that governance and institutions lag 543.3 The transition countries are making broad progress

in removing obstacles to business 553.4 The developing-country policy diamond shows progress,

but much more is needed 554.1 Faster growing countries typically have better macroeconomic policies 594.2 Better-off countries tend to restrict trade less 654.3 Overall policy on trade in services remains more restrictive

in developing countries 664.4 Heavy regulation is associated with lower productivity 674.5 Heavier regulation contributes to the informal economy

and corruption 684.6 More regulation does not necessarily produce better social outcomes 684.7 Protecting property rights is associated with more credit 694.8 Poor countries regulate the most 704.9 Low-income countries lag far behind best practice

in promoting business 714.10 Low-income countries lag the most in property rights and rule of law 724.11 Countries are improving their private business environment 73

C O N T E N T S

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 v

C O N T E N T S

vi G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

4.12 The financial sector is deepening in the developing world, but at a varying pace 75

4.13 Deposit growth has not been equally reflected in growth in private sector credit 76

4.14 Financial system strength is typically positively correlated with compliance with Basel Core Principles 77

4.15 Capital markets are shallow in low-income countries 785.1 The quality of governance is weak but improving

in developing countries 875.2 Civil liberties are gradually improving in developing countries 885.3 Participatory processes are also improving in developing countries 886.1 Gaps in infrastructure call for significantly increased spending,

which must be managed well for effectiveness 1017.1 Investment in human capital is up, but more is needed 1097.2 Developing countries are allocating more public spending to human

development 1107.3 Public spending covers more of the cost of health care in high-income

countries than in low-income countries 1117.4 Public spending on human development often benefits the rich more

than the poor 1147.5 Teachers’ salaries absorb most recurrent education spending 1168.1 Environmental policy ratings are low but improving 1269.1 A robust global economy requires orderly resolution of the large

external and fiscal imbalances 1359.2 Low-income countries receive little foreign direct investment 1379.3 Private capital flows to developing countries are recovering,

led by debt flows 13710.1 Potential income gains from trade reforms are large and can help

reduce poverty 14510.2 Escalating tariff rates discourage development 14710.3 Protection in agriculture is high—a multiple of that in manufacturing 15310.4 Is overall trade policy pro development? Mixed picture 15410.5 Potential gains from liberalization of services, especially migration,

are large 15711.1 Aid is rising but is well short of what is needed 16411.2 The increase in ODA in 2002 was concentrated

in special-purpose grants 16511.3 The proportion of aid provided in cash and more flexible forms

should be rising, not falling 16911.4 Institutions and policies matter for aid effectiveness 16911.5 More selective donors provide more aid per capita to countries

with stronger policies and institutions 17011.6 Aid fragmentation is high 17211.7 Aid flows are typically more volatile than fiscal revenues

in aid-dependent countries 17312.1 MDGs and Kyoto Protocol call for reduction of greenhouse

emissions, but results tell a different story 185

C O N T E N T S

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 vii

Tables2.1 Growth rates and decline in poverty by region, through 2015 362.2 Projected per capita growth and improvement in poverty

in Sub-Saharan Africa, 2000–2015 392.3 Primary education completion rates, progress needed by region 412.4 Distance from the goal of gender parity in primary and secondary

education, by region, circa 2000 424.1 Macroeconomic indicators for low-income countries,

by region, 1983–2008 584.2 Quality of macroeconomic policies for low-income countries,

by country characteristics 604.3 Macroeconomic indicators for middle-income countries, 1983–2008 614.4 Impact of financial crises on poverty, East Asia, 1997–98 624.5 Decline in tariffs in developing countries, late 1980s to 2003 634.6 Decline in core nontariff barriers in developing countries,

1989–94 to 2000 634.7 Major users of antidumping, developing countries, 1995–2002 644.8 Overall trade restrictiveness of developing-country groups, 2001 644.9 Investment climate, selected low- and middle-income countries, 2002 724.10 Summary of lessons learned from corporate governance assessments 744.11 Evolution of selected financial soundness indicators, 1998–2002 765.1 Improvement in public financial management in developing countries,

1999–2003 825.2 Public financial management benchmarks used in HIPC assessments 835.3 Central government tax revenue, 1990–2001 855.4 Quality of public sector governance, 1999–2003 865.5 Worldwide governance indicators, 1998–2002 876.1 Distribution of studies according to their findings on impact

of infrastructure investment on productivity or growth 946.2 Access of population to infrastructure services 966.3 Technical quality of infrastructure services 976.4 Quality of infrastructure services as perceived by commercial users 986.5 Electricity reform in selected countries, 2003 996.6 Expected annual needs for new investment and maintenance

in infrastructure, 2005–10 1006.7 Private commitments for infrastructure, 1990–2002 1027.1 Private health expenditure 1117.2 Public spending on social protection for selected countries by region 1137.3 Frequency of attended births by wealth and region, various years 1198.1 Selected outcome indicators of environmental sustainability

by region and country income category 1279.1 Macroeconomic indicators for advanced economies, 1993–2008 1349.2 Global economic environment and developing countries 1369.3 Remittances to developing countries, by region, 2001–03 13810.1 MFN tariffs, developed countries, 1990 and 2002 14610.2 Simple average NTB coverage ratios, 2001 147

C O N T E N T S

viii G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

10.3 Agricultural support, 1986–88 and 2000–02 14810.4 Trade shares of products affected by agricultural subsidies 15010.5 Antidumping investigations by selected OECD members, 1995–2002 15110.6 Average tariff imposed in final U.S. antidumping duty determinations 15210.7 Overall trade restrictiveness and import shares, high-income OECD

and Quad, 2001 15211.1 Actual ODA and post-Monterrey commitments (DAC donors),

2002 and 2006 16511.2 Framework of indicators on progress on harmonization and alignment

(provisional) 17511.3 Debt service and poverty-reducing expenditure by the 27 HIPCs that

have reached the decision point 17712.1 Aid for the environment, 1990–2000 18612.2 Performance of developed countries on global environmental

sustainability 186

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 ix

This Global Monitoring Report is thefirst in a planned series of annualreports assessing the implementation

of policies and actions for achieving the Mil-lennium Development Goals (MDGs) andrelated outcomes. These reports will under-pin regular monitoring by the DevelopmentCommittee, the joint Ministerial body of ourtwo institutions, of progress on the policyagenda; and they will aid the DevelopmentCommittee in setting priorities for actionand defining the accountabilities of the keyactors—developing and developed countriesand multilateral institutions. The GlobalMonitoring Report is being published in orderto disseminate its findings and messages morewidely. We expect to publish these reportseach year and to maintain the momentumtoward achieving the MDGs by 2015.

This report has been prepared jointly by thestaff of the World Bank and the InternationalMonetary Fund, in close collaboration withpartner agencies—other multilateral develop-ment banks, the United Nations, World TradeOrganization, Organisation for Economic Co-operation and Development and its Develop-ment Assistance Committee, and the EuropeanCommission. It is particularly encouraging thatwe now have such a broad partnership sup-porting this effort. We hope that this first report

Foreword

will spur further interaction on policies andactions to meet the agreed development goals.

The findings of the report provide asobering assessment of progress toward theMDGs, and of progress in meeting the com-mitments made at Monterrey. As the reportmakes clear, on current trends, most MDGswill not be met by most countries. This isnot to downplay the impressive progressmade in some countries, particularly largeones such as China and India. But there aremany that are falling behind, especially inSub-Saharan Africa. And we can do bettereven in countries where progress has beengood. There is an urgent need to scale upactions, building on the foundations laid bypast successes and lessons learned from pastmistakes. The report identifies where theneed for action is most critical.

The report provides a range of evidence onthe improvement of policies and institutionsin developing countries. We have seen evi-dence of this improvement directly, throughour own visits and interactions with people incountries in all parts of the developing world.In Africa, for instance, the New Partnershipfor Africa’s Development has asserted a newaccountability, and there are now manycountries that have taken credible steps tostrengthen their policies and governance with

F O R E W O R D

x G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

a clear sense that there is now an unprece-dented opportunity for accelerating progresson development results. More needs to bedone to deepen and sustain reforms, but webelieve that we have the basis for scaling upin many countries.

As the report shows, in contrast to theimproving performance in many developingcountries, the developed countries are fallingbehind on two critical fronts that are vital toaccelerating progress and scaling up results—trade and aid. We cannot overemphasize theimportance and urgency of reinvigoratingthe Doha Round, including agreement onimproved market access for agriculture andlabor-intensive manufactures that are socritical to the prospects of the poorest coun-tries. The other area for priority attention isthe quantity and quality of aid. There is aclear need to increase the volume of devel-opment assistance in support of countriesimplementing sound macroeconomic, struc-tural, and institutional policies to accelerateprogress toward the MDGs. In parallel, weneed political commitment on the agenda toimprove the quality of aid—to align aid bet-

ter with country-owned priorities, to makeit more predictable and flexible, to focus iton results, and to harmonize aid practicesand procedures.

The report also highlights the need forcontinued attention to how internationalagencies, including our own, can strengthentheir support at the country, regional, andglobal levels. There is clear recognition of theimportance of strong results orientation inwhat we do individually and collectively.Refining and strengthening institutional rolesin low-income countries, including deepen-ing the process of preparation and imple-mentation of Poverty Reduction StrategyPapers, is a priority area for attention. At thesame time, we need to continue to adaptapproaches and instruments to the evolvingneeds of middle-income countries.

In sum, this first Global Monitoring Reportis, in our view, an important and credible stepin putting in place a framework for results andaccountability that will enhance the inter-national development community’s review ofprogress on and priorities for achieving globaldevelopment goals.

James D. WolfensohnPresidentWorld Bank

Rodrigo de Rato Managing DirectorInternational Monetary Fund

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 xi

This report has been prepared jointly bythe staff of the World Bank and theInternational Monetary Fund. In

preparing the report, staff have collaboratedclosely with partner institutions—othermultilateral development banks, the UnitedNations, World Trade Organization, Organ-isation for Economic Co-operation andDevelopment and its Development AssistanceCommittee, and the European Commission.The cooperation and support of staff of theseinstitutions are gratefully acknowledged.

Zia Qureshi was the lead author and man-ager of the report. The work was carried outunder the general guidance of ShengmanZhang, Managing Director, World Bank. TheBank team of contributors included ShaidaBadiee, Gilles Bauche, Gordon Betcherman,Milan Brahmbhatt, Cecila Briceno-Garmen-dia, Barbara Bruns, Punam Chuhan, MansoorDailami, Anis Dani, Shantayanan Devarajan,Simeon Djankov, David Dollar, Mark Dorf-man, William Dorotinsky, Poul Engberg-Ped-ersen, Antonio Estache, Manuel Felix, DeonFilmer, Lucia Fort, Engilbert Gudmundsson,Christopher Hall, Jonathan Halpern, KirkHamilton, James Hanson, Bernard Hoekman,

Acknowledgments

Mary Hollifield, Philip Keefer, Steve Knack,Victoria Levin, Maureen Lewis, Pilar Maistera,Tamar Manuelyan Atinc, Roberto Martin-Hurtado, Caralee McLiesh, Aaditya Mattoo,Marcelo Olarreaga, John Panzer, Axel Peuker,Dilip Ratha, Vanessa Saldanha, Joanne Salop,Sudhir Shetty, Susan Stout, Eric Swanson,Caroline Van Den Berg, Dominique Van DerMensbrugghe, and Linda Van Gelder.

Inputs from IMF staff were coordinated byJames Boughton. Other Fund contributorsincluded Charles Blitzer, Benedict Clements,Thomas Dorsey, Peter Fallon, Judith Gold,Sanjeev Gupta, Peter Heller, Timothy Lane,Hans Peter Lankes, James Morsink, DavidRobinson, Carlo Sdralevich, Mark Swin-burne, Patrizia Tumbarello, and Kal Wajid.

Guidance received from the ExecutiveDirectors of the Bank and the Fund duringdiscussions of the draft report is gratefullyacknowledged. The report has also benefitedfrom many useful comments and suggestionsreceived from Bank and Fund managementand staff in the course of the preparation andreview of the report. The World Bank’s Officeof the Publisher coordinated book design,editing, and print production.

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 xiii

Abbreviations and Acronyms

AFD Agence Française de Développement

AfDB African Development BankAGOA African Growth

Opportunity ActAMS Aggregate measure of supportAPRM African Peer Review

MechanismARDE Annual Review of Development

EffectivenessAROE Annual Report on Operations

EvaluationAsDB Asian Development BankATC Agreement on Textiles

and ClothingAVEs Ad valorem equivalentsBank World BankBCP Basel Core PrincipleBEEPS Business Environment and

Enterprise Performance SurveysBPO Business process outsourcingCAC Collective action clausesCAE Country Assistance EvaluationCAP Common Agricultural PolicyCAS Country Assistance StrategyCBO Community-based organizationCDF Comprehensive Development

FrameworkCEDAW Convention on Elimination

of All Forms of Discriminationagainst Women

CGIAR Consultative Group onInternational AgriculturalResearch

CIPE Country Institutional and Policy Evaluation

CPA Country PerformanceAssessment

CPIA Country Policy andInstitutional Assessment

DAC Development AssistanceCommittee (OECD)

DATA Data Accountability andTechnical Assistance Project

DB Doing Business ProjectDEC Development Economics Group

(World Bank)DHS Demographic and Health

SurveyEAP East Asia and PacificEBA Everything-but-Arms InitiativeEBRD European Bank for

Reconstruction andDevelopment

ECA Europe and Central Asia regionEduco Education with the Participation

of Communities ProgramEFA-FTI Education for All–Fast-Track

InitiativeEIB European Investment BankEPR Environmental performance

review

A B B R E V I A T I O N S A N D A C R O N Y M S

xiv G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

ESAF Enhanced StructuralAdjustment Facility (IMF)

ESW Economic and sector workEU European UnionFDI Foreign direct investmentFSAP Financial Sector Assessment

Program (IMF and WorldBank)

FSLC Financial Sector LiaisonCommittee

FSSAP Female Secondary SchoolAssistance Program(Bangladesh)

Fund International Monetary FundGAC Governance and Anti-

Corruption SurveyGATS General Agreement on Trade

in ServicesGDF Global Development FinanceGDP Gross domestic productGEF Global Environment FacilityGEP Global Economic ProspectsGFATM Global Fund to Fight AIDS,

Tuberculosis, and MalariaGFS Government Finance StatisticsGNI Gross national incomeGPGs Global public goodsGSP Generalized System

of PreferencesHIPC Heavily indebted poor countryHIPC AAP HIPC Assessment and

Action PlanHS Harmonized SystemIAP Infrastructure Action PlanIBRD International Bank for

Reconstruction andDevelopment

ICA Investment climate assessmentICB/NCB International Competitive

Bidding/National CompetitiveBidding

ICRG International Country Risk Guide

IDA International DevelopmentAssociation

IDB Inter-American DevelopmentBank

IEO Independent Evaluation Office

IFC International FinanceCorporation

IFF International Finance FacilityIFI International financial

institutionIFS International Financial

StatisticsIGR Institutional and Governance

ReviewsILO International Labour

OrganisationIMF International Monetary FundINFVP Infrastructure Vice Presidency

(World Bank)IPP Independent power producersI-PRSP Interim Poverty Reduction

Strategy PaperITC International Trade CenterITU International Telecommunication

UnionJIC Joint Implementation

CommitteeJSA Joint Staff AssessmentKfW Kreditanstalt fur WiederaufbauLAC Latin America and Caribbean

regionLDCs Least developed countriesLICUS Low-income countries

under stressMDB Multilateral development bankMDGs Millennium Development GoalsMFA Multifiber ArrangementMFN Most favored nationMICs Middle-income countriesMIGA Multilateral Investment

Guarantee AgencyMNA Middle East and North Africa

regionMoU Memorandum of UnderstandingMTEF Medium-Term Expenditure

FrameworkNAFTA North American Free

Trade AgreementNCA New Capital Accord or Basel IINEPAD New Partnership for Africa’s

DevelopmentNGO Nongovernmental organizationNPV Net present value

NTB Nontariff barrierO&M Operation and maintenanceODA Official development assistanceOECD Organisation for Economic Co-

operation and DevelopmentOED Operations Evaluation

Department (World Bank)OTRI Overall trade restrictiveness

indexPARIS21 Partnership in Statistics

for Development in the 21st Century

PEFA Public Expenditure andFinancial Accountability

PFM Public financial managementPPP Purchasing power parityPREM Poverty Reduction and

Economic Management Group(World Bank)

PRGF Poverty Reduction and GrowthFacility

PRSP Poverty Reduction StrategyPaper

PSE Producer support estimateQAG Quality Assurance Group

(World Bank)ROSC Reports on Observance of

Standards and CodesRPGs Regional public goodsRTA Regional trade agreementSDT Special and differential

treatmentSME Small and medium enterprisesSPA Strategic Partnership

with AfricaSSA Sub-Saharan Africa regionSWAp Sectorwide approachTA Technical assistanceTB TuberculosisTI Transition indicatorsTRIPS Trade-related aspects of

intellectual property rightsTRQ Tariff rate quota

TSE Total support estimateTUPE Transfer of Undertakings and

Protection of EmployeesU.N. United NationsUNAIDS Joint United Nations

Programme on HIV/AIDSUNCTAD United Nations Conference

for Trade and DevelopmentUNDAF United Nations Development

Assistance Framework UNDP United Nations Development

ProgrammeUNECA United Nations Economic

Commission for AfricaUNEP United Nations Environment

ProgrammeUNESCO United Nations Educational,

Scientific, and CulturalOrganization

UNGASS United Nations GeneralAssembly Special Session on HIV/AIDS

UNICEF United Nations Children’s FundUNSD United Nations Statistics

DivisionURAA Uruguay Round Agreement

on Agriculture (World TradeOrganization)

USAID United States Agency forInternational Development

VAT Value-added taxWBI World Bank InstituteWBI GI WBI Governance IndicatorsWDR World Development ReportWDI World Development IndicatorsWEO World Economic OutlookWHO World Health OrganizationWITS World Integrated Trade SolutionWP-EFF Working Party on Aid

Effectiveness and DonorPractices

WSS Water supply and sanitationWTO World Trade Organization

A B B R E V I A T I O N S A N D A C R O N Y M S

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 xv

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 xvii

Executive Summary

MDG Prospects—Need to ScaleUp Action, Significantly andSwiftlyOn current trends, most Millennium Devel-opment Goals (MDGs) will not be met bymost countries. The income poverty goal islikely to be achieved at the global level, butAfrica will fall well short. For the humandevelopment goals, the risks are much morepervasive across the regions. Likely shortfallsare especially serious with respect to thehealth and related environmental goals—child and maternal mortality, access to safedrinking water and basic sanitation. Few, ifany, regions will achieve the mortality goals.

The implication is clear. Achievement of theMDGs requires rising above current trendsand substantially accelerating progress towardthe goals. There is an urgent need for all par-ties to scale up action. The agenda has threeessential elements:

� Accelerating reforms to achieve strongereconomic growth—Africa will need todouble its growth rate.

� Empowering and investing in poor peo-ple—scaling up and improving the deliveryof human development and related keyservices.

� Speeding up the implementation of theMonterrey partnership, matching stronger

reform efforts by developing countrieswith stronger support from developedcountries and international agencies.

Priorities for DevelopingCountriesPolicies in developing countries have improved,enhancing their capacity to make effective useof resources, domestic and external, for devel-opment. Performance varies widely, however,and reform needs to be accelerated and deep-ened in many countries, especially in Sub-Saharan Africa. The analysis suggests fourareas for particular attention:

� Improving the enabling climate for privatesector activity, by solidifying progress onmacroeconomic stability, further reducingbarriers to trade, and shifting emphasisfrom regulating business operations tostrengthening market institutions. In macro-economic policy, the main area for improve-ment is fiscal management. Strengtheningproperty rights and the rule of law are thekey areas for attention with respect tothe institutional environment. An improvedenabling economic climate is essential bothfor mobilizing domestic investment andattracting more foreign investment.

� Strengthening capacity in the public sectorand improving the quality of governance

E X E C U T I V E S U M M A R Y

xviii G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

—the biggest challenge for many coun-tries. The most serious shortcomings are intransparency, accountability, and controlof corruption. Performance is better ingeneral in public financial management—expenditure and revenue management—but needs to improve further. On average,low-income countries can increase tax rev-enue by at least 1 to 2 percent of GDP byeliminating tax exemptions and improvingtax administration. The bulk of the financ-ing needed to achieve the MDGs, however,will have to come from improving theefficiency of existing spending, economicgrowth, and external resources. In Africa,which has the weakest governance indica-tors, the New Partnership for Africa’sDevelopment (NEPAD) initiative providesa very promising foundation for reform tobuild on.

� Scaling up investment in infrastructure andensuring its effectiveness, according prior-ity to infrastructure services closely linkedto the human development goals—waterand sanitation, transport. Compared withthe levels of the 1990s, infrastructurespending (investment plus operation andmaintenance) will need to rise by 3.5 to 5percent of GDP in low-income countriesand by 2.5 to 4 percent of GDP in lower-middle-income countries, with the pace ofthe increase depending upon institutionalcapacity and macroeconomic conditions inthe country concerned.

� Enhancing the effectiveness of servicedelivery in human development, by bettertargeting education, health, and socialassistance services toward poor people,addressing governance-related impedi-ments to service quality and effectiveness,increasing community participation, andscaling up on the basis of successful pro-grams (for example, the Female SecondarySchool Assistance Program in Bangladesh,the Education with the Participation ofCommunities (Educo) program in El Sal-vador, and the Education, Health, andNutrition (Progresa) program in Mexico).

Implementation needs to be expedited ontwo key donor-supported programs—theEducation for All-Fast Track Initiative (EFA-FTI) and the Global Fund to FightAIDS, Tuberculosis, and Malaria (GFATM).As of January 2004, only $6 million hadbeen disbursed under the EFA-FTI againstinitial commitments of $170 million (totalexternal financing needs for primary edu-cation in low-income countries are esti-mated to rise to at least $3.7 billion by2005–06, compared with actual assistanceof about $1 billion in 2002). As of thesame date, only $230 million had been dis-bursed under GFATM against $3.4 billionin pledges and $1.5 billion in commit-ments. Swifter action is needed on the partof both donors in providing funds andrecipients in addressing implementationconstraints.

Cutting across the policy agenda is theneed to empower women, by removing bar-riers to their fuller participation in the devel-opment process, and the need to ensureenvironmental sustainability. These cross-cutting concerns should be fully integratedinto policymaking.

Within the foregoing agenda, specific pri-orities and sequencing of actions of coursevary across countries and must be determinedat the country level in the context of coherent,country-owned development strategies, asreflected in the Poverty Reduction StrategyPapers (PRSPs) in the case of low-incomecountries and respective national strategyframeworks in the case of middle-incomecountries.

Priorities for DevelopedCountriesOverall, developed country actions to datehave fallen well short of the Monterrey vision.Progress seriously lags commitments in mostareas. This must change, and change quickly,to help accelerate progress toward the devel-opment goals. The vision of Monterrey needs

to be translated rapidly into concrete actions.Priorities for developed countries relate totrade and aid policies. But also important arethe broad conduct of macroeconomic andfinancial policies conducive to robust growthin the world economy and increased attentionto key global public goods, including environ-mental sustainability.

� Sustaining stable and strong growth in theglobal economy. A key issue is the orderlyresolution of fiscal and external imbal-ances, especially the large U.S. externalcurrent account deficit. An abrupt adjust-ment in the large economies could retardgrowth and leave global economic condi-tions vulnerable to shocks. Further workby developed countries—working withemerging market countries and interna-tional financial institutions—is needed toimprove the international financial archi-tecture to enhance prospects for strongerand more stable capital flows to develop-ing countries and to reduce the likelihoodand severity of financial crises. Rapidprogress is being made in the use of collec-tive action clauses, but substantial workremains to improve practices in sovereigndebt restructuring.

� Ensuring a successful, pro-development,and timely outcome to the Doha Round.High-income countries, given their weightin the system, must lead by example. Theyshould aim for reform targets that are suffi-ciently ambitious. These could include:complete elimination of tariffs on manufac-tured products; complete elimination ofagricultural export subsidies and completedecoupling of domestic agricultural subsi-dies from production, and reduction ofagricultural tariffs to, say, no more than 10percent; and commitments to ensure freecross-border trade in services delivered overtelecommunications links, complementedby actions to liberalize the temporary move-ment of workers. Developing countries alsomust seize the opportunity provided by theRound to further their own trade liberal-


G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 xix

ization. For developing countries (especiallythe low-income ones) to take full advantageof improved market access, they will needsupport in dealing with the “behind-the-border” agenda. Some countries will alsoneed assistance with adjustment costs asso-ciated with trade liberalization.

� Providing more and better aid. Aid flowsneed to rise well above current levels.Although donors have made post-Mon-terrey additional commitments of about$18.5 billion a year by 2006, estimatesshow that an initial increment of at least$30 billion annually can be effectively uti-lized by developing countries. As countriesimprove their policies and governance, theamount of additional aid that can be usedeffectively will rise into the range of $50billion plus a year, which estimates sug-gest will be needed to support adequateprogress toward the MDGs. ODA rose by$6 billion in nominal amount ($4 billion inreal terms) in 2002, but the increase wasalmost wholly accounted for by special-purpose allocations—technical coopera-tion, debt relief, emergency and disasterrelief. More aid will need to be provided informs that can flexibly meet the incremen-tal costs of achieving the MDGs, includingproviding a higher proportion directly tocountries in the form of cash, supportinggood policy performance with predictableand longer-term aid commitments, andallowing for the financing of recurrentcosts where country circumstances war-rant. There is also substantial scope forincreasing the effectiveness of aid byimproving the allocation of aid acrosscountries, aligning aid with nationaldevelopment strategies and priorities (asexpressed through PRSPs in the case oflow-income countries), and harmonizingdonor policies and practices around therecipient country’s own systems. To ensuredebt sustainability in heavily indebtedpoor countries that are pursuing goodpolicies, a larger proportion of additionalaid should be provided in the form of


xx G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

grants. Timely and adequate assistance inthe event of adverse exogenous shocks isespecially important for these countries.

� Improving policy coherence for develop-ment. Increased aid and other actions needto be part of a coherent overall approachto supporting development. In many cases,there are contradictions in policies, withsupport provided in one area undercut byactions in another. Putting in placeprocesses that enable an integrated assess-ment of the coherence of policies thataffect development—trade, aid, foreigninvestment and other capital flows, migra-tion, knowledge and technology transfer,environment—would help avoid such out-comes. Recent actions by Sweden to insti-tute an “integrated global developmentpolicy,” and by Denmark and other coun-tries to prepare regular assessments oftheir contribution to the goal of establish-ing a global partnership for development,go in the right direction.

Priorities for InternationalFinancial InstitutionsReview of how the international financialinstitutions are playing their role in contribut-ing to the achievement of the MDGs andrelated outcomes shows that they have madeprogress in enhancing their development effec-tiveness. This is reflected in progress in coun-try focus and ownership, results orientation ofoperations, transparency and accountability,and partnership. But there is much more to do.There are three key areas for action to deepenand build on the progress made:

� Refining and strengthening institutionalroles in low-income countries, includingby deepening the PRSP process and har-

monizing operational programs and prac-tices around national strategies and sys-tems, while also continuing to adaptapproaches and instruments to the evolv-ing needs of middle-income countries.

� Furthering progress on the results agenda,including implementation of the action planendorsed by the sponsoring agencies at theMarrakech Roundtable on Managing forDevelopment Results held in February 2004.

� Improving selectivity and coordination ofagency programs in line with comparativeadvantage and mandate to achieve greatersystemic coherence and effectiveness.

Priorities for Strengthening the Monitoring ExerciseTo carry this agenda forward, the World Bankand the International Monetary Fund plan tofocus future Global Monitoring Reports onspecific challenges—at the country, agency, andglobal levels—for meeting these priorities. Thiswill require further work, especially in the fol-lowing areas:

� Strengthening the underlying developmentstatistics, including timely implementationof the action plan agreed among interna-tional statistical agencies at the MarrakechRoundtable.

� Conducting research on the determinants ofthe MDGs, on critical issues such as effec-tiveness of aid, and on development of morerobust metrics for key policy areas such asgovernance and for the impact on develop-ing countries of rich country policies.

� Deepening collaboration with partner agen-cies in this work, building on respectiveagency comparative advantage and ensuringthat the approach to monitoring and evalu-ation is coherent across agencies.

xxii G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

Millennium Development Goals Goals and Targets from the Millennium Declaration

GOAL 1 ERADICATE EXTREME POVERTY AND HUNGER

TARGET 1 Halve, between 1990 and 2015, the proportion of people whose income is less than $1 a day

TARGET 2 Halve, between 1990 and 2015, the proportion of people who suffer from hunger

GOAL 2 ACHIEVE UNIVERSAL PRIMARY EDUCATION

TARGET 3 Ensure that by 2015, children everywhere, boys and girls alike, will be able to complete a fullcourse of primary schooling

GOAL 3 PROMOTE GENDER EQUALITY AND EMPOWER WOMEN

TARGET 4 Eliminate gender disparity in primary and secondary education, preferably by 2005, and at alllevels of education no later than 2015

GOAL 4 REDUCE CHILD MORTALITY

TARGET 5 Reduce by two-thirds, between 1990 and 2015, the under-five mortality rate

GOAL 5 IMPROVE MATERNAL HEALTH

TARGET 6 Reduce by three-quarters, between 1990 and 2015, the maternal mortality ratio

GOAL 6 COMBAT HIV/AIDS, MALARIA, AND OTHER DISEASES

TARGET 7 Have halted by 2015 and begun to reverse the spread of HIV/AIDS

TARGET 8 Have halted by 2015 and begun to reverse the incidence of malaria and other major diseases

GOAL 7 ENSURE ENVIRONMENTAL SUSTAINABILITY

TARGET 9 Integrate the principles of sustainable development into country policies and programs andreverse the loss of environmental resources

TARGET 10 Halve by 2015 the proportion of people without sustainable access to safe drinking water and basicsanitation

TARGET 11 Have achieved a significant improvement by 2020 in the lives of at least 100 million slum dwellers

GOAL 8 DEVELOP A GLOBAL PARTNERSHIP FOR DEVELOPMENT

TARGET 12 Develop further an open, rule-based, predictable, nondiscriminatory trading and financial system(including a commitment to good governance, development, and poverty reduction, nationallyand internationally)

TARGET 13 Address the special needs of the least developed countries (including tariff- and quota-free accessfor exports of the least developed countries; enhanced debt relief for heavily indebted poorcountries and cancellation of official bilateral debt; and more generous official developmentassistance for countries committed to reducing poverty)

TARGET 14 Address the special needs of landlocked countries and small island developing states (through theProgramme of Action for the Sustainable Development of Small Island Developing States and theoutcome of the 22nd special session of the General Assembly)

TARGET 15 Deal comprehensively with the debt problems of developing countries through national andinternational measures to make debt sustainable in the long term

TARGET 16 In cooperation with developing countries, develop and implement strategies for decent andproductive work for youth

TARGET 17 In cooperation with pharmaceutical companies, provide access to affordable, essential drugs indeveloping countries

TARGET 18 In cooperation with the private sector, make available the benefits of new technologies, especiallyinformation and communication

Note: The Millennium Development Goals and targets come from the Millennium Declaration signed by 189 countries, including 147 heads of state, in September 2000. The goals and targets are related and should be seen as a whole. They represent a partnership of countries determined, as the Declaration states, “to create an environment—at the national and global levels alike—which is conducive to developmentand the elimination of poverty.”Source: United Nations. 2000 (September 18). Millennium Declaration. A/RES/55/2. New York.United Nations. 2001 (September 6). Road Map towards the Implementation of the United Nations Millennium Declaration. Report of the SecretaryGeneral. New York.

The turn of the century was marked bysome significant and promising eventsfor world development. The Millen-

nium Declaration—signed by 189 countries inSeptember 2000—led to the adoption of theMillennium Development Goals (MDGs).These goals set clear targets for eradicatingpoverty and other sources of human depriva-tion and for promoting sustainable develop-ment (see goals and targets, p. xxii). This wasfollowed in 2001–02 by major internationalmeetings in Doha, Monterrey, and Johannes-burg that contributed to the emergence of ashared understanding of the broad develop-ment strategy and policies needed to attain theMDGs. The meeting in Monterrey in March2002 ushered in a new compact betweendeveloping and developed countries. TheMonterrey Consensus stressed mutual respon-sibilities in the quest for the MDGs. It calledon the developing countries to improve theirpolicies and governance. And it called on thedeveloped countries to step up their support,especially by providing more and better aidand more open access to their markets.

With broad agreement on the goals andstrategies to achieve them, the task now isimplementation. It is time to translate visioninto action; to strive for a better balance inthe development effort so that all parties play

their part. Implementation needs to happenwithin countries and at the global level. Allparties must deliver on their part of the com-pact. This was the central message of the2003 International Monetary Fund (IMF)and World Bank (Bank) Annual Meetingsheld in Dubai. But is implementation actuallyhappening? What progress has been made?What constraints are blocking implementa-tion? How are all parties doing in deliveringon their commitments? What are the priori-ties in the agenda? This report attempts toanswer those questions.

The themes of implementation andaccountability constitute the fundamentalmotivation behind the global monitoringinitiative, launched at the request of theDevelopment Committee, the joint Ministe-rial body of the World Bank and the IMF.The annual global monitoring reports, ofwhich this report is the first, will provide anassessment of progress on policies andactions for achieving the MDGs and relateddevelopment outcomes. These reports willunderpin the Development Committee’s reg-ular monitoring of progress on the policyagenda and reinforcement of the prioritiesfor action and the accountabilities of the keyactors—developing and developed countriesas well as multilateral agencies.

G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4 1

OverviewFrom Vision to Action

Prospects for Meeting the MDGs:Reasons for Optimism, Grave Concerns

Income poverty goal: a mixed picture. At theglobal level, the world will likely meet thefirst goal of halving income poverty between1990 and 2015, thanks to stronger economicgrowth spurred by improvements in policies—especially in China and India, the world’s twomost populous countries. With current trends,most regions will achieve or come close toachieving the goal. East Asia has already metit. However, Sub-Saharan Africa is seriouslyoff track; just eight countries, representingabout 15 percent of the regional population,will likely achieve the goal. Within otherregions that will likely meet the goal at theaggregate level, a number of countries will not.Low-income countries under stress (LICUS),about half of which are in Africa, are especiallyat risk of falling far short. The trends arebroadly similar with respect to the target ofhalving the proportion of people who sufferfrom hunger, also part of Goal 1. The targetwill likely be met at the global level, but Sub-Saharan Africa and a number of countries inother regions will likely fall short.

Human development and environmentalgoals: more serious concerns. The risks asso-ciated with the human development goals aremore pervasive across the regions. While eco-nomic growth has a significant effect on edu-cation and health outcomes, just as it does onincome poverty, the magnitude of the effect istypically smaller. Prospects for progress onthe human development goals also dependimportantly on the scale and effectiveness ofdevelopment interventions directed specifi-cally toward them. The determinants of thesegoals are multiple and cut across sectors.

The prospects are brighter in educationthan in health. With current trends, severalregions will achieve or approach the goal ofproviding universal primary education. Again,shortfalls are likely in Sub-Saharan Africa,and possibly in South Asia and the MiddleEast and North Africa as well. Gender gaps in

education are most serious in the same threeregions. While the target for gender equalityin primary and secondary education is to beachieved preferably by 2005, about a thirdof developing countries appear unlikely toachieve it even by 2015. Prospects for genderparity at all levels of education, includinghigher education, are even less comforting.

But the prospects are gravest in health. Oncurrent trends, the goals of reducing child andmaternal mortality will not be attained inmost regions. Only a small proportion ofcountries (15–20 percent) currently appear tobe on track. The goal of halting and reversingthe spread of HIV/AIDS (human immuno-deficiency virus/acquired immunodeficiencysyndrome) and other major diseases (malaria,tuberculosis) appears daunting. Their incidencecontinues to rise, further aggravating condi-tions affecting child and maternal mortalityand entailing broad and serious economic andsocial consequences. The risks of failure tohalt the spread of HIV/AIDS are especiallyhigh in Sub-Saharan Africa, but are substantialin many countries in other regions as well.

The health goals are rendered more diffi-cult by the large gaps in access to safe drink-ing water and basic sanitation. The gaps arelargest in Sub-Saharan Africa for water andin South Asia for sanitation. The goal of halv-ing, by 2015, the proportion of populationwithout access to safe water and sanitationmeans providing an additional 1.5 billionpeople with water and 2 billion with sanita-tion. With current rates of progress at abouthalf what is needed, most regions will fallwell short. At those rates, only about one-fifth of countries will achieve the targetincrease in access. Among low-income coun-tries, only about one-tenth will make it.

Variation across countries. Global andregional trends hide considerable variationacross countries. East Asia, with its richdiversity, provides a good example. At oneend, the region has middle-income countries,such as China and Thailand, that havealready met or will soon meet several of theMDGs; some of those countries are develop-

O V E R V I E W

2 G L O B A L M O N I T O R I N G R E P O R T 2 0 0 4

ing “MDG-plus” agendas. At the other end,low-income countries such as Cambodia andPapua New Guinea are seriously off track, asare many poor countries in Sub-SaharanAfrica. There is also diversity within coun-tries, especially large ones. Although Chinahas already met the income-poverty MDG atthe national level, progress has been muchslower in some inland provinces that con-tinue to have large concentrations of poverty.

In general, middle-income countries aremuch better positioned to achieve the MDGsthan low-income countries. Many of themhave already met the goals or are well on theirway. Notwithstanding their progress onincome poverty, those countries remain hometo 280 million people living on less than $1 aday and 870 million people living on less than$2 a day. A number of them lag in relation tosome of the non-income MDGs. For exam-ple, China, despite its spectacular perfor-mance in reducing income poverty, is not ontrack to meet the child mortality goal, basedon current trends.

Daunting challenges, but grounds for hope.The MDGs present a difficult challenge, butpast development successes give cause forhope. Globally, adult illiteracy fell by halfover the past 30 years, while life expectancyat birth rose by 20 years over the past 40years. Some countries have advanced partic-ularly far, particularly fast. Vietnam, forexample, a low-income country, reducedpoverty from 51 to 14 percent from 1990 to2002. Even in Sub-Saharan Africa, therehave been encouraging stories of success.Botswana doubled the proportion of childrenin primary school in 15 years, nearly achiev-ing universal primary education. Beninincreased its primary enrollment rate andMali its primary completion rate by morethan 20 percentage points in the 1990s. Mau-ritania increased the ratio of girls to boys atschool from 67 to 93 percent between 1990and 1996. Uganda reduced HIV/AIDS infec-tion rates for eight consecutive years in the1990s. Zambia may soon become the secondAfrican country to slow the spread of this

scourge. These achievements demonstratethat rapid progress is possible, given goodpolicies and the support of partners.

Scaling Up on the Basis of the Monterrey ConsensusThe implications of the foregoing assessmentare clear. The achievement of the developmentgoals will require rising above current trendsand accelerating the pace of development, anddoing so swiftly. In line with the principles andpartnership established at Monterrey, all par-ties must scale up their action. The agenda hasthree essential elements:

� Accelerating and deepening reforms toachieve stronger economic growth

� Empowering and investing in poor peo-ple—stepping up action to improve thedelivery of services affecting human devel-opment

� Speeding up the implementation of theMonterrey partnership, matching strongerdeveloping country efforts to spur growthand improve service delivery to poor peo-ple with stronger support from developedcountries and international agencies.

Acting on multiple fronts. The multidimen-sionality of the MDGs, the linkages amongthem, and their multisectoral determinantsimply that the policy agenda for achieving thegoals is similarly broad. Indeed, the agendaspans the gamut of development. There is noone-to-one link between the MDG relating toa sector and policies relating to that sector.The outcome in a given sector dependsimportantly on factors outside that sector.For child survival, for example, mother’s edu-cation and access to safe water and sanitationmay be more important than access to healthfacilities. Likewise, schools and health facili-ties may exist, but girls may be preventedfrom attending if they spend much of theirtime fetching water from distant sources or ifadequate and safe means of transport arelacking. The agenda cuts across sectors, andacross policies, investments, and institutions.

O V E R V I E W


The scaling-up effort, therefore, will requireconcerted action on multiple fronts.

Promoting stronger economic growth. At thecenter of the strategy to achieve the MDGsand related development outcomes must bethe promotion of stronger economic growth.Growth directly reduces income poverty andexpands resources for use toward the non-income goals. So, first and foremost, eco-nomic growth in developing countries willneed to be stronger than recently attained orcurrently projected. Sub-Saharan Africa willneed to double its average GDP (gross domes-tic product) growth rate, to around 6 percent.This target is ambitious, of course, but somecountries in the region achieved it in the1990s: Cape Verde, Mauritius, Mozambique,and Uganda. What is needed is to acceleratepolicy and governance reforms to improvethe enabling climate for growth: macroeco-nomic stability and openness, the regulatoryand institutional environment for private sec-tor activity, physical and financial infrastruc-ture, and public sector governance.

Scaling up service delivery. Reaching the goalswill also require policies and actions toenhance the capabilities of poor people—menand women—to participate in and benefitfrom growth. For their participation to beeffective, the poor need to be empoweredthrough improved delivery of education andhealth services, as well as related infrastructureservices, such as water and sanitation and ruralroads. Stepped-up investments in those ser-vices must be accompanied by reforms in sec-tor policy and institutional frameworks thatimprove the effectiveness of delivery, includinggreater involvement of communities, especiallypoor people, in making decisions.

Enhancing the global development partner-ship. The developing countries are in the dri-ver’s seat in setting the agenda for achievingthe development goals, but they will needhelp from development partners. Implemen-tation will require increased cooperation atthe global level. The developing countries

need expanded access to markets in devel-oped countries to increase exports and spurgrowth. And they need more aid to financedevelopment programs that improve thedelivery of human development and infra-structure services. This mutualism was clearlyrecognized and affirmed at Monterrey, butprogress to date has been relatively slow. Thespirit of Monterrey needs to be translatedrapidly into action.

Priorities for DevelopingCountriesPolicies improving, but much further to go.Indicators for the past five years show improve-ment in policies in all regions, although to vary-ing degrees. On average, policy indicatorsremain the lowest in Sub-Saharan Africa, buteven there show an encouraging improvementon most dimensions, suggesting that recentreforms are beginning to take hold. Theimprovement in policies is creating conditionsthat enhance countries’ capacity to make effec-tive use of resources for development, domesticand external. While some improvement hasoccurred across all policy areas, progress isespecially notable in macroeconomic manage-ment and trade policy. Average inflation andtariff rates have been cut in half in the pastdecade. The improved policy environment hascontributed to a pick-up in economic growth.Indeed, average per capita GDP growth in low-income developing countries in the past fiveyears was higher than during any other five-year period in the past two decades. Better poli-cies pay off.

Despite this improvement, however, growthin many countries—most of them in Sub-Saharan Africa—remains below the levelneeded to achieve the MDGs. During 1998–2002, nearly 60 percent of low-income coun-tries (with a combined population of 950 mil-lion) experienced per capita growth of lessthan 2 percent, while 32 percent (with a com-bined population of 555 million) experiencednegative per capita growth. Factors such asadverse political and external circumstanceshave played a role, including the limited avail-

O V E R V I E W


ability of aid resources and impediments toaccess to export markets in developed coun-tries, but the growth response to improve-ments in the macroeconomic and trade policyenvironment has been dampened by slowerprogress on structural and institutional reformsthat are essential to improving the enabling cli-mate for private sector activity. Strongergrowth in the future will depend crucially onmore vigorous and consistent efforts to speedup reforms in these areas.

In the delivery of services—human develop-ment, infrastructure—the picture is broadlysimilar. Some areas show encouraging progress;others require stronger action. Resource allo-cation has improved somewhat, as evidencedby the increased investment in human capital.Public education and health spending increasedover the 1990s from 6.9 to 7.4 percent of GDPin those low-income countries for which dataare available. Successful innovations in servicedelivery to the poor are encouraging. Theseinclude the Education with the Participation ofCommunities (EDUCO) program in El Sal-vador; the Progresa program of conditionalcash transfers to the poor linked to school andclinic attendance in Mexico; and the FemaleSecondary School Assistance Program inBangladesh, which employs targeted financialincentives and community engagement toincrease girls’ enrollment in schools. Key ideasfrom these innovations are now being appliedin other countries, including most recentlyNepal. In many countries, however, the qualityand effectiveness of service delivery show majordeficiencies. These gaps point to the need toaccelerate improvements in the underlying pol-icy and institutional framework to raise theyield of increased spending on services.

The core of the reform agenda: institutionalreform. Cutting across the policy agenda isthe need to improve governance. Public sec-tor governance, while improving, remains theweakest area of the reform agenda in mostcountries. Institutional dimensions of reformare also paramount in the improvement of theprivate sector business climate and the per-formance of the service delivery sectors. In

macroeconomic management as well, perfor-mance is strongly correlated with the qualityof institutions responsible for policy imple-mentation. And in most developing countries,improved management of the environmentrequires building up fledgling environmentalinstitutions. Responding to these challenges,governments in more and more developingcountries have launched governance andinstitutional reforms. An important exampleis the New Partnership for Africa’s Develop-ment (NEPAD), an initiative owned and ledby African countries that places improvementin governance at the center of the reformagenda.

Country focus and ownership: central to suc-cess. The primary determinant of the prospectsfor achieving the MDGs is developing coun-tries’ own policies. Overall, progress has beenencouraging, but reforms need to be acceler-ated and deepened. The review conducted forthis report points to five areas needing particu-lar attention, as discussed below. Within thesebroad areas, policy priorities for individualcountries must be determined at the countrylevel, in the context of coherent country devel-opment strategies. Country ownership andleadership of the development strategy are cru-cial to effective implementation and achieve-ment of results.

For low-income countries, the PovertyReduction Strategy Papers (PRSPs) are the pri-mary avenue of expression of a country-ownedand -led development strategy. In middle-income countries, the policy integration andprioritization role is performed within respec-tive national strategy frameworks. By the endof March 2004, 37 countries had prepared andwere implementing full PRSPs; 16 more hadprepared Interim PRSPs (I-PRSPs). Countriesare increasingly reflecting the MDGs in theirPRSPs. The PRSP process itself is being deep-ened along various dimensions, including par-ticipatory process, growth strategies, publicexpenditure management, and poverty andsocial impact analysis. Continued strengthen-ing of the PRSP process, and deepening of thelinks with the MDGs, will ground the agenda

O V E R V I E W


for achieving the development goals in coun-try-owned strategies. Countries can spell outtheir commitments to policy and institutionalreforms in these strategies, which in turnenables donors to commit support in a coher-ent and consistent way.

Solidifying Macroeconomic Stability

Fiscal management: the main area forimprovement. Macroeconomic managementhas improved in all regions. Yet progress hasbeen uneven and remains fragile in manycountries, especially in Sub-Saharan Africa.Fiscal management is the area of most concern.Performance is much better on monetary andexchange rate management. Fiscal policy, as itrelates to the sustainability of public debt andcontainment of fiscally derived macroeco-nomic imbalances, remains unsatisfactory inabout a third of low-income countries. Thedeficiencies in structural aspects of fiscal policyare even more serious, with nearly half of low-income countries assessed as having an unsat-isfactory composition of public expenditures.For these countries, therefore, a strengtheningof macroeconomic policies, especially fiscalmanagement, remains necessary. Even in coun-tries with better performance, maintainingand building on macroeconomic stability—anessential foundation for sustained growth—will be a continuing challenge.

In middle-income countries, macroeco-nomic policy indicators are better, on aver-age, than in low-income countries. Becausemiddle-income countries typically are moreintegrated into international capital mar-kets, maintaining sound macroeconomicpolicies is especially important for reducingvulnerability to crises—which can washaway hard-won gains in reducing poverty. Inthe past two decades, average output loss indeveloping countries from currency crises isestimated at about 7.5 percent of pre-crisisGDP. Although vulnerability indicators haveimproved in the past few years, the reduc-tion of public debt, especially external debt,relative to GDP remains a key area for

improvement in several countries. Gover-nance of financial and corporate sectors alsoneeds to be strengthened to prevent thebuild-up of balance-sheet vulnerabilities.

Improving Private Sector Enabling Environment

Extending progress toward outward-orientedstrategies. Despite significant liberalization,there is substantial scope for further reduc-tions in trade barriers, especially in regionswhere they remain high. In South Asia, forexample, despite sharp declines since the late1980s, the average tariff remains around 20percent. Taking into account nontariff barri-ers (excluding technical product regulations),South Asia’s average tariff equivalent was anestimated 32 percent in 2001, the highestamong developing regions. Developing coun-tries should take advantage of the DohaRound to make further strides toward tradeopenness. Countries that derive a sizable partof government revenue from trade taxes mayneed assistance in adjusting to a regime oflower trade tariffs. In addition to reducingtrade barriers, countries should move vigor-ously on the “behind-the-border” agenda, toenable the private sector to exploit theopportunities created by lower trade barri-ers. The agenda includes the efficient supplyof services closely related to trade—customs,transport, and telecommunications, financialservices—and improvement of the broaderenabling environment for entrepreneurshipand private investment. Evidence suggeststhat full liberalization and regulatory reformin services trade could add significantly toeconomic growth.

Reducing regulation, strengthening institu-tions—especially property rights, rule of law.While improving, the regulatory and institu-tional environment for private sector activitystill needs significant reform in many coun-tries. Regulation typically is much heavierand more complex in low-income countries,notwithstanding their more limited imple-

O V E R V I E W


mentation capacities, raising the cost of start-ing and operating a business and creatingopportunities for corruption. Starting a busi-ness in high-income countries typically takes30 days and costs less than 10 percent of percapita income; in low-income countries, ittakes 74 days and costs two times per capitaincome. While regulation is heavy, the essen-tial institutions underpinning markets areweak. The most serious shortcomings are inproperty rights and rules-based governance,an area assessed as less than satisfactory inalmost four-fifths of low-income countries.Such an environment deters investors, bothdomestic and foreign. Weak creditor rightsand contract enforcement also inhibit thegrowth and deepening of the financial system.

Countries need to shift emphasis from reg-ulating business operations to building insti-tutions that facilitate business by supportingefficient and fair functioning of markets. Akey area of reform is the strengthening ofproperty rights and of institutions that estab-lish and enforce the rule of law, including legaland judicial reform and the reduction ofbureaucratic harassment. Continued strength-ening of the institutions of corporate gover-nance, especially in middle-income countries,is also important.

Upgrading Public Sector Governance

Accelerating governance reform. The need toaccelerate reform is greatest in public sectorgovernance. The quality of public sector gov-ernance has improved, especially in Europeand Central Asia and South Asia. But thereform agenda calls for more vigorous actionin many countries. In three-fourths of low-income countries, overall public sector gov-ernance is assessed to be less than satisfactory,making it the weakest area of performance.The weaknesses are most pervasive preciselyin those countries where stronger institu-tional capacities are needed to manage devel-opment interventions that will spur progresstoward the MDGs—low-income countries inSub-Saharan Africa. They are especially acute

in the low-income countries under stress.Governance ratings are higher in middle-income countries, but those ratings still arelower than their ratings in other policy areas.These findings highlight governance and insti-tution-building reforms as an area for partic-ular attention, as poor governance and weakinstitutions can seriously undermine the effec-tiveness of policies and programs throughoutan economy. Initiatives such as the NEPAD,therefore, are especially valuable and timely.

Controlling corruption. The most seriousshortcomings are in transparency, accountabil-ity, and control of corruption. Reform is com-plex in these areas, which are less amenable to“technocratic” solutions. Progress will dependon a careful nurturing of reform ownership andof needed changes in bureaucratic culture.Political will is key, as are political processesthat allow broad participation, build in checkson executive authority, and enable citizens tohold administrations accountable.

Building on progress in public financial man-agement. Performance is better in publicfinancial management, on average. A greaterfocus on public expenditure and budget man-agement in the preparation of the PRSPs andin the initiative to help heavily indebted poorcountries (HIPCs) has contributed to progressin these areas, which must be sustained anddeepened. The importance of improved man-agement of public resources is underscored bythe need to create fiscal space for increasedspending on key infrastructure and humandevelopment services (see below) within sus-tainable overall fiscal positions. In manycountries, the scope for reallocating spendingtoward development remains substantial. Onthe revenue side, analysis shows that on aver-age low-income countries can increase theirtax-to-GDP ratio by 1 to 2 percentage pointsby eliminating tax exemptions and improvingtax administration. Doing so would helpmobilize resources, although the bulk of thefinancing needed to achieve the MDGs willhave to come from improving the efficiency

O V E R V I E W


of existing spending, economic growth, andexternal resources.

Decentralized governance can improve thedelivery of services at the local level. Decen-tralization is especially important for largemiddle-income countries like Brazil andChina that need to tackle major concentra-tions of poverty at the subnational level. Tobe effective, decentralization must be under-pinned by sound intergovernmental fiscal sys-tems and local institutional capacities.

Strengthening Infrastructure

Substantial scaling up of investment needed.Infrastructure plays a dual role in the effort toachieve the MDGs. It is an important part ofthe enabling environment for economicgrowth, and it also delivers services that arekey to meeting the human development andgender equality goals. Gaps in the availabilityand quality of key infrastructure are large,especially in low-income countries and inrural areas within countries. Narrowingthose gaps will require sizable increases ininvestment and associated spending on oper-ation and maintenance. Average spending oninfrastructure (investment plus operation andmaintenance) in low-income and lower-middle-income countries may have to almostdouble from the levels of the 1990s (whensuch spending fell by 2 to 4 percent of GDP).This implies increases in infrastructure spend-ing (covering power, transport, telecommuni-cations, and water and sanitation) on theorder of 3.5 to 5 percent of GDP in low-income countries and 2.5 to 4 percent of GDPin lower-middle-income countries relative tothe low levels of the 1990s. The pace of theincrease will depend upon the institutionalcapacity and macroeconomic conditions inthe country concerned.

Financing this spending will be a majorchallenge. Private investment in infrastructurehas increased, but not as much as expected.Efforts must continue to improve the regula-tory and institutional environment for suchinvestment. Innovative instruments for riskmitigation could also help leverage more pri-

vate financing. At the same time, the decade-long decline in public spending on infrastruc-ture needs to be reversed. That will requirestronger mobilization of domestic resources,including improved cost-recovery and reallo-cation of spending, and increased externalassistance. Especially in the low-income coun-tries, external assistance must provide a largershare of total infrastructure spending than theroughly 10 percent it provided in the 1990s.Infrastructure requirements relating to waterand sanitation warrant special attention inpublic spending and foreign assistance pro-grams, given their close links to the health andgender goals, and the fact that this sector tra-ditionally attracts less private financing thanother infrastructure sectors such as power andtelecommunications.

Increased investment: not the sole answer.To ensure its effectiveness and sustainabil-ity, investment must be underpinned byimprovements in the policy and governanceframework, especially the capacity of keyinstitutions. With more and more responsi-bilities in infrastructure falling on local gov-ernments, strengthening administrative andfinancial capacities at the local level, includ-ing developing and facilitating the use ofappropriate subsovereign financing instru-ments, is increasingly important.

Accelerating Human Development

More resources complemented by more effec-tive use. Encouraging progress has been madein human development. More investment isbeing made in education and health and moreattention is being paid to the effectiveness ofservice delivery. But progress needs to beaccelerated and broadened if the humandevelopment goals are to be achieved. Thedeficiencies in service delivery are most seri-ous in Sub-Saharan Africa and South Asia,although even in these regions, individualcountries are making progress, such as Ghanaon child mortality, and Ethiopia and Rwandaon primary completion.

O V E R V I E W


In most low-income countries, the targets ineducation and health require the commitmentof more resources to these services. However,in a number of these countries, there is sub-stantial scope for increasing the impact ofexisting spending by correcting poor targetingof subsidies, lax resource management, lowefficiency and quality of service, and informa-tion failures. Examples abound. In Guinea, theshare of public spending on education andhealth accruing to the richest quintile wasfound to be seven times that accruing to thepoorest. In Uganda, 87 percent of nonwageresources intended for schools was diverted toother uses before the problem was discoveredand corrective action taken. Teacher salariesabsorbed more than 90 percent of the recur-rent education budget in Kenya. Teacherabsenteeism is 39 percent in Bihar, India.Among doctors in primary health facilities inBangladesh, absenteeism is 73 percent. Despitefree immunization, 60 percent of children arenot immunized in India, because mothers areunaware of the benefit. Many of these prob-lems can be traced to weaknesses in gover-nance and institutional capacities.

Main elements of the agenda. Concertedaction is needed on several fronts: scaling upinvestment in human capital in low-incomecountries while maximizing the impact ofexisting public spending by improving thetargeting of public services in education,health, and social assistance; paying attentionto intersectoral linkages when developing andimplementing programs (it is hard to reducechild mortality when only 10 percent of thepoor households have access to an improvedwater source, as in Ethiopia); addressinggovernance-related impediments to servicequality and effectiveness; and piloting and eval-uating empowerment options to strengthen theinvolvement of stakeholders, especially poorpeople, in the design and delivery of services(and scaling up on the basis of successful pro-grams, such as Educo and Progresa).

Community involvement is particularlyimportant to the goal of reducing gender dis-parities in education. Since the success of

interventions to educate girls is fundamen-tally embedded in the sociocultural context,community involvement can help ensure thatinterventions are responsive to needs. Effec-tive improvement of female access to educa-tion—and to other key services—requiresthat the design of services reflects gender con-cerns. Indeed, the goal of empowerment ofwomen calls for gender concerns to be fullyintegrated into policymaking more broadly.

Donor support: EFA-FTI and GFATM. Thescaling up of human development in low-income countries will require that moredonor support come in forms that promotebroad sector reform, encompassing the pol-icy and institutional dimensions of the sectorand moving away from past practicesfocused more on earmarked expenditures orvertical programs that delivered a narrowpackage of interventions. The Education forAll–Fast Track Initiative (EFA-FTI) is helpingto support a shift in that direction. Disburse-ments under the program, slow to take offbecause of agency programming and budget-ing cycles, need to be expedited. As of Janu-ary 2004, only $6 million of the first $170million committed to the initial group ofcountries had been disbursed. World Bankprojections suggest that as the FTI scales upto all low-income countries, at least $3.7 bil-lion per year will be needed in externalfinancing for primary education by 2005–06,compared with about $1 billion in 2002.Implementation has also been slow under theGlobal Fund for AIDS, Tuberculosis andMalaria (GFATM). As of January 2004, outof $3.4 billion in pledges, $1.5 billion hadbeen committed but only $230 million hadbeen disbursed. Expediting progress in thispriority area requires better donor coordi-nation and the alleviation of institutionalcapacity constraints in recipient countries.

Priorities for Developed CountriesActions well short of the Monterrey vision.As agreed in Monterrey, if the MDGs are to be achieved, stronger reform actions by

O V E R V I E W


developing countries must meet with strongersupport from developed countries in an en-hanced global development partnership. Pri-orities for developed countries relate to tradeand aid policies. Also important is the broadconduct of macroeconomic and financial poli-cies in a way that is conducive to strong globaleconomic growth and stable private capitalflows. Attention to key global public goods isalso needed. How well are developed coun-tries doing in living up to their commitments?The assessment carried out for this reportshows that actions seriously lag commitmentsin most areas. Accelerating progress towardthe MDGs requires much stronger actions ofsupport from the developed world than wit-nessed so far. As for developing countries, theagenda can be grouped under five heads.

Fostering a Robust Global Economic Recovery

Through their impact on trade and capitalflows, global economic conditions exercise amajor influence on prospects for growth andpoverty reduction in developing countries.Growth in the developing countries cannotthrive in the absence of strong and sustain-able growth in the advanced economies.Although the prospects for recovery in worldeconomic growth appear to be reasonablybright over the near term, sustaining a strongglobal economy will require the major coun-tries to address some outstanding issues andimbalances.

Orderly resolution of imbalances. Disorderlyadjustment in the largest economies couldretard growth or leave global economic con-ditions vulnerable to shocks. Most notably,the United States is running a large externalcurrent account deficit. Such large externalimbalances, financed increasingly with debtinstruments, are difficult to sustain for a longperiod. As economic growth in the UnitedStates gathers steam, a gradual tightening offiscal and monetary policies could help bringabout an orderly adjustment. In Europe, the

central challenge is to implement neededstructural reforms, especially in labor mar-kets and social security systems, to returneconomic growth to a sustainable 2 to 3 per-cent range over the medium term. In Japan,economic policy needs to continue to focuson countering deflationary tendencies, stabi-lizing public sector debt, and addressing theaccumulation of imbalances in the financialand corporate sectors. A common, longer-term structural challenge is to address thefiscal impact of the demographic changesbuilding up in these countries.

The ongoing global economic recovery,buttressed by low interest rates in theadvanced economies, is reflected also in somerecovery in private capital flows to develop-ing countries in 2003. The outlook for sus-taining these flows in the longer term wouldimprove if the large fiscal and external imbal-ances in the advanced economies werereduced, freeing up financing for developingcountries, and the recipient countries contin-ued to improve their policy and institutionalenvironment to make sound and sustainableuse of external financing. Prospects for pri-vate capital flows would also benefit fromimprovements in the international financialarchitecture to make those flows more stableand reduce the likelihood and severity offinancial crises, including more extensive useof collective action clauses and improvedpractices in sovereign debt restructuring.

Moving Forcefully on the DohaDevelopment Agenda

Trade barriers: a major impediment to devel-opment. Improving market access for devel-oping country exports would provide a majorboost to economic growth and accelerateprogress toward reducing poverty and meet-ing other MDGs. At present, trade barriers indeveloped countries effectively discriminateagainst developing countries in many cases.They are highest on products of major exportinterest to developing countries. Protection inagriculture is a multiple of that in manufac-

O V E R V I E W


turing. Taking into account both tariff andnon-tariff barriers—including domestic sub-sidies but excluding technical product regula-tions—average protection in agriculture inhigh-income OECD (Organisation for Eco-nomic Co-operation and Development) coun-tries in 2001 was 25.6 percent, compared to3.6 percent in manufacturing. Both borderbarriers (tariffs) and domestic subsidies con-tribute significantly to the high protection inagriculture, but the former have a muchlarger impact. Protection is particularly highon key individual products. OECD countries’protection rates for sugar are frequentlyabove 200 percent, and their support to sugarproducers of $6.4 billion a year roughlyequals developing country exports. In theEuropean Union, producer support for beef isas high as 84 percent of the value of domesticproduction. U.S. subsidies to cotton growerstotaled $3.6 billion in 2001–02, twice asmuch as all U.S. foreign aid to Africa, andcost West African cotton growers an esti-mated $250 million by depressing prices.

Within manufacturing, while average pro-tection is low, tariff peaks and escalation dis-criminate against developing country exportsand efforts to move up the value chain. Inclothing, for example, tariff peaks average 16to 17 percent in Canada, Japan, and the UnitedStates. More than 60 percent of imports sub-ject to tariff peaks originate in developingcountries. The incidence of contingent protec-tion—antidumping actions—also is higheragainst developing countries, on average.

Trade policy reform: source of large gains.Gains from a significant removal of thesetrade barriers would be substantial, both fordeveloping and developed countries. Strongergrowth resulting from a pro-developmentoutcome of the Doha Round could increasereal income in developing countries by $350billion by 2015 (roughly equivalent to theentire GDP of Sub-Saharan Africa), and liftan additional 140 million people out ofpoverty by that year (a decline of 8 percent).The bulk of these potential income gains—as

much as 70 percent—would arise from liber-alization in agriculture.

Liberalization of services trade, includingmigration: source of substantial additionalgains. Gains from liberalization of trade inservices, especially the temporary move-ment of workers, could be a multiple ofthose from liberalization of merchandisetrade, according to some estimates. Servicesoverall are the fastest growing componentof developing country exports. Services pro-vided over telecommunications links andthrough migrant workers show particulardynamism. Workers’ remittances, estimatedat $93 billion in 2003, are now the second-largest source of private external fundingfor developing countries, behind foreigndirect investment. Against this background,the recent build-up of protectionist pressureagainst services imports in some developedcountries is a cause for concern. This build-up is exemplified by new legal norms in theEuropean Union (EU) and pending legisla-tion in the United States that could limitoutsourcing of government contracts, forexample.

Timely and pro-development outcome of theDoha Round: a critical need. Putting theDoha Round back on track must be accordedthe highest priority. Developed countries,because of their weight in the system, need tolead by example. Bilateral or regional agree-ments are a poor alternative to a forwardmovement on the multilateral front. Agree-ment on some focal points or targets for tradepolicy reform would provide a needed impe-tus. Such focal points could include completeelimination by high-income countries of tar-iffs on manufactured products by a targetdate; complete elimination of agriculturalexport subsidies and complete decoupling ofall domestic agricultural subsidies from pro-duction, and reduction of agricultural tariffsto, say, no more than 10 percent, by a targetdate; and commitments to ensure free cross-border trade in services delivered through

O V E R V I E W


telecommunications networks, complementedby actions to liberalize the temporary move-ment of service providers. At the same time,reform should aim to achieve greater trans-parency and predictability in trade policy, bylimiting the use of less transparent instrumentssuch as specific tariffs, simplifying regulatoryrequirements, and imposing greater disciplineon the use of contingent protection.

Any incorporation of rules relating todomestic regulations such as competition andinvestment policies (the so-called Singaporeissues) into World Trade Organization(WTO) trade agreements needs to ensure thatthe rules support development and take intoaccount the different implementation capaci-ties of developing countries. A flexibleapproach is warranted. An example of suchflexibility is the agreement reached in 2003 toclarify the Trade-Related Intellectual Prop-erty Rights (TRIPS) Agreement to expandpoor countries’ access to essential drugs atlow cost. Relatedly, support to developingcountries should be stepped up to build theirinstitutional capacities to deal with the trade-related agenda and take advantage of bettermarket access opportunities. The IntegratedFramework for Trade-Related TechnicalAssistance is a useful initiative in that con-text. “Aid for trade,” and complementarymeasures to facilitate technology transfers todeveloping countries, can have high impact,and will be needed to enable poor countriesto realize the potential gains from globaltrade reforms discussed above. Some of thesecountries will also need assistance in adjust-ing to a reduction in trade preferences fol-lowing further nondiscriminatory tradeliberalization, and to the potential effects of asignificant increase in world food prices—should that materialize.

Providing More and Better Aid

Need for a substantial increase in ODA. Offi-cial development assistance (ODA) needs torise well beyond current commitments. At cur-rent levels of ODA, there is a large gap between

the development ambitions of the internationalcommunity and the resources provided. Anincrease in aid is critical for low-income coun-tries, to support their reforms and enhancetheir prospects for achieving the developmentgoals. Aid also plays an important role in mid-dle-income countries by reinforcing domesticefforts to tackle concentrations of poverty andcountering negative shocks. Against this back-ground, it is encouraging to see aid volumesbegin to reverse the decline of the past decade.ODA rose in 2002 and, according to prelimi-nary estimates, again in 2003. Indications ofincreased assistance from the donor commu-nity in follow-up to Monterrey, if realized,would raise ODA by about $18.5 billion by2006 from the 2002 level of $58 billion. Thiswould increase total net ODA to 0.29 percentof donors’ gross national income (GNI), upfrom the 2002 level of 0.23 percent. This isindeed welcome, but well short of what isneeded as part of the global compact to achievethe MDGs.

Country-level analysis conducted recentlyby the World Bank indicates that, as a con-servative estimate, an initial increment of atleast $30 billion could be used effectively.Early commitment of this additional sumwould help create a virtuous circle by encour-aging developing countries to undertake andsustain deeper reforms, which would makeaid still more productive. As countriesimprove their policies and governance andupgrade their capacities, the amount of addi-tional aid that can be used effectively wouldrise into the range of $50 billion plus a yearthat estimates suggest is likely to be necessaryto support adequate progress toward theMDGs. Ongoing work to examine the meritsof various options, such as an internationalfinancing facility, to mobilize the substantialadditional resources that are needed and canbe effectively used to achieve developmentresults is therefore important and timely.

It is useful to view these numbers in con-text. An additional $50 billion would raiseODA relative to donors’ projected GNI in thelatter half of the 2000s to roughly the same

O V E R V I E W


level as at the turn of the 1990s (levels in ear-lier decades were still higher). Since then, con-ditions for effective use of aid in developingcountries on average have improved, thanksto better policies. Donors’ income levels havealso risen. Ironically, as aid has become moreproductive and donors’ capacity to give hasgrown, aid amounts have declined sharply.This does not mean that all donors havereduced their assistance. The aid effort varieswidely across members of the OECD Devel-opment Assistance Committee (DAC), rang-ing from a high of 0.96 percent of GNI in thecase of Denmark to a low of 0.13 percent inthe case of the United States in 2002 (how-ever, the United States has more recentlyincreased its aid commitments, which wouldraise its net ODA in 2006 by about 50 per-cent over the 2002 level).

While aid volumes are rising again, there issome concern that much of the increase maybe dominated by strategic considerations—the war on terrorism, conflict and recon-struction in Afghanistan and Iraq. Largeamounts have recently been committed forthese purposes, but it is unclear whether all ofthese commitments represent an increase intotal aid or are in part a reallocation of aidfrom other countries. In the period ahead, itwill be important to ensure that developmentaid is not crowded out by aid influenced bysuch strategic objectives.

Improving the allocation of aid. Most donorstoday are more selective than they were abouta decade ago; they are allocating more aid tocountries with better policies and morepoverty. However, there is considerable vari-ation among donors. On average, multilat-eral institutions are more than three times asselective as bilateral donors, based on a newlydeveloped index that measures both policyand poverty selectivity in aid allocation. Mul-tilateral assistance is much more sharply tar-geted to good policies and to poverty, withthe International Development Association(IDA) being the most selective. About two-thirds of total ODA is bilateral aid. Among

bilateral donors, the Nordic countries, theNetherlands, and the United Kingdom are themost selective (with Denmark the highest).The index shows that some of the largestdonors in absolute size, such as France andthe United States, have not been particularlyselective along either the policy or povertydimension. Japan is selective on policy butnot on poverty, reducing the overall selectiv-ity of its aid. Thus while the typical donor hasimproved its aid quality in terms of targetingmore funds to poor countries with better poli-cies and governance, this cannot be said forthe typical aid dollar, as the largest donors inabsolute amount are less selective. Lookingahead, actions now being taken by some ofthese donors, such as the establishment of theMillennium Challenge Account (MCA) in theUnited States, are expected to contribute tofurther improvements in aid allocation. TheMCA aims to improve aid effectiveness bytying increased assistance to performance.

Efforts to target aid better need to takeaccount of the special needs of conflict-affectedand other low-income countries under stress.The challenge is to balance issues of weak poli-cies and institutions with the need to maintaincritical engagement. Appropriately timed anddirected aid, sensitive to local efforts to rebuildand the institutional capacity constraints, canplay a very useful role in these situations. Well-timed aid can also be quite productive follow-ing adverse exogenous shocks, helping to limitthe diversion of development resources intoshort-term relief efforts.

Increasing the effectiveness of aid throughimproved alignment and harmonization.Related to better allocation of aid across coun-tries, the effectiveness of aid depends cruciallyon its alignment with national developmentpriorities within country programs and on har-monization and coordination of donor policiesand procedures around the recipient country’sown systems. In low-income countries, thePRSP provides the framework for strategicalignment with country-owned and -led prior-ities and for achieving better coherence and

O V E R V I E W


coordination in donor support activities. Thecountry-led alignment and harmonizationefforts in Tanzania and Vietnam, centered onthe PRSP, provide good examples. Aid align-ment and harmonization efforts were given animpetus by the High-Level Forum on Harmo-nization held in Rome in February 2003.Important follow-up work is now being carriedout by the donor community jointly under theauspices of the DAC Working Party on AidEffectiveness and Donor Practices (WP-EFF),including elaboration of a set of indicators ofprogress. Results from this work will be impor-tant to widening the application of good prac-tices and better monitoring progress.

Providing aid in forms that are responsive tocountry circumstances and needs. As coun-tries build a track record of policy perfor-mance, their efforts should be supported withtimely, predictable, and longer-term aid com-mitments. Such commitments would enablethem to embark upon sustained reforms andinvestments necessary to meet the MDGs,with assurance that needed support would beforthcoming. Aid should be provided informs that can flexibly meet countries’ needsfor incremental financing. Currently, onlyabout a third of bilateral ODA is available forprogram and project expenditures in recipi-ent countries. The rest is allocated to specialpurposes such as technical cooperation, debtrelief, emergency and disaster relief, food aid,and costs of aid administration. These spe-cial-purpose grants accounted for almost allof the $6 billion nominal increase in ODA in2002. (In real terms, the increase was about$4 billion.)

Going forward, a much higher proportionof additional aid will need to be provideddirectly to countries in the form of cash sothat it can be deployed in accordance withcountry priorities to finance the costs of meet-ing the MDGs. Where country circumstanceswarrant, and budget frameworks are sound,more aid could be provided in forms thatallow for the financing of recurrent costs,either through budget or sectorwide support,or through targeted assistance to well-

designed sectoral programs. Many activitiesin education and health that are crucial toprogress toward the MDGs involve in majorpart an expansion of recurrent spending. Toensure debt sustainability in heavily indebtedcountries that are pursuing good policies,consideration should be given to providing alarger share of additional aid in the form ofgrants.

Debt relief and debt sustainability. Muchprogress has been made under the HIPC Ini-tiative in reducing heavily indebted poorcountries’ debt and debt service burden andcreating fiscal space for much-neededincreases in poverty-reducing spending.While most acute in the case of the HIPCs,the issue of achieving and maintaining debtsustainability is of broader concern to low-income countries. Work is under way at theIMF and World Bank on a debt sustainabilityframework that is intended to provide guid-ance on issues relating to financing strategiesfor low-income countries, including the rangeof indicators for assessing debt sustainability,the role of policies in determining appropri-ate debt thresholds, the importance of includ-ing domestic debt in such assessments, andthe appropriate mix of grants and new cred-its. These issues are becoming increasinglyimportant in the light of the need for largeincreases in external financing to meet theMDGs and implications for country debt sus-tainability. Debt sustainability is not only aresource flow issue, however; it also dependscrucially on increasing growth, expandingand diversifying exports, improving access toglobal markets, and mitigating the effects ofexogenous shocks.

Stepping Up Action on Key Global Public Goods

As globalization has advanced, and aware-ness has grown of the international spilloversof local actions and conditions, there hasbeen a welcome increase in attention to areasfor global collective action. Reference hasbeen made in the foregoing to several such

O V E R V I E W


areas—control of infectious diseases, promo-tion of education and dissemination ofknowledge, opening up of the internationaltrade regime, promotion of a more stableinternational financial system. In all of theseareas, there is progress to report, but, asnoted, there is also need to step up action.

Stronger resolve needed to address environ-mental concerns. One key area for global col-lective action, and directly related to theMDGs, is environmental sustainability. Devel-oped countries bear much of the responsibilityfor the preservation of the global environmen-tal commons, as they are the largest contribu-tors to the degradation of the commons andpossess the financial and technical resourcesneeded for prevention and mitigation. Devel-oping countries must also play their part byimproving their environmental management,including by increasing regional cooperationamong themselves that donors could support.While there has been good progress on pro-tecting the ozone layer, thanks to implementa-tion of the Montreal Protocol, much lessprogress has been made in most other areas—greenhouse gas emissions, biodiversity, fish-eries. Aid to developing countries to supportimproved environmental practices, both bilat-erally and through multilateral vehicles, hasdeclined after a short-lived increase followingthe 1992 Rio Convention. Not all advancedcountries have shown weak resolve in address-ing the environmental challenges; there aregood global citizens such as Sweden andSwitzerland. Looking ahead, priorities includestronger and more concerted action on green-house gas emissions and increasing aid todeveloping countries in support of environ-mental sustainability, including through theGlobal Environment Facility (GEF).

Improving Policy Coherence for Development

Cutting across the policy areas is the need toimprove the overall coherence of policies inhigh-income countries in terms of theirdevelopment impact. All too often, there are

contradictions in policies, with support todevelopment provided in one area defeatedby actions in others. There are both collec-tive and individual country examples: $58billion in ODA by the OECD countries isundermined by five times as much protectionto domestic agricultural producers; advocacyof and support for private sector develop-ment and export diversification in develop-ing countries are blunted by systematicescalation of tariffs on higher-value importsfrom those countries; Norway’s stellar per-formance as an aid donor coexists with themost restrictive agricultural trade policyregime among the OECD countries; a similarcontradiction between aid and trade policiesapplies to the European Union; the AfricanGrowth and Opportunity Act of the UnitedStates was undercut by its 2002 Farm Billand its higher protection against importsfrom low-income and least-developed coun-tries than from the rest of the world.

Institutionalizing policy coherence. Therealization that development policy extendswell beyond aid and specific trade prefer-ences is leading to welcome indications thatthe developed countries are willing to lookbroadly at the policy areas that affect devel-opment—trade, aid, foreign investment andother capital flows, migration, knowledgeand technology transfer, environment—andto put in place institutional arrangementsthat would help ensure coherence. A note-worthy development in this context is thepassing into law of an “integrated globaldevelopment policy” in Sweden in January2004 that calls for the country’s aid, trade,agriculture, environment, migration, secu-rity, and other policies to be aligned withthe objective of reducing poverty and pro-moting sustainable development. Anothernotable development has been the issuanceby Denmark of the first in a planned seriesof reports on how it is contributing to thegoal of establishing a global partnership fordevelopment (the aim of MDG 8). Prepara-tion of similar reports is being consideredby some other OECD members, including

O V E R V I E W


Belgium, Canada, Finland, Germany, theNetherlands, Norway, and Sweden.

Two related ongoing initiatives at the OECDare a “Horizontal Project” on policy coherencefor development that looks at the impact ondeveloping countries of a broad range of devel-oped-country policies, and increased attentionin DAC peer reviews to issues of policy coher-ence. Work on these issues is also being under-taken by private think tanks and civil society,among them the Center for Global Develop-ment and the World Economic Forum. Theseencouraging efforts would prove very valuableif they were instrumental in bringing aboutmore systematic attention to issues of develop-ment impact and coherence in policymaking inrich countries.

Priorities for InternationalFinancial InstitutionsHow are the international financial institu-tions (IFIs) contributing to the achievement ofthe MDGs and related outcomes? This reportassesses the IFI contribution along fourdimensions: country programs, global pro-grams, partnership, and results. Applying theframework across the IFIs suggests that therehas been progress in recent years, especiallysince Monterrey. But the evidence is incon-clusive on the critical questions of compara-tive performance and whether the whole ofthe IFI contribution is larger (or smaller) thanthe sum of the individual IFI contributions.Going forward, greater availability and com-parability of evaluation data will facilitatemonitoring, and in turn improve the qualityof reporting to the taxpaying public in allcountries. The joint work program on resultsendorsed by the multilateral developmentbanks (MDBs) in February 2004 at the Mar-rakech Roundtable on Managing for Devel-opment Results should provide a key vehiclefor progress. Future assessments of the IFIrole will also benefit from greater use ofexternal evaluations.

Individual IFIs. Within each of the institutions,the evidence points to enhanced attention to

client focus and country ownership, trans-parency, results, and accountabilities. Natu-rally, the degree of progress varies acrossinstitutions and areas. Highlights include thestrong start of the IMF’s Independent Evalua-tion Office—whose creation means that all theIFIs now have independent evaluation officesreporting directly to their executive boards—and its reviews of prolonged use of Fundresources, the Fund’s role in capital-accountcrises, and fiscal adjustment in Fund-supportedprograms; the World Bank’s commitment tothe results agenda and its focus on actions andimplementation in countries, with partners andwithin the Bank; and the regional developmentbanks’ respective commitments to enhancedoperational quality, development effectiveness,and results, as demonstrated recently in Mar-rakech. But there is clearly no room for com-placency. The recent progress needs to continueand be deepened.

� IMF. For the Fund, the priority is to con-tinue to refine its role in assisting low-income countries, in several ways: byadapting its instruments of financial andtechnical support to enable its low-incomemembers to catalyze other donor assistance,deal with postconflict situations, respond toexogenous shocks, absorb the cost ofadjustment to multilateral trade liberaliza-tion, and establish institutions that willenable them to gain access to private financ-ing. The Fund’s work agenda also aims atstrengthening the design of Fund-supportedeconomic programs in low-income coun-tries, while enhancing alignment with thePRSP. A third element of the Fund’s ongo-ing work, together with the World Bank, isto develop an effective and flexible frame-work for assessing debt sustainability inlow-income countries.

� World Bank. The Bank’s country supportpriorities are to continue to work withcountries and partner agencies to deepenthe PRSP process as a basis for the designof its assistance strategies in low-incomecountries; to adapt approaches and instru-ments to the evolving needs of middle-

O V E R V I E W


income countries; and to complete themajor agenda the Bank has set out onresults, harmonization, and simplification.Supporting and complementing the deep-ening of country-led approaches in theBank’s assistance strategies is the strength-ening of its analytic, knowledge, and advo-cacy work. A priority for Bank-supportedglobal and sectoral programs is the imple-mentation of an effective framework forappraisal, statistics, monitoring, and eval-uation that is every bit as strong as theframework for country programs.

� Regional Development Banks. The otherMDBs also have large agendas beforethem—in their country programs and intheir support for regional public goods. Allneed to complete their ongoing reformsassociated with the results agenda, as theyset out at the Marrakech Roundtable. Inaddition, like the World Bank, they need toapply greater efforts to strengthen theoverall governance and accountabilityframework for their regional and sectoralprograms.

Systemic coherence. Looking across the IFIs,the evidence also points to progress collec-tively—both institutionally and in day-to-day work at the country level. Bank-Fundcollaboration and coordination among theMDBs are smoother and more productivethan five years ago. In tandem with theincrease in partnership and coordination,there also is a healthy movement towardgreater specialization in line with institu-tional comparative advantage. This reversesthe trend of the early 1990s when overlaps inIFIs’ and other agencies’ capacities increased,as the consensus on the comprehensivenessof the development paradigm was beginningto grow. But the “gains from trade” betweenand among IFIs have not all been harvestedyet. Opportunities include increased selectiv-ity of agency programs in line with compar-ative advantage, harmonization of agencypractices around national poverty reductionstrategies and systems, and joint evaluationsof their support.

Progress is needed also in the ongoingwork related to IFI governance. Strengthen-ing the voice and participation of developingcountries in the IFIs is part of the Monterreycompact.

Priorities for Strengthening the Monitoring ExerciseFully developing the potential of the globalmonitoring exercise will require a strongfocus on specific challenges at the country,agency, and global levels for meeting the pri-orities set out above. This will in turn requirecontinuing work to strengthen the statisticaland analytic foundations of the exercise andto deepen collaboration with partner agen-cies. Three areas for further work are partic-ularly important:

� Data. Timely statistics on the desireddevelopment outcomes and good metricsfor the key policy drivers are critical foreffective monitoring. At present, there aremajor gaps in data, especially with respectto human development and infrastructureservices and outcomes in developing coun-tries. The World Bank and its partneragencies in the U.N. system, workingtogether and in consultation with clientcountries, have developed a time-boundand costed action plan to strengthen abroad range of data in developing coun-tries and build their statistical capacities.Presented and agreed at the MarrakechRoundtable, the plan will need timely andcoordinated donor support for its objec-tives to be realized.

� Research. More research is needed tostrengthen the analytic underpinnings ofthe monitoring framework, especially thelinks between policies and outcomes.While there is a broad consensus on themain policy and institutional determi-nants of growth, poverty reduction, andthe other MDGs, less is known about theprecise transmission mechanisms and therelative weights of the various determi-nants and interrelationships among them.

O V E R V I E W


Collaborative research is under way at theWorld Bank, the IMF, and the U.N. Mil-lennium Project to better model and quan-tify some of these relationships, especiallythrough in-depth work at the countrylevel. More research is also needed on crit-ical issues such as aid effectiveness and ondevelopment of more robust metrics forkey policy areas such as governance andfor the impact on developing countries ofrich country policies.

� Partnership. In these and other areas, col-laboration developed with partner agen-cies—the United Nations, other MDBs, theWTO, OECD, the European Commisson—in the preparation of this first global mon-

itoring report will be deepened further,building on respective agency mandates andcomparative advantage and ensuring thatthe approach to monitoring is coherentacross agencies. This could include broad-ening the framework for reflecting thecontribution of the multilateral agencies,currently focused on the IFIs, to encompassother agencies as well, even as efforts aremade to strengthen the evaluation of theIFI contribution, including through betterharmonizing the self- and independent-evaluation criteria used by the IFIs. Col-laboration will also be expanded with civilsociety, which has become increasinglyinvolved in monitoring activities.

O V E R V I E W


IFramework

The purpose of the Global MonitoringReport is to assess progress on thepolicies and actions needed to achieve

the Millennium Development Goals andrelated outcomes. Covering developing anddeveloped countries as well as multilateralagencies, the assessment aims to facilitate theDevelopment Committee’s monitoring of theaccountabilities of the entities jointly respon-sible for reaching the development goals.1 Italso aims to inform the international com-munity more broadly on progress on thedevelopment policy agenda and the prioritiesfor action.

The report is organized as follows. Part Isets out the context and analytic frameworkfor monitoring progress on the policy agendafor achieving the MDGs. It also assessesprospects for attaining the goals based oncurrent trends. That assessment informs thesubsequent presentation of the need to scaleup policies and actions. Part II moves on toevaluate policies in developing countries. Thefocus is on low-income countries, where thechallenge to achieve the development goals isgreatest, but the discussion also covers issuesimportant to middle-income countries—among them reducing economic volatility andpromoting sustainable growth, and address-ing concentrations of poverty and deprivationat the subnational level. Part III assesses the

policies of developed countries, suggestingactions they need to take to fulfill their com-mitments in support of the development goals.Part IV looks at the role of multilateral agen-cies, asking how they are supporting coun-tries’ progress toward the development goals,both individually and collectively.

Background on GlobalMonitoringDiscussions at major international meetingsat the beginning of the new millennium pro-duced a broad consensus on key developmentgoals and on the strategies and partnershipsneeded to achieve them. The 2000 U.N. Mil-lennium Summit led to the adoption of theMillennium Declaration and the MillenniumDevelopment Goals, 8 clear goals and 18 tar-gets (p. xxii) by which the international com-munity could measure progress on keydimensions of development.2 Since theiradoption, the MDGs have rallied public sup-port for development. Reinforcing the agree-ment on the development goals, meetings inDoha, Monterrey, and Johannesburg in 2001and 2002 contributed to the emergence of ashared understanding of the broad develop-ment strategy and policies needed to attainthe goals, as well as of the roles of the differ-ent development partners.3 The meeting in


Monitoring Framework1

Monterrey ushered in a new compact betweendeveloping and developed countries—theMonterrey Consensus—that stressed theirmutual responsibilities and accountabilities inthe development effort.

With broad agreement on the goals andstrategies, attention soon focused on imple-mentation, to convert the ideas and sharedapproaches into concrete actions and to mon-itor progress on the delivery of commitmentsunder the compact. It was against this back-ground that the Development Committee, atits meeting in September 2002, expressed itsintent to regularly monitor progress on thepolicies and actions needed to achieve theMDGs and related development outcomes.To reinforce the accountabilities of the keyactors—developing countries, developedcountries, and multilateral agencies—theCommittee asked the World Bank and theInternational Monetary Fund to prepare pro-posals for monitoring MDG-related policies,while recognizing the role of the UnitedNations in monitoring the MDGs.4

In response to the request, the staffs ofthe Bank and Fund proposed a monitoringframework that the Development Commit-tee discussed at its April 2003 meeting.5 Aprogress report on implementation of themonitoring framework, including issuesrelated to the measurement and assessmentof relevant policies and actions and collab-oration with partner agencies, was pre-sented to the Development Committee at itsSeptember 2003 meeting.6 The frameworkenvisages an annual Global MonitoringReport—to be prepared jointly by Bank andFund staff in advance of the Committee’sspring meeting—supplemented by interimreports on selected issues.

This first annual report is part of a broadarchitecture of monitoring and reporting thathas emerged since the Millennium Declarationand the Monterrey Consensus. Individualinternational agencies lead the monitoringwork in the areas of their respective mandatesand expertise, while supporting and comple-menting the work of partner agencies. With itsfocus on policies and actions, the Global

Monitoring Report complements the UnitedNations’ monitoring of the MDG targets andindicators. The latter includes an annual reportby the Secretary General to the General Assem-bly on the implementation of the MillenniumDeclaration and an annual report with a bien-nial comprehensive review by the SecretaryGeneral to the General Assembly on the imple-mentation of the agreements reached at theInternational Conference on Financing forDevelopment in Monterrey.7 Complementingthese reports, the U.N. Millennium Project is athree-year research project—to which theBank and the Fund are also contributing—designed to identify approaches for achievingthe MDGs. The Millennium Campaign isaimed at mobilizing public support for theMDGs and fostering country ownership.

In preparing this report, Bank and Fundstaff have worked closely with the UnitedNations and other partner agencies, includingthe other multilateral development banks(MDBs), the World Trade Organization(WTO), the Organisation for Economic Co-operation and Development (OECD), and theEuropean Commission (EC). The collabora-tion has spanned the initial design of themonitoring framework and the subsequentassembling of the supporting data and analy-sis. All of the partner agencies participated inan Inter-Agency Workshop on Global Moni-toring organized by the Bank and the Fund inWashington in June 2003.8 Going forward,collaboration will be deepened further, build-ing on respective agency strengths and ensur-ing that the approach to monitoring iscoherent across agencies. Collaboration alsowill be expanded with civil society organiza-tions that have become increasingly involvedin monitoring activities.9

To establish a baseline for future assess-ments, this report covers policies and actionsfor achieving the MDGs and related out-comes in a comprehensive manner. Futurereports may be more selective and thematic,focusing on identified priorities and majorchanges from the baseline established here.

Timely statistics on desired outcomes andgood metrics for key policies are critical for

P A R T I : F R A M E W O R K


effective monitoring. But there are importantgaps in both areas, especially in countries mostat risk of not meeting the MDGs. Filling inthose gaps calls for systematic investment inmore robust and timely data and more preciseindicators of the relevant policies and theirimpact. Research to strengthen the analyticalunderpinnings of the monitoring framework isalso needed—especially into the links betweenpolicies and outcomes. In particular, whilethere is broad consensus on the main policyand institutional determinants of growth,poverty reduction, and the other MDGs, asdescribed in the next section, less is knownabout the precise transmission mechanisms,interrelationships among the determinants,and their relative weights and elasticities withrespect to the development goals.

The Bank and the Fund are working withtheir partners on initiatives to address thesegaps.10 Progress has been made in developinga time-bound and costed action plan forinternational support for strengthening abroad range of data in developing countriesand building their statistical capacities (box1.1). Progress and needed follow-up work inmeasuring specific policies and their impactand supporting research are noted in therespective sections of the report. The ongo-ing work will strengthen the statistical andanalytic foundation for monitoring. How-ever, results will take time to materialize. Theglobal monitoring exercise, therefore, shouldbe seen as a phased process, the full potentialof which will appear only in the mediumterm. With more timely and reliable underly-ing data, better metrics for key policies, andmore clarity on the links between policiesand outcomes, the findings and recommen-dations reported to the Development Com-mittee will become increasingly specific andquantified.

Complementing the annual monitoringreports to the Development Committee, anopen information platform accessible throughthe Development Gateway will present thedata underlying the reports for monitoring bythe broader international community on anongoing basis.11

Framework Linking Policies to Development GoalsThe MDGs reflect the multidimensionalnature of development. They span incomeand important non-income dimensions ofdevelopment, which are interlinked. Higherincomes and less poverty mean better humandevelopment outcomes. Better health andeducation contribute to increased productiv-ity and higher incomes.12 There are importantlinkages among the human developmentgoals. Perhaps the strongest determinant ofchild mortality, for example, is the mother’seducation.13 Health and nutritional statusaffect the probability that a child will enrolland succeed in school.14 And the humandevelopment outcomes are affected greatly byconditions in other sectors. Child mortality isclosely linked to access to safe water andbasic sanitation.15 Roads and other transportfacilitate access to education and health facil-ities, electrification of those facilitiesimproves the quality and effectiveness of theservices they provide, and refrigeration net-works help preserve vaccines.16

The multidimensionality of the MDGs,interlinkages among them, and their multi-sectoral determinants imply that the policyagenda for the achievement of these goals issimilarly broad. Indeed, it spans the gamut ofdevelopment. The central element of theagenda is promotion of stronger economicgrowth. Growth directly reduces incomepoverty and expands resources available forprogress toward the other MDGs. No coun-try has made rapid progress toward thesedevelopment goals without robust growth. Inmost low-income countries, reaching thegoals will require improvements in policiesand governance to achieve stronger economicgrowth than recently attained or currentlyprojected. It will also require policies specifi-cally targeted to enhancing the capabilities ofpoor people to participate in growth, espe-cially through improved access to education,health, and related key services. While eco-nomic growth has a significant effect on edu-cation and health outcomes, the magnitude of

M O N I T O R I N G F R A M E W O R K




The focus on measurable development goals, as embodied in the MDGs, and clear accountabilityfor policy, as reflected in the Monterrey Consensus, implies the need for more and better statisticsto inform policy and monitor progress. But the new demands placed on national statistical systemsexceed the capacity of many countries. External support has helped, but often progress has not beensustained. Today many national statistical systems underperform, in part because they receive inad-equate domestic funding and uncertain, sometimes conflicting, donor support.

Recognizing the need for action, the statistics community is responding, at both the national andinternational levels. Learning from experience, the new initiatives emphasize responsiveness tocountry needs and priorities and country-level capacity building. Most encouraging are nationalefforts embedded in broader national development strategies, such as the PRSP. Notable initiativesat the international level include the PARIS21 (Partnership in Statistics for Development in the 21stCentury) consortium that brings together users and producers of data, the U.N. MDG IndicatorsExpert Group to pool agencies’ data and expertise, improved international data standards, and trustfunds and lending facilities to support investment needs in national statistical capacity building.

Building on these initiatives, and reflecting collaborative work across agencies, a global actionplan was discussed and agreed at the second Roundtable on Managing for Development Resultsheld in Marrakech on February 4–5, 2004. The plan recommends six short- and medium-term setsof actions aimed at achieving tangible and sustainable improvements in national and internationalstatistical capacity.

The first set of actions addresses national needs:

� Action 1. Mainstream strategic planning of statistical systems and prepare national statisticaldevelopment strategies for all low-income countries by 2006

� Action 2. Begin preparations for the 2010 census round � Action 3. Increase financing for country-level statistical capacity building.

The second set addresses international responsibilities:

� Action 4. Set up an international Household Survey Network� Action 5. Undertake urgent improvements needed for MDG monitoring by 2005� Action 6. Increase the accountability of the international statistical system.

Cost estimates have been prepared for the implementation of the action plan in 2004–06. Theannual incremental cost of improvements to national statistical systems is estimated at $115 mil-lion to $120 million. These costs are extrapolated from a limited number of countries based onrecent experience or expert opinions. For many of the poor countries, external financing will be nec-essary. Additional spending by development agencies for improvements in the international systemis estimated at $24 million to $28 million a year.

There is need for a clear and strong commitment on the part of the international community toa shared work program over the next three to five years for improving development data and build-ing statistical capacity in developing countries. The tasks specified in the action plan are not exhaus-tive, nor are they intended to preclude other initiatives. But they do require commitment on the partof the international community to work together, to share resources, and to keep the needs and pri-orities of developing countries at the forefront. In keeping with the spirit of the Monterrey com-pact, countries that set realistic goals for improvements in their statistical systems and make areasonable commitment of their own resources should receive commensurate assistance from theinternational community.

Source: A full description of the action plan is available at:http://www.managingfordevelopmentresults.org/Agenda-Seminar2.html.

BOX 1.1 An action plan for improving development statistics

the effect is typically smaller than on incomepoverty. Prospects for progress on the humandevelopment goals also depend on the scaleand effectiveness of development interven-tions targeted directly toward these goals.17

And policies and programs need to take intoaccount implications for environmental sus-tainability. Although the developing countriesthemselves are in the driver’s seat, imple-menting this agenda will require increasedcooperation at the global level, in whichstronger reform efforts by developing coun-tries are matched with enhanced supportfrom developed countries and internationalinstitutions.

The consensus forged at Monterrey inMarch 2002 was based on a recognition thata strong global development partnership wasneeded to achieve the MDGs. It called ondeveloping countries to improve their policiesand governance and on developed countries tostep up their support, noting that the goalscalled for scaling up actions to acceleratedevelopment on the part of both of these

groups. Are the parties to the consensus deliv-ering on their commitments? It is the purposeof this report to provide an answer. Anchoredin the Monterrey Consensus, the global mon-itoring framework translates the agenda forachieving the MDGs and related developmentoutcomes into a set of policies and actions, oraccountabilities, of developing and developedcountries, and a corresponding set of respon-sibilities of international agencies, that will bemonitored on a regular basis.

The analytic framework underpinning theglobal monitoring exercise is outlined in fig-ure 1.1. It depicts the results chain leading tothe MDGs, running from the inputs of poli-cies and actions of developing and developedcountries and international agencies to theintermediate outcomes of economic growthand delivery of human development andrelated key services to poor people and on tothe final outcomes—the MDGs themselves.The main elements of the framework, and thepolicies and actions to be monitored, are dis-cussed below.



Improved delivery of servicesto poor people

Key intermediate outcomesin developing countries

Policies and actionsAgenda anchored in theMonterrey Consensus

Stronger and sustainableeconomic growth M

DGs

Developed countries• Enhance support to developing

country reform efforts and global public goods

Developing countries• Create a good enabling climate for

economic activity• Empower and invest in poor people

Development agencies• Better align support to delivery

of development results

FIGURE 1.1 Framework linking policies and actions with development outcomes

Developing Countries

Research and experience have led to a broadconsensus on an effective strategy for develop-ment, one that is country owned and countryled, promotes economic growth, and ensuresthat poor people participate in it and benefitfrom it—and that produces maximum progresstoward the MDGs. The strategy has two inter-linked and mutually reinforcing pillars. Thefirst is an enabling climate for economic activ-ity that encourages private firms and farms toinvest, create jobs, and increase productivity.The second is empowerment of and investmentin poor people. Improvements in the economicclimate spur growth but also expand opportu-nities for the poor. Empowerment of poor menand women through improved access to edu-cation and health fosters social inclusion butalso promotes growth as these groups increasetheir participation in economic activity.

Enabling climate for economic activity. Thefirst pillar focuses on the process of thegrowth. Entrepreneurship, investment, andinnovation by the private sector drive eco-nomic growth. But they will not occur with-out the right environment. A good economicclimate spurs domestic economic activity andattracts investment from abroad. Such a cli-mate is based on several policy elements:

� Sound macroeconomic policies—sustain-able fiscal, monetary, and exchange-ratepolicies and prudent debt management—promote sustainable economic growth withlow inflation and instill confidence amonginvestors. By fostering a stable economicenvironment, sound policies encourage sav-ing and investment. Good macroeconomicpolicies also play an important role in avert-ing the high costs of financial crises in termsof lost output and increased poverty.Emerging markets in particular need tomaintain policies that sustain investor con-fidence and avoid the risk of capital flight.

� Openness to trade expands opportunitiesfor growth and diversification and spurs

innovation. Development success has gen-erally been associated with an outward ori-entation with respect to trade policies, withtrade acting as a crucial engine of growth.18

� Good governance and institutions are essen-tial to competitive and efficient functioningof markets, providing a level playing fieldand predictability and enforceability of con-tracts. This is especially important forsmaller firms, which are less able than largefirms to finance the costs of dealing withexcessive and arbitrary regulations.

� Stronger and deeper financial markets chan-nel savings to productive uses and broadenaccess to finance. A sound financial sectorcomplements good macroeconomic man-agement in maintaining economic stabilityand strengthening an economy’s resilience toshocks.

� Investment and productivity also dependon the availability of key physical infra-structure—transport, power, telecommuni-cations. In addition to facilitating growth,infrastructure services contribute directlyto the development goals, as when accessto safe water and sanitation reduces childmortality.

Empowerment of and investment in poorpeople. The second pillar of the developmentstrategy is concerned with increasing thecapabilities of poor people, by widening theiraccess to key services and fostering socialinclusion:

� Better access of the poor to education andhealth care, and better quality of these ser-vices, expand opportunities for them toimprove their own well-being. This callsfor allocating adequate public resourcesto spending on human capital develop-ment and, equally important, improvingpublic sector management and gover-nance, including building the capacity ofrelated institutions, to ensure the effectivedelivery of these services. Good gover-nance is key to the effectiveness of publicservices more generally.



� Also important is the poor’s access tosocial protection—well-targeted safetynets that protect poor and vulnerable peo-ple from unforeseen shocks and disloca-tions occasioned by necessary reforms.

� The effectiveness of any strategy toempower and invest in the poor willdepend to an important degree on mecha-nisms that foster their participation indecisions that affect them. Enabling thepoor to have an increased voice needs to bean integral element of the strategy.

� Cutting across this agenda is the empow-erment of women by removing barriers totheir fuller participation in the develop-ment process. Promoting gender equalityis not only an important social goal itself,but is also essential for the achievement ofthe development goals more broadly.19

Environmental sustainability. Both the en-abling economic climate and empowermentof the poor have environmental dimensions.The principles of sustainable developmentmust be fully integrated in economic poli-cies. Environmental concerns are addressedthrough an agenda similar to that outlinedabove, spanning policies that generate theright incentives for private agents (using andcreating markets, through instruments suchas taxes, user charges, concessions, and trad-able permits), regulations (particularly fortoxic substances or where market-basedinstruments are impractical), capable insti-tutions, and engagement of the public. Theimportant linkages between environmentaland other development outcomes, such aspoverty alleviation and human develop-ment, should inform policymaking. Propermanagement of natural resources protectsrural incomes, just as access to water andsanitation reduces child mortality.

It is clear from the foregoing that attain-ment of the MDGs will depend on progress indeveloping countries across a broad range ofpolicies and institutions. The framework formonitoring developing-country policies clas-sifies the policy agenda into four clusters: eco-

nomic and financial policies; public sectorgovernance; human development; and policiesfor environmental sustainability. These arerepresented by the four facets of the develop-ing-country “policy diamond” in figure 1.2.

Developed Countries

In calling for a global partnership for devel-opment (MDG 8), the international commu-nity recognized that attaining the agreeddevelopment outcomes would require, inaddition to redoubled reform by developingcountries, enhanced support from the devel-oped world. Developed countries help bymaintaining macroeconomic and financialpolicies that promote stability and growth inthe world economy. But the Monterrey Con-sensus also envisages increased support intwo key areas that affect outcomes in thedeveloping world more directly—trade andaid. In addition, developed-country policiesgreatly affect outcomes in areas of global col-lective action, such as the preservation of theglobal environmental commons. The globalmonitoring framework focuses on these key



Environmental management• Policies and institutions

Public sector governance• Public expenditure and revenue management • Public administration • Transparency, accountability, control of corruption

Humandevelopment• Education and health • Social protection • Voice/inclusion• Gender

Economic and financial policies• Macro• Trade

• Private sector regulatory and institutional environment• Financial sector • Infrastructure

FIGURE 1.2 Monitoring: dimensions of developing-country policies

aspects of developed-country policies, asdepicted by the developed-country “policydiamond” in figure 1.3.

Macroeconomic policies. Macroeconomicpolicies and outlook in developed countriesexert a major influence on developing coun-tries’ growth prospects. Growth in developedcountries stimulates activity in developingcountries through trade and creates thepotential for raising aid flows. Macroeco-nomic policies and outcomes also affectdeveloping countries through their impact onprivate capital flows, which include not onlyforeign direct investment and portfolio debtand equity flows but, increasingly, workers’remittances. Fluctuations in real GDP indeveloping countries tend to be synchronizedand positively associated with GDP fluctua-tions in large developed countries. Macro-economic policies in advanced countries thatsupport a stable and growing world economy,therefore, make an important contribution toprogress toward desired development out-comes in developing countries.20

Trade policies. Better access to developed-country markets is crucial to the develop-ment of poor countries. Many developingcountries that have opened their traderegimes are prevented from reaping the bene-fits because of barriers maintained by devel-oped countries, especially in agriculture,textiles, and clothing—labor-intensive sectorswhere developing countries typically have acomparative advantage. Reducing trade pro-tection can yield large benefits for developingand developed countries alike.21 Liberalizingtrade in services, including through the so-called Mode 4 trade (temporary migration ofworkers), also promises large potential gains.

The Doha Development Agenda providesan opportunity for major progress on marketaccess for developing-country exports, aswell as for further liberalization by develop-ing countries themselves. Liberalization mustbe supported by increased trade-relatedfinancial and technical assistance to helpdeveloping countries with the “behind-the-border” investments they need to make totake advantage of wider market access and toenhance their institutional capacity to makeand implement effective trade policy.

Aid. The parties to the Monterrey Consensusagreed that more and better aid would benecessary to meet the MDGs. Various esti-mates prepared in the past year indicate thatthe additional aid requirements are sizable.A major factor determining whether aid iseffective is of course the recipient country’sown policies; hence the focus on developing-country actions, including sound macroeco-nomic management, structural reform, andimproved institutions and governance. Suchpolicies not only enhance the effectiveness ofaid, but also help alleviate the macroeco-nomic instability that can result from a largeinflux of aid. Recent and ongoing reforms inmany developing countries have improvedthe setting for effective use of aid.22 A recentstudy of better-performing developing coun-tries showed that, as a conservative estimate,additional assistance of $30 billion or more



Macrofinancial policies• Policies supportive of economic

growth and stable capital flows

Aid• Quantity and

quality of aid• Debt relief

Trade policies• Agriculture and

manufactures• Services (including

temporary movement of workers)

• Trade-related assistance

Global public goods• Environment• Others

FIGURE 1.3 Monitoring: dimensions of developed-country policies

can be productively absorbed.23 The effec-tiveness of aid also depends importantly onthe policies of donors. As underscored at theHigh Level Forum on Harmonization held inRome in February 2003, and in the follow-up work of the Development AssistanceCommittee Working Party on Aid Effective-ness and Donor Practices, there is large scopefor increasing the impact of aid through astronger focus on selectivity in aid allocation,tighter alignment of aid with nationalpoverty reduction strategies and priorities,provision of aid in forms that are responsiveto recipients’ needs, and closer harmoniza-tion of donor policies and practices.24 Whatis needed now is consistent implementationof that consensus and concerted monitoring.

Heavily indebted poor countries need debtrelief to create the fiscal space to allocatemore resources to social spending and otherprograms to reduce poverty and accelerategrowth. That need underscores the impor-tance of continued progress on the HeavilyIndebted Poor Country (HIPC) Initiative. Ifheavily indebted countries are to avoid slip-ping back into unsustainable debt positions,they will have to continue to strengthen theirpolicies to promote increased efficiency inresource use and stronger economic growth,and new external assistance will have to beprovided on appropriately concessional terms.Stronger export growth is especially impor-tant to such countries, underscoring the needfor better market access and improved trade-related infrastructure.

Global public goods. Developed countriesalso need to step up action in support of keyglobal public goods and areas for global col-lective action. In some areas, notably the pre-vention and treatment of infectious andcommunicable diseases, especially HIV/AIDS,investment in global public goods will provideessential support for national programs.Major issues relating to the global environ-mental commons—mitigation of globalwarming (tropical countries are particularlyvulnerable to climate change and environ-

mental degradation) and preservation of bio-diversity—also call for stepped-up efforts atthe global level. Developed countries have aspecial contribution to make to protecting theglobal environment, as they dominate energyuse, consume the most natural resources, andproduce the most pollutants. Of course,developing countries also need to play theirpart, including through increased regionalcooperation among themselves that aiddonors should support. International financialstability, another important global publicgood, can be enhanced by strengthening theinternational financial architecture to improvecrisis prevention and resolution and avoid thespread of contagion between countries.

International Financial Institutions

The international financial institutions (IFIs)have an important role to play in helpingcountries achieve the MDGs and relateddevelopment outcomes. How are they play-ing their part? The monitoring framework forassessing their contribution is focused on fourkey dimensions: country programs—strategicalignment with country-owned and country-led strategies for poverty reduction and otherMDGs, and design of programs and instru-ments that support those priorities; globalprograms—support for global public goodssuch as international financial stability,health, environmental sustainability, an opentrade system, and spread of knowledge; part-nership—effectiveness in working together,through harmonization of policies and prac-tices, strategic selectivity in country supportprograms, and coordination; and results—orientation to quality and results in opera-tions and transparency in evaluation andreporting. These four monitoring dimensionsare represented in the IFI “policy diamond”in figure 1.4. The assessment also looksacross individual agencies and addressesissues of systemic performance and coher-ence—whether the overall effectiveness of themultilateral agencies is larger or smaller thanthe sum of individual agency contributions.



Country Focus and OwnershipThe elements of the global compact and theimplied policy agenda for achieving theMDGs must come together at the countrylevel within coherent country developmentstrategies that respond to local conditionsand priorities and are nationally owned. Pol-icy priorities vary across countries and time,and policies also vary in terms of the relativeweights they carry with respect to the differ-ent development goals. It is in the process ofarticulating and integrating a country devel-opment strategy that priorities are identifiedand coherence achieved. Country ownershipand leadership of the development strategyare crucial to effective implementation.

For low-income countries, the PovertyReduction Strategy Papers (PRSPs) are the pri-mary avenue of expression of such a country-



Country programs• Strategic alignment–with country-owned

priorities for poverty reduction and other MDGs• Relevance and selectivity in program design

Partnership• Harmonization of

policies and practices

• Coordination of support programs

Global programs• Support to country

capacity building for RPGs/GPGs

• Anchor role in international system for key RPGs/GPGs

Results• Focus on quality and results• Systems for monitoring,

reporting, and evaluation

FIGURE 1.4 Monitoring: dimensions of development agency support

The MDGs are global, but their implementation must occur at the country level, through country-owned and -led development strategies. In low-income countries, the Poverty Reduction StrategyPaper is the vehicle through which country policies and programs and resource requirements arelinked to the MDGs.

Although there is no required set of indicators or goals that must be included in PRSPs, the tar-gets specified are expected to be framed against the backdrop of the Millennium DevelopmentGoals. They also need to take account of initial conditions and national priorities so that they setout credible policies and programs for making progress toward the goals. As they prepare PRSPs,therefore, countries need to balance their long-term aspirations linked to the MDGs with the needto formulate concrete measures that can be feasibly implemented in the context of their annual plan-ning cycles. Thus, for example, Vietnam’s PRSP (called the Growth and Poverty Reduction Strat-egy) articulates the government’s commitment to the MDGs but specifies goals and targets (theVietnam Development Goals) that reflect extensive work to adapt progress toward the MDGs tolocal conditions.

Overall, most PRSPs do reflect fairly good coverage of the MDG topics, and that coverage hasimproved over time. All PRSPs completed by July 2003 included targets related to poverty head-count, school enrollment, and maternal health. Over 90 percent of the strategies covered child mor-tality and water access. On the other hand, coverage of the MDG indicators for women’s voice,income distribution, biodiversity, and housing was only around 20 percent. Moreover, while theindicators chosen in PRSPs are often consistent with the MDGs, they frequently differ from the spe-cific indicators chosen to track the MDG targets.

Data limitations and methodological issues associated with the time spans covered by PRSPs(three to five years) and the MDGs (up to 2015) make it difficult to compare the respective targets.However, a recent review by World Bank staff suggests that in some areas (malnutrition, access towater) targets set out in full PRSPs tend to be at least as ambitious as the MDG targets. On the otherhand, for child and maternal mortality targets, PRSPs often set less ambitious targets than the MDGtargets. At the same time, PRSP targets are generally quite ambitious compared to historical trends.

Box 1.2 Strengthening the links between PRSPs and the MDGs



owned and -led development strategy. Inmiddle-income countries, the policy integra-tion and prioritization role is performedwithin respective national strategy frame-works. By the end of March 2004, 37 coun-tries had prepared and were implementingfull PRSPs; 16 more had prepared InterimPRSPs. Increasingly the MDGs are being in-corporated in national PRSPs, and the PRSPprocess itself is being deepened along variousdimensions—participatory processes, growthstrategies, public expenditure management,poverty and social impact analysis. Contin-ued strengthening of the PRSP process, anddeepening of the links with the MDGs, will bekey to the grounding of the policy agenda forthe achievement of the development goals atthe country level, in country-owned strategies

in which countries spell out their commit-ments to policy and institutional reforms andwhich donors can commit to support in acoherent and consistent way (box 1.2).

Notes1. The Development Committee is the Joint

Ministerial Committee of the Boards of Gover-nors of the World Bank and the InternationalMonetary Fund on the Transfer of RealResources to Developing Countries.

2. United Nations 2000, 2001.3. WTO 2001; United Nations 2002a, 2002b.4. Development Committee 2002.5. Development Committee 2003a, 2003b,

2003c.6. Development Committee 2003f, 2003g.7. United Nations 2003.

For poverty, malnutrition, immunization, child mortality, and access to water, most countries forwhich data are available have set PRSP targets at levels that entail a substantial acceleration ofimprovement relative to current rates of progress.

Looking ahead, it is important that low-income countries (with support from their developmentpartners) continue to strengthen the links between their PRSPs and the MDG targets. One way ofdoing this, suggested in the 2003 PRSP Progress Report, is for countries to develop alternative sce-narios that illustrate how improved policies and governance and more and better aid would helpthem in accelerating progress toward the MDGs.

Linking the PRSP approach more closely to the MDGs will require work on several fronts, andcountries will need support from external partners in the process. First, PRSPs will need to reflectmedium-term goals that are linked to the MDG targets for 2015. These goals should be sufficientlyambitious, relevant to country circumstances, and based on national consultations undertaken aspart of the PRSP process. Second, the goals should be related to a medium-term program of policy,institutional, and financing requirements that will include, in addition to the normal growth agenda,necessary infrastructure investments, service-delivery improvements, capacity upgrades, and mar-ket-access enhancements. Third, the medium-term program will need to be linked to a medium-term budget, including a public expenditure program and financing requirements. Fiscal constraintswill need to be considered, together with options for increasing aid volumes and changes in aidmodalities. Fourth, donors will need to be willing to make upfront commitments to close themedium-term financing gap in ways that are consistent with a realistic assessment of the country’sspending priorities and debt sustainability. As long as the country’s implementation of its PRSPremained on track, its program would be supported by higher and more predictable aid flows pro-vided in ways appropriate to meeting the MDGs.

An enhanced approach along these lines could initially be applied in a group of focus countriesbased on their relative readiness. The experience in implementing the approach would providelessons, which could then be used to extend the approach to other countries.

Sources: Development Committee 2003d; Harrison, Swanson, and Klugman 2003.

8. For the proceedings of this workshop, seethe World Bank’s global monitoring Web site athttp://www.worldbank.org/.

9. See the papers submitted to the March 2003and September 2003 Development Committeemeetings (Development Committee 2003a, 2003b,2003f) for a fuller account of the related monitor-ing activities of the partner agencies and collabo-ration with them, as well as the relevant activitiesof the civil society.

10. Details of this work were set out in theglobal monitoring implementation report pro-vided to the September 2003 Development Com-mittee meeting (Development Committee 2003f).

11. Development Gateway Web site:www.developmentgateway.org.

12. Ends in themselves, better education andhealth also are means to stronger economicgrowth because they raise workers’ productivity.One study estimated that a 10 percent improve-ment in life expectancy at birth can raise the percapita income growth rate by 0.4 percentage point(WHO 2001). Barro (2001) estimated that anadditional year of schooling can raise the growthrate by about 0.5 percentage point.

13. A large part of this effect might not be gen-eral education but specific health knowledge, butthis is usually acquired using literacy and numer-acy skills learned in school, as a study in Moroccofound (Glewwe 1999).

14. A study in the Philippines found that aone–standard deviation increase in early-age childhealth increased subsequent test scores by aboutone-third of a standard deviation (Glewwe andKing 2001). An evaluation of school-based masstreatment for deworming in rural Kenya foundthat student absenteeism fell by one-fourth(Miguel and Kremer 2001).

15. An eight-country study found that goingfrom no improved water to “optimal” wateraccess was associated with a 6-percentage-pointreduction in the prevalence of diarrhea in childrenunder three years of age (Asrey 1996). Lack ofaccess to water supply and sanitation kills an esti-mated 2.1 million to 3.5 million children a year(WHO 2002).

16. The interlinked nature of the MDGs andtheir multisectoral determinants are illustrated bya recent study of child mortality, which found that,holding other factors constant, child mortalitydeclines by 3 to 4 percent if access to drinking

water improves by 10 percent, by 3 percent if yearsof schooling among women rise by 10 percent, by0.8 to 1.5 percent if government health spendingrises by 10 percent, by 1 to 1.5 percent if the den-sity of paved roads rises by 10 percent, and by 2 to3 percent if per capita income growth rises by 10percent (Wagstaff 2003).

17. The growth elasticity of income poverty istypically between 1 and 2, whereas that of childmortality and primary enrollment is around 0.5(Filmer and Pritchett 1999; Pritchett and Summers1996; Schultz 1987). These elasticities of courseare not fixed; they can be higher with better poli-cies and governance.

18. For a recent review of evidence on the roleof trade in growth and poverty reduction, see Bergand Krueger (2003).

19. Studies find that gender equality con-tributes to better education and health outcomes(see, for example, Behrman and others 1999).Research on Sub-Saharan Africa shows thatgreater gender equality in farm inputs couldincrease farm output by up to 20 percent (Udry1996). More recent cross-country research hasfound that gender inequalities in education impedeeconomic growth (Klasen 2002).

20. It is estimated that a 1 percent increase inoutput growth in G-7 countries has on averagebeen associated with a 0.4 percent increase indeveloping country growth (IMF 2001).

21. Results from modeling of a pro-developmentoutcome of the Doha Round show that strongergrowth associated with the removal of significantimpediments to merchandise trade and slashing oftariff peaks could mean as many as 140 millionfewer people living in dire poverty in 2015—adecline of 8 percent. Agricultural subsidies also havea huge cost for the developed countries. It is esti-mated that these subsidies cost the average work-ing family in the European Union, Japan, and theUnited States more than $1,000 a year (WorldBank 2003).

22. For evidence on increased productivity ofaid in developing countries, see Goldin, Rogers,and Stern (2002).

23. Development Committee 2003e. See chap-ter 11 for more discussion of additional aidrequirements.

24. See Rome Declaration on Harmonization,February 25, 2003, at http://www1.worldbank.org/harmonization/romehlf.



Past the halfway mark to the 2015 targetdate, prospects for achieving the Mil-lennium Development Goals raise both

optimism and grave concerns. Based on cur-rent trends and growth forecasts, the goalof halving extreme poverty and hunger willbe achieved—although Sub-Saharan Africaremains highly vulnerable. A few countries,notably China, already have reached the goal;some others, including India, are on track tomeet it. But progress has stalled in somecountries, and others are at severe risk offalling short of the goal. In Sub-SaharanAfrica, on current trends, only eight countriesare projected to halve extreme poverty by2015.

Progress on non-income poverty goals hasbeen considerably slower.1 The least progresshas been made in child and maternal mortal-ity and in providing access to safe water andsanitation. In hardly any region are thosegoals likely to be met by the target date of2015. Only 16 percent of countries, home to22 percent of the developing world’s popula-tion, are on track to meet the under-five mor-tality target.

Prospects for reaching the goal of univer-sal primary education are better, but the goalremains a challenge. The East Asia and Pacificregion has nearly met the goal; Europe andCentral Asia and Latin America and theCaribbean are on target to achieve it. The

Education for All Fast Track Initiative islikely to accelerate progress toward universalprimary school completion, but shortfalls arestill likely in Sub-Saharan Africa. While therehas been global progress toward gender par-ity in education, particularly at the primarylevel, the 2005 goal of parity in primary andsecondary education may not be met in abouta third of developing countries even by 2015.

Accelerating economic growth is central toachieving the MDGs. The link between eco-nomic growth and income poverty is particu-larly strong, but growth is also an importantdeterminant of human development out-comes. For faster growth, developing coun-tries will need to implement substantial policyand institutional reforms. Particularly impor-tant will be actions to improve the regulatoryand institutional framework for private sec-tor activity, strengthen infrastructure, andenhance the quality of governance.

Making progress on the human develop-ment goals will also require efforts directed atimproving the delivery of key services—health, education, and water and sanita-tion—especially to poor people. Investmentin human development will need to be scaledup, together with reforms in public sectorgovernance and capacity building to improveservice delivery.

The effectiveness of interventions will alsodepend on taking into account cross-effects


MDG ProspectsReasons for Optimism, Grave Concerns

2

among goals. Health and nutritional status,for example, directly affect a child’s proba-bility of school enrollment, and access to safewater and sanitation is critical for child sur-vival. The existence of such cross-effectsmeans that improvements in one goal dependalso on progress on other goals. Interlinkagesare also evident across sectors. Sectoral inter-linkages imply that isolated interventionsmay do little to achieve goals if bottlenecksremain in other sectors. Progress on the goalof universal primary education, for example,may depend importantly on better trans-portation, as well as increased spending onschools. Accordingly, national reform pro-grams must be built on multisectoral analysisand anchored in coherent country develop-ment strategies.

As envisioned in the Monterrey Consen-sus, national efforts will need to receivestronger support from external partners. Pri-ority areas of support are more open access toexport markets and more and better aid.

The human development goals present adaunting challenge, but they are not unreach-able. Past successes, such as halving adultilliteracy in the last 30 years and extendinglife expectancy by 20 years over the last 40years, suggest that much can be achieved.2

Growth and Prospects for Achieving the MDGsEconomic growth is central to achieving theMDGs, particularly the goal of reducingincome poverty. Economic growth reducespoverty because average incomes of the poortypically tend to rise proportionately with theaverage income of the population.3 Thisresult is robust over time and across countriesand regions.

Higher national income contributes tohuman development goals as well. At themacro level, growth generates increased pub-lic resources to improve the quantity andquality of education, health, water supply,sanitation, and other services. At the house-hold level, as growth reduces income poverty,demand for schooling, health care, and other

services rises. Thus income, health, and edu-cation reinforce one another. More incomeleads to better human development out-comes, while better health and education canlead to increased productivity and higherincomes.4

Growth Outlook

The outlook for growth over the next decadeis promising—although with continuing widedisparities across regions.5 Should the currentfavorable trend in growth continue, mostregions will see an acceleration in per capitaincome growth—some substantial (figure2.1). East Asia is likely to continue its stronggrowth in the medium term, with an eventualmoderation in growth as some countries inthe region begin to approach the level ofincome of OECD countries. The robust turn-around in Europe and Central Asia shouldcontinue—although at a more moderate pacethan that of the last five years—with theaccession of several countries of the region tothe European Union and also as the transitionfrom planned economies advances.

There is some evidence that countries inSub-Saharan Africa have turned the corner.Between 1997 and 2002, 24 of 46 countriesin Sub-Saharan Africa—representing 53 per-cent of the region’s population—experiencedpositive per capita income growth.6 Reformsundertaken over the last few years, and a shifttoward more manufacturing exports, shouldbenefit African economies—although contin-ued conflict and the HIV/AIDS epidemic willweigh on their prospects. To achieve theMDGs, growth in the region will have toaccelerate much further (see below).

The other regions also will benefit frompast and ongoing reforms—a more stablemacroeconomic environment and greateropenness—to boost their long-term growthrates, with South Asia leading the pack. LatinAmerica is the only developing region to havefaltered in the late 1990s and early 2000s—mostly weighed down by the lengthy crisis inArgentina, with related contagion to the restof the region. A hopeful sign is that popula-



tion growth is slowing in the developingworld: average growth is expected to beabout 1.1 percent over the next decade(2005–15), below the 1.4 percent rate of the1990s.

Drivers of growth. Underlying the growthscenario are structural shifts—first fromsubsistence agriculture to low-skilled man-ufacturing and later, as investment acceler-ates and skills are upgraded, to intermediateand finished goods of increasing technicalsophistication.7 A dramatic decrease in bar-riers to trade and cross-border investmentin recent decades has made these structuraltransformations possible.8 Greater open-ness has been accompanied by improve-ments in investment climate. Inflation hasdropped dramatically across the world aspolicymakers have come to understand thenegative impacts of high inflation. Govern-ment balances have improved, contributingto both macroeconomic stability and moreeffective use of public resources in supportof development. Workers have become bet-ter educated and healthier; infrastructurehas grown; and institutions have improved.

The contribution of productivity to growthwill become more important in the future.First, millions of low-productivity agriculturalworkers—particularly in Asia and Africa—constitute a labor reserve for industrial and ter-tiary activities. Second, productivity in mostdeveloping countries lags far behind that inindustrial countries—a significant opportunityfor convergence as global trade and investmentlinkages convey the knowledge and technolo-gies needed for greater productivity.

The scenario presented above is useful pri-marily as a reference point. Hypotheticaldeviations from it provide insights into thepotential range of outcomes—particularlythose driven by changes in specific policies.For example, full global merchandise tradereform has the potential to boost the averagegrowth rate in developing countries by up tohalf of a percentage point—depending onhow trade reforms deal with the factors thataffect productivity.9 But the converse is alsotrue. Numerous factors could lead to slowergrowth than that assumed in the referencescenario—either on a cyclical or sustainedbasis, and with a global, regional, or morelocalized impact. Examples include a rise in

M D G P R O S P E C T S : R E A S O N S F O R O P T I M I S M , G R A V E C O N C E R N S


8

6

4

2

0

–2

1990–20002001–032004–15

East Asia& Pacific

Perc

ent p

er y

ear

SouthAsia

LatinAmerica

&Caribbean

Europe &Central Asia

MiddleEast &

North Africa

Sub-SaharanAfrica

Middle-income

countries

Low-income

countries

FIGURE 2.1 Growth prospects improve, but not enoughReal per capita growth by region, 1990–2015

Source: World Bank staff estimates.

protection, failure to address long-term struc-tural imbalances, regional conflicts, andfinancial crises.

Implications for the Poverty MDG

The scenario outlined above would meet theMDG of halving global poverty by 2015.10

The poverty headcount index would fall fromits 1990 level of 27.9 to 12.5 percent. Thenumber of poor people (on a $1-a-day basis)would fall below 735 million, from about1.22 billion in 1990 (table 2.1).11 China andIndia, which had the largest number of poorin 1990 but have achieved high per capitagrowth rates, would account for most of theglobal success. China already has achievedthe poverty goal; India is well on its way.

But even if the global poverty goal is met,many individual countries, particularly inSub-Saharan Africa, will fail to achieve thetarget (figure 2.2). Even in regions withstrong overall performance, some countriesare seriously off track. It is also likely thatsome countries that meet the income povertytarget at the national level may leave someareas or groups behind. Special effort willneed to be made to address the situation fac-ing vulnerable groups (box 2.1).

The Sub-Saharan Africa region is the leastlikely to achieve the income poverty MDG,having made little or no progress on reducingthe incidence of poverty over the 1990s. Theincidence of poverty is much lower in LatinAmerica and the Caribbean, but progress infurther reducing poverty has been slow.Faster improvement will be needed if theregion is to reach its target. In Europe andCentral Asia, poverty rates rose in the 1990sfrom low average levels as the region experi-enced a sharp drop in income in the early partof the decade. These rates are now declining,and by 2015 will be about a third of currentlevels. In these regions, either growth willneed to accelerate or the poverty elasticity ofgrowth will need to rise to boost prospects formeeting the MDG target. The challenge ismuch greater for Sub-Saharan Africa. There,in order to reach the income poverty goal,GDP growth will have to double relative tothe base scenario, or the poverty elasticity ofgrowth will have to rise by a large factor.

Sub-Saharan Africa faces low per capitagrowth and a relatively low poverty elasticity.Of the 43 countries in the regional sample,28, representing two-thirds of the region’spopulation in 2000, had per capita growthrates of 2 percent or less (table 2.2). On cur-



TABLE 2.1 Growth rates and decline in poverty by region, through 2015

Annual averagegrowth rates (2004–15) Population living under $1 per day

Scenario Headcount (percent) Number of persons (millions)

Region Per capita GDP GDP 1990 2001 2015 1990 2001 2015

East Asia and Pacific 6.0 6.3 29.6 15.6 2.3 472 284 44China 6.4 7.0 33.0 16.6 3.0 377 212 41

Europe and Central Asia 3.9 3.8 0.5 3.7 1.3 2 18 6Latin America and Caribbean 2.3 3.7 11.3 9.5 7.6 49 50 46Middle East and North Africa 2.3 4.1 2.3 2.4 1.2 6 7 4South Asia 4.0 5.3 41.3 31.1 16.4 462 428 268Sub-Saharan Africa 1.8 3.8 44.6 46.5 42.3 227 314 366

Total 3.7 4.7 27.9 21.3 12.5 1,219 1,101 734Excluding China 3.0 4.1 26.1 22.8 15.4 841 888 692


rent trends, only four countries—with some16 percent of the region’s population—arelikely to see per capita growth of more than 3percent. Mirroring the growth figures, onlytwo countries in the region are expected to seean improvement in the headcount index of 50percent or more between 2000 and 2015,12

with six more countries achieving an improve-ment between 30 and 50 percent. Together,these eight countries represent somewhat lessthan 15 percent of the 2000 population.

The two countries with the best povertyperformance have relatively modest percapita growth rates (between 2 and 3 per-cent), but have a relatively high poverty elas-ticity. Both also have the more equal incomedistribution, with Gini coefficients under

0.2.13 Pro-growth policies would help theregion, but pro-growth policies that favor thepoor would have a much greater impact onthe incidence of poverty.

Why some countries and regions have suc-ceeded in reducing poverty while others havelagged was the subject of a major conferenceon scaling up poverty reduction held in May2004 in Shanghai. The review of a number ofcountry experiences, based on case studies,provided new insights into what works andwhat doesn’t and what countries can learnfrom the experience of others. The main pur-pose of the conference, cohosted by the Gov-ernment of China and the World Bank, wasto encourage South-South learning aboutscaling up poverty reduction. Many of the



Europe & Central Asia

50

40

30

20

10

01990 1995 2000 2005

Middle East & North Africa

2010 2015 1990 1995 2000 2005

South Asia

2010 2015

50

40

30

20

10

01990

29.6

14.815.6

2.3 3.70.30.5 1.3

2.31.2

2.4 1.2

Perc

ent o

f to

tal p

opul

atio

nPe

rcen

t of

tota

l pop

ulat

ion

1995 2000 2005

East Asia & Pacific

2010 2015 1990 1995 2000 2005 2010 2015

Latin America & Caribbean

1990 1995 2000 2005

Sub-Saharan Africa

2010 2015

1990 1995 2000 2005 2010 2015

31.1

20.6

41.3

16.4

22.3

46.544.6

42.3

9.511.3

5.6

7.6

Goal2015

Poverty rate at $1 a dayActual and forecast Path to goal20011990

FIGURE 2.2 Most regions will reach the goal of halving poverty by 2015, but Sub-Saharan Africa is seriously off trackActual and projected decline in number of people living on less than $1 a day by region, 1990–2015




East Asia has done remarkably well in meeting several MDGs. Seven of the nine countries for whichadequate information is available either have already achieved the poverty reduction goal or are ontrack to achieve it. East Asia is also broadly on track to meet the goal on universal primary educa-tion, with enrollments already comparable to many high-income countries.

But progress has been uneven: across goals, across countries, and across socioeconomic groups andgeographical areas within countries. The region’s performance on health, particularly child immuniza-tion and maternal mortality, does not match its progress on income poverty and education. The regionis not on track to cut the child mortality rate by two-thirds. In China, under-five mortality is estimatedto have fallen from 49 per 1,000 live births in 1990 to 39 in 2001; however, if there is no accelerationin this pace of improvement, China will not meet the child mortality target of 16 for 2015. On the otherhand, middle-income Southeast Asian economies such as Malaysia and the Philippines saw significantdecreases in child mortality over the last decade and appear on track to meet MDG targets for 2015.

The region’s low-income countries are lagging behind in several dimensions (see accompanyingtable). And the persistence of disparities within countries with respect to most MDGs remains a con-tinuing challenge, even where goals have been or are about to be attained in the aggregate.

The region needs a differentiated approach that combines accelerated efforts to reach the MDGswhere they are in danger of being missed with a will to exceed particular targets in countries thatare on track. The so-called MDG-plus agenda could focus on faster reduction of national poverty,specify targets with respect to a particular geographic region or segment of the population, or spec-ify different indicators to capture the remaining challenges relating to a goal. In preparing its PovertyReduction Strategy Paper, the Government of Vietnam adapted the MDGs to the particular contextof Vietnam. In many respects, the adaptations represent an ambitious agenda, including reducingpoverty faster than envisaged in the MDGs, eliminating the gender gap not only in primary but alsoin secondary education and improving the quality of education not just access to it, and eliminat-ing the gap with ethnic minorities. Thailand, too, is developing an MDG-plus agenda for incorpo-ration into its Ninth Five-Year Plan. This agenda calls for reaching the poverty goal in the northeastof Thailand, the poorest region in the country.

BOX 2.1 East Asia and Pacific: Despite solid performance on MDGs, challenges remain

Prospects for reaching the MDGs in East Asia

Note: Prospects in the above table are assessments based on the current levels of these indicators and recent progress toward the targets.Outcomes in the future could change markedly from past trends.Source: UNESCAP/UNDP 2003; World Bank 2003a and 2003b.

CambodiaChinaFijiIndonesiaLao PDRMalaysiaMongoliaPapua New GuineaPhilippinesThailandVietnam

CountryPoverty Child

malnutrition

Primaryschool

completion

Genderequality in

school

Childmortality

Birthsattended

Immmunizationfor measles

Access tosafe water

* *

* *

* ** *

* *

* *

* *

* ** ** *

* * * *

* *

* *

* *

* *

* *

Seriously offtrack

Off track On track Achieved Insufficientdata



TABLE 2.2 Projected per capita growth and improvement in poverty in Sub-Saharan Africa, 2000–2015

Projected growth of consumption 2000 2000per capita, 2000–2015 Number of countries Population (millions) Population share (percent)

<1 12 233 37.21–2 16 190 30.32–3 11 103 16.43–4 4 101 16.1>4 0 0 0.0

Percent improvement in headcount 2000 2000index, 2000–2015 Number of countries Population (millions) Population share (percent)

<10 9 223 35.610–20 16 244 38.920–30 10 69 11.030–50 6 48 7.7>50 2 43 6.8


basic ingredients of a successful povertyreduction strategy are relatively well known.How countries and communities have appliedthem and with what results differs widely,however. There is no one size that fits all, andthe point of the conference was to enablepractitioners to learn about successful casesand draw their own conclusions about whatmay or may not be transferable to their ownunique situation.14

Hunger. Halving the proportion of the world’speople who suffer from hunger is part of Goal1. The target of halving the proportion ofunderweight children under five years of agebetween 1990 and 2015, the key measure ofprogress toward the goal, requires an averageannual reduction of 2.7 percent. Availabledata show that, as for income poverty, thetarget is achievable at the global level, butSub-Saharan Africa is seriously off track, asare a number of countries in other regions. InSub-Saharan Africa, on current trends, only17 percent of countries expect to reach thetarget.15 In some countries, notably in Europeand Central Asia, slippages are occurring,resulting in rising malnutrition, particularlyin the former Soviet Union.16

Prospects for Other MDGs

Economic growth has a significant effect onnon-income development goals—education,health—just as it does on income poverty.However, the magnitude of the effect on theformer is typically smaller than on the latter.Prospects for progress on the human devel-opment goals also depend importantly on thescale and effectiveness of development inter-ventions specifically directed toward them,such as policies that improve the delivery ofkey services. And the determinants of thesegoals are multisectoral. Compared to the typ-ical estimate of the elasticity between growthand income poverty of 1 to 2, the typical esti-mate of the elasticity between growth and theprimary completion rate or under-five mor-tality is around 0.5. However, these elastici-ties are not fixed and can be higher withpolicies and institutions that are better able totranslate growth into improved human devel-opment outcomes.17

Another reason that gives some cause foroptimism is that over time, education andhealth outcomes have been improving for thesame income level, reflecting advances inknowledge and technology, such as break-

throughs in immunizations against infectiousdiseases (figure 2.3). The changes are signifi-cant, even among poorer countries. Forexample, at a national income of $600 percapita, predicted child mortality fell from 100per 1,000 births to 80 between 1990 and2000—a full 20 percent.18

In the discussion that follows, the prospectsfor the achievement of non-income povertygoals are reviewed, based on the growth out-look described in the preceding section andcurrent assessments with respect to the statusof other determinants of these goals—relevantpolices and institutions, multisectoral linkages.

Primary education. The target for complet-ing primary school is 100 percent by 2015.While a few regions are close to the target, asmeasured by a new database on primarycompletion rates jointly developed by theWorld Bank and the United Nations Educa-tional, Scientific, and Cultural Organization(UNESCO), others are at risk of falling short

if current trends persist. East Asia and Pacificand Europe and Central Asia were fairlyclose to the target in 2000, with population-weighted completion rates of 97 and 93 per-cent, respectively (figure 2.4). Although bothregions would need to accelerate progressslightly over the trend observed in the 1990s,the required pace appears achievable.

Of the other four regions, Latin America hasregistered the fastest progress and is on tracktoward the goal. In South Asia and the MiddleEast and North Africa, however, progress onprimary completion rates is slower than onewould expect from the regions’ average percapita GDP growth. A sharp acceleration overrecent trends will be needed to meet the goal.Sub-Saharan Africa could miss the target by awide margin. Because the annual improvementin completion rates was just 0.2 percent in the1990s, an annual improvement of more than4.5 percent will be needed for the period2000–15 to achieve the target (table 2.3). Evenif the economic growth rate doubles, othermechanisms and policies will be needed toachieve the education target in the region. TheEducation for All Fast Track Initiative specifi-cally addresses the reform and financial needsof countries that are struggling to move towardthe education goals.

Gender equality. The third MDG seeks to pro-mote gender equality and empower women.The targets include achieving gender parity inprimary and secondary education preferablyby 2005 and at all levels of education by 2015.Progress at the primary level has been rela-tively good, with the ratio of girls to boysenrolled improving from 88 to 94 percentbetween 1990 and 2000. Girls’ enrollmentshave increased faster than boys’ in all regions,and in the three regions where gender in-equalities are greatest—Sub-Saharan Africa,the Middle East, and South and West Asia—disparities have eased. Nonetheless, the targetof gender parity in primary and secondary edu-cation by 2005 will not be met. Moderate toserious gender disparities against girls at theprimary level in about 35 percent of developing



6

5

4

3

2

14 5 76 8 9 10 11

Und

er-f

ive

mor

talit

y (p

er 1

,000

, log

)

1990

2000

GDP per capita (1995 US$, log)

1990 2000

FIGURE 2.3 Mortality at a given level of national income has beendecliningChanges in the association between GDP per capita andchild mortality, 1990 to 2000

Note: Lines show outcome as predicted by a nonlinear function of GDP per capita.Source: GDP per capita data from World Development Indicators database. Under-five mortality from UNICEF (2002).




1990 1995 2000 2005


2010 2015 1990 1995 2000 2005

South Asia

2010 2015

100

90

80

70

60

50

401990

100

83

51

100

8310010093

97100

79

74

100

68

50

69

9096

Perc

ent o

f re

leva

nt a

ge g

roup

Perc

ent o

f re

leva

nt a

ge g

roup

1995 2000 2005

East Asia & Pacific

2010 2015 1990 1995 2000 2005 2010 2015


1990 1995 2000 2005

Sub-Saharan Africa

2010 2015

1990 1995 2000 2005 2010 2015

1990 2000 2015Primary completion rate

Actual Path to goal

100

90

80

70

60

50

401990 1995 2000 2005 2010 2015

Goal

FIGURE 2.4 A few regions are close to the target on primary education; others are off trackPrimary education completion rates by region, 1990–2015

Source: United Nations; World Bank staff estimates.

TABLE 2.3 Primary education completion rates, progress needed by region (percent)

Annual improvement

Required Observed RequiredRegions 1990–2015 1990–2000 2000–15

East Asia and Pacific 0.16 0.10 0.20Europe and Central Asia 0.42 0.33 0.48Latin America and Caribbean 1.50 1.86 1.25Middle East and North Africa 0.95 0.50 1.25South Asia 1.55 0.85 2.03Sub-Saharan Africa 2.81 0.20 4.59


countries and at the secondary level in about 43percent of countries suggest that a number ofcountries may not reach this target even by2015 (table 2.4).19 Prospects for reaching the2015 goal of gender parity at all levels of edu-cation, including higher education, are remote.

Girls’ primary school completion rates andsecondary-school enrollments, especially, willneed to grow even faster than they did in the1990s. The fact that girls’ enrollments aregrowing at a faster pace than boys’ in mostcountries is encouraging, and suggests thatgender parity will hinge importantly on gen-eral actions by governments, particularly inSub-Saharan Africa and South Asia, to expandthe coverage and increase the effectiveness ofprimary and secondary schooling. But incountries where girls’ participation is lowest,targeted demand-side interventions may alsobe required, to reduce the direct and indirectcosts of girls’ schooling for households in need

of girls’ labor or concerned about their safety,particularly where schools are far away. Coun-tries such as Bangladesh and Guatemala haveshown the effectiveness of targeted stipends forgirls.

The prospects for achieving the broaderMDG on gender equality and women’sempowerment are even weaker than for theeducation targets. Almost one-quarter ofdeveloping countries are not expected toachieve parity in literacy by 2015, and limiteddata make it difficult to evaluate progress inother areas. Unemployment rates, for exam-ple, are higher for women than for men in 21of 37 countries with available data, and in nocountry with available data do women earnthe same as men.20 Women are also vastlyunderrepresented in national parliaments inall regions of the world. In only 14 countriesdo women hold more than 30 percent of seatsin national parliaments.21



TABLE 2.4 Distance from the goal of gender parity in primary and secondary education, by region, circa 2000(number of developing countries)

Have achieved Close to goal Medium positionGPI between GPI between GPI between Far from goal Total number

Region 0.97 & 1 0.95 & 0.96 0.80 & 0.94 GPI < 0.80 of countries

Primary educationEast Asia and Pacific 11 2 4 17Europe and Central Asia 18 2 2 22Latin America and Caribbean 19 4 2 25Middle East and North Africa 4 4 5 2 15South Asia 2 2 1 5Sub-Saharan Africa 10 5 14 12 41

Subtotal 64 17 29 15 125

Secondary education East Asia and Pacific 2 2 8 4 16Europe and Central Asia 17 2 2 1 22Latin America and Caribbean 4 4 13 2 23Middle East and North Africa 2 1 8 2 13South Asia 1 1 3 5Sub-Saharan Africa 3 16 17 36

Subtotal 28 10 48 29 115

Note: The Gender Parity Index (GPI) is the ratio of the gross enrollment ratio (GER) for girls divided by that for boys. The GER is the number ofpupils enrolled in a given level of education, regardless of age, expressed as a percentage of the population in the relevant official age group. Forthis table, in cases where girls’ GER exceeds that of boys (which would produce a GPI greater than 1), the inverse of GPI was used in order to allowGPI values to be compared on a scale with a maximum value of unity. Thus, this table registers gender gaps against boys as well as against girls.While at the primary level, all of the moderate to serious gender disparities are against girls, this is not true at the secondary level, where gendergaps are observed in both directions. Source: Adapted from UNESCO 2003.

Child mortality. Progress on child mortalityhas been particularly weak, and prospects forreaching the under-five child mortality goalare dim (figure 2.5). For developing countrieson average, under-five child mortality wasmeasured at roughly 80 deaths per 1,000 in1990. The goal is to lower this by two-thirdsby 2015. Progress is highly uneven acrossregions and countries, as well as within coun-tries, often reflecting differences in income.As of 2000, only 16 percent of developingcountries, with 22 percent of the developingworld’s population, were on track to meetthe target.22 At the regional level, only oneregion—Latin America and the Caribbean—is on track. The target is unlikely to bereached in other regions, including East Asia,

which will comfortably attain the incomepoverty target. Progress has been particularlyslow in Sub-Saharan Africa, where conflictand the HIV/AIDS epidemic have driven uprates of infant and child mortality in severalcountries. Not a single country in this regionis on track to meet the child mortality target.

Meeting the child mortality target calls foran average annual mortality reduction of 4.3percent during 1990–2015. Because improve-ments in this indicator have been slow, areduction of this magnitude will be challeng-ing in the best of environments.23

In some countries the gap in child mortal-ity is widening between the poor and the bet-ter-off. Results on the disparities betweenrich and poor (comparing the bottom and





1990 1995 2000 2005


2010 2015 1990 1995 2000 2005

South Asia

2010 2015

180

150

120

90

60

30

0

180

150

120

90

60

30

0

1990

26

54

171

59

341815

3644

20

77

99

43

129 178

5344

59

Dea

ths

per

1,00

0 liv

e bi

rths

Dea

ths

per

1,00

0 liv

e bi

rths

1995 2000 2005

East Asia & Pacific

2010 2015 1990 1995 2000 2005 2010 2015


1990 1995 2000 2005

Sub-Saharan Africa

2010 2015

1990 1995 2000 2005 2010 2015

Goal Under-five mortalityActual201520011990 Path to goal

1990 1995 2000 2005 2010 2015

FIGURE 2.5 Prospects for reaching the child mortality goal are dimMortality rate for children under age five by region

top quintiles) for both infant and child mor-tality for four countries—Armenia, Bolivia,Cambodia, and Central African Republic—suggest that poverty underlies much of thelack of progress in improving the health statusof children.24 Low incomes, ignorance, andlimited access to effective services contributeto the problem. Advances on intermediatefactors that affect well-being (immunizationcoverage, availability of effective diagnosisand treatment, water availability, responsiveprovider behaviors) will be required to makea serious dent in child mortality.

Maternal health. Progress on reducing mater-nal mortality has been insufficient, andprospects are rather bleak. The MDGs callfor a full three-quarters reduction in maternalmortality between 1990 and 2015, equivalentto an annual reduction of 5.4 percent. Withan average annual decline of 3.2 percent inthe 1990s, the developing world as a whole isoff target.25 Only 17 percent of countries,with 32 percent of the developing world’spopulation, are on track to achieve the target.Poverty, distance, limited hospital facilities,and uneven quality of services inhibitprogress and require investment in a broadrange of areas to make meaningful advancesin saving mothers’ lives. The cost of suchinvestment is high and likely out of reach forthe poorest countries, where maternal mor-tality rates are highest.

HIV/AIDS, malaria, and other diseases. Themost prevalent and problematic communica-ble diseases—HIV/AIDS, malaria, and tuber-culosis—have proven difficult to contain;achieving a decline in incidence is even morechallenging. HIV prevalence continues toincrease—with but a few exceptions. The epi-demic remains most severe in SouthernAfrica, where pregnant women 15 to 24 yearsof age display infection rates of 18 to 39 per-cent. In East Africa, there are some encour-aging signs that the prevalence of infection isdeclining, with the rate dropping in Uganda,for example, from nearly 30 percent in 1990

to 9 percent in 2002. In West and CentralAfrica, prevalence remains low (1 to 3 per-cent), although Cameroon reports a preva-lence of 12 percent.

Although the risks of failure to halt thespread of HIV/AIDS are especially high inSub-Saharan Africa, they are substantial inmany countries in other regions as well. Eventhough current estimates suggest overall lowprevalence rates in India and China, theabsolute number of cases is estimated at morethan 4 million in India and is approaching 1million in China. In Latin America and theCaribbean, where the epidemic is well estab-lished, 12 countries report HIV prevalence of1 percent or more among pregnant women.The highest rates of increase in HIV/AIDSinfection are found in Europe and CentralAsia, where the absolute number of reportedcases is about 1.5 million. Denial, stigma, andthe institutional challenges of providing ser-vices to marginalized and vulnerable sub-populations, such as injecting drug users,jeopardize progress in combating HIV/AIDS.

Declines in the prevalence of HIV/AIDSlag considerably behind efforts to contain thedisease, making the achievement of the HIV/AIDS-related goal especially challenging.Those newly diagnosed were infected 8 to 10years earlier. The long latency period sug-gests the imperative of mass education cam-paigns and other early preventive measures.

Malaria has proven a stubbornly endemicdisease in many parts of the world. Failedefforts to stop mosquito breeding with pesti-cides have left limited intervention options,among them the use of treated bed nets inendemic areas and drug prophylaxis andtreatment. Behavioral and economic factorslimit progress, since the affordability and useof bed nets and pharmaceuticals require newbehaviors and spending on the part of house-holds and public and private providers of ser-vices and goods.

The global incidence of tuberculosis (TB)is on the rise, with 9 million estimated casesin 2001. The increase is most affected by theHIV epidemic in Sub-Saharan Africa. TB



infection and disease are especially prevalentamong those with impaired immune systemsand poor nutrition. There has been a majorexpansion in the application of the approachfor TB control recommended by the WorldHealth Organization (WHO), which wasadopted by more than 155 countries with anaverage cure rate of 82 percent in 2001. It isestimated that only one-third of TB caseswere reported in 2001. The interim target is70 percent reporting for 2005, suggesting thatthe goals for TB will be difficult to achieve.Progress in Sub-Saharan Africa is slowest, butthere are encouraging signs of progress else-where, even in low-income countries.

Water and sanitation. The MDGs call forhalving the proportion of the world’s popu-lation without access to safe drinking waterand sanitation between 1990 and 2015. Tomeet that target, 1.5 billion additional peo-ple (1 billion in urban areas and 0.5 billionin rural areas) will have to be provided withsustainable access to safe water and about 2billion more with access to basic sanitation(1.1 billion in urban areas and 0.9 billion inrural areas) in the period from 2000 to 2015.At current rates of service expansion, about20 percent of countries are on track, butfewer than 10 percent of low-income coun-tries appear to be on track. Much fasteraccess rates will be required if countries areto achieve the water and sanitation goals.Only one region—Latin America and theCaribbean—currently appears likely to attainthe MDG for drinking water. Water accessrates will need to grow much faster, especiallyin Sub-Saharan Africa—from a current paceof 0.43 percent a year to 2.53 percent.26

Accelerating Progress towardthe Development GoalsThe implication of the foregoing assessmentis clear. The achievement of the developmentgoals will require rising above current trendsand accelerating the pace of development,and doing so swiftly. In line with the princi-

ples and partnership established at Monter-rey, all parties must scale up their action. Theagenda has three essential elements:

� Accelerating and deepening reforms toachieve stronger economic growth thanrecently experienced or currently projected

� Stepping up action on the delivery of humandevelopment and related key services—education, health, water and sanitation

� Speeding up the implementation of theMonterrey partnership, matching strongerdeveloping country efforts to spur growthand improve service delivery to poor peo-ple with stronger support from developedcountries and international agencies.

Substantial reforms of domestic policy andinstitutions will be necessary to boost growth.Particularly important are reforms to improvethe enabling environment for private sectoractivity—by extending progress on macroeco-nomic stability, improving the regulatory andinstitutional framework governing markets,and strengthening basic infrastructure. Alsokey are reforms to enhance the quality of gov-ernance and the capacity of the public sector.Such reforms will improve the effectiveness ofexisting resources and help mobilize moredomestic resources. Moreover, they will allowincreases in aid to be productively utilized.

Policy and institutional reforms will alsoboost MDG prospects by improving the deliv-ery of key services, especially to the poor.Investment in services such as education,health, and water and sanitation needs to bescaled up, in many cases substantially. But theeffectiveness of the delivery of these serviceswill depend on improvements in the under-lying policy and governance framework. Toooften, basic services fail to reach poor people.27

Since societies often view basic services such aseducation and health as largely the responsi-bility of governments, reforms that strengthenpublic sector institutions—making them moreresponsive to the poor, more accountable, andless corrupt—will be critical to improvingaccess to and quality of services.



The acceleration and deepening of reformsin developing countries must be comple-mented by increased support from developedcountry partners. Priorities include improv-ing market access to developing countryexports and providing more and better aid.Also important is the broad conduct ofmacroeconomic and financial policies in away that is conducive to strong global eco-nomic growth and stable capital flows.

A recent World Bank study of 18 low-income countries with relatively good poli-cies found that reforms that substantiallystrengthen policies and institutions, sup-ported by additional external resources, canbe a very powerful combination.28 It showedthat if better policies and institutions werecombined with more aid, the number of coun-tries in the sample that would achieve MDGs1 to 7 would at least double—with all of themachieving the poverty goal (figure 2.6).

The rest of this report assesses the variouselements of this powerful combination: howdeveloping countries are doing on improvingpolicies and governance (Part II); how devel-oped countries are fulfilling their commit-ments (Part III); and how the internationalfinancial institutions are contributing to thedevelopment effort (Part IV).

Notes1. United Nations 2003.2. World Bank 2002.3. Dollar and Kraay 2002; Adams 2002. In-

come inequality, conversely, changes more slowly(Ravallion 1995, 2001; Adams 2002). All thesestudies used household survey data to examinehow growth affects inequality. The studies con-cluded that income distributions are rather stableover time, and that economic growth does notaffect inequality.



20

18

16

14

12

10

8

6

4

2

0

Likely to reach the goal with reform and extra resourcesLikely to reach the goal with current policies

Primaryeducation

completion

Genderequality

Childmortality

Maternalmortality

HIV/AIDS WaterPoverty

Num

ber

of c

ount

ries

FIGURE 2.6 Reform combined with stronger partner support can substantially boost prospects for achieving the MDGsResults for 18 low-income countries undertaking reforms

Source: Development Committee 2003.

4. Studies that have tried to disentangle theserelationships typically have found income to be arobust and strong determinant of outcomes.

5. A word of caution regarding the economicscenarios described here. The key word is sce-nario. Unlike short- and medium-term forecasts—such as those presented regularly in the WorldBank’s annual Global Economic Prospects or theIMF’s World Economic Outlook—long-termforecasts are predicated on informed assump-tions, particularly as regards supply-side factors.In this sense, the scenarios are more akin togrowth-accounting exercises, where more attentionis paid to underlying growth fundamentals—employment, productivity, savings, investment, andcapital accumulation. These scenarios also try todeduce long-term structural changes, such as therising share of services in economic output, increas-ing urbanization, and terms-of-trade impacts, par-ticularly as they affect commodity-dependentexporters and importers. Some of these structuralchanges can also be used to deduce potentialimpacts on achieving the nonmonetary MDGs. Forexample, increasing urbanization may make iteasier to provide health and educational services.Unlike the short- and medium-term forecasts, long-term outlooks typically are not linked to econ-ometrically validated economic relations withstandard confidence intervals. While this may beperceived as a weakness, the purpose of scenarioanalysis is different from the short- and medium-term forecasting exercise, and providing confidenceintervals would be either imparting a false sense ofcertainty or producing such a wide band as to makethe intervals meaningless.

6. Nigeria almost eked out positive growth, los-ing 0.1 percent per capita over the five-year span.The addition of Nigeria to the list of countries withpositive growth would have increased the propor-tion of the region’s population experiencing posi-tive growth to nearly 75 percent.

7. One of the most striking outcomes of thisstructural transformation is the change in the com-position of developing-country exports since the1970s. Three decades ago, exports of primaryproducts (including processed foods), accountedfor three-fourths of total developing-countryexports. Manufactured products now account forthe same percentage, or more, and trade growthhas been phenomenal over a broad range of sec-tors. Though there are regional variations, this pic-

ture holds broadly over all developing countries,including Sub-Saharan Africa, even though in thelatter region manufacturing exports are still asmall share because of the initial starting point.

8. Three factors have driven the decline in barri-ers. First, successive rounds of multilateral tradenegotiations have simultaneously lowered barriersacross a broad set of countries. Since the end ofWorld War II, eight multilateral rounds have beenconcluded, with the ninth, the Doha Round, sched-uled to be concluded by the end of 2004. The sec-ond factor has been the creation of several largeregional free-trade areas—for example, the Euro-pean Union, North American Free Trade Agreement(NAFTA), and Southern Cone Common Market(Mercosur). The third factor has been the wave ofunilateral liberalization following the debt crisis ofthe early 1980s—affecting mostly countries in LatinAmerica. Freer trade enabled producers to purchasetechnology embodied in imported inputs and capi-tal goods, contributing to productivity.

9. World Bank 2001b, 2003c.10. The poverty forecast for most countries is

based on an estimated functional form for theLorenz curve. The parameters of the Lorenz curveare estimated from the latest available householdsurveys, which are available for over 80 develop-ing countries. For most regions, distribution neu-trality is assumed. In technical terms, this meansthat the shape of the Lorenz curve is assumed to beconstant between 2000 and 2015. The two excep-tions are China and India, where some specificassumptions are made regarding change in the pat-tern of regional income disparities. Typical valuesof the elasticity of the headcount index relative togrowth are between –1 and –2. The elasticityapproach has limitations as a tool for making pro-jections over very long periods since the elasticity(assuming distribution neutrality) increases overtime (in absolute terms).

11. While the income poverty MDG is definedin terms of proportion of people living on less than$1 per day, that definition underestimates theextent of poverty in regions where the true povertyline would be higher, such as in Europe and Cen-tral Asia. Numbers in table 2.1 should be inter-preted with that caveat in mind.

12. The MDG is a 50 percent improvementbetween 1990 and 2015, not 2000 and 2015. Atthis time, the 1990 headcount by country is notavailable.



13. Only 8 countries in the sample of 43 havepoverty elasticities above 1.25, 22 have elasticitiesbetween 1 and 1.25, and the remainder are under1. Four countries have poverty elasticities under0.5. The average Gini coefficient is around 0.5,lower than in Latin America but significantlyhigher than in most of Asia.

14. For more information on the conference, seehttp://www.worldbank.org/wbi/reducingpoverty/index.html.

15. World Bank 2004.16. Rokx, Galloway, and Brown 2002.17. The World Bank and its partner institu-

tions are conducting research into a frameworkthat will allow a better modeling of the relation-ship between growth, other determinants, and theMDGs. A two-track approach has been laid. Thefirst track involves pilot studies (initially in Ghanaand Ethiopia). Researchers will develop a frame-work for forecasting the MDGs at the countrylevel, with the intention of extending the approachto other countries. The second track will focus onglobal analysis, using the findings of the pilot stud-ies to inform regional and global level assessments.Preliminary results will be available for discussionby Spring 2005.

18. Preston 1980. 19. Seventy-seven countries have moderate to

serious gender gaps at the secondary level. In 49 ofthese countries the gender disparity is against girlsand in 28 the disparity is against boys.

20. United Nations Millennium Project TaskForce on Education and Gender Equality 2004.

21. Based on data from Inter-ParliamentaryUnion (IPU) 2004: www.ipu.org. Also see WorldBank 2001a.

22. World Bank 2004.23. Monitoring health goals is challenging

because quality data are difficult to obtain. This ispartly due to the large element of judgment neededto classify causes of death, as well as constraints todata collection by public and private providers.The Partnership in Statistics for Development inthe 21st Century works with DAC members andother donors to strengthen statistical capacitybuilding in developing countries. See OECD 2004.

24. World Bank 2003d. 25. World Bank 2004.26. World Bank 2004. 27. World Bank 2003d.28. Development Committee 2003.



IIDeveloping-Country Policies

As discussed in Part I, achieving theMDGs will require stronger economicgrowth in developing countries than

recently attained or currently projected. Itwill also require actions to empower poorpeople and enhance their capabilities to par-ticipate in growth, through improved deliveryof key services to them. Progress on thisagenda calls for improvements in policies andgovernance in developing countries andstronger support to these reforms from devel-oped countries and international agencies, asrecognized in the Monterrey Consensus. Toachieve the goals, each of these three groupsof actors must scale up their efforts. Againstthis background, how are their policies andactions actually evolving? The rest of thisreport attempts to provide an assessment.

As the quest for better development out-comes must begin with developing countriesthemselves—their policies and institutions—this assessment likewise begins there. Howthe developed countries and internationalagencies are playing their part is reviewed inParts III and IV, respectively. This chapterprovides an overview of progress on reformin developing countries, drawing on broadcountry policy assessments prepared by thevarious international agencies that covercountry policies and institutions in a rela-tively comprehensive manner. The following

chapters in Part II then focus on specific pol-icy areas, using a host of indicators and dataspecific to those policy areas and drawn froma multiplicity of sources.

Overall, there is an improving trend indeveloping-country policies. This progress isreflected in the World Bank’s Country Policyand Institutional Assessment (CPIA) ratings,which have improved on average over the pastfive years (figure 3.1). This is true for both low-and middle-income countries. Average ratingshave improved in all regions. And in mostregions they have improved across the fourpolicy clusters that are assessed in the CPIA:economic management (comprising macro-economic policies); structural policies (broadlycovering trade, financial sector, and regulatorypolicies that determine the enabling climate forthe private sector, as well as policies for envi-ronmental sustainability); policies for socialinclusion and equity (covering policies forhuman resource development, gender, socialprotection, and equity of resource use); andpublic sector management and institutions(comprising public financial management,quality of public administration, control ofcorruption, and rules-based governance).1

While the ratings have improved, thechange in general has been relatively modest,though encouraging nonetheless. Going for-ward, the scope for improvement remains


Overall Picture3

large. Notwithstanding the rise in the ratings,average ratings remain in the 3 to 4 range, ona scale of 1 (low) to 6 (high). Ratings for low-income countries on average are appreciablylower than those for middle-income coun-tries. And, across regions, average ratingsremain the lowest in Sub-Saharan Africa,though improving. The regional averages ofcourse also hide large differences amongcountries within a region.

Across country groups, there is a fairlyconsistent pattern in terms of policy relativi-ties. Ratings are highest for macroeconomicmanagement; those for structural and socialpolicies are in the middle; and those for pub-lic sector governance are the lowest. This istrue across regions and across income cate-

gories. This pattern provides part of the expla-nation of the developing countries’ growthperformance reviewed in chapter 2. Whilemacroeconomic management has improved,providing a more stable enabling economicclimate, this progress has not been equallyreflected in an improved growth performance.Part of the explanation is that progress onstructural and governance reforms has beenslower, limiting private sector response.

The low ratings for public sector gover-nance are also reflected in the relative weak-ness of the more institutional dimensions ofpolicy in other areas. Within the nexus ofstructural policies affecting private businessenvironment, the weakest area is propertyrights. Within social sector policies, again the

P A R T I I : D E V E L O P I N G - C O U N T R Y P O L I C I E S


Public sector management/institutionsPolicies for social inclusion/equityStructural policiesEconomic managementOverall rating

CPIA

rat

ings

Low-income countries

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

Sub-Saharan Africa

Middle-income countries

East Asia & Pacific

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

FIGURE 3.1 Developing countries’ policies have improved; governance and institutions lagTrends in developing-country policies and institutions, 1999–2003

weakest link in most cases is the quality andcapacity of the underlying institutionalframework. These findings serve to focusattention on governance and institution-building reforms as an area for particularattention, as poor governance and weak insti-tutions can seriously undermine the effective-ness of policy initiatives throughout aneconomy. Fortunately, developing-countrygovernments recognize this, and more andmore of them have launched governance andinstitutional reforms. This is reflected in theimproving ratings for governance, but thepace of reform needs to be accelerated.

These trends and findings are corrobo-rated by similar country policy assessmentsconducted by other multilateral development

banks for their respective regions. The assess-ments conducted by the African DevelopmentBank (AfDB), also called Country Policy andInstitutional Assessment, and the AsianDevelopment Bank (AsDB), called the Coun-try Performance Assessment (CPA), are quitesimilar to the World Bank’s CPIA in terms ofthe policy criteria and clusters. Their countrypolicy ratings for recent years show broadlythe same trend and pattern. There is a grad-ual improving trend in policies across policyareas. In terms of policy relativities, gover-nance and institutions are the weakest area(figure 3.2).

In the context of its concessional resourceallocation system, the Inter-American Devel-opment Bank (IDB) also initiated in 2002 a

O V E R A L L P I C T U R E


CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

South Asia

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

CPIA

rat

ings

20032002200120001999

4.4 4.2 4.0 3.8 3.6 3.4 3.2 3.0 2.8

Europe & Central Asia Latin America & Caribbean


Public sector management/institutionsPolicies for social inclusion/equityStructural policiesEconomic managementOverall rating

Note: World Bank CPIA ratings range from a low of 1 to a high of 6, with an increase denoting improvement.Source: World Bank CPIA database.

FIGURE 3.1 (continued)

Country Institutional and Policy Evaluation(CIPE) exercise for the 12 eligible countries.The CIPE does not as yet provide a time serieson the ratings, but the pattern indicated by the2002 CIPE is consistent with that of the otherMDB ratings. The strongest policy area ismacroeconomic management and the weakestis public sector management and institutions.

The European Bank for Reconstruction andDevelopment (EBRD) uses Transition Indica-tors (TI), which focus on policies related totransition to a market-based economy. These

indicators show sustained progress in the tran-sition process in Europe and Central Asia, withthe ratings on average higher for market liber-alization measures than for institutions under-pinning markets: financial institutions; and theregulatory and institutional framework for theprovision of infrastructure.2

Progress on transition is also indicated bythe results of surveys of regional enterprisesin Europe and Central Asia jointly conductedby the EBRD and the World Bank in 1999and 2002 (Business Environment and Enter-



1999

Africa - AfDB CPIA

2001 2002

Asia - AsDB CPA

2003

Europe & Central Asia - EBRD TI Latin America - IDB CIPE

Macroeconomic policiesGovernance and publicsector performance

Overall ratingStructural policiesPolicies for growth with equityand poverty reduction

Government and public sectormanagementSustainable economic growth

Overall ratingSocially inclusive development

Markets andtrade

EnterprisesInfrastructure

Overall ratingFinancial institutions

3.4 3.2 3.0 2.8 2.6 2.4 2.2 2.0 1.8

Public sector management/

institutions

Policiesfor

socialinclusion/

equity

Structural policies

Economicmanagement

Overallrating

4.5 4.0 3.5 3.0 2.5 2.0 1.5 1.0 0.5

0

3.6

3.5

3.4

3.3

3.2

3.1

3.0

2.9

3.6

3.5

3.4

3.3

3.2

3.1

3.02000 2001 2002 2002

200320022001200019991998

FIGURE 3.2 Other ratings corroborate that developing-country policies have improved but that governance and institutions lag

Note: AsDB and IDB ratings cover member countries eligible for concessional borrowing. AfDB and EBRD ratings cover all of their developing-country members. IDB ratings shown relate to 2002. AfDB, AsDB, and IDB ratings range from 1 to 6 and EBRD ratings from 1 to 4+. An increase denotes improvement in all cases.Source: For Africa, African Development Bank; for Asia, Asian Development Bank; for Europe and Central Asia, European Bank for Reconstruction and Development; for Latin America, Inter-American Development Bank.

prise Performance Surveys, or BEEPS). Thesesurveys aim to provide more in-depth assess-ment of key elements of the business environ-ment and governance. Comparison betweenthe 1999 and 2002 surveys shows broadprogress. Areas where obstacles to businessare perceived to be greater are those involv-ing firm-state interactions and the judicialsystem (figure 3.3).

Figure 3.4 maps the World Bank’s CPIAratings into the four policy clusters identifiedfor assessing progress on developing-countrypolicies as part of the global monitoringframework set out in chapter 1: economic andfinancial polices; public sector governance;human development; and policies and institu-tions for environmental sustainability. As thepolicies improve and the ratings rise, the “pol-icy diamond” defined by these four clustersincreases in size. In the five-year periodbetween 1999 and 2003, the policy diamondhas grown. The increase has been greater inthe middle-income countries than in the low-income ones. It is also clear that the diamond,

O V E R A L L P I C T U R E


Finance

Infrastructure

Taxes

RegulationsJudiciary

Crime

Corruption

2002 1999 4

3

2

1

0

FIGURE 3.3 The transition countries are making broad progress in removing obstacles to business Index of progress for 26 transition countries in Europe andCentral Asia

Note: Numerals refer to scores on the BEEPS survey, where 1 indicates low obstacles to business and 4 indicates high obstacles, so a decrease denotes improvement.Source: EBRD and World Bank Business Environment and Enterprise Performance Surveys.

Humandevelopment

Low-incomecountries

Policies forenvironmentalsustainability

Public sectorgovernance

19992003

Economic andfinancial policies

6.0

5.0

4.0

3.0

2.0

1.0

Humandevelopment

Middle-incomecountries

Policies forenvironmentalsustainability

Public sectorgovernance

19992003

Economic andfinancial policies

6.0

5.0

4.0

3.0

2.0

1.0

FIGURE 3.4 The developing-country policy diamond shows progress, but much more is neededProgress in developing countries, 1999–2003

Note: The diamond corresponds to the global monitoring policy diamond for developing countries (see figure 1.2). The scale is based on World Bank CPIA ratings, rangingfrom a low of 1 to a high of 6, with an increase denoting improvement.Source: World Bank CPIA database.

especially in low-income countries, is stillsmall relative to the boundary defined by bestpractice (a rating of 6). The shape of the dia-mond, as outlined by the four facets, showsthe policy relativities. These reflect the picturediscussed earlier, with performance strongerin economic and financial policies and weakerin public sector governance. Policies and insti-tutions for environmental sustainability alsoare a relatively weak area.3 These have shownsome improvement, however, especially inmiddle-income countries.

In summary, a review of assessments ofdeveloping-country policies conducted by theWorld Bank and other development agenciesleads to the following main conclusions:

� Policies in developing countries areimproving and creating a more conduciveenvironment for development, includingmore effective use of assistance providedby partners.

� While this is encouraging, the pace ofreform will have to be accelerated if thedevelopment goals are to be achieved,especially in the low-income countries andin Sub-Saharan Africa.

� The core of the reform agenda is institu-tional—reform of public sector gover-nance and the building of institutionsunderpinning private sector activity.

Notes1. There are 20 policy and institutional criteria

in all, divided into these four clusters, that capturethe different dimensions of an effective growth andpoverty reduction strategy and that are ratedannually on the basis of staff assessments. Theindividual criteria are assigned equal weights inaveraging them. The CPIA methodology is period-ically reviewed as part of a continuing effort toincrease its robustness, for example, throughimproving the documentation of the evidence andindicators on which the assessments are based.Increased disclosure of the CPIA ratings for IDA-eligible countries is currently under consideration.For a more detailed description of the CPIA, seehttp://siteresources.worldbank.org/IDA/Resources/FinaltextIDA13Report.pdf.

2. Comparison between the ratings of otherMDBs and World Bank CPIA ratings for therespective regions shows broad consistency. Whilethere inevitably are some differences at the level ofdetailed ratings for individual policy criteria orcountries, the overall correlation is high. Forexample, the estimated correlation coefficientsbetween the AfDB and the World Bank CPIA rat-ings for 2003 and between the EBRD TI and theWorld Bank CPIA for the same year exceed 0.9.The high correlation serves to enhance confidencein the robustness of these ratings.

3. These are included among structural policiesin the CPIA but shown separately in the globalmonitoring policy clusters.



The preceding chapter makes clear thateconomic growth in developing coun-tries needs to accelerate if the Millen-

nium Development Goals are to be achieved.Strong growth requires an economic climatethat is conducive to investment, job creation,and higher productivity. Key elements of suchan enabling economic climate are macroeco-nomic policies that help maintain economicand financial stability; openness to trade thatpromotes access to world markets for goods,services, and knowledge; a regulatory andinstitutional environment for private sectoractivity that facilitates rather than impedesentrepreneurship and competition; and afinancial sector that efficiently and sustain-ably mobilizes resources and channels themto their most productive uses. How are thedeveloping countries doing with respect tothese determinants of growth?

In general, there has been a notableimprovement in macroeconomic and tradepolicies, which needs to be consolidated anddeepened. There is also an improving trend inremoving regulatory impediments to privateactivity and strengthening supporting legalframeworks, and in placing financial systemson a more solid footing and building theircapacity to play their role in contributing todevelopment. Progress on these more struc-tural and institutional dimensions of reform,however, has been less in many countries. This

has acted to limit economies’ supply responseto the improvement in the macroeconomicand trade policy environment. Going for-ward, stronger growth will depend cruciallyon more vigorous and consistent efforts toease the burden of regulation of private enter-prise and build supportive institutions.

Macroeconomic PoliciesSound macroeconomic policies are an essentialfoundation for sustained growth. The impactof macroeconomic policies on poverty reduc-tion works largely through growth: policiesthat contribute to macroeconomic stabilityhelp sustain growth, and growth is typicallyassociated with lower poverty. Policies thatimpede the attainment of macroeconomic sta-bility, such as expansionary monetary policiesthat encourage high inflation and lax fiscalpolicies that produce large budget deficits, canhurt growth by reducing investment and therate of increase in productivity.

Low-Income Countries

Macroeconomic policies in low-income coun-tries have improved over the past decade. Bet-ter macroeconomic management in turn hascontributed to the recent improvement inthese countries’ growth performance and out-look, as reviewed in chapter 3. Inflation has


Improving Enabling Climate for Growth

Economic and Financial Policies

4

slowed on average to half the rate of a decadeago. Other indicators also show improvements,though less spectacular. Fiscal and external cur-rent account deficits have narrowed, and exter-nal debt service ratios typically have improved.Looking ahead, this trend is expected broadlyto continue (table 4.1).

While all regions show some improvementin the macroeconomic policy indicators,progress has been uneven and remains fragilein many countries, especially in Sub-SaharanAfrica. In these countries, therefore, the needremains for a broad strengthening of macro-economic policies. Even in countries with bet-



TABLE 4.1 Macroeconomic indicators for low-income countries, by region, 1983–2008(annual averages)

2003 2004–08Region 1983–87 1988–92 1993–97 1998–2002 (estimated) (projected)

Inflation (median annual %)East Asia and Pacific 6.9 9.7 8.0 7.1 6.6 3.4Europe and Central Asia 1.3 196.2 453.5 11.1 4.8 4.0Latin America and Caribbean 5.2 11.7 7.4 3.7 3.3 3.2Middle East and North Africa 13.7 17.8 26.3 6.4 6.4 4.6South Asia 7.7 10.5 8.8 4.7 4.7 4.3Sub-Saharan Africa 10.8 10.3 13.5 4.7 6.6 3.8All low-income 7.1 10.0 12.4 5.1 5.0 3.8

Current account balance (% of GDP)East Asia and Pacific –1.3 –6.3 –2.2 –1.0 –3.1 –3.7Europe and Central Asia –0.7 –5.6 –13.2 –8.4 –8.3 –5.2Latin America and Caribbean –10.6 –14.2 –14.0 –13.0 –13.1 –11.1Middle East and North Africa –7.4 –5.3 –2.2 0.3 –2.8 –9.8South Asia –8.8 –5.4 –6.9 –2.0 –2.0 –1.9Sub-Saharan Africa –7.0 –8.8 –9.3 –9.5 –9.1 –8.5All low-income –6.0 –8.3 –8.7 –7.4 –7.6 –7.0

Debt service (% exports of goods/services)East Asia and Pacific 22.2 14.2 12.9 13.4 10.9 9.4Europe and Central Asia 0.0 2.0 8.5 14.3 14.5 11.7Latin America and Caribbean 18.5 39.4 37.6 12.3 11.5 9.2Middle East and North Africa 6.7 18.8 25.3 7.2 3.8 4.3South Asia 20.4 19.5 20.8 15.9 13.2 11.2Sub-Saharan Africa 17.5 21.4 21.2 24.5 17.7 16.6All low-income 16.5 20.1 20.5 18.8 14.7 13.2

Fiscal balance (% of GDP)East Asia and Pacific –8.7 –6.7 –4.4 –4.3 –4.5 –3.5Europe and Central Asia –1.3 –6.8 –11.1 –4.4 –1.8 –1.2Latin America and Caribbean –8.5 –6.0 –2.4 –4.5 –5.7 –3.7Middle East and North Africa –17.6 –11.8 –8.1 –0.5 –1.8 –1.9South Asia –6.1 –5.1 –4.8 –5.9 –5.5 –4.5Sub-Saharan Africa –6.2 –6.9 –5.8 –5.7 –4.8 –2.6All low-income –6.6 –6.7 –5.7 –5.0 –4.4 –2.9

Memo Item: Real GDP per capita growth (%)All low-income 0.7 –0.8 0.7 1.2 1.7 3.1

Note: Averages are calculated as unweighted means of country values. Median inflation is calculated from the annual medians and then averaged over five-year periods.Source: IMF World Economic Outlook database, staff calculations.

ter performance, maintaining and building onprogress on macroeconomic stability, espe-cially in the face of possible shocks, will be acontinuing challenge.

To get a better perspective on the qualityof macroeconomic policies in low-incomecountries, IMF staff have assessed recentcountry experience with a view to evaluatingand categorizing them on several key policydimensions.1 The results show that there aresubstantial differences in the quality of poli-cies across the different policy areas (figure4.1). Broadly, fiscal areas seem to be of themost concern. Overall fiscal management, inrelation to the goals of public debt sustain-ability and containment of fiscally derivedinflationary pressures, remains unsatisfactoryin about one-third of low-income countries.

The deficiencies are more serious with respectto structural aspects of fiscal policy: almosthalf of low-income countries are assessed tohave unsatisfactory composition of publicexpenditures. In contrast, there are seriousconcerns in only a relatively small proportionof countries in monetary areas—monetarypolicy, foreign exchange regime. Looking atthe consistency of macroeconomic policy as awhole, the picture is unsatisfactory in aboutone-fifth of the countries.

Staff assessments suggest that faster-growing low-income countries typicallyhave better macroeconomic policies (see fig-ure 4.1). While this result does not reveal thedirection of causality, it is consistent withstudies that suggest that better policies aregood for growth.2 Faster-growing countries

I M P R O V I N G E N A B L I N G C L I M A T E F O R G R O W T H : E C O N O M I C A N D F I N A N C I A L P O L I C I E S


Perc

ent

Good Satisfactory Unsatisfactory

100

80

60

40

20

0Consistency

of macropolicies

Liberalityof exchange

regime

Monetarypolicy

Compositionof publicspending

51

29

22

78

13

9

77

12

12

18

32

49

Fiscalpolicy

Consistencyof macropolicies

Compositionof publicspending

Fiscalpolicy

47

19

34

42

29

29

54

30

16

30

32

38

8

33

59

57

27

16

36

Low HighHighLowHighLow

13

51

Percentage distribution of all countriesPercentage distribution of countriesdistinguished by growth performance

FIGURE 4.1 Faster growing countries typically have better macroeconomic policiesGrowth performance and quality of macroeconomic policies, low-income countries

Note: Low and high refer to countries with GDP growth below or above the median rate for the whole group. Some numbers do not add up to 100 because of rounding. Source: IMF staff assessments.

appear to have markedly better fiscal poli-cies, as well as public sector spending com-positions that are more consistent withgrowth and poverty reduction. They alsohave superior consistency in their overallmacroeconomic frameworks. Differences inmonetary policy areas are less pronounced.

Analysis of the factors in terms of differentcountry characteristics that underlie the vari-ation in the quality of macroeconomic poli-cies shows some interesting patterns (table4.2). The quality of fiscal policy is typicallyhigher in countries with a higher general levelof economic development (proxied by percapita income). The association is muchstronger with respect to government effec-tiveness (quality of institutions for policyimplementation), as suggested by World Bank

governance indicators. The quality of thecomposition of public expenditures and theoverall consistency of macroeconomic poli-cies are particularly strongly correlated withgovernment effectiveness (for example, overtwo-thirds of countries with low governmenteffectiveness have unsatisfactory publicexpenditure compositions compared withhalf of all low-income countries).

Countries with undiversified commodityexports generally have significantly weakerpolicy performance.3 Differences in policyperformance associated with the degree ofexport diversification are broadly similar tothose linked to government effectiveness. Thisfinding is consistent with the view thatreliance on just a few exports tends to be asso-ciated with weak governance, resulting from



TABLE 4.2 Quality of macroeconomic policies for low-income countries, by country characteristics(percent of countries)

Composition of Monetary Liberality of Consistency ofCountry characteristic Fiscal policy public spending policy exchange regime macro policies

Income per capitaBelow median

Unsatisfactory 41 51 16 16 24Good 43 22 68 73 43

Above medianUnsatisfactory 28 48 8 3 21Good 49 38 85 82 54

ExportsUndiversified


DiversifiedUnsatisfactory 27 40 10 4 21Good 56 17 81 85 60

Government effectivenessBelow median


Above medianUnsatisfactory 23 31 5 5 11Good 59 28 85 87 66

All low-income countriesUnsatisfactory 34 49 12 9 22Good 47 18 77 78 51

Note: Policies are assessed good, satisfactory, or unsatisfactory. Percentages in the table do not add up as the intermediate category—satisfac-tory—is not shown.Source: IMF staff assessments.

nontransparency of payment of revenues.Lack of export diversification also renderscountries more prone to shocks that canadversely affect fiscal revenues and macro-economic performance. Small economies andlow-income countries are particularly vulner-able to natural disasters and terms-of-tradeshocks. IMF staff calculations show thatlosses per large natural disaster averaged morethan 6 percent of the GDP of the affected low-income countries from 1997 to 2000. Lossesfrom terms-of-trade shocks averaged morethan 4 percent of the GDP of the affected low-income countries during the same period.

Broadly, governance factors seem to beimportant in determining the quality of fiscalpolicy and public expenditure managementand the consistency of policies. The quality ofmonetary policy and the liberality of theexchange regime also vary with these countrycharacteristics, but proportionally less. Vari-ations in these policy areas can be as readilyexplained by the general level of economicdevelopment.

Middle-Income Countries

Middle-income countries have enjoyed con-sistently higher average growth than low-income countries since the late 1970s; theindicators suggest that such growth has beensupported by better macroeconomic policies

(table 4.3). This advantage, however, hasbeen blunted by the risk of financial crisis.These countries are typically more integratedwith international capital markets, and finan-cial globalization has increased the vulnera-bility to fluctuations in these markets.4 Theconsequent risk of economic volatility hasalso made sustained reductions in povertymore difficult. The 1990s have witnessed sev-eral financial crises, of which the East Asiancrisis is perhaps the best known. In East Asia,poverty increased markedly (table 4.4) as realwages fell and unemployment rose.

A number of measures are needed to makethe economies of middle-income countries,and particularly those of emerging marketcountries, more resilient to economic shocks.First, reducing public debt, and particularlyexternal debt, relative to GDP is a central itemon the agenda as high debt levels increase thelikelihood of debt crises. Increased flexibilityin exchange rates can also help, both bydampening destabilizing capital flows and byfacilitating adjustments between tradable andnontradable sectors. Other structural reformsare also important, particularly in the finan-cial and corporate sectors, to prevent thebuild-up of balance-sheet vulnerabilities. Sev-eral of these countries also need to improvetheir social protection systems to cushion theimpact of shocks and economic adjustmentson the poor and vulnerable.



TABLE 4.3 Macroeconomic indicators for middle-income countries, 1983–2008(annual averages)

2003 2004–08Indicator 1983–87 1988-92 1993-97 1998-2002 (estimated) (projected)

Inflation (median annual %) 7.9 14.4 11.5 5.0 4.5 3.6Fiscal balance (% of GDP) –4.4 –2.7 –2.7 –3.6 –3.0 –1.4Current account balance (% of GDP) –2.6 –1.2 –4.1 –3.4 –1.5 –1.3Debt service

(% exports of goods/services) 25.4 22.5 19.0 21.7 20.1 18.0External debt (% of GDP) 44.0 43.5 42.1 44.8 44.4 38.1Broad money (% of GDP) 52.8 54.2 46.3 51.9 55.9 57.0Real GDP per capita growth (%) 1.4 0.0 1.7 2.1 2.8 3.3

Note: Averages are calculated as unweighted means of country values. Mean inflation is calculated from the annual medians and then averaged over five-year periods.Source: IMF World Economic Outlook database and staff calculations.

Some progress has been made in all of theseareas. On average for middle-income coun-tries, fiscal deficits and debt and debt serviceratios have fallen since the late 1990s. Therehas also been a clear shift among emergingmarket economies toward more flexibleexchange rate regimes. For example, in 2001,half of these economies had floating regimes,compared with only 17 percent in 1990. Evenwithin emerging market economies that main-tain exchange rate regimes that are intermedi-ate between hard pegs and floating regimes,there has been a shift toward greater flexibil-ity. As a cushion against volatility, foreignreserve positions have in general been appre-ciably strengthened. As discussed later in thischapter, the financial soundness of banks indeveloping countries has also improved, andreforms have been initiated to improve corpo-rate governance. And some countries havetaken steps to strengthen their social protec-tion systems, while taking care to ensure theirsustainability.

Trade PolicyComplementing sound macroeconomic poli-cies, openness to international trade con-tributes to an environment fostering growth,by giving firms and households access toworld markets for goods, services, andknowledge and encouraging investments in

areas where countries have a comparativeadvantage. The primary determinant of thebenefits from trade is a country’s own poli-cies, though, as discussed in chapter 10,restrictions imposed by trading partners cansignificantly limit the possibilities for trade.Progress in trade reform by developing coun-tries in the past several years has been encour-aging. However, while many developingcountries have done well in using trade aspart of a growth strategy, several others havenot. Thus, the benefits of trade are distributedasymmetrically across (and within) countries.

Merchandise Trade Liberalization

The average nominal tariff in developingcountries was cut by half during the pastdecade, dropping to 13.5 percent (table 4.5).Particularly notable progress has been madein reducing nontariff barriers (NTBs). Theprevalence of core NTBs (quotas on trade,nonautomatic licensing, minimum prices, andsimilar policies), as measured by the numberof tariff lines covered, has fallen dramaticallyin almost all regions during the past decade,dropping from over 50 percent in someregions (South Asia) to the 2 to 15 percentrange (table 4.6). Currently, core NTBs coversome 11 percent of imports in developingcountries. Despite this significant liberaliza-tion, trade policies continue to generate sig-



TABLE 4.4 Impact of financial crises on poverty, East Asia, 1997–98

Poverty headcount index (percent)

Country Crisis year Overall Urban Rural

Indonesia 1997 11.0 9.2 12.41998 19.9 15.8 23.0

Korea 1997 2.6 7.5 —1998 7.3 10.0 —

Malaysia 1997 8.2 — —1998 10.4 — —

Thailand 1997 9.8 1.2 11.81998 12.9 1.5 17.2

— Not available.Source: Fallon and Lucas 2002.

nificant anti-export bias in some regions; forexample, even after the significant decline, theaverage tariff in South Asia remains around20 percent.

The above picture is confirmed by the IMF’strade restrictiveness index. The average ratingfor low-income countries was 3.8 (relativelyopen) in 2003, down from 5.4 (moderatelyrestrictive) in 1997. Trade liberalization hascontributed to a rise in developing countries’share of world trade, which rose from 17 per-cent in 1995 to 22 percent in 2002. The shareof low-income countries also rose, though inabsolute terms it remains small, at around 3percent.

While traditional trade policy barriershave been lowered, there has been a tendencyfor technical product regulations (mandatoryhealth and safety standards) to rise. Althoughnot generally protectionist in intent, they areincreasingly becoming a factor in interna-tional trade and have been argued to be animpediment for low-income countries to pen-etrate export markets.5 The incidence of suchstandards is four times higher in middle-income countries than in low-income coun-tries. Countries in Latin America and in theMiddle East and North Africa subject around20 percent of all imports to product stan-dards, compared with less than 5 percent inthe least developed countries.

In a number of (mostly) middle-incomecountries, there has also been an increase in theuse of contingent protection—antidumpingand safeguard actions. Measured on a “perdollar of imports affected” basis, many of thelargest developing countries are now moreintensive users of antidumping than theadvanced countries (table 4.7). Although it isoften argued that such “safeguard” instru-ments are needed to allow countries to pursuetrade liberalization, the country incidence ofthese actions is biased against other developingcountries; it also imposes substantial costs onthe economies invoking such protection.

The spread of bilateral and regional tradeagreements (RTAs), of both South-South andNorth-South types, has continued. Manycountries are involved in both bilateral and

regional trade negotiations with differentpartners at the same time. As of March 2003,only four WTO members—Macao SAR;Mongolia; Taiwan, China; and Hong Kong,China—were not members of any RTA. Thequestion has often been posed whetherregionalism is a building block toward multi-lateral liberalization or whether it is morelikely to be a stumbling block. There is noclear consensus on this issue. Much dependson the specifics of the agreements. Overall,however, discrimination can be costly; it not



TABLE 4.5 Decline in tariffs in developing countries,late 1980s to 2003(simple averages of MFN tariffs, percent)

Region Late 1980s 2003

East Asia and Pacific 18.8 10.4Europe and Central Asia 10.2 8.9Latin America and Caribbean 22.4 12.0Middle East and North Africa 17.3 14.8South Asia 68.9 19.8Sub-Saharan Africa 25.1 17.6

All developing countries 25.4 13.5Least developed 28.4 16.4Low income 31.7 15.8Middle income 21.8 12.7

Source: World Bank, World Development Indicators, various years; IMF tariff data.

TABLE 4.6 Decline in core nontariff barriers in developing countries,1989–94 to 2000(percent of tariff lines affected)

Region 1989–94 2000

Latin America and Caribbean (13, 17) 18.3 15.3Europe and Central Asia (11) N.A. 3.4East Asia and Pacific (7, 9) 30.1 5.5Sub-Saharan Africa (12, 17) 26.0 2.3Middle East and North Africa (4, 8) 43.8 8.5South Asia (4, 3) 57.0 13.3

Note: Numbers in parentheses indicate the number of countries for which data are availablefor 1989–94 and 2000, respectively. Due to differences in country coverage, data are notstrictly comparable across years by region.Source: World Bank.

only hurts those excluded but also imposescosts on those included. By reducing tradebarriers on a subset of partners, countries gen-erally increase the real cost of their imports,reduce the flow of technology from nonmem-ber countries, and increase dependence onparticular export markets. If liberalizationproceeds on a discriminatory basis, there is adanger of stronger pressure against furtheropening up of the regional market through theWTO. Finally, given the limited administra-tive resources of poor countries, divertingthese away from multilateral negotiations canentail a substantial opportunity cost.

The policy indicators reviewed above—tar-iffs, core NTBs, technical standards, and pref-erential arrangements—cannot be readilycompared across countries. To enable a com-parison that encompasses tariffs and differenttypes of NTBs, an index has recently beendeveloped at the World Bank that measuresthe overall policy restrictiveness implied bythese barriers. It measures the uniform tariffequivalent of the actual structure of tariffs andNTBs that would generate imports at theobserved levels in a given year.6 The index,called the Overall Trade Restrictiveness Index(OTRI), has been estimated for some 100developing and developed countries for whichunderlying data are currently available.

The average OTRI across all developingcountries in the sample was 15.1 percent in2001 (table 4.8). The most restrictive regionis South Asia. On average, trade barriersagainst low-income countries imposed bymiddle-income countries are somewhat lowerthan the average restriction on overall trade,and barriers against the least developed coun-tries (LDCs) are somewhat lower still (12.7percent). However, low-income countries as agroup impose higher barriers against LDCsthan against low-income countries. LDCsimpose, on average, lower barriers againsttrade with other LDCs than on trade with therest of the world. As is often the case, theregional and income-based aggregations hideconsiderable variance across countries.

OTRIs are negatively correlated with percapita income, indicating that richer coun-tries tend on average to have less restrictivetrade regimes (figure 4.2).7 Developing coun-tries with the highest OTRIs include India,Tunisia, Jordan, Morocco, Bangladesh, Alge-ria, and Lebanon (OTRIs above 20 percent).India is the most restrictive country in thesample, with an OTRI of 36 percent. Coun-tries in the Middle East and North Africatend to have more restrictive trade regimes: ofthe 10 most restrictive countries in the sam-ple, 5 are in that region—Tunisia, Jordan,Morocco, Algeria, and Lebanon. In LatinAmerica, Mexico and Argentina have themost restrictive trade regimes. Among Sub-



TABLE 4.7 Major users of antidumping, developing countries,1995–2002

Against all economies

Initiations per US dollar Number of of imports Index

Country initiating antidumping initiations (USA=100)

Argentina 176 2,549Brazil 98 580India 273 2,197Indonesia 39 90Korea 48 126Malaysia 22 106Mexico 56 144South Africa 157 2,006

Source: World Bank staff based on notifications to the WTO.

TABLE 4.8 Overall trade restrictiveness of developing-country groups, 2001(in percent)

OTRI toward low-income OTRI toward

Region Overall OTRI countries LDCs

East Asia and Pacific 12.1 12.6 11.4Europe and Central Asia 11.6 12.0 12.1Latin America and Caribbean 16.8 16.5 18.3Middle East and North Africa 17.7 14.0 12.4South Asia 32.2 26.4 29.9Sub-Saharan Africa 10.7 13.1 11.8

All developing countries 15.1 14.8 16.6Least developed 18.2 14.4 11.6Low income 23.3 20.0 24.2Middle income 14.2 13.5 12.7


Saharan African countries, Nigeria and Mau-ritius have the highest OTRIs. China andThailand are the most protective countries inEast Asia, and Romania and Poland in Europeand Central Asia. Developing or transitioneconomies with the lowest OTRIs include theBaltic states, Costa Rica, Guatemala, Indo-nesia, Philippines, South Africa, Trinidad andTobago, Ukraine, and the low-income coun-tries of Sudan, Madagascar, and Uganda (withOTRIs below 7.5 percent).

Services Liberalization and “Behind-the-Border” Agenda

Developing countries’ services trade has beenrising more rapidly than merchandise trade,fueled by falling communications and trans-port costs and the associated rise in global out-sourcing. The share of services in totaldeveloping-country exports rose from 20 per-cent in 1990 to 26 percent in 2000. India is awell-known example of a country that hasgreatly expanded distance business servicesexports, but several other countries alsohave increased such exports. Business serviceexports from Brazil, Dominica, India, andMauritius grew at rates around or above 20percent per year from 1995 to 2000. Thepotential for further growth of services exportsfrom developing countries is substantial.

The benefits of services trade extendbeyond services exports and can generatesubstantial gains on the import side as well.Services, such as customs clearance, logisticsrelating to product labeling and standards,transport and telecommunications, andfinancial services, are an important input intothe production process and affect the com-petitiveness of an economy more broadly.Progress on this “behind-the-border” agendais crucial for countries to take full advantageof the increased market access resulting fromreduction in border barriers (tariffs, NTBs).The availability of these services at low costis becoming increasingly important as inter-national competition becomes sharper andefficient and just-in-time supply chain man-agement more crucial. A more open services

trade regime can contribute to the lowering ofsuch key behind-the-border costs. It can alsogenerate broader spillover benefits arisingfrom increased competition and new technol-ogy and know-how brought by foreign ser-vice providers.

Research provides supporting evidence thatopenness to services trade contributes to betterlonger-term growth performance. A studyfinds that, controlling for other determinantsof growth, countries that fully reformed thefinancial services sector grew, on average,about 1 percentage point faster than othercountries. An even greater impetus to growthwas found to come from fully reforming boththe telecommunications and the financial ser-vices sectors, with countries that fully liberal-ized both sectors achieving average growth ofabout 1.5 percentage or higher than others.8

While these estimates are only suggestive, theydo indicate the scope for substantial gainsfrom liberalizing key service sectors.

Developing countries have made progresson liberalization, but policy remains relativelyrestrictive in many of them. Services tend to beless tradable across borders than goods,implying the need to open up to foreign direct



1210864Log GDP per capita

40

30

20

10

0O

TRI/

Fitt

ed v

alue

s, p

erce

nt

Tunisia

JordanMorocco

India

Bangladesh

Sudan

Uganda

Madagascar

UkrainePhilippinesIndonesia

GuatemalaCostaRica

ChinaNigeria

Lithuania

AlgeriaRomania

Lebanon

MexicoMauritius

Argentina

PolandThailand

South AfricaTrinidad & Tobago

Hong KongEstonia

Latvia JapanSingapore

NorwayEuropean Union

CanadaUnitedStates

FIGURE 4.2 Better-off countries tend to restrict trade lessOverall Trade Restrictiveness Index and GDP per capita, 2001

Source: World Bank staff.

investment. Many countries restrict trade inservices through barriers to labor movementor to entry of foreign-owned firms. Withrespect to financial services and telecommuni-cations, for example, the picture varies sub-stantially across regions. While some regionshave made appreciable progress, overall pol-icy remains significantly more restrictive thanin developed countries (figure 4.3).

Successful reform of trade policy in servicesrequires more than lowering entry barriers.They need to be underpinned by complemen-tary changes in domestic regulation. Badlydesigned reform can seriously undercut thebenefits of liberalization. For example, if liber-

alization of financial services is not supportedby strengthening of the regulatory and institu-tional framework for prudential regulation andsupervision, the soundness of the financial sys-tem could be placed at risk. If privatization ofstate-owned service monopolies is conductedwithout attention to creating conditions forcompetition and appropriate regulatory over-sight, the result may largely be transfers ofmonopoly rents to private owners—perhapsforeign ones. Also, if policies to ensure broadaccess are not put in place, liberalization maynot improve access to essential services forsmaller businesses and the poor. Getting theelements and sequencing right, therefore, isimportant to the realization of benefits.

Private Sector Regulatory andInstitutional Environment

A vibrant private sector requires a favorableregulatory environment. But in many coun-tries, especially low-income ones, businessesare hampered by bureaucratic regulationsand weak protection of property rights. Theresult is slower development. The more timeentrepreneurs spend dealing with red tape ordefending property rights, the less time theyspend building their business. A host of evi-dence shows that heavy regulation of busi-ness is associated with poor economicoutcomes—less investment (both domesticand foreign), employment, and productivity(figure 4.4).9 The ultimate effect is slowergrowth and more poverty.

As an example, compare China and SouthAsia. Recent studies show that a sizable partof the gap between their economic perfor-mance is the result of differences in invest-ment climates.10 Similarly, by improvinginfrastructure and regulatory conditions tothe levels of Shanghai, studies show that firmsin Dhaka could increase productivity by 43percent; those in Calcutta, 78 percent; andthose in Karachi, 81 percent. Even citieswithin China would experience dramaticgains by emulating Shanghai. For instance, ifChengdu could attain investment climateindicators identical to those of Shanghai, pro-



South Asia (3)

South Asia (5)

Sub-Saharan Africa/Middle Eastand North Africa (17)

Sub-Saharan Africa/Middle Eastand North Africa (42)

Telecoms

Financialservices

East Asia and Pacific (5)

East Asia and Pacific (8)

Europe and Central Asia (3)

Europe and Central Asia (3)

Latin America andCaribbean (18)

Latin America andCaribbean (21)

High Income (26)

High Income (21)

0 2 4 6 8 10Less restrictive

FIGURE 4.3 Overall policy on trade in services remains more restrictivein developing countriesLiberalization indices for financial services and telecommunications, most recent years

Note: Numbers in parentheses show the number of countries covered.Source: Mattoo, Rathindran, and Subramaniam (2001).

ductivity of its firms could be increased byabout 30 percent, their investment rate from14 percent to 19 percent, and their averagesales growth by up to one-half.11

The cost of poor business regulation is notjust economic. Heavy business regulationshurt the very people they are supposed toprotect. Faced with burdensome rules andlittle benefit from registering officially, manyentrepreneurs tend to operate in the informalsector (figure 4.5). There, there are noworker safety or product standards, there isno social security, access to credit and use ofthe courts is more difficult, and firms do notcontribute to government and social servicesby paying taxes. This is the case for well overhalf of business activity in many developingcountries. Another social cost is corruption,a regressive tax that hits poor people thehardest. Without checks and balances, com-plex rules breed discretion. Each encounterbetween a bureaucrat and an entrepreneurbecomes an opportunity to extract a bribe.Not surprisingly then, more bureaucratic

procedures are associated with higher levelsof corruption (figure 4.5).

Burdensome regulation excludes vulnera-ble groups from the opportunity to partici-pate in markets. The wealthy and connectedcan find ways around cumbersome rules.They may even be protected by them. Butnew entrants and the disadvantaged are shutout. For example, rigid employment regula-tion is associated with higher female andyouth unemployment. There is no evidencethat countries with stronger employment pro-tection have more equal income distributions.If anything, the rigidity of employment lawsmay be associated with more inequality (fig-ure 4.6). Relatedly, poor credit laws particu-larly hurt small firms’ access to credit.12

Complex and rigid business regulationthus brings both poor economic and socialoutcomes. But this does not of course suggestthat all regulation is harmful. All countriesregulate business activity. The most success-ful ones do so in a simple and transparentmanner. Rather than intervene heavily inbusiness operations, successful governmentschannel their resources into regulations andinstitutions that protect property rights andensure an even playing field for firms. Doingso encourages investment, facilitates tradebeyond a narrow circle of established busi-ness networks, and enhances access to credit.As one example, protection of creditor rightsand efficient enforcement of contracts throughcourts is associated with deeper credit markets(figure 4.7).

Current Picture and Agenda

The countries that most need a vibrant privatesector—low-income countries—are the veryones that place the most obstacles in front ofentrepreneurs. Indicators developed under theWorld Bank’s Doing Business Project (DB)show that on average it takes 30 days andcosts less than 10 percent of income per capitato start a business in high-income countries.13

In low-income countries, it takes 74 days andcosts twice income per capita—prohibitivefor the ordinary person. Among low-income



Source: World Bank Doing Business database 2003.

40

30

20

10

0Labo

r pr

oduc

tivi

ty (U

S$ th

ousa

nds

per

wor

ker)

1Less More

2

Countries ranked by regulatorydifficulty of starting a business (quartiles)

3 4

FIGURE 4.4 Heavy regulation is associated withlower productivityEase of starting a business and labor productivity



Gini

inde

x 19

95– 2

000

0.50 0.70 0.90

Employment laws index 1999

1.10 1.30 1.50 1.70 1.90 2.10 2.30 2.50

0.7

0.6

0.5

0.4

0.3

0.2

More restrictive


FIGURE 4.6 More regulation does not necessarily produce better social outcomesRestrictiveness of employment laws and income inequality

Info

rmal

eco

nom

y (p

erce

nt in

com

e pe

r ca

pita

)

Corr

upti

on

High

Low High

High

Less MoreCountries ranked by rigidity of

employment law index (quintiles)Countries ranked by procedures to

register a business (quintiles)

Business registration and corruptionEmployment laws and the informal economy

1 2 3 4 5 1 2 3 4 5

FIGURE 4.5 Heavier regulation contributes to the informal economy and corruption


countries, the time varies from 22 days in Pak-istan to 215 days in the Democratic Republicof Congo, and the cost ranges from 9 percentof income per capita in Armenia to as much as10 times income per capita in Sierra Leone. Bycontrast, in Australia it takes only two days toregister a business and in Denmark there is nocost to register a business.

A similar picture emerges when an entre-preneur uses the courts to resolve disputes, orgoes through bankruptcy procedures. Delaysare much shorter in high-income countries.For example, insolvency procedures on aver-age take around four years in low-incomecountries, compared with two in high-incomecountries. It costs on average 8 percent ofincome per capita to enforce contractsthrough courts in the latter but 64 percent inthe former. In Cambodia, Indonesia, KyrgyzRepublic, and Malawi, for example, feescould amount to more than twice income percapita. Faced with such high fees, most busi-nesses avoid the courts and opt for informalmeans of enforcement, and limit trade to onlya narrow circle of known partners.

Regulation typically is heavier and morecomplex in low-income countries than inhigh-income countries, contrary to what onewould expect given the greater enforcementcapacity and more developed regulatoryregimes for protecting workers, consumers,and the environment in the latter (figure 4.8).Most strikingly, in low-income countries, ittakes on average 30 procedures to enforce acontract through the court. By contrast, ittakes only 18 procedures in the high-incomecountries. Similarly, an entrepreneur mustundergo an average of 11 procedures to starta business in low-income countries, comparedwith only 7 in high-income countries. Indexesof the rigidity of employment law and courtinvolvement in bankruptcy are over 20 per-cent greater in low-income countries.

In which areas do lower-income countrieslag the most? The costs to start a business andenforce a contract are areas where thereappear to be large opportunities for reform.Average costs of business in these respects area large multiple of those in best-practicecountries (figure 4.9).



Priv

ate

cred

it (p

erce

nt o

f GD

P)

Priv

ate

cred

it (p

erce

nt o

f GD

P)Less

More

More MoreLess

More

Countries ranked by days toenforce a contract (deciles)

Countries ranked by legalcreditor rights (deciles)

Contract enforcement and availability of private credit Creditor rights and availability of private credit

FIGURE 4.7 Protecting property rights is associated with more credit


These findings are corroborated by evi-dence gathered through the World Bank’sinvestment climate assessments (ICAs).14

These studies find bureaucratic harassment tobe a major constraint to business. For exam-ple, customs clearance typically takes muchlonger in low-income countries. Firms inthese countries in general also experiencegreater production losses from power failuresand have less access to credit (table 4.9).Related research finds that these deficienciesin investment climate tend to offset the low-wage advantages of the low-income countriesand make it much harder for them to competewith better-off middle-income countries.15

Economic growth will have greater impact onreducing poverty if countries do not have torely on low-wage strategies to make up forinvestment climate deficiencies.

A similar picture emerges from the WorldBank’s Country Policy and InstitutionalAssessment ratings (figure 4.10). Average

ratings for low-income countries are lowerthan those for higher-income countries forall dimensions of private sector regulatoryand institutional environment covered bythese ratings. The area of property rightsand rules-based governance emerges as theweakest element of the business environ-ment. This is an area of relative weaknessfor all developing countries, low or middleincome, but the ratings are particularly lowfor the former.

Although the developing countries regu-late more heavily on average, the picturevaries appreciably across countries and evenwithin the same country across different typesof business regulation. This is encouraging, asit suggests that good practices do exist indeveloping countries and that partial reformsare possible as a step toward more compre-hensive, longer-term reform. Over the pastdecade, countries such as Jamaica, Latvia,Pakistan, Serbia and Montenegro, Slovakia,



Moreregulation

Lessregulation

Court powers in bankruptcy indexEmployment laws index

Contract proceduresEntry procedures

High-incomecountries

Upper-middle-incomecountries

Lower-middle-incomecountries

Low-incomecountries

3066

11

53

12

55

27

56 1053

18 43 7 43

6327

FIGURE 4.8 Poor countries regulate the mostDegree of regulation, developed and developing countries, 2003

Note: The indicators for high-income countries are used as benchmarks. The average value of the indicator is shown above each column.Source: World Bank Doing Business database 2003.

South Africa, Thailand, Tunisia, and Vietnamhave greatly improved certain aspects of busi-ness regulation. Following reforms to its judi-cial procedures in 1996, Tunisia is nowamong the most efficient countries in resolv-ing commercial disputes. The 1999 Thaibankruptcy reforms have achieved great suc-cess, as have revisions to collateral law in Slo-vakia. And recent streamlining of businessregistration in Latvia, Pakistan, Serbia andMontenegro, and Vietnam has significantlyreduced the regulatory requirements to starta business. Although time-series data on busi-ness regulations are scant, evidence from theCPIA suggests a trend toward improvement.While average CPIA ratings for business envi-ronment remain relatively low (on a scale of1 to 6), indicating continuing weaknesses,they are rising (figure 4.11).

Low-income countries are making particu-larly notable progress in the competitive envi-ronment for the private sector, narrowing thegap with the average for all developing coun-tries (including middle-income ones). Lessprogress has been made in property rights andrules-based governance, although the overalltrend in this respect also is positive.

Reforms are also taking place in the area ofcorporate governance. This is a particularlyimportant area of reform in middle-incomecountries; weak corporate governance con-tributed to the corporate financial vulnerabil-ities that underlay the East Asian crisis in thelate 1990s. But it is also relevant to low-income countries seeking to attract capital forgrowth. For the past three years, the WorldBank has been conducting assessments of cor-porate governance in client countries, and bythe middle of 2004, 38 such assessments(known as Reports on Observance of Stan-dards and Codes, or ROSCs) will have beencompleted.16 In almost all of the countries sur-veyed, there is now a wide acceptance ofimportance of corporate governance reform.However, reforms are only just beginning topenetrate the business cultures in most coun-tries. Table 4.10 summarizes some of thelessons learned from the corporate gover-nance ROSC assessments.

The foregoing assessment suggests twobroad themes for the reform of the privatesector regulatory and institutional environ-ment. First, countries need to simplify regula-tions related to the start and conduct ofbusiness. Eliminating unnecessary proceduresand simplifying those that are necessary willreduce business costs, minimize chances ofcorruption, and boost firms’ competitive-ness—while reducing costs for government.Second, emphasis should shift from regulat-ing business operations to strengthening insti-tutions that facilitate business by supportingthe efficient and fair functioning of markets.A key area is strengthening property rightsand the legal and judicial system underpin-ning contract enforcement. Another priority,especially in middle-income countries, is thecontinued building of the institutions of cor-porate governance.



0 30 60 90 130

Cost to go through insolvency

Days to go through insolvency

Creditor rights index

Procedures to enforce a contract

Cost to enforce a contract

Days to enforce a contract

Rigidity of employment law index

Procedures to start a business

Cost to start a business

Days to start a business

7.4

4.4

2.2

2.1

52.2

5.7

2.1

3.3

10.7

128.2

FIGURE 4.9 Low-income countries lag far behind best practice in promoting businessLow-income country average relative to the 10 best-practicecountries

Note: Numbers denote multiples by which low-income country indicators fall short of thosein the top 10 countries.Source: World Bank Doing Business database 2003.



TABLE 4.9 Investment climate, selected low- and middle-income countries, 2002

Firms with a loan Avg. days to clear % of production lost Have own Firms with overdraft from a bank or

Country customs (exports) from power loss generator (%) or line of credit (%) financial institution (%)

Middle-income countriesAlgeria 8.6 5.3 29.5 38.6 50.4Brazil 8.4 2.5 17.0 74.4 34.7China 5.5 1.8 17.0 26.6 57.0Ecuador 7.1 5.7 33.8 71.5 72.8Honduras 1.9 5.2 33.3 43.1 51.6Malaysia 2.6 — 22.6 71.7 87.1Moroccoa 1.7 — 16.7 78.9 45.1Peru 5.2 3.3 27.6 47.7 44.6Mean 5.1 4.0 24.7 56.6 55.4

Low-income countriesBangladesh 8.8 3.3 71.5 66.2 58.8Eritrea 3.2 5.5 43.0 47.4 44.9India 5.1 6.5 68.5 57.4 11.6Mozambique 17.0 5.1 23.3 12.1 29.0Nicaragua 2.0 6.5 18.4 29.5 44.0Pakistan 9.2 5.4 41.8 22.8 19.5Mean 7.6 5.4 44.4 39.2 34.6

— Not available.a. Data from a year 2000 survey.Source: World Bank investment climate assessment surveys.

Betterinstitutions

Weakerinstitutions

Upper-middle-income

countries

Lower-middle-income

countries

Low-incomecountries

2.6

3.23.4

3.1

3.6 3.6

4.04.2 4.2

Property rights and rules-based governanceCompetitive environmentFactor and product markets

FIGURE 4.10 Low-income countries lag the most in property rights andrule of lawProperty rights, rule of law, and competitive environment,developing countries

Source: World Bank CPIA ratings 2003.

Financial Sector PoliciesAn effective financial system promotes eco-nomic growth and reduction of poverty bymobilizing savings and allocating them to pro-ductive uses. It can also reduce povertydirectly: it protects savings, allows migrants toremit funds, and provides small entrepreneursaccess to credit, including women throughmicrofinance programs responsive to theirneeds. Higher incomes and financial assetsfeed back into gains on the non-incomepoverty MDGs as well, such as education andhealth. A sound financial system also strength-ens an economy’s resilience to shocks, therebyprotecting growth and the hard-won gains inreducing poverty.

Financial Deepening

For most of the last decade, banking sectorpenetration, as measured by the ratio of bankdeposits to GDP, has been on the rise fordeveloping countries as a whole. The ratio

rose from 27 percent in 1990 to 34 percent in2002 (figure 4.12). The aggregate indicators,however, mask wide differences betweengroups of countries both across income levelsand across regions. Looking across incomelevels, countries with higher income havehigher banking sector penetration, with theratio ranging from 21 percent of GDP in low-income countries to 42 percent in middle-income countries in 2002 (49 percent inupper-middle-income countries).

The overall trends in banking penetrationalso mask considerable inter- and intra-regional differences. The process has pro-gressed most in emerging market countries inSoutheast Asia, and recently in the develop-ing and transition countries in Europe andCentral Asia. In Latin America, penetrationrates were on the rise during most of the1990s, reflecting economic reform and pricestability, but have receded since 2000, fol-lowing the ebbing of capital flows and a waveof financial crises. In Sub-Saharan Africa,bank penetration continues to be low, owingto long-standing legal, judicial, and institu-tional weaknesses.

Bank credit to the private sector has alsogrown, but more slowly than deposits. Fordeveloping countries as a whole, bank creditto the private sector increased from 24 per-cent of GDP in 1990 to 27 percent in 2002.As for deposits, there are large variationsacross income levels, with the ratio rangingfrom 13 percent for low-income countries to35 percent for middle-income countries in2002 (40 percent for upper-middle-incomecountries). Private credit to GDP ratios ingeneral stagnated during this period for low-income countries, and rose only marginallyfor lower-middle-income countries, whiledisplaying a more dynamic trend in upper-middle-income countries.17

In many countries, the slower growth inprivate sector credit reflects increases in netcredit to government arising from restructur-ings of banking systems and fiscal deficits, andin central bank debt related to the conduct ofmonetary policy. A rise in banks’ own hold-ings of foreign assets (in part to hedge their



3.7

3.5

3.3

3.1

2.9

2.7

2.5

2.3

Competitive environment (low income)Competitive environment (all)Factor and product markets (low income)Factor and product markets (all)Property rights and governance (low income)Property rights and governance (all)

20032002200120001999

FIGURE 4.11 Countries are improving their private business environmentBusiness environment indicators, developing countries

Source: World Bank CPIA ratings 2003.

foreign liabilities) also has contributed to theoutcome (figure 4.13). Analysis of a sample of14 developing countries with relatively largefinancial systems and 10 African countriesshows that, in most cases, relatively little ofthe growth in deposits was reflected in anincrease in credit to the private sector. In thecase of the African countries in the sample, inaggregate, none of the deposit growth wastranslated into growth in private credit.

Access to credit remains a difficult issue inmost developing countries, reflecting not onlycrowding-out by public sector credit, but alsoweaknesses in the informational and legaland judicial frameworks. Attempts simply tochannel low-cost funds to small borrowers toovercome these and other obstacles are likelyto undermine sustainable finance for ruraland other small and medium enterprises(SMEs), as occurred in the era of repressedfinance. A sustainable improvement in access

to credit requires improvement of creditinformation and strengthening of creditors’rights, to increase incentives for sound lend-ing to small borrowers. Attempts are underway in many countries to improve informa-tion on borrowers. For example, 23 develop-ing countries have established public creditbureaus since 1994. In several other countriesthe functioning of existing public and privatecredit bureaus is being improved.18

The legal and judicial framework, particu-larly with regard to execution of collateralbut also bankruptcy, is important to creditaccess. In many countries, the legal systemfavors debtors, making collateral executionalmost impossible. Furthermore, even if thelegal system provides for creditors’ rights intheory, weaknesses in the judicial system maymake collateral execution difficult in practice.

Many countries are now reforming their lawswith respect to collateral execution and bank-ruptcy, but, unless accompanied by a strongerjudicial system, enforcement will remain weak.Another key area for improvement is titling,particularly for small mortgages and land.

There are institutions in developing coun-tries that have successfully sustained creditaccess to small borrowers and that providegood examples. Microbanking institutionsare of all sizes, ranging from Bank Rakyat,Indonesia’s unit desa program (with a small-loan portfolio of $2.5 billion) to Bangladesh’sGrameen Bank ($275 million), Peru’s Mi-Banco ($100 million), Ecuador’s CrediFe($35 million), India’s SEWA ($10 million),and Uganda’s FOCCAS ($1 million). Whetherlarge or small, successful institutions follow afew key principles: collect on loans; charge



TABLE 4.10 Summary of lessons learned from corporate governance assessments

Successes Improvements needed

Basic shareholder rights• Improving shareholder record-keeping infrastructure • Several items of interest to international investors remain unresolved • Improved shareholder meeting mechanics in many countries (such as share blocking, shareholder meeting • Powers of board and shareholders being clarified notification period)

• In some countries fundamental decisions can be taken by board (not shareholders)

Ownership and related party transaction disclosure• Basic rules in place in many countries • Compliance with ownership and related party transaction disclosure is

still spotty in many countries• Understanding ownership remains difficult

Financial and nonfinancial disclosure• Greatly improved rules, standards, and formats • Compliance with new rules is growing more slowly• Gradual progress on accounting and auditing reform • Quality of financial reporting is still mixed in many countries• Audit committee introduced in many countries• Electronic information dissemination now standard

Duties of board of directors• Board role being strengthened and clarified • True independence remains rare• Duties better described in law • Most director training is in early stages• Idea of “independence” introduced in many countries• Many efforts at improving director training

Shareholder redress and enforcement• Lowered thresholds for shareholder action • Results in basic legal reform are limited• More active regulators and exchanges • Implementation of new laws is lacking

• Enforcement remains key challenge

Source: Corporate Governance ROSCs, World Bank.

enough to cover costs; contain costs and selectborrowers well; mobilize deposits and reducedependence on donors; and provide transpar-ent information on accounts, clientele, andany subsidies.19

Systemic Soundness and Stability

A sound financial system is essential formacroeconomic stability. Crises in financialsystems during the 1990s spilled over intoslower economic growth and led to a costlyoverhang of debt. Some improvements haveoccurred, notably an increased role of rep-utable international banks in Eastern Euro-pean and Latin American banking systems.

Yet much remains to be done to reduce therisk of crises. The variation in the health ofthe banking systems across developing coun-tries is reflected in key indicators of financialsoundness, such as the return on assets, theratio of nonperforming loans to total loans,and the ratio of capital to assets (table 4.11).A key indicator of financial soundness, theaverage ratio of performing loans to totalloans, does show some improvement in mostregions recently. However, improper loanclassification and regulatory forbearancemean that the figures likely understate the sizeof problem loans. The average capital to assetratios, on the other hand, have not increasedmuch.



Source: IMF, International Financial Statistics.

Mea

n va

lues

, per

cent

of

GDP

80

70

60

50

40

30

20

10

0

1990 (1995 for Europe and Central Asia) 2002

Asia Middle East&

North Africa

Latin America&

Caribbean

Europe&

Central Asia

Sub-SaharanAfrica

Low-income

countries

Middle-income

countries

Deposits Private sector credit

FIGURE 4.12 The financial sector is deepening in the developing world, but at a varying paceTrends in financial intermediation in developing countries, 1990–2002



Source: IMF, International Financial Statistics.

FIGURE 4.13 Deposit growth has not been equally reflected in growth in private sector creditChanges in ratios of bank assets and liabilities plus capital to GDP in the 1990s

Liabilities + capital

Latin America (7) East Asia (4) Transition countries (3) Africa (10)

20

15

10

5

0

–5

–10

Perc

ent o

f GD

P

Private sector credit Net public sector credit Net foreign assets

TABLE 4.11 Evolution of selected financial soundness indicators, 1998–2002 (percent)

Nonperforming loans Return on assets to total loans Capital to assets

Region 1998 2000 2002 1998 2000 2002 1998 2000 2002

Latin AmericaMean 1.2 0.3 0.7 7.2 10.3 9.5 11.9 11.0 10.3Standard deviation 2.0 1.5 2.3 3.1 7.1 5.3 4.2 2.5 4.7

Eastern and Central Europea

Mean –0.5 1.0 1.3 11.6 10.4 8.9 11.2 10.6 10.5Standard deviation 1.8 0.9 0.5 8.3 7.6 6.7 5.1 3.7 2.4

Asiab

Mean –2.7 0.3 0.8 20.9 14.8 12.9 5.0 6.6 7.4Standard deviation 6.8 0.9 0.4 15.2 5.8 8.5 7.1 3.4 3.8

Middle East and North AfricaMean 1.4 1.1 1.1 10.5 13.7 16.3 8.8 9.6 9.3Standard deviation 0.6 0.5 0.6 3.7 5.6 8.2 1.9 2.7 2.5

Sub-Saharan AfricaMean 2.6 3.9 3.2 20.4 18.9 16.5 9.7 9.8 9.9Standard deviation 1.8 2.3 1.5 10.5 11.4 14.0 1.8 1.9 1.8

a. Includes Israel and Turkey. b. Excludes Japan.Source: National authorities, EDSS, Bankscope, Moody’s, IMF staff estimates.

Evidence shows that the financial strengthof a country’s banking system goes hand inhand with the quality of its regulatory andinstitutional framework (figure 4.14). A sim-ple comparison of indices of financial strengthand compliance with Basel Core Principles forEffective Banking Supervision (BCPs) shows aclose association between the two indexes (thesimple correlation is 0.65).20 It also showsthat the financial strength of institutions isgenerally lower in developing countries. Byand large, this is associated with weaker com-pliance with BCPs—although in some coun-tries, institutions remain financially weakdespite a higher degree of compliance.

A more detailed comparison of individualBCPs across developing countries reveals pat-terns that suggest areas for particular atten-tion. In banking, close to half of the countriesfor which supervisory assessments have beencompleted have weak arrangements for man-aging credit risks (credit policies, connectedlending, loan classification and provisioning)and market risks. Similarly, half of the coun-tries assessed have weak supervisory indepen-dence, including the adequacy of resources toconduct a proper supervision, and institutionalarrangements for consolidated supervision areweak in 70 percent of the countries assessed.Weak accounting and auditing, limited avail-ability of corporate financial information, andwidespread government guarantees limit mar-ket discipline.

Efforts are under way in many countries toremedy these deficiencies. An important focusof these efforts is to strengthen the regulatoryand supervisory framework. Measures are alsobeing taken to improve market discipline. Boththe IMF and the World Bank are providingtechnical assistance to countries in these areas.Looking ahead, the New Capital Accord(Basel II, or NCA), which proposes to improvethe risk sensitivity of capital requirements, is aparticularly important initiative, even thoughsome of its elements have proved controver-sial, including the concern that it may not beappropriate to the less sophisticated bankingsystems in many developing countries.

An area important to both financial sectordeepening and stability is the developmentand sound functioning of domestic capitalmarkets. In most cases, secondary markets indeveloping countries are shallow and illiquid.This prevents the diversification of risks awayfrom the banking system and limits the meansavailable to entrepreneurs to raise finance forinvestment (figure 4.15). As a precondition forthe development of domestic capital markets,countries need to improve their corporategovernance regimes, regulatory frameworkfor financial information and reporting, andclearance and settlement infrastructure.

Notes1. For this purpose, IMF staff assessed each low-

income country according to a common set of cri-teria. For example, a country with an unsustainablelevel of public debt and a large fiscal deficit wouldbe judged to have an unsatisfactory fiscal policy.

2. See, for example, Barro 1997 and Berg andKrueger 2003.

3. For the definition of export diversity, see IMF2003, p. 188.



90

80

70

60

50

40

30

20

10

00 10 20 30

Basel Core Principles compliance index

40 50 60 70 80

Developing countries Developed countriesM

oody

's fi

nanc

ial s

tren

gth

inde

x

FIGURE 4.14 Financial system strength is typically positively correlatedwith compliance with Basel Core PrinciplesCorrelation between financial strength and compliance withBasel Core Principles

Sources: Moody’s and IMF Basel Core Principle Assessment, staff calculations.

4. One recent study, focusing on currency crisesin 30 developing countries and 20 advanced coun-tries over 1975–97, found that the developingcountries experienced an average of one crisis percountry every five years—roughly twice the aver-age for advanced economies (Aziz, Caramazza,and Salgado 2000).

5. Maskus and Wilson 2002; Wilson 2002.6. The measure calculated is a variant of the

trade restrictiveness index developed by Andersonand Neary (2003). The index is a theoretically well-grounded measure of protection that allows forconsistent comparisons of the level of restrictive-ness of trade policy across countries and time. TheOTRI does not suffer from the well-known draw-backs of simple or import-weighted tariff averagesand allows the impact of both tariffs and NTBs tobe estimated. The methodology used for the mea-sure comprises four steps. First, import demandelasticities by country and by product were esti-mated at the 6-digit level of the Harmonized Sys-tem of commodity classification (HS) (some 4,200product categories). Second, an estimate was madeof the impact on imports, again at the 6-digit HSlevel, of core NTBs (such as quotas, nonautomaticlicensing, and minimum prices) and domestic sup-port granted to agriculture. Third, using thesedemand elasticity and impact estimates, the prod-uct-level ad valorem equivalent (AVE) of the twotypes of NTBs (core NTBs and domestic agricul-



tural support) was calculated for each country inthe sample. Finally, tariffs (ad valorem and AVEs ofspecific tariffs) and AVEs of NTBs at the productlevel were aggregated to produce an overall mea-sure of trade restrictiveness. The index is calculatedbilaterally for every country and its partners. Dataon technical product regulations (such as labelingand testing requirements, product requirements toprotect human and animal health) were not used inthe estimation of the OTRIs, because of incompletereporting by countries. However, estimates of theAVEs of such measures are available on request andcan be found in the background document (Kee,Nicita, and Olarreaga 2004).

7. The simple correlation coefficient betweenOTRI and GDP per capita is –0.32, and is statisti-cally significant at the 1 percent level.

8. Mattoo, Rathindran, and Subramaniam2001. The measure of reform included not just lib-eralization but also regulatory improvements.

9. World Bank 2003; Alesina and others 2003;Scarpetta and others 2002; Dollar and others2003; Batra 2003.

10. Dollar and others 2003.11. Dollar, Hallward-Dreimeier, and Mengistae

2003; Hallward-Dreimeier, Wallsten, and Xu 2003.12. World Bank 2003.13. These indicators of business regulation cover

more than 130 countries and are based on datagathered through study and assessment of laws andregulations in force, with input and verification bylocal experts. For more on the Doing Business Proj-ect, see http://www.worldbank.org/doingbusiness.

14. Country investment climate assessments arebased on standardized firm surveys that cover arange of potential constraints to business opera-tions. For more details on the ICAs, see http://www.worldbank.org/privatesector/ic.

15. For example, Dollar, Hallward-Dreimeier,and Mengistae (2003) find that despite low wages,low productivity resulting from a poorer invest-ment climate makes it hard for Cambodian gar-ment firms to compete in open markets with thosein China, India, and Pakistan.

16. The corporate governance ROSCs arebenchmarked against OECD Principles of Corpo-rate Governance. For more on these assessments,see http://www.worldbank.org/ifa/rosc_cg.html.

17. Even these figures may overstate the growthof private sector credit. Many countries, such asChina, report all nongovernment credit, includingpublic enterprise credit, as private sector credit.

Perc

ent o

f GD

P

80

70

60

50

40

30

20

10

0High-income

countriesMiddle-income

countriesLow-income

countries

FIGURE 4.15 Capital markets are shallow in low-income countriesStock market capitalization, 2002

Source: Standard & Poor’s 2003; World Bank.

18. World Bank Doing Business Project database.19. Yaron, Benjamin, and Piprek, 1997.20. The first index is the Moody’s indicator of

financial strength, with a scale of 0 to 100. It is acomposite of the financial strength of individualbanks measured by key financial ratios. The secondindex is based on the qualitative assessments of com-

pliance with BCPs. (Since 2001, almost 60 countryassessments have been conducted in Financial SectorAssessment Programs, or FSAPs), of which 41 havebeen for developing countries.) The index is con-structed by mapping the qualitative assessment ofeach of the 25 BCPs on a scale from 0 to 100 andtaking the average across BCPs for each country.



Public sector governance is related to theMillennium Development Goals in twocritical ways. First, alongside the eco-

nomic and financial policies reviewed inchapter 4, governance is a key element of theenabling environment for economic growth.Better governance produces better growthoutcomes.1 Growth, in turn, contributes tothe reduction of income poverty and to otherMDGs.2 Higher incomes increase citizens’effective demand for health and education, aswell as resources with which to satisfy thatdemand.

Second, better governance contributes tomore efficient and effective delivery of educa-tion, health, and other services that are keyto the achievement of non-income MDGs.Indeed, it is a critical element in ensuring thatpublic spending translates into quality ser-vices that reach target groups. Where corrup-tion is prevalent, funds may be diverted fromprograms that would benefit the poor towardactivities that benefit well-connected peopleor public officials themselves. Resources arewasted if procurement contracts or jobs areallocated on the basis of connections, ratherthan on merit.3

Governance consists of the manner in whichthe state acquires and exercises the authority toprovide and manage public goods and ser-vices.4 Its two pillars—capacity and account-ability—are anchored on sets of rules called

governance institutions. Those institutionsinteract to shape the incentives of state actorsand thus the quality of governance—the extentto which high-quality public goods and servicesare provided to the citizenry and the businesscommunity. Capacity means that the state hasappropriately skilled personnel, sufficientfinancial resources, and adequate supportingprocesses, such as a working budget system.Accountability means that the state is respon-sible to citizens for delivering public goodsand services; its stewards (politicians, policy-makers, bureaucrats, judges) are prepared toexplain and face the consequences of failuresoccurring within their jurisdiction. Weakcapacity and low accountability—poor gov-ernance—translate into inefficient and inade-quate provision of public goods and services.

Progress in improving public financialmanagement—expenditure management,revenue mobilization—in developing coun-tries is assessed in the next section. There isslow but steady improvement in developingcountries on these dimensions of public sec-tor governance, but reforms need to go muchfurther. Transparency, accountability, andcontrol of corruption in government areassessed in the subsequent section. Progress ismore modest in these areas than in thosemore amenable to “technocratic” reforms—public financial management—and seriousweaknesses persist in many countries.


Upgrading Public Sector Governance

5

The analysis underscores the need to accel-erate governance reform and capacity build-ing in the public sector, with special attentionto improving the quality of governance—reduction of corruption, enhancement ofaccountability. As compared to other policyareas, performance in developing countries isin general weakest in public sector gover-nance. And within public sector governance,it is weakest with respect to transparency,accountability, and control of corruption.The weaknesses are most pervasive preciselyin countries where stronger institutionalcapacities are needed to manage developmentinterventions that will spur progress towardthe MDGs—poor countries. Governancereform is of crosscutting significance, sincepoor governance and weak institutions canseriously undermine the effectiveness of poli-cies and programs throughout an economy.

Fortunately, recognizing the centrality ofgood governance, governments in more andmore developing countries are stepping upreform efforts. The New Partnership forAfrica’s Development initiative is a particu-larly encouraging development, especially asit manifests clear and strong African owner-ship of the governance reform agenda.

Measuring the quality of public sector gov-ernance is difficult. While recent years haveseen much progress in this area, more work isneeded. Efforts to improve the indicators of

the quality of governance are reviewed in thelast section.

Public Financial ManagementThe importance of improved public resourcemobilization and use is underscored by theneed to support stronger programs in infra-structure and human development as part ofthe effort to achieve the MDGs, as discussedin chapters 6 and 7. While increased externalassistance will help, more resources will alsohave to be found domestically to supportthese priorities—through raising more rev-enues and improving expenditure allocationand management.

Public Expenditure and BudgetManagement

Developing countries’ performance in publicfinancial management improved slowly from1999 to 2003, according to the World Bank’sCountry Policy and Institutional Assessment(table 5.1). Not surprisingly, average ratings arelower for low-income countries, and are lowerstill for low-income countries under stress.

Improved expenditure tracking and man-agement is an important element of reform inthe heavily indebted poor countries coveredby the HIPC initiative. The HIPC Assessmentand Action Plan (AAP) assesses the qualityand capacity of public expenditure systems inthese countries and formulates action plans tostrengthen those systems. The assessmentfocuses on government capacity to managepriority public spending as defined in eachcountry’s PRSP. It covers a minimum of 15measures indicative of the quality of a coun-try’s public expenditure management system(table 5.2).

The first major assessment under the AAPwas completed in 2002, based on field assess-ments carried out during 2001–02. Covering24 countries, the assessment concluded thatpublic financial systems of 15 countriesneeded substantial upgrading; those of the 9other countries also needed some improve-ment. Twenty-one countries had inactive or



TABLE 5.1 Improvement in public financial management in developing countries, 1999–2003(average CPIA ratings)

Quality of budgetary Efficiency ofand financial revenue management mobilization

Income category 1999 2003 1999 2003

All developing countries 3.21 3.43 3.27 3.56LICUS 2.05 2.30 2.50 2.73Low income 2.95 3.18 3.09 3.32Middle income 3.54 3.73 3.27 3.56

Note: Table entries are mean values for indicated country groups. Ratings are on a rising scale of 1 to 6.Source: World Bank CPIA database.

ineffective audit arrangements, and 20 lackeda medium-term perspective integrated intotheir budget formulation process. On thebrighter side, in 16 of the countries it waspossible to establish a broad mapping of bud-get data to the poverty-reducing spending cat-egories defined in final or interim PRSPs.Fifteen of the 24 countries had conductedpublic expenditure tracking surveys to moni-tor whether budget allocations were reachingthe intended service providers.

A 2003 update on the assessment indi-cated that most countries were makingprogress in implementing action plans thatwere developed to address the weaknessesidentified in the earlier assessment. More thanthree-fourths of the measures described in theaction plans had been implemented or wereunder implementation. An increasing numberof countries were able to report on poverty-

reducing spending—and to report that suchspending was rising in relation to GDP andtotal expenditures.5

The 2004 HIPC AAP reassessment willyield results for 28 countries in mid-2004,although fewer than seven assessments hadbeen completed as of February 2004. Prelim-inary results suggest mixed progress, withsome countries improving and others fallingin the number of benchmarks met. The for-mer, unsurprisingly, were the ones that the2003 update found as more active in imple-menting their action plans.

The IMF prepares Reports on the Obser-vance of Standards and Codes that have beenagreed upon for fiscal transparency. Thereports examine transparency practices inrelationships between levels of government,adequacy of budget preparation, executionand reporting, and strength of oversight. The

U P G R A D I N G P U B L I C S E C T O R G O V E R N A N C E


TABLE 5.2 Public financial management benchmarks used in HIPC assessments

Budget management Benchmark description

Comprehensiveness1. Composition of the budget entity Meets GFS definition of general government2. Limitations to use of off-budget transactions Extra- (or off-) budget expenditure is not substantial3. Reliability of budget as guide to outturn Level and composition of outturn is “quite close” to budget4. Data on donor financing Both capital and current donor-funded expenditures included

Classification5. Classification of budget transactions Functional and/or program information provided6. Identification of poverty-reducing expenditure Identified through use of classification system (such as a virtual poverty fund)

Projection7. Quality of multiyear expenditure projections Projections are integrated into budget formulation

Internal control8. Level of payment arrears Low level of arrears accumulated9. Quality of internal audit Internal audit function (whether effective or not)10. Use of tracking surveys Tracking use on regular basis

Reconciliation11. Quality of fiscal/banking data reconciliation Reconciliation of fiscal and monetary data carried out on routine basis

Reporting12. Timeliness of internal budget reports Monthly expenditure reports provided within four weeks of end of month13. Classification used for budget tracking Timely functional reporting derived from classification system

Final Audited Accounts14. Timeliness of accounts closure Accounts closed within two months of year end15. Timeliness of final audited accounts Audited accounts presented to legislature within one year

Source: IMF–World Bank 2002.

Rep

orti

ngEx

ecut

ion

Form

ulat

ion

reports find that there has been progress inthese areas, but much remains to be done in anumber of countries (box 5.1).

Building on the assessment work under theHIPC AAP, efforts are under way at the IMFand the World Bank—working with the mul-tidonor Public Expenditure and FinancialAccountability Program (PEFA)6 and in coop-eration with the OECD-DAC Joint Ventureon Public Financial Management—to developa broader public financial management per-

formance measurement framework that couldbe used in a wider group of client countries.The HIPC AAP indicators are the foundationof this system, but it will include some addi-tional indicators as well, for example, thoserelating to fiscal risk management, govern-ment procurement, external audit, and parlia-mentary scrutiny. The proposed frameworkwill be a central element of an integrated diag-nostic assessment of public financial manage-ment, undertaken increasingly in the context



As of the end of April 2004, fiscal reports on standards and codes (ROSCs) had been completed for46 developing countries; 41 are published on the IMF Web site. Analysis of the completed ROSCspoints to a number of common areas for attention.

� Reporting and dissemination of fiscal information tend to be incomplete and subject to consid-erable lags. Many developing countries do not meet adequate standards of timeliness or relia-bility in reporting. Coverage of general government expenses is weak in many countries becauseof inadequate subnational data. Moreover, a wide range of governments resort to off-budget fis-cal activity, such as quasi-fiscal activities and unreported contingent liabilities (for example,Malawi, Mali, Poland, Romania, and Tanzania).

� Unrealistic budgeting is a widespread phenomenon. Because of the absence of clear legal andregulatory provisions governing the budget (for example, in Azerbaijan, Mali, and the Philip-pines) or weak compliance with such provisions (such as in Bangladesh and Papua New Guinea),actual spending often differs substantially from the budget. This, together with poor reportingof expenditures, limits the usefulness of the budget as an indicator of fiscal policy.

� Many of the above issues arise because government activities are not clearly distinguished fromthose of nongovernmental public sector agencies. This factor is often the genesis of excessivequasi-fiscal activity (for example, in Bangladesh, Iran, and Malawi). In addition, the roles andresponsibilities of the different levels and branches of government are not yet clearly defined insome transition countries (such as Azerbaijan and the Kyrgyz Republic).

� Widespread tax exemptions, excessive discretion in tax administration, and inadequate enforce-ment are also common in many developing countries (such as Bangladesh, the Kyrgyz Republic,and Malawi).

� External audit mechanisms are weak in a number of countries (Azerbaijan, Benin, Honduras,and Mauritania, for example).

Many countries participating in fiscal ROSCs have since undertaken or are undertaking signifi-cant fiscal reforms that will lead to improved fiscal transparency practices. These reforms includestrengthening budget execution and reporting (Cameroon, Honduras, Papua New Guinea, Zam-bia), implementing program budgeting (Brazil, Mexico, South Africa), developing modern budgetlaws (Brazil, the Czech Republic, Hungary, Poland, South Africa), and reducing the scope of quasi-fiscal activities through privatization and price liberalization (Benin, Burkina Faso, Mozambique,Nicaragua, Tunisia, Uganda, Zambia).

Source: IMF Fiscal Affairs Department.

BOX 5.1 Improving fiscal transparency through ROSCs

of country-owned and -led developmentstrategies. The approach going forward isanchored on the following key principles: gov-ernment articulation of a public expenditurereform strategy in the PRSP or other country-owned document; an integrated and well-sequenced program of diagnostic work bydevelopment partners; well-coordinated tech-nical and financial support from developmentpartners for implementation of the countries’public expenditure reform strategies; and peri-odic reporting by countries of performance inpublic expenditure policy, financial manage-ment, and procurement.7

Revenue Mobilization

Excessive taxation can impede private sectordevelopment, but in most developing coun-tries the problem is collecting enough revenueto provide essential public infrastructure andhuman development services. Tax revenue inlow-income countries as a share of GDP isabout 14 percent, compared with about 19percent in lower-middle-income countriesand 23 percent in upper-middle-income coun-tries (table 5.3). The ratios, however, conceallarge variations across and within regions.

Overall, it appears that there is scope in manycountries for raising additional revenues.

Policies to enhance domestic revenuemobilization need to focus on a few taxes thatcan generate significant revenues at relativelylow cost. In view of the high share of agricul-ture and informal economic activity in manycountries, corporate and personal incometaxes are unlikely to be a major source ofdomestic revenues in the short to mediumterm. The value-added tax (VAT) perhapsoffers the greatest potential. To maximize rev-enues from a VAT, countries need to ensurethat the tax base is as broad as possible andthe rate structure is simple. Excise taxesalso have significant potential, although careneeds to be taken in their design, as some ofthem can be regressive. Currently, excise rev-enues amount to less than 2 percent of GDPin low-income countries, compared withabout 3 percent in high-income countries.Levied on products such as alcohol, tobacco,petroleum products, vehicles, and spareparts, this tax has a potentially buoyant baseand can be administered at a low cost.

Reducing tax exemptions and tax incen-tives can generate substantial revenues. Inmany developing countries, tax incentives



TABLE 5.3 Central government tax revenue, 1990–2001(percent of GDP)

Total tax Taxes on General Social Countries/regions revenue international trade Excises sales tax security

Low income (21) 14.1 4.2 1.8 3.1 1.8Lower middle (26) 18.5 3.4 2.3 4.9 3.7Upper middle (20) 23.1 3.1 2.5 5.9 6.4High income (31) 26.7 1.4 2.7 6.0 7.4

Asia (12) 13.4 3.7 2.1 2.6 0.3Latin America and Caribbean (16) 17.0 2.0 2.4 4.7 2.6Middle East and North Africa (6) 17.9 4.5 2.3 3.3 1.7Sub-Saharan Africa (12) 16.6 6.8 1.4 3.1 0.5Europe and Central Asia (20) 23.9 1.6 2.6 6.4 7.7OECD (23) 29.1 0.6 3.0 6.0 8.4

Note: Figures are unweighted averages over the indicated period. Some taxes are omitted; hence, columns may not add to the total. Numbers inparentheses represent the maximum number of countries in each group for which data are available. Because of data availability, sample sizevaries across taxes.Source: IMF Government Finance Statistics; IMF World Economic Outlook, various years.

and exemptions are a major source of revenueloss. They not only shrink the tax base butalso complicate tax administration. Taxincentives are often used to attract invest-ment. But studies have shown that investmentdecisions depend on a host of other factorsthat can be more important than tax incen-tives.8 Once established, tax incentives andexemptions are difficult to abolish. In addi-tion, their budgetary cost is not made explicit.Eliminating or reducing tax incentives andexemptions can enhance the efficiency of thetax system by enlarging the tax base and sim-plifying tax administration.

Efforts to mobilize tax revenues shouldinclude steps to strengthen tax administra-tion. Policies should focus on strengtheningthe technical capacity and organization ofrevenue authorities through computerizationand improved operating procedures. Stricterenforcement mechanisms and improved taxaudits and inspections could increase tax-payer compliance. Countries could increasetheir tax-to-GDP ratio by an estimated 1 to2 percentage points through elimination oftax incentives and exemptions and betteradministration.9

But the potential for generating additionalrevenues is almost certainly less than what isneeded to achieve the MDGs in most develop-ing countries. The bulk of the financing will

have to come from improving the efficiency ofexisting spending, economic growth, and exter-nal resources.

Quality of Public Administrationand Control of CorruptionPolicy ratings for developing countries on thequality of governance—quality of publicadministration, transparency, accountability,and control of corruption—are lower onaverage than their ratings for public financialmanagement (table 5.4). This is not alto-gether surprising. Changing public expendi-ture or revenue management or budgetprocesses, where technocratic improvementscan bring quick returns, is likely to be easierthan changing the culture of public bureau-cracies and strengthening accountability.There may be powerful interests with a stakein blocking reform in any policy area. How-ever, opposition to reforms that embraceknown international best practices is moreeasily exposed as self-interested. Best practiceis clearer in budget management and tax pol-icy than in other areas, where accountabilitymechanisms are weak. Less is known abouthow to change the culture of dysfunctionalpublic bureaucracies, or how to implementthe checks on executive power necessary toensure an open and accountable administra-tion. Progress in these areas depends on acareful understanding of the underlying con-text and behavior and fostering of reformownership.

Although public sector governance ratingsare low on average compared to the ratingsfor other policy areas—making it the weakestpolicy area—there is a trend toward improve-ment. All developing country regions showedsome improvement in the quality of gover-nance from 1999 to 2003, as measured by theWorld Bank CPIA (figure 5.1). The largestimprovement was in Europe and CentralAsia, followed by South Asia. The lowestaverage ratings are in Sub-Saharan Africa,but there is evidence that increased attentionto governance issues is beginning to pay off,as reviewed in the next section.



TABLE 5.4 Quality of public sector governance, 1999–2003(average CPIA rating)

Transparency, accountability,

Quality control ofof public corruption

administration in public sector

Income category 2001 2003 1999 2003

All developing countries 3.13 3.19 2.92 3.18LICUS 2.03 2.25 2.05 2.23Low income 2.87 2.95 2.73 2.97Middle income 3.33 3.49 3.16 3.44

Note: Table entries are mean values. Ratings are on a rising scale of 1 to 6. For the quality ofpublic administration, the base year shown is 2001 rather than 1999, because the CPIA defini-tion of that criterion was revised significantly in 2001.Source: World Bank CPIA database.

The patterns and trends in governancerevealed by CPIA assessments are corrobo-rated by similar assessment exercises under-taken by the regional development banks. Asreviewed in chapter 3, country ratings by theAfrican Development Bank, Asian Develop-ment Bank, and Inter-American DevelopmentBank show public sector governance to be theweakest policy area on average across coun-tries. Where ratings are available across years—as in the AfDB and ADB assessments—theyalso show that governance is improving, albeitgradually. The particularly notable improve-ment in governance in Europe and Central Asiais confirmed by the results of the Business Envi-ronment and Enterprise Performance Surveys(BEEPs), conducted jointly by European Bankfor Reconstruction and Development and theWorld Bank. For example, the BEEPS surveysfound that in a majority of the developing andtransition countries in the region, the incidenceof payment of bribes by firms fell substantiallybetween 1999 and 2002.10

The worldwide governance indicators com-piled by the World Bank Institute (WBI) and theBank’s Development Economics Group (DEC)are an important source of cross-country assess-ments of the quality of governance. They takethe form of indices of various aspects of thequality of governance, aggregated from severalsources.11 Consistent with the CPIA results,average ratings on the WBI/DEC indicators arelower for low-income countries than for all

developing countries, and lowest for the LICUSgroup (table 5.5). The gaps relative to the indi-cators for advanced countries are large. How-ever, there is no clear trend over time in countrygroup ratings relative to those of other groups.The indicated changes between 1998 and 2002are too small to be statistically significant.



FIGURE 5.1 The quality of governance is weak but improving in developing countriesScores for overall quality of governance by region,1999–2003

5

4

3

2

1

0

Aver

age

gove

rnan

ce r

atin

g

1999 2003

SouthAsia

MiddleEast &NorthAfrica

LatinAmerica

&Caribbean

Europe &Central

Asia

East Asia& Pacific

2.963.19 3.22 3.35 3.22

3.61 3.49 3.603.20 3.41 3.39

3.68

Sub-SaharanAfrica

Note: Because the differences across regions are similar for each of the CPIA public sector governance criteria, ratings for criteria 4, 17, 18, and 20 are averaged in this figure. Criterion19 was omitted because comparable data are available beginning only in 2001. In all, CPIAhas 20 criteria, of which these 5 are more directly related to public sector governance.Source: World Bank CPIA database.

TABLE 5.5 Worldwide governance indicators, 1998–2002

Control of Voice and Government corruption accountability effectiveness

Income category 1998 2002 1998 2002 1998 2002

All developing countries –0.44 –0.48 –0.19 –0.24 –0.34 –0.39LICUS –0.93 –1.10 –0.95 –1.08 –1.16 –1.24Low income –0.72 –0.80 –0.40 –0.49 –0.54 –0.65Middle income 0.44 0.41 0.27 0.31 0.39 0.44

OECD DAC members 2.02 1.77 1.35 1.43 1.81 1.71

Note: Table entries are mean values for country groups. Values represent the number of standard deviations above or below (indicated by a minussign) the overall mean value (of 0). The approximate range of values for each index is +2 to –2.Source: World Bank, WBI/DEC governance database.

Some of the data sources used in theWBI/DEC governance indicators provide agood basis for a comparison of the level of rat-ings over time.12 The Freedom House indica-tors of civil liberties, for example, have been

published since the mid-1970s and have useda fixed methodology since 1990. The indica-tors show a gradually improving trend in civilliberties in developing countries (figure 5.2).13

Objective indicators corroborate the trendin Freedom House indicators. Over time,there has been an increase in the number ofchief executives in developing countries whohave been elected rather than appointed byother means. Moreover, the number of chiefexecutives who have been elected in contestsfeaturing candidates from multiple partieshas increased. There is some indication ofincreasing fairness and integrity in electoralprocedures. For example, the percentage ofchief executives who have been chosen inmultiparty elections in which the winnerreceived less than 75 percent of the officialvote has increased since 1993 (figure 5.3).14

The data on executive elections, together withthe Freedom House data, indicate a trendtoward broader participatory processes bywhich citizens can influence policymakingand hold their leaders accountable.

Increased Focus on Governancein AfricaGood governance is an increasingly impor-tant topic in Sub-Saharan Africa, which his-torically has been a weak performer in thisarea. A powerful consensus is evolving acrossAfrica on the critical role of good governanceas a precondition for Africa to meet its devel-opment challenges. The MDGs are beinginstrumental in focusing the attention of pol-icymakers on the need to accelerate reformsto deal with governance-related impedimentsto the achievement of poverty reduction andother development goals. The PRSP process ishelping to integrate these reforms into over-all country development strategies.

A key manifestation of the strengthenedresolve in the region to come to grips with thegovernance agenda is the New Partnership forAfrica’s Development (NEPAD). Launched in2002, this initiative, owned and led by Africancountries, places improvement of governanceat the center of the reform agenda. To give



5.0

4.5

4.0

3.5

3.0

2.5

2.0

Free

dom

Hou

se in

dex

20022001200019991998

Middle-income countriesLow-income countriesLow-income countries under stress

199719961995

FIGURE 5.2 Civil liberties are gradually improving in developing countriesTrend in civil liberties in developing countries, 1995–2002

Source: Freedom House.

20001999199819971996199519941993

80

70

60

50

40

30

20

10

Perc

ent

Middle-income countriesLow-income countriesLow-income countries under stress

FIGURE 5.3 Participatory processes are also improving in developingcountriesPercentage of country leaders elected in multiparty elections with less than 75 percent of the vote

Source: Beck and others (2000).

concreteness to its reform focus and objectives,NEPAD is putting in place an African PeerReview Mechanism that aims to improve gov-ernance policy and practice through mutuallearning from national governance assess-ments prepared by the African countries them-selves (box 5.2).

An increased focus on governance is alsoreflected in the work of regional developmentagencies. An example is a major new reporton governance in Africa prepared by theUnited Nations Economic Commission forAfrica (UNECA). The report, which draws ondetailed individual country studies, indicates

that while the reform agenda is huge, Africancountries are making progress toward goodgovernance. It emphasizes African achieve-ments but recognizes the occurrence of set-backs and the depth of the challenges ahead(box 5.3).

Improving the Monitoring of GovernanceThe centrality of good governance to devel-opment effectiveness underscores the impor-tance of reliable measures of the variousdimensions of the quality of governance.



The New Partnership for Africa’s Development (NEPAD), an organization of African leaders whohave pledged to place their countries on a path of sustainable growth and development, haslaunched the African Peer Review Mechanism (APRM) “to foster the adoption of policies, stan-dards, and practices that lead to political stability, high economic growth, sustainable development,and accelerated subregional and continental economic integration through sharing of experiencesand reinforcement of successful and best practice, including identifying deficiencies and assessingthe needs of capacity building.”

The APRM addresses all dimensions of governance, organized under four headings: democracyand political governance; economic governance and management; corporate governance; and socio-economic development. By the end of 2003, 16 countries had formally joined the APRM: Algeria,Burkina Faso, Cameroon, Republic of Congo, Ethiopia, Gabon, Ghana, Kenya, Mali, Mauritius,Mozambique, Nigeria, Rwanda, Senegal, South Africa, and Uganda. An independent panel of seveneminent Africans, working with the countries, is responsible for producing country reviews to be sub-mitted to NEPAD and the African Union. The first reviews are scheduled to begin in April 2004.

In March 2003, NEPAD agreed on a set of objectives, standards, criteria, and indicators to beapplied in each country review. The set includes 25 broad governance and development objectivesand 76 African and international standards. Among the objectives are the following:

� An objective on constitutional democracy, political competition, and the rule of law that takesas its standard the African Charter on Human and Peoples’ Rights

� An objective on transparent, predictable, and credible economic policies that includes interna-tional accounting and auditing standards and codes of good practices

� An objective on accountable, efficient, and effective public office holders and civil servants thatrefers to the African Union’s Convention on Preventing and Combating Corruption.

The APRM process will build upon in-country self-assessments that involve extensive consulta-tions with all branches of government, civil society, and the private sector. The challenge is to buildupon existing processes, such as the PRSPs, and add value with governance assessments made byAfrican stakeholders, their governments, and eminent independent opinion leaders from the region.Expanding the APRM to all NEPAD member countries and ensuring wide discussion of its findingswill be important to the success of the APRM initiative.

Source: NEPAD 2003a, 2003b.

BOX 5.2 The African Peer Review Mechanism: self-assessing governance



The United Nations Economic Commission for Africa (UNECA) has launched a multiyear projectto monitor progress toward good governance in Africa. Twenty-eight country studies have beenundertaken, each based on a household survey, a survey of national experts, and a desk-study witha range of governance indicators. The project explores three components of good governance: polit-ical representativeness, institutional effectiveness and capacities, and economic governance and man-agement. Summarizing the country reports, UNECA’s African Governance Report (forthcoming)finds that:

� African countries have made significant strides toward political representation. The political sys-tem has been liberalized, more political parties are being allowed to operate, civil society hasgained a place in social and political bargaining, elections form the basis of political leadership,and the integrity of the political system is on the rise in many countries. Most countries haveembarked upon the process of constitutional review, to promote a culture of constitutionalismwith adherence to the rule of law, due process, and political accountability. The issue of leader-ship succession and change through the electoral process constitutes a significant step in theprocess of democratic renewal and a new culture of governance in Africa.

� Despite progress on social inclusiveness in the elected organs of government, drawbacks remainin two major areas: domination of the political process by an urban-based political elite, andunfair treatment of minority groups. Although the political regime is becoming more inclusivethrough more robust public participation achieved by the media, civil society, and high votingrates, political parties remain weak.

� Democracy, good political governance, effective economic governance, and prudent public finan-cial management are intertwined. Management weaknesses persist, especially in public pro-curement, and in equity in resource mobilization and public expenditure. Several Africancountries are using the Medium-Term Expenditure Framework (MTEF) to improve macroeco-nomic stability, resource allocation, and prioritization of public expenditures, leading to greateraccountability for the outcomes of public expenditures.

� Dominance of the executive branch is gradually decreasing. Laws, rules, and regulations pro-posed by the executive are debated in standing or select committees of the legislature and sub-sequently enacted. Other emerging checks and balances are the laws and regulations managingthe ways in which a government conducts its business: conflict of interest laws, codes of conductfor elected and appointed officials, and asset declarations by politicians. The judiciary remainsin a formative stage, as many African governments are still defining judicial independence. Inmany countries, an overburdened court system denies poor and disadvantaged citizens access tojustice, encourages corruption, and undermines the rule of law.

� African governments are making headway in terms of institutional capacity, accountability,transparency, and policy coherence, whereas service delivery is still inadequate. In many coun-tries, criteria other than merit still take precedence in recruitment, promotion, and discipline ofpublic servants. While government business is often shrouded in official secrecy, some countrieshave passed legislation on administrative justice and access to information. Although the statebureaucracy exercises growing institutional and professional competence in policy design andimplementation, there is considerable variation among African countries in the adequacy, qual-ity, and efficiency of basic public services.

Source: UNECA forthcoming.

BOX 5.3 Governance in Africa—progress on a difficult agenda



More and better measures of governance areavailable today than was the case about adecade ago. But work continues to improvethe indicators to enhance their robustness.

Effective monitoring of governance mustideally meet the following requirements: agree-ment on a set of international benchmarks forgood governance; identification of specific gov-ernance measures that indicate change towardor away from the agreed benchmark; and for-mulation of indicators that are cost-effectiveand replicable and that can be applied andmanaged by the countries themselves. Espe-cially difficult challenges attend the search forrobust indicators in areas of governance notrelated to the public budget—accountability,corruption—where agreed international bench-marks and good quantitative measures arelacking in comparison.

At the World Bank, efforts include bothimproving indicators that allow cross-countrycomparisons to be made (for example, im-proving the quality of the CPIA governanceindicators by underpinning them with moredetailed and specific assessment criteria) anddeepening the analysis of governance issues atthe individual country level, for example,through in-depth Institutional and Gover-nance Reviews (IGRs)15 and Governance andAnti-Corruption (GAC) Diagnostic Surveys.16

Several partner institutions also are involvedin similar efforts, individually as well as withothers. The work being carried out in theframework of PEFA and by NEPAD APRMand UNECA, as noted above, are examples.The Africa work is especially promising in viewof its strong ownership and leadership by thecountries themselves, and its use of self-assess-ment and peer review mechanisms. Gover-nance is also an area where private think tanksand civil society are active. Their measures pro-vide a useful supplement to those developed byofficial entities.

Future assessments of governance will ben-efit from an increased emphasis on the need toground governance measures as much as possi-ble in objective, quantifiable indicators.17

Nonetheless, qualitative assessment will remain

an inevitable element of these measures, giventhe difficulty of quantifying behavioral aspectsof governance. There is also an increasing useof citizen surveys, which would complement—and provide a check against—assessmentsbased on expert opinion.

Another area of emphasis in the futurework on public sector governance is thedevelopment of better information at the sub-national level of governance. In an effort toimprove service delivery at the local level,many governments in developing countriesare devolving more responsibilities to localgovernments. Effective public resource man-agement and adequate institutional capacitiesat the local level will become increasinglyimportant to development effectiveness andto the effort to achieve the MDGs. Monitor-ing at the subnational level is at presenthampered by major data gaps relating to sub-national finances and administrative arrange-ments. Filling these gaps should receive moreattention in the work ahead.

Notes1. Mauro 1995; Knack and Keefer 1995; Evans

and Rauch 1999; IMF 2003. Knack and Keeferfound that each 10-point increase in a 50-pointindex of the quality of governance was associatedwith a 1-percentage-point increase in growth ofper capita income. The Spring 2003 IMF WorldEconomic Outlook found that an improvement inSub-Saharan Africa’s level of institutional devel-opment from its current average to the mean ofdeveloping Asia would imply an 80 percentincrease in the region’s per capita income.

2. Dollar and Kraay 2002; Bruno, Ravallion,and Squire 1998. Pritchett and Summers (1996)showed that each 10 percent increase in incomereduces infant and child mortality in developingcountries by between 2 and 4 percent.

3. Kaufmann, Kraay, and Zoido-Lobaton (1999)showed that countries with better governance—asmeasured by measures of corruption, voice, andaccountability—have lower infant mortality andhigher literacy rates. They found that an improve-ment of 1 standard deviation in the quality of gov-ernance produces an increase of 15 to 25 percentagepoints in literacy. Much of this effect is indirect,

through the impact of governance on income levels;but part of it is attributable to the direct effect ofbetter governance on more accountable and effec-tive delivery of public services. The quality of gov-ernance has also been linked to volatility in incomegrowth (IMF 2003), which can affect service deliv-ery quality by increasing the unpredictability ofboth revenues and demand for services.

4. World Bank 2003.5. World Bank 2003. 6. The Public Expenditure and Financial

Accountability Program was established in 2001.PEFA is a partnership of U.K. Department forInternational Development, the European Com-mission, France, Norway, SECO (Switzerland), theStrategic Partnership with Africa, and the IMF andWorld Bank. PEFA supports integrated and har-monized approaches to assessment and reform ofpublic expenditure, procurement, and financialaccountability systems. See http://www.pefa.org.PEFA cooperates with the OECD-DAC Joint Ven-ture on Public Financial Management, for whichthe establishment of a means to monitor publicfinancial management performance in partnercountries is a priority objective.

7. IMF–World Bank 2003.8. OECD 1994; Tanzi and Zee 2000.9. World Health Organization 2002.10. Not all data sources, however, show the

same large improvements in Europe and CentralAsia as shown by the CPIA and BEEPS.

11. The advantage of aggregating indicatorsfrom different sources is that it reduces the impactof measurement error in any single source, pro-ducing more accurate comparisons at a point intime. However, index values for each country aredefined relative to the mean for all countries in a

given year, so the indexes are not designed to trackchange over time on an absolute scale (Kaufmann,Kraay, and Mastruzzi 2003).

12. The benefit of disaggregated indexes comesat the potential cost of increased measurementerror, which results from relying on a smaller num-ber of sources. However, because the purpose hereis to describe the performance of large groups ofcountries, errors in measuring the performance ofindividual countries are of less relevance.

13. Freedom House assigns a rating of 1 forcountries with the most civil liberties and 7 for thosewith the least. For figure 5.2, the index was reversedso that higher values indicate more civil liberties.

14. These data are from the Database on Polit-ical Institutions (Beck and others 2000). The 75percent threshold used is meant to capture com-petitive elections with candidates from multipleparties; a greater than 75 percent vote share oftenis associated with questions about the fairness ofthe election.

15. Fifteen Institutional and Governance Reviewswere completed during 2001–03, and a similar num-ber is planned for 2004. More information on theIGRs is available at: http://www.worldbank.org/prem/PREMNotes/premnote 75.pdf.

16. The Governance and Anti-Corruption Sur-veys cover citizens, firms, and public officials, andare used to diagnose areas of particular strengthsand weaknesses in governance, such as the inci-dence of corruption across government agencies orpublic services. Surveys have been completed orare under way in 25 countries, about half of themin Europe and Central Asia.

17. These are sometimes referred to as “secondgeneration governance indicators.” See Knack,Kugler, and Manning 2003.



Infrastructure plays a dual role in the effortto achieve the MDGs. It is an importantelement of the enabling environment for

economic growth—energy to power industry,transport to support commerce. Reducingpoverty by promoting growth is one channelthrough which infrastructure contributes tothe achievement of the MDGs. But the avail-ability of reliable and affordable infrastruc-ture also contributes more directly to theattainment of the MDGs by providing or sup-porting the delivery of key services. SomeMDGs relate to particular infrastructure ser-vices, such as those that aim to reduce the pro-portion of people without sustainable accessto safe drinking water and basic sanitation,and to make housing and shelter more acces-sible. The MDGs related to human develop-ment (education, health, empowerment ofwomen) rely on services whose effectivedelivery depends greatly on supportive infra-structure: improved water and sanitation toprevent disease and free up women’s timefrom daily chores; electricity to serve schoolsand health clinics; and roads to access them.

But there are large gaps in the availabilityand quality of the key infrastructure servicesin developing countries. Although particu-larly great in low-income developing coun-tries, those gaps remain sizable in mostmiddle-income countries. Within countries,coverage rates are typically much lower in

rural areas, where the majority of the poorpopulation lives; but coverage in urban areasalso is under pressure from rapid rural-to-urban migration in many countries.

Narrowing gaps in access and quality willrequire sizable increases in investment and inassociated spending on operation and mainte-nance (O&M). Efforts must continue to im-prove the enabling environment for privateinvestment, which has increased but not asmuch as expected. At the same time, publicinvestment in infrastructure will need toreverse the decline of the past decade. This willrequire stronger domestic resource mobiliza-tion, as well as increased foreign assistance.

Increased investment alone, however, is notthe answer. Investment must be underpinnedby improvements in policy and governance toensure effectiveness and sustainability. Manycountries have made progress in this respect byimplementing policy and regulatory reform.Progress has been uneven across regions andincome groups. Africa and low-income coun-tries have tended to lag behind other groups,although not on all dimensions. Progress hasalso been uneven across sectors. In general,telecommunications is well ahead of the curve;electricity, transport, and housing are at inter-mediate stages of reform; and water and sani-tation lag behind.

Assessment of policies and needs in infra-structure is seriously constrained by gaps in


Strengthening Infrastructure6

data. A better information base is needed, notonly for effective monitoring but also toinform policymaking. Filling the data gapsshould be a priority area for donor support.

Infrastructure and theMillennium Development GoalsA growing body of evidence, across andwithin countries, supports the importance ofinfrastructure in facilitating growth, reducingpoverty, and achieving related developmentgoals.1 In a sample of 102 studies conductedover the last 15 years, few report a negativeeffect of infrastructure investment on pro-ductivity or growth (table 6.1). The sampleincludes 30 multicountry studies (includingdeveloping countries); 41 studies on the UnitedStates; 19 on Spain; and 12 on specific devel-oping countries (Argentina, Brazil, Colombia,India, and the Philippines).

Studies of the United States and of multi-ple countries offer mixed results: the effect ofinfrastructure investment does not appear tobe significant in about half of the cases. Theresults for the United States reflect the effectsof differences in infrastructure endowmentsin a mature economy. Findings of negativeeffects often can be explained by saturation ofsome types of infrastructure or by negativeexternalities. For example, a new transportproject in a specific state may facilitate anexodus of workers or industries from otherstates, hence slowing down growth in those

states. Similar conclusions emerge from thecross-country data sets.

The studies of Spain make a much strongercase for investment in infrastructure. In theSpanish context, infrastructure has generallybeen found to be a major determinant ofgrowth and productivity convergence acrossregions. Studies of developing countries con-firm the role of infrastructure in reducing dis-parities between rich and poor regions.Research on Argentina, Brazil, Colombia,India, and the Philippines demonstrates thesignificant impact on growth and productiv-ity of various types of infrastructure. Takenjointly, the results suggest that the returns oninvestment in infrastructure are probablyhigh in the early stages of development, wheninfrastructure is scarce and basic networkshave not been completed. Returns on infra-structure investment tend to fall, sometimessharply, as economies reach maturity.

Measuring the elasticity of output toimprovements in infrastructure quantity orquality gives a more concrete sense of thepotential impact of infrastructure investmentand associated policies. Depending on thesector, country (or country groups), and theperiod covered, the elasticity estimates rangefrom 0.14 to 1.12, but the lower bound is notas small as it may seem. Consider the case ofLatin America. In the 1990s, the elasticitiesestimated for that region imply that a 10 per-cent increase in infrastructure stocks wouldhave enabled an increase of 1.4 to 1.6 percentin output—quite significant, since a rise of 1percentage point in per capita income wouldreduce the share of people living in poverty byhalf a percentage point.2 Another recentstudy of Latin America estimated that lack ofinvestment in infrastructure during the 1990sreduced long-term growth by 1 to 3 percent-age points, depending on the country.3 Thisassessment suggests that infrastructure insuf-ficiencies account for about one-third of thedifference in output per worker betweenLatin America and East Asia.

The story for Africa is similar. One of themost extensive multicountry studies suggeststhat if Africa had enjoyed growth rates in



TABLE 6.1 Distribution of studies according to their findings on impactof infrastructure investment on productivity or growth

Percent of studies showing

Number Positive No significant Negative Country/group of studies effect effect effect

Multiple countries 30 40 50 10United States 41 41 54 5Spain 19 74 26 0Developing countries 12 100 0 0Total 102 53 42 5

Source: De la Fuente and Estache 2004.

telecommunications and power generationinfrastructure comparable to those observedin East Asia in the 1980s and 1990s, itsannual growth rate would have been about1.3 percent higher.4 The authors’ overall con-clusion is that for developing and transitioneconomies, the quantity and quality of infra-structure are critical components of thegrowth equation, even if the relative impor-tance of each subsector may vary by countryand over time.

A more limited body of literature looks atthe direct impact of improvements in infra-structure on the poor. Datt and Ravallionfound that rural poverty rankings of Indianstates in 1990 were very different from thoseof 1960.5 States starting with better infra-structure and human resources saw signifi-cantly higher long-term rates of povertyreduction. Deininger and Okidi obtained sim-ilar results in exploring factors underlyinggrowth and poverty reduction in Uganda dur-ing the 1990s.6 In their work, the importanceof improving access to basic education andhealth care emerges clearly, but benefits alsodepend on complementary investments inelectricity and other infrastructure. A studyby Fan, Zhang, and Zhang documents thecritical role of infrastructure development,particularly roads and telecommunications,in reducing rural poverty in China between1978 and 1997.7 The authors show that thereduction in poverty is due to growth in ruralnonfarm employment following expansion ofinfrastructure.

Access to infrastructure can have littleeffect, however, if the service is not afford-able.8 Governments must bear this in mindwhen setting tariffs, particularly in caseswhere average tariffs must cover averagecosts to support private sector participation.Two main solutions are being pursued. Alarge body of experience across regions—par-ticularly with water supply and sanitation,rural electrification, and secondary roads—demonstrates the effectiveness of offeringhouseholds (and businesses) a range of servicelevels at a range of costs, allowing them tochoose according to their preferences and

ability to pay.9 A second solution is to designtariff structures that include an explicit, well-targeted subsidy to ensure that the share oftheir income users spend on infrastructureservices stays within reasonable boundaries.A common rule of thumb is that expenditureson utilities and transport should not exceed15 percent of a poor individual’s income.10

Other recent research emphasizes theimportance of core infrastructure inputs forthe achievement of the MDGs. Leipziger, Fay,and Yepes, for example, estimated from asample of 43 countries that differences inaccess to water explained about a quarter ofthe difference in infant mortality between thepoorest and richest quintiles and 37 percentof the difference in child mortality. Similarly,the difference in access to sanitation accountsfor 20 percent and 10 percent, respectively, ofthe difference between the poorest and rich-est quintiles in the prevalence of malnutrition.In rural India, Jalan and Ravallion found thatthe prevalence and duration of diarrheaamong children under five were significantlylower on average for families with pipedwater than for those without it.11

Supply Gaps and Policies in Infrastructure Services: The Current PictureThe preceding section gave a sense of theimplicit demand for infrastructure; this sec-tion provides an assessment of supply. Themonetary value of the world’s infrastructurestock, at best-practice average prices andexcluding housing, is approximately $15 tril-lion.12 Of the total, about 60 percent is foundin high-income countries (16 percent of totalpopulation), 28 percent in middle-incomecountries (45 percent of total population),and 13 percent in low-income countries (39percent of total population). Electricity androads account for about 80 percent of thetotal asset value.

The assessment of the current status ofinfrastructure services in developing coun-tries in terms of access to and quality of ser-vices and progress on sector policy reform is

S T R E N G T H E N I N G I N F R A S T R U C T U R E


constrained by major gaps in data. Indeed,obtaining better infrastructure data should beamong donors’ priorities for improvingdeveloping-country data. The World Bank, incollaboration with partner agencies in theUnited Nations system and elsewhere, hasinitiated an infrastructure indicators projectthat aims to develop a more systematic data-base of core infrastructure indicators.13 Themost serious data gaps are found in the trans-port sector, where the poorest spend half ofthe resources they allocate to public services.For lack of an appropriate international data-base, it is not possible to monitor access ratesto transport services on a regular basis.

Access to Infrastructure Services

City dwellers in low-income countries have farless access to infrastructure services of all typesthan do urban residents of middle-incomecountries (table 6.2). This means that any effortto catch up will require major investments.Rural populations, which make up more than60 percent of the population of low- and mid-dle-income countries, have significantly lowerrates of access than do urban populations—

roughly 30 percent lower (table 6.2). The dis-crepancy does not mean that investmentrequirements are significantly larger for ruralareas than for urban areas, however, becausecosts are lower in rural areas.

Similar differences can be seen in access tohousing. The percentages of the populationwith authorized housing are 36, 73, and 91for low-, lower-middle-, and upper-middle-income countries, respectively.14 Behind theseoverall figures lie major disparities withincountries between low-income slums and bet-ter-off areas. For example, whereas 33 per-cent of households in Mysore, India, havewater connections, only 8 percent of those ininformal settlements do. The result of suchdisparities is evident in health outcomes:whereas the under-five mortality rate in urbanKenya averages 84 (and 62 in Nairobi), in thetwo largest slums it averages 187 (Kibera)and 254 (Embakasi).15

Quality of Infrastructure Services

Although data on access are more meaning-ful when combined with data on service qual-ity, most current access data are not “quality



TABLE 6.2 Access of population to infrastructure services

Country Electricity Water Sanitation Telecoms Transportincome category (1997–2001)a (2000)a (2000)a (1999–2002)b (1994–2001)c

Urban

Low 62.4 (24) 76.9 (56) 74.6 (55) 4.6 (65) 90 (7)Lower middle 95.1 (8) 90.8 (40) 90.5 (37) 22.0 (52) —Upper middle — 92.3 (23) 92.5 (22) 53.9 (36) —

Rural

Low 20.3 (24) 52.7 (54) 41.8 (52) 1.7 (55) 61 (21)Lower middle 67.3 (8) 75.0 (40) 59.8 (37) 8.7 (43) —Upper middle — 76.4 (23) 81.3 (22) 22.5 (24) —

— Not available.Note: Figures in parentheses indicate the number of countries for which data are available.a. Percent of populationb. For urban populations: telephone subscribers per 100 people; for rural populations: main lines per 100 peoplec. For urban populations: percent of population within 20 minutes of public transport; for rural populations: percent of population within 2 km ofan all-season roadSource: USAID, Demographic and Health Surveys; World Bank, World Development Indicators, various years; International TelecommunicationsUnion; Roberts 2003.

adjusted.” Access data usually include indi-viduals with all-day access, as well as indi-viduals with access for just a few hours a day.But quality matters. A recent study coveringseven Latin American countries suggests thatthe effectiveness of public infrastructure inthat region is only about 74 percent of thatin industrialized countries because of poorquality.16 According to the same study, thelong-run cost of this underperformance isequivalent to about 40 percent of real GDPper capita. Raising infrastructure effective-ness to industrialized-country levels wouldreduce the per capita income differencebetween Latin America and the United Statesfrom tenfold to about sevenfold.

Despite growing recognition of the eco-nomic and social significance of service qual-ity, there is currently no indicator of thecombined dimensions of quality. So in theshort run, the only option is to rely on partialindicators. In some cases, for example, qual-ity must be inferred from associated healthindicators. While this is clearly not satisfac-tory, analysis based on partial indicators pro-vides a global sense, for each sector and eachcountry group, of the importance of objectivequality indicators.

The amount of information available ontechnical quality varies widely across sectors

(table 6.3). Reasonable technical indicatorsare widely available for electricity and tele-communications. The usual approximationfor transport, the ratio of paved roads to totalroads, is useful, but it may be somewhat mis-leading, since paving roads is not necessarilya priority for the poorest countries. Mostworrisome is the water supply and sanitationsector, which does not systematically generateindicators such as water losses, number ofhours of service per day, water quality, or vol-ume and quality of treated sewage.

The available data suggest that the corre-lation between income levels and technicalquality is extremely high. The main practicalconsequence of that correlation is that the dif-ference in investment requirements acrossincome groups is probably much more signif-icant than the access data alone would sug-gest. Not only must access be improved, butmajor rehabilitation efforts and capacitybuilding are probably also needed to addressweak technical performance.

With new investment, policy changes arerequired to ensure that service quality isimproved and maintained at a reasonablelevel—and that assets are effectively operatedand systematically maintained. Preventivemaintenance translates into lower operatingcosts, reduced adverse external impacts, and



TABLE 6.3 Technical quality of infrastructure services

Electricity Water Telecoms Transport Housing

Percentage ofurban households Permanent

Transmission with water access Reported structures:and distribution that get water phone faults Percentage percentage oflosses as percent from piped or well per 100 of total structures built

Country of total output water source main lines roads paved to last 20 yearsincome category (1999) (1997–2001) (1997–2000) (1997–2000) (1994–2000)

Low 24.1 (33) 89.4 (48) 77 (48) 28.6 (59) 76Lower middle 16.2 (31) 84.5 (18) 42.9 (38) 46.9 (46) 94Upper middle 13.6 (23) — 25.3 (27) 55.1 (33) 97Best practice in OECD 8–12 100 <5 >80 100

— Not available.Note: Figures in parentheses indicate the number of countries for which data are available. No information was available on sanitation as measured by volume and quality of treated wastewater.Source: USAID, Demographic and Health Surveys; World Bank, World Development Indicators; International Telecommunications Union. Housingestimates are from Angel 2000.

extended asset life—savings that are generallyunderestimated when maintenance budgetsare cut to meet specific fiscal targets.

Available data on the perceived quality ofinfrastructure services reflect commercial andindustrial perspectives (table 6.4). There is nosystematic effort to collect comparable datafrom residential users—a clear informationgap.

The available evidence suggests three con-clusions:

� Perceived quality is higher in all sectors incountries with higher income levels, exceptfor roads.

� Even for upper-middle-income countries,perceived quality is lower than best prac-tice.

� In all country groups, the water sector isperceived by commercial users to be themost effective. The favorable perception ofthe water sector illustrates the bias intro-duced by the use of indicators collectedfrom the business community. Water is nota major issue for many businesses, many ofwhich have their own water supplies anddo not depend on urban utilities for theirsupply.

Progress on Policy Reform

Reliable data on infrastructure policies are noeasier to come by than data on infrastructure

outcomes. A review of available indicators,however, shows that the policy and regula-tory environment for infrastructure has beenimproving—although much remains to bedone. The picture varies appreciably acrossregions, countries, and sectors.

A 1998 scorecard on the electricity sectorin 115 countries consisted of questions oncommercialization, restructuring, regulation,legal reform, and private investment in thesector.17 At the time, less than half the stepsidentified as necessary for fully effectivereform had been taken in the energy sectors ofthe countries studied. Reforms have pro-gressed significantly since then, as the forth-coming update of the scorecard is expected toshow. Updated results on the scorecard indi-cators are available for a selected number oflow-income countries (table 6.5). Expertrespondents in each sampled country wereasked whether a variety of reforms had beencompleted and to grade reforms in progresson a scale of 1 to 5 (from low to highprogress). All but one of the countries in thesample have a new electricity law; all but threehave or will soon have an independent regu-lator. Progress on corporatization and restruc-turing is less uniform. Even more disparate isthe presence of the private sector, except per-haps in the form of independent power pro-ducers (IPPs) in the generation subsector.

A survey of urban water issues showedthat 65 percent of countries had achieved a



TABLE 6.4 Quality of infrastructure services as perceived by commercial users(ratings scaled 1 to 7, with 7 indicating highest quality)

Quality ofwater Quality of

Quality of pollution telephone delivered services regulation infrastructure Quality of delivered services

Country Electricity Water Sanitation Telecoms Roads Railroads Ports Airportsincome category (2001–02) (2001–02) (2001–02) (2001–02) (2000) (2001–02) (2001–02) (2001–02)

Low 2.6 (9) 4.0 (27) 2.3 (9) 3.4 (9) 3.4 (27) 2.7 (9) 2.6 (9) 3.6 (9)Lower middle 4.2 (25) 4.8 (24) 3.1 (25) 4.9 (25) 4.2 (24) 2.6 (25) 3.5 (25) 4.2 (25)Upper middle 5.1 (20) 5.0 (18) 4.0 (20) 5.6 (20) 4.1 (18) 2.9 (26) 3.8 (20) 4.5 (20)

Note: Figures in parentheses indicate the number of countries for which data are available.Source: World Economic Forum 2003; World Bank 2000.

degree of separation between operator andgovernment, whereas only 17 percent had afunctioning regulatory body in place.18 Anincreasing number of countries were usingmore appropriate policies, governance, andmanagement arrangements. But the challengeremains formidable, especially in low-incomecountries. Out of a maximum score of 7, low-income countries averaged 1.8, compared to3.1 for lower-middle-income countries and3.9 for upper-middle-income countries. Theseresults show that although progress has beenmade in the past decade, most countries havea long way to go in establishing an environ-ment that will stimulate increased investment.

According to the International Telecommu-nications Union, by the end of 2002 more thanhalf the countries of the world had fully orpartly privatized their telecommunicationoperators. A further quarter (24 percent) ofcountries—although retaining state-owned

incumbents—had introduced private sectorparticipation through licensing of new fixed,international, or mobile operators. Today,fewer than one-fifth of countries have no formof private participation in their telecommuni-cations sector. Most are low-income countries.

The introduction of competition has beenuneven, with a majority of countries retainingmonopolies in fixed-line services such as localand long-distance telephony. An overwhelm-ing majority of countries, however, nowallow competition in the mobile and Internetaccess markets. International long-distancemarket competition also grew dramaticallythroughout the 1990s. By mid-2003, 73countries were allowing users a choice ofmore than one facility-based operator forinternational telephone calls—up from just8 countries in 1992.

There has been progress in regulation aswell. The number of telecommunications



TABLE 6.5 Electricity reform in selected countries, 2003

Commercialized/ Independent Private PrivateCountry corporatized Law regulator IPPs Restructureda generation distribution

Albania Y Y Y N IP-1 N NBangladesh Y Y IP-4 Y IP-2 N NBenin N Y N N Y N NBolivia Y Y Y Y Y Y IP-4Burkina Faso N Y N N N N NEthiopia IP-3 Y Y N N N NHonduras N Y Y Y N N NIndiab N Y Y Y Y N NIndonesia Yc Y IP-4 Y Y Y NKyrgyz Republic Y Y Y Y Y Nd NMadagascar IP-4e Y IP-5 Y N N NMali N Y Y IP-1 IP-4 Y YMauritania N Y Y N N N NMozambique Y Y IP-2 Y IP-2 N IP-1Pakistan Y Y Y Y Y IP-3 IP-3Tanzania IP-4 IP-3 IP-4 Y IP-3 N IP-4Uganda Y Y Y Y Y Y IP-4Vietnam Y N N Y IP-3 N N

Note: Y and N stand for yes and no, respectively, to the question about whether significant reform has been completed. Reforms in progress areindicated by IP, where IP-1 indicates an early stage of implementation and IP-5 indicates that implementation is nearing completion. a. A state-owned utility is deemed to have been restructured if it has successfully completed accounting separation.b. Some of India’s states have done much more than others. c. The state utility, PLN, is a limited liability company. d. The government does not intend to privatize existing strategic generation assets but is seeking private investment in small hydro schemes.e. Moving toward a management contract.Source: World Bank, Infrastructure Vice Presidency.

regulatory agencies has increased dramati-cally over the last decade. In 1990, only 12countries had regulatory agencies that func-tioned separately from telecom operators.The number had increased to 123 by mid-2003. A further 28 countries have indicatedtheir intention to establish a separate regula-tor in the coming years. In Latin Americaand Sub-Saharan Africa, more than three-quarters of countries have regulatory agen-cies. South Asia and Europe and Central Asiafollow closely with more than 60 percent.The East Asia and Pacific Region remainsbehind, with only 18 percent of countrieshaving a separate regulator.19

The Agenda AheadAddressing the large gaps in access and qual-ity identified in the preceding section willrequire action across a broad front. Substan-tial new investment will be needed. But thatis only part of the challenge. Policy and gov-ernance regimes must continue to improve,building on past gains, to ensure that newinvestments translate into improved infra-structure services for underserved segmentsof the population and economy.

Gauging the Need for New Investment

Estimating the amount of investment neededto fill gaps in access and quality is difficult.The estimate should reflect the needs of boththe household sector and businesses—in agri-

culture, commerce, and industry. This meansthat requirements need to be related to thebroad growth prospects of the economy. Oneset of estimates developed on this basis showsthat the investment requirements are large.For the first decade of the new millennium,preliminary estimates of new investmentneeds vary from a high of 3.2 percent of GDPfor low-income countries to a low of 0.4 per-cent of GDP for upper-middle-income coun-tries, with an average of 2.7 percent for alldeveloping countries.20

Operation and maintenance requirementsfollow a similar pattern. Total resourcesneeded, for both investment and mainte-nance, are estimated at roughly 6.9 percent ofGDP in low-income countries and 5.1 percentin lower-middle-income countries, with anaverage of about 5.5 percent of GDP for alldeveloping countries (table 6.6). In terms ofsectoral allocation, three sectors (electricity,telecommunications, and roads) account forfour-fifths of the total new investment.

The needed investments and maintenanceexpenditures amount to roughly $465 billiona year for the years between 2005 and 2010.Most are related to telecommunications(about $185 billion), followed by power gen-eration (about $140 billion), roads (about$90 billion), and water and sanitation (about$30 billion).

The numbers presented in table 6.6 under-estimate the actual investment requirementsbecause they do not take into account theneeds in transmission and distribution.



TABLE 6.6 Expected annual needs for new investment and maintenance in infrastructure, 2005–10

New investment Maintenance Total

Billions Percent Billions Percent Billions Percent Country income category of dollars of GDP of dollars of GDP of dollars of GDP

Low 50.0 3.18 58.6 3.73 108.6 6.92Lower middle 183.2 2.64 173.0 2.50 356.2 5.14Upper middle 136.0 0.42 248.0 0.76 384.0 1.18All developing countries 233.1 2.74 231.7 2.73 464.8 5.47

Note: These estimates exclude the needs in electricity transmission and distribution as well as in ports and airports.Source: World Bank staff estimates.

Including these two business lines couldroughly double the estimates of the needs forenergy and add, for low-income countries, upto 3 percent of GDP to the estimates of theinvestment needs in the table. Another exclu-sion from these estimates is investment inports and airports. However, as these types ofinfrastructure represent a relatively small pro-portion of the total, it is unlikely that includ-ing them would change the estimates greatly,especially for poor countries.21

While there are no formal internationalstatistics on past investment levels in infra-structure, investment estimates can be gener-ated from past changes in stocks. Five-yearaverages of past investment levels can be veryroughly approximated by pricing changes instocks (at a constant price). Doing so yields amonetary value that can then be normalizedto GDP.22 Such a rough approximation basedon the evolution of stocks suggests thatinvestments in the past 5 to 10 years wereabout 4 percent of GDP in low-income coun-tries, 2.9 percent in lower-middle-incomecountries, and 2.6 percent in upper-middle-income countries—a decline of 2 to 4 per-centage points from previous decades.23

Comparing the past investment and O&Mrates to the projected requirements, andallowing for efficiency improvements, the gapcould be 3.5 to 5 percent of GDP for low-income countries and 2.5 to 4 percent forlower-middle-income countries (figure 6.1).The gap is likely to be less than 0.5 percent ofGDP for upper-middle-income countries—most of it ascribable to O&M and majorrehabilitation. This suggests that sizable in-creases in investment from recent levels areneeded. In water and sanitation, for example,investment would need to double from his-torical levels.

Traditionally, most investment in infra-structure has been publicly funded. Accordingto a recent estimate, 70 percent of all infra-structure spending in developing countries inthe 1990s was financed by governments orpublic utilities’ own resources. The privatesector contributed 20 to 25 percent, whileofficial development assistance financed 5 to

10 percent.24 These proportions of coursevary across sectors, with telecommunications,for example, attracting more private invest-ment and relying less on official assistancethan water and sanitation.

Private sector commitments to infrastruc-ture in developing countries during the 1990samounted to some $805 billion (table 6.7).The bulk of investments went into energy andtelecommunications in Latin America, EastAsia, and, to a lesser extent, Eastern Europe.The commitments increased sharply duringthe 1990s before declining rapidly after 1997,affected by the series of major crises startingin East Asia and ending in Argentina in 2001.Indeed, as uncertainty grew in emerging mar-kets, the cost of capital shot past the pointwhere most new projects could generate ade-quate rates of return. In 2002, private invest-ments totaled $46.7 billion, the lowest levelof investment since 1994.

During the 1990s the public sector in manydeveloping economies reduced its participation



Perc

ent o

f GD

P

3

5.5–7

Low-incomecountries

Lower-middle-income countries

Upper-middle-income countries

7.5–9

Needed

Actuals4

2.9 2.6

FIGURE 6.1 Gaps in infrastructure call for significantly increased spending, which must be managed well for effectivenessAnnual spending on infrastructure services, actual vs. needed

Note: Actual = levels in the 1990s. Needed = estimated requirements for 2005–10. Annualspending includes investment plus spending on operations and maintenance.Source: World Bank staff estimates (work in progress).

in infrastructure. This unexpected developmentwas the result of anticipated increases in privatesector participation and fiscal adjustment pro-grams. There is growing evidence that fiscalretrenchment usually leads to a disproportion-ate reduction in public capital expenditures,particularly in infrastructure. Easterly andServén found that during the 1990s, reductionsin public infrastructure investment in LatinAmerica made up a sizable share of fiscaladjustment in the region—ranging from 31.5percent in Mexico to 174.3 percent in Brazil.25

But the problem is more generalized. In India,total investment in infrastructure was 5.4 per-cent of GDP in 1990, including 4 percent ofpublic sector money. By 1998, total investmentin infrastructure had dropped to 4.6 percent,with a decline in public investment of 1 per-centage point of GDP. Private investmentincreased, but not enough to offset the drop inpublic investment.

That most urban infrastructure (secondaryroads, drainage, sanitation) is the responsi-bility of local governments has contributed tothe decline in public investment. Local gov-ernments are increasingly dependent on localtaxation since, with fiscal decentralizationand consolidation, many countries havereduced central government fiscal transfers.

Evidence suggests that changes in ODAalso may have contributed to the decline. Forexample, the level of commitment to infra-structure of the multilateral development

banks has declined since 1995, reaching a lowof $13.5 billion in 1999 before rebounding to$16 billion in 2002. Even at their peak, com-mitments were small in relation to needs.

The Conditions for Increased Investment

If developing countries are to reach investmentlevels consistent with their needs in a contextof fiscal constraints and limited private invest-ment, external assistance to infrastructure willneed to increase. This is particularly true forlow-income countries, but foreign assistanceplays an important catalytic role in middle-income countries as well. The impact of aidcan be enhanced through reallocations thatreflect the relative effectiveness of the variousforms and uses of support. Moreover, withinnovation and coordination efforts across thedevelopment community, ODA could be usedto better leverage private and local publicresources. Better coordination between bilat-eral and multilateral donors would help max-imize leverage of funding from all sources,including the private sector.

The use of resources, both foreign anddomestic, needs to be guided by clear priori-ties at the country level. Access and qualitydeficiencies are typically much more seriousin rural areas, implying the need for a specialfocus on the needs of rural and underservedareas. At the same time, the rapid increase inurbanization will require attention to the



TABLE 6.7 Private commitments for infrastructure, 1990–2002(US$ billions)

Region Telecoms Electricity Natural gas Airports Railways Seaports Toll roads W & S Total

East Asia and Pacific 56.2 68.3 6.8 2.8 10.3 10.3 26.8 17.0 198.4Europe and Central Asia 68.1 21.1 11.3 1.5 0.3 0.7 2.6 3.5 109.0Latin America and Caribbean 182.9 100.4 19.5 7.5 18.3 6.9 40.6 21.3 397.4Middle East and North Africa 10.6 8.4 3.9 0.9 0.2 1.2 — 1.3 26.5South Asia 19.7 22.6 0.2 0.2 — 2.1 0.8 0.2 45.8Sub-Saharan Africa 18.5 5.0 1.3 0.4 0.3 0.1 2.0 0.2 27.8Total 355.9 225.7 43.0 13.2 29.4 21.2 72.8 43.6 804.9

— Not available.Source: World Bank, Private Participation in Infrastructure Database.

infrastructure needs of the migrating pop-ulation. Better planning of investment ininfrastructure networks (secondary roads,sanitation, drainage) and protection of greenspaces can help ensure economic, social, andenvironmental sustainability in the context ofurban expansion.

Cutting across the various infrastructuresectors is the need to emphasize quality andto make adequate provisions for O&M in theallocation of resources. It is not uncommonto find underinvestment in O&M because theincentive structure is such that it is easier toraise resources to finance new investment ormajor rehabilitation than to design tariffs torecover O&M expenses. Such a bias in incen-tives must be corrected, as the impact andproductivity of investment depend cruciallyon its quality and upkeep.

Affordability is a key determinant of theeffective access of the poor to infrastructureservices. It is important to consider the extentto which the pricing of services and the qual-ity options offered to consumers are consis-tent with ability to pay. Without affordability,expanded access is of only limited use to thepoorest. As illustrated in the recent debate onpublic utilities in Argentina, the many indi-rect taxes levied on infrastructure services,many of which are regressive, must be docu-mented. While such taxes may reflect fiscalconcerns, governments need to recognize thatthey may thwart the desired improvement inthe poor’s access to public services.

Institutional capacity building is a keycross-cutting element of the reform agenda.The problems of quality deficiencies, thedeclining interest of potential privateinvestors, misallocations of resources, andexcessive costs observed in many countriesoften can be traced to insufficient institu-tional capacity. Capacity constraints can beespecially acute at the local level. With urbangrowth and fiscal decentralization, public sec-tor responsibilities for infrastructure plan-ning, financing, and management areincreasingly falling to local governments,many of which lack the requisite capacity.Effective delivery of infrastructure services at

the local level calls for adequate intergovern-mental arrangements for sharing the invest-ment costs of local public-good infrastructure,enhanced resource mobilization by local gov-ernments, and stronger efforts at local capac-ity building.

Increased private sector participation ininfrastructure requires an adequate regula-tory framework, including competent regula-tory agencies. Such agencies now exist inmany countries but often lack adequatecapacity. Coordinated efforts by governmentsand donors to provide the necessary technicalassistance for capacity building would have ahigh payoff. Governments and donors canalso bring pressure to bear on the private sec-tor to contribute to the effort to improve gov-ernance—for example, by adopting a code ofethics to commit to due process in regulatoryinteractions. Increasing the transparency ofdecisionmaking processes can minimize therisk of corruption.

To better assess and monitor infrastructureneeds and policies along these and relateddimensions, the availability of informationwill need to improve markedly. The monitor-ing of service quality, affordability, fiscal cost,and governance quality is handicapped bymajor information gaps, and data are lackingeven on basic access to some key infrastruc-ture, especially transport. The internationalcommunity should commit to supporting asystematic effort to develop a better informa-tion base, including statistical capacity build-ing in developing countries.

Countries, supported by their developmentpartners, are responding to these challenges.Declining public investment in infrastructureand overoptimistic expectations of privatesector participation, combined with growingappreciation of infrastructure as a foundationfor achievement of the MDGs, have forced theinternational development community tobegin reassessing its priorities. At the WorldBank, this reassessment has resulted in theInfrastructure Action Plan (IAP).26 The IAP isdesigned to respond to client-country demandfor infrastructure with a broad menu ofoptions for public and private infrastructure




The MDG for water supply and sanitation aims to halve by 2015 the proportion of people who lackaccess to safe drinking water and adequate sanitation. Currently, one in five people in the develop-ing world lacks access to clean water; one in two has inadequate sanitation. For four out of five—4 billion people—wastewater disposal is inadequate. Every day, 25,000 people die from diseasesassociated with poor water supply, sanitation, and hygiene. Morbidity is even higher.a

Improving water supply and sanitation is known to reduce child mortality and protect health,while promoting gender equality, educational opportunity, and environmental sustainability. Yet atcurrent rates of service expansion, only about 20 percent of countries are on track to meet the goal,and fewer than 10 percent of the low-income countries are on track. The actual number may wellbe lower, due to huge gaps in data on access.

To meet the goal, about 1.5 billion people (1 billion in urban areas, and 0.5 billion in rural areas)will have to be provided with sustainable access to safe water, and about 2 billion to basic sanita-tion (1.1 billion in urban areas and 0.9 billion in rural areas). Various analyses suggest that invest-ment will need to double, from $15 billion to $30 billion annually.b Existing assets must be properlyoperated and maintained, and new investments will have to be efficient. Additional investment willbe needed for wastewater treatment, rehabilitation of existing infrastructure, and infrastructure forwater storage and conveyance to deal with urbanization and growing climatic variability.

The bulk of the financing needed will have to come from the public sector and official external assis-tance, even if efforts succeed in mobilizing increased private participation in the sector. In the pastdecade, 43 percent of the funding for water supply and sanitation came from government budgets andabout 28 percent from external assistance. Another 17 percent was funded by small-scale providers,communities, and households. Internal cash generation from utilities provided only 7 percent, in partreflecting inappropriate pricing policies. The international private sector contributed 5 percent.c

Focusing solely on investment will not generate sustainable service improvements, however, asthe experience of the International Drinking Water Decade has shown. Success in achieving the mil-lennium goal requires a combination of sound policies, improved governance, capable institutions,increased investment and financing, and more effective financing modalities.

A rapid water sector survey carried out in 2002 showed that while 65 percent of the countrieshad achieved a degree of separation between operator and government, only 17 percent had a func-tioning regulatory body in place. And 13 percent of the countries still had to embark on any formof urban sector reform.d A recent survey of 246 water utilities in 51 countries showed that overallquality was unacceptable, although with significant exceptions.e That 25 percent of the operatorswere performing well shows that efficient service delivery is possible where sound policies and gov-ernance are in place.

BOX 6.1 Water supply and sanitation in the MDGs

Access to water and sanitation by region, 2000

Percentage of population with access

To improved water supply To improved sanitation

Region Urban Rural Total Urban Rural Total

East Asia and Pacific 93 67 76 73 35 48Europe and Central Asia 95 82 91 97 81 91Latin America and Caribbean 94 66 86 86 52 77Middle East and North Africa 95 77 87 93 70 83South Asia 94 80 85 67 22 34Sub-Saharan Africa 83 44 57 73 43 53Developing countries 95 71 82 77 35 52Least developed countries 82 55 62 71 35 44

Source: JMP 2000.


An increasing number of countries are nowusing more appropriate policies and institutionalarrangements, but the challenge ahead is still for-midable, especially in low-income countries.Effectively managing existing assets and sizablenew investments will require stronger efforts atpolicy reform and institutional upgrading, sup-ported by increased international assistance. Abroad consensus on the main elements of the pol-icy agenda at the country level, and of interna-tional support, is beginning to emerge. These are:

� Providing incentives to invest and operate effi-ciently and promoting sustainable access for thepoor. Existing capacity must be optimally usedto meet demand through tariffs that recovercosts while taking into account the ability ofconsumers to pay; and policies that emphasizemanagerial autonomy and accountability.

� Strengthening the capacity of local institutionsto scale up activities. Efforts need to focusmore on building the financial and adminis-trative capacities of subnational institutionsresponsible for service delivery. This should beaccompanied by institutional reforms thatenable stakeholders, especially poor people, to monitor and discipline service providers.

� Creating and disseminating knowledge needed to set priorities and improve resource use. Policy-makers and stakeholders need reliable information to make good decisions, to use resources effi-ciently, and to hold providers accountable for the quality of service they deliver.

� Mobilizing financing for investments to expand access and improve service quality. Appropriatecost-recovery policies and practices are a prerequisite for raising the large amounts of investmentfinancing needed in the sector. Increased domestic resource mobilization should be supported byincreased external assistance.

� Enhancing the range and reliability of financial instruments. Most IFIs provide contractual andregulatory risk coverage, credit risk coverage, and foreign exchange risk coverage. Yet demand forrisk mitigation instruments is low because few international investors are willing to invest in watersupply and sanitation in developing countries. Only a few IFIs provide subsovereign lending on asignificant scale. The IFIs are now stepping up the deployment of risk mitigation instruments andconsidering how to increase financial and technical support to subsovereigns, as proposed by theCamdessus Panel in 2003.f

� Aligning development assistance programs with country priorities and capacities, pooling aid atthe country level, and ensuring continuity of support. Toward these ends, several consultativemeetings have been held in the past year to increase donor coordination—among them an infor-mal consultative meeting between bilateral donors and the World Bank in May 2003 and a Feb-ruary 2004 consultative meeting of IFIs and bilateral donors to coordinate support for millenniumgoals in water supply and sanitation.

In the Water Action Plan adopted at the Evian Summit in June 2003, the G-8 governments requestedthe World Bank and other IFIs to study and recommend measures to facilitate subsovereign lending andimprove risk mitigation. A comparative review of IFIs showed that only EBRD and the Islamic Devel-opment Bank presently have a sizable subsovereign lending program; others reported legal or policy con-straints on subsovereign lending. Four more IFIs (AfDB, EIB, IDB, and IFC) are planning to increase thistype of financial support; IFC and IBRD are piloting a municipal fund to enable such lending.

Notes: (a) JMP 2000. (b) Global Water Partnership 2000; Winpenny 2003; United Nations Millennium Project Task Force on Water and Sanitation 2003. (c) Development Committee 2003. (d) World Bank 2002. (e) Kingdom and Tynan 2002. (f) Camdessus 2003.

Source: Global Water Partnership 2000 and World Bank.

US$

bill

ions

per

yea

r

35

30

25

20

15

10

5

0

Water Sanitation

Present MDG needs

Investment in water and sanitation will need to doubleHistorical and required MDG investment levels

provision and a scaling up of the projectpipeline, with a better integration of infra-structure into country assistance strategiesand poverty reduction strategy papers. TheIAP also aims at rebuilding the knowledgebase for infrastructure by strengthening ana-lytic work to support policy reforms and insti-tuting more systematic collection of data.

The scale of the challenge ahead is huge.The response needs to span all the elements ofthe agenda outlined above. And it needsincreased international support—from bilat-eral donors and multilateral agencies—andmore effective coordination of that support.Efforts under way to scale up investment, pol-icy reform, and support in water and sanita-tion provide a good example of the wayforward (box 6.1). But these efforts are still atan early stage, and success will require sus-tained commitment to the goals.

Notes1. The 1994 World Development Report on

infrastructure (World Bank 1994) highlighted anemerging debate on the relationship betweenaccess to infrastructure, productivity, and growth.

2. Estache, Foster, and Wodon 2002.3. Caldéron, Easterly, and Sérven 2003.4. Ramirez and Esfahani 2000.5. Datt and Ravallion 1998.6. Deininger and Okidi 2003.7. Fan, Zhang, and Zhang 2002.8. Estache, Foster, and Wodon 2002.9. Estache, Foster, and Wodon 2002.10. Households in Guatemala spend around 10

percent of their household budget on water, energy,and telecommunications services (Foster and Tre2000). More than half is spent on energy for cookingand heating, and more than 25 percent on energy forlighting and powering appliances. Barely 0.5 percentof income is spent on water services. The overall bud-get share is relatively constant across consumptionquintiles, although the composition of the budgetshifts away from cooking fuels and toward telecom-munications for richer households. Although only atiny fraction of the poorest households have accessto telephones, those that do so spend as much as 5percent of their income on the service. Such break-downs, however useful for designing tariff structuresto ensure affordability, are seldom available.

11. Jalan and Ravallion 2001; Jayasuriya andWodon 2003; Leipziger, Fay, and Yepes 2003;Development Committee 2003.

12. Fay and Yepes 2003.13. The project is part of the Bank’s Infrastruc-

ture Action Plan, which is described later in thischapter.

14. Angel 2000.15. APHRC 2002.16. Roja 2003.17. Bacon 1999.18. World Bank 2002.19. Qiang 2004.20. These estimates are calculated with an

econometric model estimating the elasticity ofdemand with respect to GDP growth. For details,see Fay and Yepes (2003). This type of approach ismuch better suited to producing aggregate results,which usually provide reasonable ballpark esti-mates, than it is to producing individual countrypredictions. Also, it cannot adequately capture theimpacts of improvements in efficiency, differentia-tion in service levels, and cross-sectoral linkagesbetween infrastructure sectors.

21. Extending the forecast to 2015—the MDGhorizon—requires much more analysis of the sen-sitivity of various determinants of investmentdemand and would be much less robust. Simula-tions for 2010–15 imply that for low-income coun-tries, annual investment needs would be roughlythe same as for the previous period with a marginof +/–0.5 percent of GDP. They could be as muchas 1 percent lower for middle-income countries.

22. This is clearly a rough approximation, sinceit ignores the effects of technological changes onprices. This effect is particularly important, forexample, for telecommunications, but much lessso for water and roads.

23. Note that a major difference between theforecast and the past is that past investment levelsinclude major rehabilitation requirements resultingfrom weak allocation of resources to O&M. In asense, it could be argued that ex post O&M is “cap-italized” and that the best approximation for it isthe rehabilitation component of past investments.On the other hand, the forecast bets on a commit-ment to adequate O&M. This is why it is more use-ful to rely on the sum of investment and O&M toget a fuller sense of the resource needs of the sector.

24. DFID 2002.25. Easterly and Servén 2003.26. http://www.worldbank.org/infrastructure/

files/iaPPublic.pdf.



The Millennium Development Goalsrelated to human development—achievement of universal primary edu-

cation; elimination of gender disparities ineducation; reduction of child and maternalmortality; control of major diseases such asHIV/AIDS, tuberculosis, and malaria—require a multidisciplinary and cross-sectoralapproach if they are to be achieved. This isparticularly so because a key factor in the suc-cess or failure of human development pro-grams is improvement of governance—thepolicy and institutional framework govern-ing public resource management, institu-tional capacities, control of corruption,community involvement, and empowermentof stakeholders. While substantial addi-tional resources will be needed to financeexpanded and improved access to education,health, and related services (requiring bothstronger domestic resource mobilization andincreased foreign support), the effectivenessof those investments in providing qualityservices, especially to the poor, will hinge onimprovements in the underlying policy andgovernance regime.

Achieving the human development goalswill require efforts on multiple fronts todeliver more and better services. Those effortsmust address both supply- and demand-sideconstraints within each program. It is notenough to build more schools and clinics, for

example, if the quality of service is so poor asto discourage their use, or if the poor do nothave the resources to afford the service orlack the knowledge of its availability andvalue, or if the services are not properlyfunded and providers require “unofficial” pay-ments. Moreover, those efforts must addressthe complex connection between progress inother development areas and human develop-ment programs.

Outcomes in education and health may bedetermined as much by what happens in othersectors as within these sectors. Do people haveaccess to safe drinking water and basic sanita-tion? Do rural roads provide access to schoolsand clinics? Is there electricity to power thoseschools and clinics? Thus policies and pro-grams need to be framed within an integratedcountry strategic framework that internalizessuch linkages.

If current worldwide trends (reviewed inchapter 2) continue, the human developmentMDGs are unlikely to be achieved by the tar-get dates. Progress is being made on the pol-icy agenda. Investment in human capital indeveloping countries is on the rise, and theunderlying policy framework bearing on itseffectiveness is improving (see chapter 3 forratings by multilateral institutions of socialsector policies). Some countries are providingencouraging examples of successful innova-tion in service delivery to the poor. Overall,


Accelerating Human Development7

however, progress has been slow and uneven,and many gaps remain, particularly withrespect to the institutional aspects of theagenda. The deficiencies are most serious inSub-Saharan Africa and South Asia, but evenin these regions individual countries (CapeVerde in education, health, and gender;Ghana in child mortality; and Ethiopia inprimary enrollment) are making progress andare likely to meet at least some of the ambi-tious MDG targets.

The unreliability of information and widegaps in data complicate the task of assessingprogress toward the human developmentMDGs and of making policy to achieve them.Social sector data in many countries are par-ticularly weak. Administrative data sufferfrom gaps and low quality, while survey dataare available only periodically and captureonly some characteristics. The need to moni-tor improvements and determine prioritiesfor action increases the urgency of strength-ening the information base.

To accelerate progress toward the goals,countries and their partners in the interna-tional community will need to scale up theirefforts substantially. Priorities include:

� Expanding investment in human capital inlow-income countries, while maximizingthe impact of existing public spending byimproving the targeting of public servicesin education, health, and social assistance

� Focusing on governance issues in servicedelivery to determine priority interven-tions to address service quality problems

� Piloting and evaluating empowermentoptions to strengthen the role of clients,especially poor people, in the design,scope, and delivery of key services.

Much recent and ongoing work analyzesaspects of the challenges posed by the humandevelopment MDGs:

� Working groups on education, gender,health, and major diseases of the UnitedNations’ Millennium Project are reviewing

recent findings and recommending direc-tions for the international community.1

� The United Nations Development Pro-gramme (UNDP) is monitoring progresson the millennium goals; its latest HumanDevelopment Report assesses the prospectsfor attaining them.2

� The World Bank’s 2004 World Develop-ment Report focuses on service delivery tothe poor, paying considerable attention toeducation and health services.3

Drawing on those sources and others, thischapter reviews trends in the financing ofhuman development programs, assesses theoverall effectiveness of those programs, andfinally, sets out an agenda for acceleratingprogress toward the human developmentMDGs.

Trends in Spending for Human DevelopmentAvailable data show that spending on educa-tion, health, and other human developmentactivities in developing countries has risenover the past decade, although it remains lowrelative to needs in many countries, especiallythe low-income ones.4 The rising trend signi-fies a growing commitment on the part ofgovernments to human development goals.Analysis shows that further increases inspending on these sectors, along with spend-ing on related priority infrastructure such aswater and sanitation, will be necessary toachieve the MDGs. To that end, public spend-ing will need to shift further toward those sec-tors; and the shift must be accompanied bymeasures to enhance efficiency and effective-ness. Higher spending alone will not deliverthe desired results.

Rising Public Spending on Education and Health

Public spending on education in low-incomecountries rose from 4.3 percent of GDP in1990 to 4.7 percent in 2000 (figure 7.1).



After dipping in the mid-1990s (to an aver-age of 3.7 percent in 1995) it recovered laterin the decade. Spending rose in all regions,though in some it is well below the low-income country average. In 2000 publicspending on education relative to GDP was3.8 percent in Europe and Central Asia and4 percent in Sub-Saharan Africa and SouthAsia, implying the need for special effortsthere. Spending also rose in middle-incomecountries, to reach an average of 5 percent ofGDP in 2000. The biggest leap occurred inLatin America, where historical underspend-ing in education has contributed to relativelyhigh economic inequality and lagging pro-ductivity. Interestingly, middle-income Sub-Saharan African countries spent more oneducation, as a share of GDP, than any otherregion in both 1990 and 2000. In compari-son to the numbers for the developing world,OECD countries spend an average of 5.0 to5.5 percent of GDP on education.

Public spending on health also rose relativeto GDP, but modestly, and mainly because ofthe increase in middle-income countries (seefigure 7.1). Public spending on health in devel-oping countries as a share of GDP remains at

about half the average level in high-incomecountries. In some low-income countries,spending on health is particularly low. In 24of the countries for which data are available,including 14 in Sub-Saharan Africa, publicspending on health is less than $5 per capita.

The rise in education and health spendingrelative to GDP in developing countries hasbeen supported by the allocation of a highershare of total government spending to theseservices (figure 7.2). In 2000, on average,developing countries allocated 15.6 percentof total public spending to education and 8percent to health, up from 13 percent and 6.2percent, respectively, in 1990.

Significant progress has been made in mobi-lizing international and domestic resources tocombat HIV/AIDS, tuberculosis, malaria, andother major diseases, although financing gapsremain. Spending on HIV/AIDS in low- andmiddle-income countries in 2003 was approx-imately $4.7 billion, a 20 percent increase over2002 and almost five times the level of spend-ing in 1996. International spending in 2003included a projected $1.6 billion in bilateralassistance. Bilateral assistance is enhanced bythe U.S. Emergency Plan for AIDS Relief, an

A C C E L E R A T I N G H U M A N D E V E L O P M E N T


Middle-income countriesLow-income countries

20001995

Education Health

1990

4.3 4.44.7

5.0

3.7

4.6

2.9

2.8

2.7

2.6

2.5200019951990

6

5

4

3

2

1

0

Perc

ent o

f GD

P

2.6

2.7

2.6 2.6 2.6

2.8

FIGURE 7.1 Investment in human capital is up, but more is neededPublic spending on education and health, 1990–2000

Source: Compiled by IMF staff on the basis of country authorities’ estimates.

initiative that promises $10 billion over thenext decade, with $2.4 billion approved forFY04. The sources of most multilateral fund-ing, estimated at $600 million in 2003, arethe Global Fund to Fight AIDS, Tuberculo-sis, and Malaria, a multidonor initiativelaunched to expedite the channeling of criti-cal assistance to needy countries, and theWorld Bank. The Joint United Nations Pro-gramme on HIV/AIDS (UNAIDS) is provid-ing leadership and financing for technicalassistance and international coordination.UNDP and others are also providing fundingfor specific activities. Currently, more thanone-half of the total international assistanceflows to Sub-Saharan Africa.

Domestic spending on HIV/AIDS also isincreasing, to an estimated $1 billion for 58low- and middle-income countries—doublethe amount documented in 1999. This esti-mate is rough, however, as many countrieslack adequate systems for tracking domesticspending on HIV/AIDS.

Private Supplements to Public Spending

Although greater public resources for educa-tion and health are helping to fill critical gapsin some countries, in most low-income coun-tries, public revenues remain inadequate toprovide the full range of needed services.Moreover, efficiency of resource use isuneven. One reflection of these shortcomingsis the sizable reliance on private financing forbasic social services, even in poor countries.In education, a study found that fees werecommon at all levels of schooling in 97 per-cent of countries, with unofficial (often ille-gal) fees charged in almost 40 percent ofcountries.5 Some types of contributions exist,even in widely publicized no-fee countriessuch as Malawi and Uganda.

With public resources inadequate to pro-vide access to quality education for all citizens,private spending plays a useful supplementaryrole. Community contributions can help sup-plement low teachers’ salaries and support the



Europe &Central

Asia

LatinAmerica

&Caribbean


SouthAsia

EastAsia &Pacific

Sub-SaharanAfrica

Europe &Central

Asia

LatinAmerica

&Caribbean


SouthAsia

EastAsia &Pacific

Sub-SaharanAfrica

25

20

15

10

5

0

25

20

15

10

5

0

Perc

ent o

f to

tal p

ublic

spe

ndin

g

20001990

Public spending on education, low- and middle-income countries

Public spending on health, low- and middle-income countries

FIGURE 7.2 Developing countries are allocating more public spending to human development

Source: Compiled by IMF staff on the basis of country authorities’ estimates.

cost of books, supplies, and basic mainte-nance.6 The key is to ensure that publicresources that do exist are sufficiently targetedto the poor and that private financing is mobi-lized from those with the ability to pay.

The same situation exists—and the sameprinciples apply—in health care, where pri-vate expenditures often exceed public spend-ing, especially at low levels of GDP per capita(figure 7.3). In India, for example, more than80 percent of health spending is private. Muchof private spending in the poorest countries isout-of-pocket payments to private providers,whereas in some middle-income countriesinsurance payments make up an importantshare of private spending.

Where services are “free” but not suffi-ciently funded by national or local govern-ment, informal side arrangements withproviders tend to emerge. In the transitioncountries of Europe and Central Asia, wherefree services are “guaranteed” and the pri-vate sector is in nascent stages, and in SouthAsia, where the private sector is vibrant,under-the-table fees may finance up to 85percent of all health spending. Since informalor illegal payments are common in manycountries, official data can seriously under-estimate the extent of reliance on privateexpenditures. In a study of seven Latin Amer-ican countries, perceptions of informal pay-ments were found to be the highest in CostaRica, the country considered to have thebroadest free system in the region.7

The poor tend to be affected more by suchinformal or illegal payments. As a propor-tion of household income, private expendi-ture on health is much higher for the poorersegment of the population in many countries(table 7.1).

Although governments around the world,and particularly in Africa, have committedthemselves to strengthening public programsto combat HIV/AIDS and help those affectedby it, much of the burden of paying forHIV/AIDS services continues to fall on indi-viduals (box 7.1). Major gaps in the coverageof services translate into inequity in access

and financial ruin for poor householdsaffected by the disease.

Despite the increased commitment ofdonor support, projected funding continuesto fall short of what UNAIDS and others esti-mate is required for HIV/AIDS prevention,



10

9

8

7

6

5

4

3

2

1

0Pe

rcen

t of

GDP

High-incomeOECD

High-income

non-OECD

Uppermiddleincome

Lowermiddleincome

Lowincome

30

25

20

15

10

5

Per

capi

ta in

com

e at

PPP

(US$

thou

sand

s)

ODA to health sector

Private health spending

Domestically financedgovernment health spending

GDP per capita

FIGURE 7.3 Public spending covers more of the cost of health care in high-income countries than in low-income countriesPublic and private spending on health, developing versusdeveloped countries

Source: World Bank 2004.

TABLE 7.1 Private health expenditure(as a percentage of annual household income)

Income level (quintiles) Russia South Africa Brazil Tanzania

Poorest 11.9 34.6 20.6 4.4Second poorest 5.3 20.3 11.6 2.7Middle 3.3 13.3 7.2 2.0Second richest 2.1 9.2 4.1 1.3Richest 0.8 3.2 1.2 0.5

Source: World Bank 2003c.

care, and support efforts, including access toantiretroviral therapies. Current spendingrepresents less than half the $10.7 billion thatUNAIDS estimates is required to finance acomprehensive health benefit package andless than a third of the $14.9 billion thoughtto be needed by 2007—a major challenge tomobilize additional resources, strengtheninstitutional implementation capacities, andscale up programs.

Social Safety Nets

Social assistance programs cushion theincome of the poor from major shocks, suchas an AIDS-related illness or a job lost in aneconomic crisis. Social assistance enablespoor families to continue financing criticalinputs, such as schooling and health clinic vis-its. The mounting HIV/AIDS orphan crisis inAfrica is increasing the need for such assis-tance programs.

Wherever children contribute to house-hold income, schooling has an opportunity

cost. Social transfers can help poor familieskeep their children in school. Creative mea-sures to compensate poor families throughconditional cash transfers linked to schoolingand health center visits (as in Brazil, Colom-bia, Mexico, and Nicaragua) have raisedattendance at schools and induced a moreregular use of preventive health care services.

Systematic information on the scale ofsocial assistance programs is not available formany countries, but it appears that spendingaverages less than 1 percent of GDP in mostregions other than Europe and Central Asiaand Latin America. There is a great deal ofvariation among countries, not alwaysrelated to national income (table 7.2). InEurope and Central Asia, the socialist legacyhas left in place large social support programsthat absorb significant public resources. Forexample, social assistance spending in Bul-garia and Serbia-Montenegro is estimated at4.5 and 6.3 percent of GDP, respectively.Total spending on social protection, includingsocial insurance, is typically much higher,



Recognizing the importance of documenting the overall flow of health funds—and funds related toHIV/AIDS—the Rwandan Ministry of Health reviewed the country’s national health accounts. Theanalysis showed that the growing HIV/AIDS burden was a significant challenge in an already under-funded health system.

The study estimated that 40 percent of total per capita spending on health was funded by house-holds. Of the total spending, about 10 percent was used for HIV/AIDS-related costs (prevention,treatment, and mitigation). HIV/AIDS expenditures were financed primarily by households (93 per-cent), followed by donors (6 percent), and government (1 percent). About 80 percent of HIV/AIDSexpenditure was for treatment of related symptoms and infections in about 400,000 patients, 14percent for antiretroviral treatment for a very small number of patients (202), and about 7 percentfor prevention. Prevention was to a large extent financed by donor funds, whereas treatment costswere the financial burden of households. Thus, access to treatment of HIV/AIDS-related diseaseswas virtually defined by the patient’s socioeconomic background and ability to pay the full cost oftreatment, posing serious equity issues.

Few countries have completed a similar analysis. This makes it difficult to identify and addressthe serious fiscal implications of the epidemic, slowing consideration of future options, and cloud-ing questions concerning the relative roles of households, government, and donors in developingHIV/AIDS policies and programs.

Source: Schneider and others 2000.

BOX 7.1 Rwanda: HIV/AIDS and health expenditures

averaging around 10 percent of GDP for thetransition countries in the region for whichdata are available. By contrast, Pakistan andMalawi come in at the low end, spendingonly 0.1 and 0.2 percent of GDP, respectively,on social assistance. Typically, in poor coun-tries, public resource constraints mean thatspending on social assistance falls well shortof the levels needed.

Quality and Effectiveness of ProgramsThe implication of the foregoing analysis isclear. Public spending on human developmentservices, particularly in low-income coun-tries, must rise if the development goals are tobe achieved; and this rise must be supportedin part by additional donor assistance. Butincreased spending will not deliver results if itis poorly directed and inefficiently imple-mented. A vital part of the agenda consists ofactions to enhance the quality and effective-ness of spending at two levels:

� Allocation of resources must be betteraligned with development priorities andbetter directed at target populations.

� The institutional framework throughwhich resources pass must be improved tostrengthen capacity for effective imple-mentation and accountability for results.

In general, the policy and institutionalregime in developing countries has beenimproving, both broadly and in the social sec-tors. This progress is confirmed by the WorldBank and other MDB ratings reviewed inchapter 3. The improvements should enhancethe effectiveness of programs. A recent studyfound that the elasticity of government spend-ing in terms of its impact on desired outcomesrose with the score on the Country Policy andInstitutional Assessments (CPIA). For exam-ple, at a CPIA score of 4 (a satisfactory rating),a 10 percent increase in public health and edu-cation spending was associated with a 7.2 per-cent decline in the maternal mortality rate.8

Are Resources Reaching Targets?

A review of the delivery of human develop-ment services across countries shows, how-ever, that there remains substantial scope formore effective use of resources. One keydimension of effectiveness is how well publicprograms are reaching and meeting the needsof the poor. Analysis of the incidence of pub-lic education and health spending in a rangeof developing countries shows that, contraryto the declared objectives of the programs,spending is often skewed toward better-offhouseholds and not the poor (figure 7.4).

In some countries, public subsidies on edu-cation and health do benefit low-incomehouseholds disproportionately—for example,in Colombia, Costa Rica, Honduras, Jamaica,and Romania. But the overall pattern suggestsconsiderable scope for improved targeting ofsubsidies. In education, for example, thiswould involve redirecting subsidies towardlower levels of education, which typically ben-efit the poor more, and improving servicedelivery to enhance the poor’s effective accessto and use of available services.

A commitment to spend does not necessar-ily translate into actual expenditures on servicedelivery. Recent evidence from several coun-tries indicates that funds are not always reach-ing the front line (box 7.2) due to bureaucratic



TABLE 7.2 Public spending on social protection for selected countriesby region (percent of GDP)

Social Social protection Region assistance overall

Europe and Central Asia (21) 2.50 10.4Latin America and Caribbean (18) 2.30 4.6East Asia and Pacific (6) 0.70 2.3Middle East and North Africa (5) 0.73 2.4Sub-Saharan Africa (1) 0.24 1.9South Asia (1) 0.90 1.9

Note: Numbers in parentheses indicate the number of countries for which data are available.Social protection includes social assistance plus social insurance (such as pensions, sicknessand disability benefits, unemployment benefits).Source: Blank, Grosh, and Weigand 2003.



FIGURE 7.4 Public spending on human development often benefits the rich more than the poorShare of public spending for health and education that accrues to the richest and poorest quintiles, selected countries

Guinea 1994

India (UP) 1995/96

Armenia 1999

Ecuador 1998

Ghana 1994

India 1995/96

Côte d'Ivoire 1995

Madagascar 1993

Tanzania 1992/93

Indonesia 1990

Vietnam 1993

Bangladesh 2000

Bulgaria 1995

Kenya 1992 (rural)

Sri Lanka 1995/96

Nicaragua 1998

South Africa 1994

Colombia 1992

Costa Rica 1992

Honduras 1995

Argentina 1991

00 1010 2020 3030 4040 5050

00 1010 2020 3030 4040 5050

Nepal 1996

Guinea 1994

Madagascar 1993/94

FYR Macedonia 1996

Kosovo 2000

Tanzania 1993/94

South Africa 1994

Côte d'Ivoire 1995

Nicaragua 1998

Lao PDR 1993

Guyana 1993

Bangladesh 2000

Uganda 1992/93

Indonesia 1989

Cambodia 1996/97

Pakistan 1991

Armenia 1996

Kyrgyz Rep 1993

Kazakhstan 1996

Brazil 1997 (NE&SE)

Malawi 1995

Ecuador 1998

Morocco 1998/99

Peru 1994

Yemen 1998

Azerbaijan 2001

Vietnam 1998

Mexico 1996

Panama 1997

Kenya 1992

Ghana 1992

Costa Rica 2001

Romania 1994

Jamaica 1998

Colombia 1992

Richest Poorest

All health spending All education spending

Percent

Percent

Source: World Bank 2003c.



bottlenecks, leakage, or lax management. Sim-ilarly, budgeted funds are not always availableto be spent. In Pakistan, for example, becauseof poor public expenditure management, only73 percent of the national health budget for1999 was executed, whereas in Tanzania,spending reached 114 percent of the recurrentbudget because of off-budget additions fromuser-fee revenue and donor transfers. In bothcases, budget management becomes difficult,with unforeseen issues shifting the actual fund-ing of public programs.

Issues in Education Quality

In education the quality of teaching is directlylinked to the effectiveness of learning, butoften quality is so low that the poor aredeterred from attending school in the firstplace. One indicator of a problem is a high stu-dent-teacher ratio that reflects congestion inthe classroom. It is no coincidence that the stu-dent-teacher ratio at the primary level in SouthAsia and Sub-Saharan Africa—the two regionswith the weakest education outcomes—is the

In the early 1990s the Ugandan government dramatically increased spending on primary education.Yet school enrollments stagnated. Could it be that the money was not reaching schools? To answerthe question, a public expenditure tracking survey started collecting data in 1996 on governmenttransfers to schools. It found that 87 percent of the nonwage resources intended for the schools werediverted to other uses.

This information was made public, prompting a vigorous information campaign from thenational government, which began publishing data on monthly transfers to local governments innewspapers. The central government also required primary schools and local administrations topost notices on receipt of funds. The information campaign brought large improvements: captureby interests along the way was reduced to less than 20 percent in 2001. Because poor people hadbeen less able than others to claim their entitlement from the local officials before the informationcampaign, they benefited most from it.

Tracking surveys can find problems in unexpected places. A survey in Peru tracking a participa-tory food supplement program (Vaso de Leche, or “glass of milk”) revealed that less than a thirdof each dollar transferred from the central government reached intended beneficiaries. Most of theleakage occurred below the municipal level—in the mothers’ committees and households. Theresults challenged the belief underlying the program that local community organizations werealways more accountable than public agencies. Authorities have decided to merge all nutrition pro-grams into a social fund that will transform Vaso de Leche into a conditional, multipurpose, cash-transfer program with stronger accountability. These and other tracking surveys in Ghana,Honduras, Madagascar, Mozambique, Papua New Guinea, Rwanda, Senegal, Tanzania, and Zam-bia suggest several lessons. They confirm that budget execution is a major problem and show thatprocedural clarity and due process are often missing.

The surveys find that poor resource management is often a result of too much discretion inresource allocation under conditions of limited information, weak controls, and strong vested inter-ests. They provide insights into the actual (rather than the formal) operation of schools and healthclinics and allow comparisons of public, private, and nongovernmental providers. Tracking surveyscan be highly cost-effective. But they need an authorizing environment: unless there is a solid polit-ical commitment for more transparency, government agencies may be reluctant to open their books.The Uganda case and a similar experience from neighboring Tanzania demonstrate the power ofinformation in enhancing “client power” and ensuring the money will not go missing.

Source: Reinikka and Svensson 2004; Dehn, Reinikka, and Svensson 2003; World Bank 2003a.

BOX 7.2 The case of the missing money: monitoring public expenditure

highest, more than twice the level in East Asiaand Europe and Central Asia.9

One reason for the high student-teacherratios is the relatively high salaries paid toteachers. Average annual teacher salaries inAfrica are 4.5 times per capita GDP, whereasin East Asia the ratio is 1.5 and in Europe andCentral Asia only slightly above 1. If the wagebill consumes the lion’s share of the recurrentbudget, little is left for textbooks and otherinstructional material. Evidence from Braziland India shows that dollar-for-dollar spend-ing on instructional materials is about 15times more productive (in terms of increasesin test scores) than spending on teachersalaries.10 In Africa, most countries spendmore than two-thirds of their recurrent edu-cation budget on salaries, with some reaching90 percent or higher (figure 7.5). The prob-lem is not limited to Africa.

But again, money does not guarantee ser-vice. Despite the high proportion of the bud-get devoted to teachers’ salaries, absenteeismamong teachers is high. Teacher absenteeism

rates have been estimated at about 25 percentin India and Uganda. Within states of India,a study found absenteeism rates ranging froma low of 15 percent in Gujarat to a high of 39percent in Bihar.11 In Uttar Pradesh, India,surprise visits to 16 schools found a stillhigher rate of absenteeism—about two-thirdsof teachers.12

Many governments provide education andhealth services in a top-down manner: the cen-tral government (with the help of donors) pro-vides the financing, hires the teachers anddoctors, and provides the complementarymaterials, such as textbooks and medicines. Intoo many cases, this system has failed to workfor poor people—in part because they have lit-tle say in it. In a few cases where poor peoplehave had a say—for example, by giving thema choice of service providers, involving par-ents in the running of the school, as in the ElSalvador Education with the Participation ofCommunities program, or simply increasingthe availability and transparency of informa-tion—the outcomes have been better. Possiblythe biggest challenge in accelerating progresstoward the MDGs—and also the biggestscope for reform—lies in strengthening thepower and capacity of poor citizens to affectthe delivery of services critical to their educa-tion and health.13

Gender, Education, and Women’sEmpowerment

Since the success of interventions to educategirls is fundamentally embedded in the socio-cultural context of households and theirdecisionmaking processes, it is particularlyimportant that family and community mem-bers be involved in the design and implemen-tation of those interventions. Increasing girls’access to education facilities alone will notachieve the goal of reducing gender dispari-ties in school attendance and achievement.Depending on the particular socioculturalcontext, other factors exert significant influ-ence, such as the presence of female teachingstaff, financial support to empower female



Perc

ent

100

80

60

40

20

0

Keny

aCo

ngo

Dem

.

Tanz

ania

Mal

awi

Ghan

a

Ango

laEt

hiop

ia

Zam

bia

Côte

d'Iv

oire

Suda

nZi

mba

bwe

Nig

erM

ozam

biqu

eU

gand

a

Burk

ina

Faso

Mal

iCa

mer

oon

Mad

agas

car

FIGURE 7.5 Teachers’ salaries absorb most recurrent education spendingPercentage of recurrent education spending paid to teachersfor selected Sub-Saharan African countries

Source: Bruns, Mingat, and Rakatomalala 2003.

students, school safety, gender-sensitivity ofthe curriculum, the availability of toilet facil-ities for girls, and information on the value ofgirls’ education. Involving households and thecommunity in the design of services wouldhelp in making them responsive to the partic-ular needs of the female students.

By strengthening the ability of the studentor client to influence the design of the service,choose among service providers, and monitorand discipline them, some countries have beenable to accelerate progress toward increasingenrollment, improving education quality, andreducing gender disparities. A good exampleis Bangladesh’s Female Secondary SchoolAssistance Program (FSSAP), which givesschools a stipend based on the number of girlsthey enroll and which has been very effectiveis raising girls’ attendance (box 7.3).

The empowerment of women of coursecalls for more than ensuring equal access toeducation opportunities. It also encompassesprovision of adequate and equal access to

health care, income-earning opportunities,and social and political rights. It requires thatwomen have control over the resources theyneed, and the rights and voice to participatein decisions that affect them. Promoting gen-der equality and women’s empowerment callsfor a strong political commitment from gov-ernments to be responsive to the different ser-vice needs of men and women and toeliminate all forms of discrimination. Someprogress has been made in this respect, butthere is much more to do. Ratification of theU.N. Convention on the Elimination of AllForms of Discrimination against Women(CEDAW) and submission of regular reportson compliance is one indication of a country’scommitment to ending gender inequalities intheir laws and regulations and to creatinginstitutions to protect women against dis-crimination. As of June 2002, all but fourdeveloping countries had ratified CEDAW,but only 71 percent had submitted at leastone annual report.14



Bangladesh’s Female Secondary School Assistance Program (FSSAP) was launched in 1993. Itsobjective was to improve rural girls’ enrollment and retention rates in secondary education.

The program cost of $88.4 million supported a package of interventions to increase the prospectsthat rural girls would enroll and remain in secondary school. A central component was the provi-sion of cash stipends to girls. The stipend is transferred through the commercial banking system toindividual bank accounts held in each girl’s name. The program also transfers rural girls’ tuitionand examination fees directly to schools.

Complementary efforts to increase the proportion of female teachers, provide water and sanita-tion facilities, improve community engagement, and provide occupational skills make schools moreattractive to girls and their families.

All unmarried girls of secondary school age who reside in rural areas were eligible to benefit fromthe program. Continuance in the program, however, required that girls attend school for at least 75percent of the days in the school year, maintain their academic performance above a set minimum,and remain unmarried.

The program has been very successful in increasing rural girls’ enrollment and retention rates.Stipend beneficiaries increased from 187,320 in 1994 to almost 900,000 in 2000. Girls now accountfor 54 percent of secondary enrollments in areas covered by FSSAP. Girls’ attendance rates haveimproved significantly and at 91 percent, are above boys’ 86 percent rate. The proportion of girlsachieving pass marks in school examinations, at 89 percent, is now above the boys’ rate of 81 per-cent. Girls’ dropout rates from the program for noncompliance are low.

Source: Khandker, Pitt, and Fuwa 2003.

BOX 7.3 The Bangladesh Female Secondary School Assistance Program

Health: A More Complex Agenda

A multiplicity of factors that affect healthoutcomes, and cross-sectoral links, make theattainment of health-related goals—reducingchild and maternal mortality and combatingmajor diseases—a still tougher challenge.Often policies focus narrowly on spending onhealth facilities and providers, but that is onlyone element of a broader and more complexagenda.

Female education and economic growthare the factors most closely correlated withlow under-five mortality. Educated girlsbecome better-informed women and mothers,who, in turn, are healthier, are better pre-pared when they have children, have fewerchildren, and care for them better. Unfortu-nately, progress in female education has beenrelatively slow, especially in primary comple-tion rates and secondary and postsecondaryenrollment rates. Gender gaps in educationare largest in Sub-Saharan Africa, the MiddleEast, and South and West Asia. Efforts suchas the Bangladesh FSSAP are encouraging,but unless they are scaled up, countries willfind it difficult to achieve the MDG of reduc-ing gender disparities in education. At thesame time, the goal of reducing child mortal-ity will become more difficult to achieve.

The second most important determinantof child survival is access to safe water andsanitation. (Challenges in developing this keyinfrastructure are discussed in chapter 6.)While overall access to safe water is increas-ing in most regions of the world, the distri-bution of safe water is quite unequal betweenrich and poor households. In many countries,almost everyone in the richest quintile of thepopulation enjoys access to an improvedwater source, but fewer than half of house-holds in the poorest quintile do. For example,in Ethiopia, only about 10 percent of thepoorest households use an improved watersource. Sanitation, too, is inequitably distrib-uted. In Accra, Ghana, 70 percent of thehouseholds from the poorest quintile sharetoilets with 10 or more households; the com-parable figure for households from the rich-

est 20 percent is 10 percent. Since high childmortality is more prevalent among poor peo-ple, major progress in reducing it is unlikelywithout improved access to safe water andsanitation for the poorest households.

Lack of access to health services is oftenexacerbated by poor service delivery or infor-mation failures. It is not uncommon to findhigh levels of absenteeism in clinics and hos-pitals. Absenteeism rates of medical person-nel in primary health facilities vary from 19percent in Papua New Guinea to 43 percentin India. Within India, the rates are as high as53 percent in the state of Bihar. The absen-teeism rate among doctors in primary healthfacilities in Bangladesh is 73 percent.15

Lack of knowledge can be another con-straint. For example, in Bolivia, many poorbabies are not delivered by a trained atten-dant simply because mothers are unawarethat they are eligible for free care. In India,where immunization is free, 60 percent ofchildren have not been fully immunized.Asked why they had not immunized theirchild, 30 percent of mothers said it wasbecause they were not aware of the benefitsavailable, while another 30 percent said itwas because they did not know where to goto have their child vaccinated.16 Public actioncan play a significant role in disseminatinginformation.

Maternal survival is strongly associatedwith the presence of a skilled attendant dur-ing birth. In general, the trend in attendedbirths is positive, but the levels remain dis-turbingly low, especially for low-incomecountries in Africa and South Asia—Nepaland Pakistan especially (table 7.3). Even somemiddle-income countries, such as Morocco,have particularly low rates of attended births.Nonetheless, the rate of improvement in thefrequency of attended births is highest for thepoor, an encouraging finding.

A related measure of progress is the preva-lence of contraception, which captures thepossibility of birth spacing and other aspectsof maternal health. Rates of contraceptionuse are only about 27 percent on average indeveloping countries, and the rate for the



poorest 20 percent of the population is about19 percent. Furthermore, especially for thepoor, the constraint may not necessarily bethe availability of contraceptives. Often,power relationships within the householddetermine contraceptive use. About 70 per-cent of the women from the poorest 20 per-cent of the population in the Central AfricanRepublic, Côte d’Ivoire, Tanzania, and Ugandareported that their husbands did not approveof their contraceptive use. Similar findingshave emerged in Central America.17

In the fight against HIV/AIDS and othermajor infectious diseases, financing issues arebeing addressed with the support of the inter-national community. However, attention tothe management and implementation capac-ity needed to make effective use of financinghas lagged in many countries. Progress isbeing made on the institutional front. TheProgress Report prepared for the U.N. Gen-eral Assembly Special Session on HIV/AIDSnoted that 90 percent of the 103 countriesreporting had established national AIDScommissions or authorities, many of themstructured to include representation from keystakeholder groups.18 But in many cases suchinstitution-building efforts are at an early stage.

Managing assistance to HIV/AIDS pro-grams as “business as usual” will not work.Encouraging successes in Brazil, Thailand,Uganda, and some other countries point theway:

� First, governments need actively to mobi-lize nongovernmental organizations (NGOs)and community-based organizations (CBOs)to help overcome social and political obsta-cles to reaching target groups. Effective pre-vention and treatment of HIV/AIDS requireseducating the public about the disease andpromoting behavior change among popula-tion groups that are normally far beyond thereach of government.

� Second, to ease the burden on limited insti-tutional capacities, governments, especiallyin severely affected countries, need to exploitopportunities to contract out services. Whilesome progress is evident—for instance,Kenya’s National AIDS Commission hashired a financial management agency that inturn has helped channel resources to morethan 750 local CBOs and NGOs—the timerequired to establish new ways of doingbusiness often conflicts with the urgent needfor rapid and extensive action.

� Third, capacity building within govern-ment agencies should include the strength-ening of monitoring and evaluation toassess performance and impact and informthe design and implementation of publicand private programs.

� Fourth, with respect to the health systemcapacity, efforts need to focus on criticalbottlenecks. Numerous NGOs in WestAfrica report that the binding constrainton expansion of the antiretroviral program



Table 7.3 Frequency of attended births by wealth and region, various years(percent)

Level Average annual change

Regional average Poorest quintile Population average Poorest quintile Population average

East Asia and Pacific 28.8 59.3 — —Europe and Central Asia 88.4 95.1 2.3 0.5Latin America and Caribbean 44.1 68.0 8.5 2.6Middle East and North Africa 33.6 52.5 10.6 6.4South Asia 7.0 21.5 11.6 7.7Sub-Saharan Africa 24.6 46.5 1.1 –0.4

— Not availableSource: USAID, Demographic and Health Survey estimates, various years from 1995 to 2002.

is a shortage of trained doctors, nurses,and laboratory technicians. They are beingforced to recruit staff from existing healthservice delivery programs, exacerbatingshortages there.

Social Assistance: Successes and Challenges

Finally, social assistance programs in severalcountries have been instrumental in improv-ing and sustaining the access of the poor tobasic education and health services. Piloted insome Latin American countries, programsproviding transfers conditional on schoolattendance or regular health center visits havebeen scaled up nationally (for example, Mex-ico’s Progresa program and Brazil’s BolsaEscola) and adopted in some other countries(for example, Bangladesh and Turkey).

Nonetheless, social assistance spending inmany countries is fragmented, redundant,and poorly targeted. In Paraguay, for exam-ple, about 90 social assistance programs werebeing implemented recently by 22 agencies. InArgentina, 60 federal social protection pro-grams have overlapping objectives and targetgroups. Bulgaria has 34 programs. Evalua-tion of these programs often is weak. A recentstudy of a sample of such programs foundthat just over 20 percent had well-developedevaluation arrangements.19 Thus while therehas been successful innovation in social assis-tance that merits scaling up, there is alsoscope for better targeting in existing socialassistance spending.

Agenda for Accelerating ProgressProgress toward human development goalswill require efforts on multiple fronts. Moreresources are needed to scale up programs inpriority areas. But just as important, there isneed to accelerate policy and governancereforms to make programs more effective.The reform agenda is multidimensional,spanning investment, policies, and gover-nance. It is also multisectoral, with humandevelopment outcomes affected in important

ways by improvements in other sectors, suchas access to safe drinking water, basic sanita-tion, roads, and electricity.

Education

The challenges facing individual countriesand the international community are wellillustrated by developments relating to theprimary education goal. While several analy-ses point out that this goal is attainable by2015, public spending on education in low-income countries has risen slowly, constrainedby limited resources, competing demands, andlow institutional capacity. The allocation ofspending, such as the funding bias toward sec-ondary and higher education, typically bene-fits the nonpoor. In contrast, the quality ofprimary education is often so low (high stu-dent-teacher ratios, inadequate allocations tononsalary expenditures, and frequent teacherabsenteeism) as to discourage school atten-dance. The effective functioning of educationservices is underpinned by good governance—functioning civil service rules and processes,effective financial management, control ofcorruption, and direct participation of com-munities. Reforms in these areas are the key toprogress, but they have lagged.

Community involvement in the design ofeducation services is particularly important tothe goal of reducing gender disparities in edu-cation. Girls’ attendance at school can beinfluenced strongly by sociocultural factors,and community involvement can help ensurethat services respond to those factors. Effec-tive improvement in female access to educa-tion—and indeed to other services—requiresthat gender concerns be fully integrated intopolicy and service design.

Overall, while there has been progress,governments have been slow to undertake thepolicy and institutional reforms necessary toimprove the coverage and quality of primaryschooling, reflecting in part the political diffi-culty of some of the agenda. A stronger polit-ical commitment is required to deepen andaccelerate reforms if the MDGs in educationand gender are to be achieved.



Donors also need to improve the quality oftheir support. The record of the donor com-munity has been mixed as well. Traditionally,donors have concentrated their aid onaddressing supply constraints, such as build-ing schools, rather than the underlying insti-tutional impediments to effective delivery ofprimary education services.

Recognizing the limitations of the pastapproach, in terms of both providing ade-quate resources and supporting broaderreforms in developing countries’ educationsystems, donors coalesced around the Educa-tion for All–Fast Track Initiative, agreeing toprovide additional support to primary educa-tion explicitly linked to countries’ policy per-formance and increased accountability forresults. The criteria for support emphasizeadequacy and sustainability of domesticfinancing, service delivery standards consis-tent with education quality and efficient useof resources, and transparent monitoring ofresults through annual joint sector reviews.The initiative seeks to improve donor per-formance through concrete actions to har-monize aid.

While the EFA-FTI has been a positiveforce for spurring country action, donor pro-gramming and budgeting cycles have trans-lated into slow disbursement, with only $6million of the first $170 million committedactually disbursed as of January 2004. Also,the number of countries touched so far by theprogram has been small. At their meeting inOslo in late 2003, donors agreed to main-stream the FTI approach, and 25 more coun-tries are expected to join in 2004. WorldBank projections suggest that as the FTIscales up to all low-income countries, at least$3.7 billion per year will be needed in ex-ternal financing for primary education by2005–06, compared with about $1 billion in2002. How effectively the FTI contributes toaccelerating progress toward the educationMDGs will hinge on donor performance inactually delivering increased and more pre-dictable financing in support of tangiblecountry reforms and close monitoring of theachievement of results on the ground.20

Health

The goal of reducing child mortality is stillmore complex and difficult to attain. First,some of the public policies and actions thatcan make the greatest contribution lie outsidethe health sector (female education, cleanwater and sanitation) and so require greatercross-sectoral coordination. Second, actionswithin the health sector to improve the supplyof services, such as access to health clinics orimmunization, confront major challenges.These include underfunding, weak institu-tional capacities and governance underminingservice quality, and limited availability oftrained personnel (aggravated by incentiveproblems reflected in high absenteeism amongdoctors, especially in rural areas). Third, ser-vice provision often must also contend withchallenges on the demand side, such as lack ofmoney or knowledge, that limit poor people’seffective access to or use of services.

Accelerating progress on child survival,therefore, requires a concerted effort infemale education, provision of safe waterand sanitation facilities, and expansion ofaccess to effective health facilities comple-mented by incentives, governance arrange-ments, and accountability for better deliveryof services. It also requires incentives andinformation to stimulate demand, especiallyfor preventive services.

As with public actions in primary education,political factors work to slow the progress ofactions to reduce child mortality. In most coun-tries the share of public health spending thatbenefits the richest quintile exceeds that goingto the poorest. While high child mortality is pri-marily a problem of poor people, ill healthaffects everybody. In countries without ade-quate health insurance markets (almost all low-income and many middle-income countries),public spending on health becomes a substitutefor the insurance against catastrophic healthexpenditures that everybody needs.

Since they have greater political power, therich are often able to capture public healthexpenditures to serve their needs. At the sametime, the evidence is clear that if systems are



set up exclusively for the poor, they tend to beless effective and of lower quality, because thepoor lack the resources, clout, and voice ofthe middle or upper classes. Unless the powerof communities to monitor and discipline ser-vice providers is enhanced—including possi-bly through co-payments—increasing publicspending in health by itself is unlikely to havea significant impact on child mortality, or forthat matter on maternal mortality.

Public resources in most countries are inad-equate to provide free basic health or educa-tion to all citizens, and there is a clear role forprivate spending. Private out-of-pocket spend-ing is part of health care financing every-where—if not formally, then through informalor illegal payments, with the poor tending tobear the brunt of such payments. A moretransparent and judicious use of fees can notonly help alleviate the public resource con-straints, but also contribute to improved qual-ity of service by motivating greater stakeholderoversight of service delivery. It would alsoallow scarce public subsidies to be targetedmore effectively to poor people.

As in primary education, the donor com-munity has had a mixed record in strengthen-ing public actions to reduce child mortality.On the one hand, sizable amounts of aid gotoward fighting the diseases of children andsupporting health expenditures in developingcountries. On the other hand, most of that aidis provided for earmarked expenditures or forvertical programs that deliver a narrow pack-age of interventions on a small scale, withinsufficient attention to the incentives thesecreate for undertaking the difficult public sec-tor reforms that are needed to make a sub-stantial dent in child mortality. Advocates forincreasing foreign aid in health tend to focusalmost exclusively on per capita spending onhealth by various countries—neglecting issuesrelating to the effectiveness of that spending orthe fact that spending in other sectors maymake a bigger difference to child-health out-comes. Aid would be more effective if themodalities of its provision were more flexibleand supported broader reforms that include

the institutional health-sector issues and inter-sectoral linkages that must be addressed inorder to scale up impact.

The maternal mortality goal may be evenharder to achieve than the child mortality goal.In addition to the resource and institutionalconstraints on the delivery of related health ser-vices, the outcomes depend greatly on under-lying social, cultural, and behavioral factors.Progress in the longer term may hinge as muchon efforts to broaden educational opportunitiesand raise awareness as on specific health serviceinterventions. With respect to the latter, theintervention that seems to be most important(and most easily measured) is the presence of askilled attendant during birth. Again, improv-ing poor people’s access to skilled attendants ismore than a matter of resource availabilities, asit suffers from limited political support—thepoor are the only segment of the populationthat relies on such paraprofessionals—andpoorly functioning institutions.

Communicable diseases. Similarly, effectivelycombating the spread of major diseases—HIV/AIDS, tuberculosis, and malaria—requiresaction across a range of sectors and actors,including civil society and the private sector.The target of halting and beginning to reversethe spread of HIV/AIDS by 2015 is particu-larly challenging, given the long incubationperiod of this disease. Effectively combatingthis epidemic requires action at three levels(and this may also be similar for TB):

� Political—strong and explicit leadershipand commitment at the highest politicallevels to help generate and sustain the kindof exceptional responses that the epidemicrequires

� Strategic—a national HIV/AIDS strategyto guide the evolution of a multisectoralresponse by multiple actors

� Implementational—translation of the strat-egy into action at the necessary scalethrough mobilization of resources, domes-tic and foreign, and building of institu-tional capacities.



The progress made in these respects so farhas been modest compared with the scale ofthe challenge.

The GFATM has focused on raisingresources to contain the spread of HIV/AIDS,TB, and malaria. While substantial commit-ments have been made to the fund, imple-mentation has been slow, as evidenced bythe delays in signing grant agreements forapproved projects and the limited volume ofdisbursements. Out of $3.4 billion in pledges,roughly $1.1 billion had been committed andonly $236 million disbursed as of January2004. In addition to slow follow-up at thedonor end, implementation capacities ofrecipient governments have been a constraint.Problems stem partly from the fund’s opera-tional model, with its insufficient appraisalof financial issues, donor coordination, andinstitutional capacity prior to grant approval.The focus of country proposal reviews hastended to be on technical issues rather thanimplementation constraints—capacity ofresponsible agencies, service delivery net-works, fiduciary environment. There is a needto find ways to better align programs andcapacity, to build capacity simultaneouslywith the allocation of additional resources,and to better coordinate donor support toexpedite implementation.

Scaling up on the basis of successful models.Well-designed social assistance programs tar-geted to the poor and linked to school atten-dance and health center visits can play animportant role in enhancing the poor’s effec-tive access to and use of critical human devel-opment services, as well as cushioning thepoor against major income shocks. Innovativeconditional cash transfer programs success-fully implemented in some Latin Americancountries provide models that can be consid-ered for scaling up and replication elsewhere.At the same time, to make room for effectiveinterventions such as these, countries need toreview their overall social assistance frame-works, which often suffer from poor target-ing, fragmentation, and fiscal unsustainability.

Notes1. Papers produced by these working groups

are available at the project Web site: www.unmil-lenniumproject.org.

2. UNDP 2003.3. World Bank 2003c.4. Data on public spending allocated to

human development suffer from many gaps andare not available for all countries. The analysispresented here draws on two sets of data: datacompiled by IMF staff, on the basis of countryauthorities’ estimates, that provide informationon the composition of public spending for low-and middle-income developing countries; anddata assembled from various World Bank coun-try Public Expenditure Reviews and joint IMF-World Bank expenditure tracking work relatingto PRSPs and HIPC programs. Improved data onthe level and composition of social sector spend-ing are a priority area in the strengthening of sta-tistics in developing countries.

5. Kattan 2003.6. Indeed, fees, for those who can pay, could

offer incentives to providers, which can serve toenhance service quality (such as lower teacherabsenteeism), contribute to more flexible fundingfor schools and health centers, and help combatcorruption by reducing the incidence of infor-mal/illegal payments. Fees could also motivate ser-vice users to become more involved in monitoringand disciplining service providers. For some coun-try evidence, see Lewis (2000).

7. Di Tella and Savedoff 2000; Lewis 2000.8. World Bank 2004. 9. World Bank 2003b.10. Filmer and Pritchett 1999.11. World Bank 2003c.12. Dreze and Gazdar 1996. 13. World Bank 2003c.14. For more information, see the Web site of

the U.N. Division for the Advancement of Women(DAW) at http://www.un.org/women watch/daw/cedaw.

15. World Bank 2003c.16. World Bank 2003c.17. USAID various years.18. UNAIDS 2003.19. Rubio and Subbarao 2003.20. For a fuller discussion of progress and

issues relating to the EFA-FTI, see DevelopmentCommittee 2004.



Improving environmental policies and mit-igating the environmental impacts of sec-toral policies are an integral element of the

agenda for achieving the Millennium Devel-opment Goals. Goal 7—ensure environmentalsustainability—aims to mainstream the envi-ronment in policy and programs, reverse theloss of environmental resources, and improveaccess to environment-related services. Theimportance of the environment to the MDGsis reinforced by its strong links to the rest ofthe goals. It is difficult to imagine reducingincome-poverty in rural areas where land isdegraded. Reductions in child mortality willbe more likely if households have access toadequate water supply, sanitation facilities,and modern fuels. And climate change result-ing from unchecked environmental degrada-tion would favor the spread of vector-bornediseases and increase the likelihood of nat-ural disasters—disasters that, in turn, reduceincome and destroy the infrastructure foreducation and health. If environmental sus-tainability is not ensured, progress towardthe other MDGs may be short-lived.

Progress on EnvironmentalPoliciesEnvironmental policies are more difficult tomonitor than most because data are scarce.

Data directly measuring specific environmentpolicies—pricing, regulations, institutions—on a cross-country basis are not availablewith any consistency. One source of infor-mation on policies is the World Bank’s Coun-try Policy and Institutional Assessment,which includes an assessment of countries’overall environmental management. TheCPIA ratings show that environmental man-agement is a policy area where the bulk ofthe agenda in developing countries is still tobe addressed. The average country ratingsare relatively low (figure 8.1). The ratings areimproving, however, although they seem tohave stalled in Asia and Latin America. Theimprovement in the past five years is particu-larly notable in Europe and Central Asia. Theratings rise with the income level; middle-income countries show appreciably betterenvironmental management than low-incomecountries.

Monitoring Policies through Outcomes: Evidence of Limited ProgressData are available on key environmental out-comes that indicate countries’ progresstoward environmental objectives, includingthe MDG related to the environment. Infor-mation on environmental outcomes, compiled


Promoting EnvironmentalSustainability

8

from a variety of sources and published inWorld Development Indicators, includesland use (agricultural land, forests, protectedareas), access to water and sanitation, waterresources (available water resources, totalwithdrawal, agriculture withdrawal), energy(energy intensity, traditional fuel use, urbanair pollution), and transport (passenger cars,fuel prices). But coverage and quality aremixed, as are the prospects of obtainingimproved data. Although satellite technologyis helping in some quantitative aspects (forexample, in monitoring forest cover), infor-mation on environmental quality (water qual-ity, land degradation) is hard to obtain. Evenmonitoring of an “easy” issue (urban air qual-ity) has worsened since the GEMS/Air pro-gram closed.1

Environmental outcome indicators sug-gest that progress toward environmental sus-tainability is limited (table 8.1). For example,despite small gains in forest cover in Europe

and Central Asia and the Middle East andNorth Africa during the 1990s, large losseswere experienced in the rest of the develop-ing world. As a consequence of economicand demographic growth, most developingregions are witnessing significant increases incarbon dioxide emissions—the impressivereductions in Europe and Central Asia areattributable largely to economic restructur-ing. Although generally decreasing acrossregions, the use of traditional (highly pollut-ing) fuels remains extremely high in Sub-Saharan Africa, South Asia, and East Asiaand the Pacific. The “adjusted net saving”measure, a useful indicator of sustainability,suggests that, based on current trends, wel-fare is likely to decline over time in Sub-Saharan Africa and the Middle East andNorth Africa, drawing attention to the needfor improved economic, environmental, andresource policies in the countries of thoseregions.



FIGURE 8.1 Environmental policy ratings are low but improvingRatings of environmental management policies for developing countries by region and countryincome category, 1999 and 2003

6.0

5.0

4.0

3.0

2.0

1.0

0.0

Rati

ng

Middle-income

countries

Low-income

countries

Sub-SaharanAfrica

SouthAsia

Middle East& NorthAfrica

LatinAmerica &Caribbean

Europe& Central

Asia

East Asia& Pacific

3.14 3.06 3.12

3.94

3.43 3.412.94

3.283.64 3.57

2.86 3.06 2.93 2.993.37

3.83

20031999

Note: World Bank CPIA ratings range from a low of 1 to a high of 6, with an increase denoting improvement. Source: World Bank CPIA database.

Ensuring EnvironmentalSustainability and Achieving the MDGs: Need for Policy IntegrationA desirable development outcome in itself,environmental sustainability also contributes toother outcomes such as economic growth,poverty alleviation, and better health. In par-ticular, the environmental MDG has stronglinks with those related to poverty and health.For example, research in Zimbabwe showsthat environmental resources make a signifi-cant contribution to rural incomes, represent-ing some 40 percent of average income for poorhouseholds.2 The World Health Organizationestimates that 40 percent of under-five mortal-ity is traceable to diseases strongly associatedwith environmental factors—among them diar-rhea and acute respiratory infections.3 But

improving environmental conditions also willbe important in achieving the education andgender goals. In many poor communities,women and girls spend vast amounts of timefetching water and collecting fuel wood, whichprevents them from engaging in productiveactivities and attending school (box 8.1).

Because policy choices in different sec-tors—agriculture, water, energy, transport,trade, public finance—affect environmentalsustainability, better integration of sectoralpolicies may be the fastest route to environ-mental sustainability, as highlighted by theUnited Nations Task Force on EnvironmentalSustainability (box 8.2). Effective environ-mental management means going beyondindividual policy actions, such as implement-ing particular regulations, and ensuring thatenvironmental concerns are fully integratedinto development policymaking more broadly.

P R O M O T I N G E N V I R O N M E N T A L S U S T A I N A B I L I T Y


TABLE 8.1 Selected outcome indicators of environmental sustainability by region and country income category

Adjusted net Carbon dioxide Solid biomass use savinga Deforestation emissions in householdsc

Percent Forest cover lostb Metric tons Percent of total of GNI 1990–2000 per capita energy use

Area Percent PercentIncrease (thousands increase increase

Region and country income category 2000 1990–2000d of km) Percente 2000 1990–2000 2001 1994–2001

RegionEast Asia and Pacific 21.3 5.8 116 2.7 2.1 8.7 71.7 –0.9Europe and Central Asia — — –81 –0.9 6.5 –37.6 8.2 0.3Latin America and Caribbean 4.4 0.5 459 4.6 2.6 19.7 37.6 –4.8Middle East and North Africa –10.0 0.9 –2 –1.4 4.1 33.1 12.5 –1.6South Asia 11.3 1.6 9 1.1 0.9 31.8 84.1 –1.5Sub-Saharan Africa –3.3 1.3 530 7.6 0.7 –21.6 88.3 0.0

Income levelLow 4.7 0.0 731 7.5 0.9 2.6 84.9 –1.9Lower middle 13.1 6.0 –147 –0.8 3.0 –17.3 46.0 –0.3Upper middle 6.9 1.9 151 5.9 6.2 9.4 23.4 1.9High 12.8 2.5 –79 –1.0 12.4 4.7 5.7 –0.7

World 12.0 2.5 950 2.4 3.8 –6.9 37.6 0.1

— Not available.a. A negative number indicates dissaving.b. A negative number indicates that forest cover has increased.c. Solid biomass is defined as any plant matter used directly as fuel or converted into fuels (such as charcoal) or electricity and/or heat. Includedhere are wood, vegetal waste (including wood waste and crops used for energy production), and animal materials and wastes.d. Calculated as average for 1995–2000 minus average for 1985–90.e. As percentage of 1990 forest cover.Source: World Bank, World Development Indicators; World Bank staff.

Priorities for ActionWhen establishing priorities for action toachieve environmental sustainability, one sizedoes not fit all. OECD standards for environ-mental quality are not appropriate for manydeveloping countries. In fact, the nature ofenvironmental problems varies considerablyby income level.

In middle-income countries, many envi-ronmental problems are the byproducts of

growth—urban pollution, toxic emissions,and agricultural runoff. In low-income coun-tries, the issues are closely tied to livelihoods,particularly soil quality and water availabil-ity, and to health outcomes (diarrheal diseaseand acute respiratory infections from dirtycooking fuels).

Because environmental issues are locationspecific, different clusters of countries havedifferent priorities:



In 1999 some 10 million children under the age of five died in low-income countries—2.1 millionin India alone. Using Indian health survey data, World Bank researchers have concluded that invest-ments targeted at improving environmental conditions can substantially reduce child mortality. Inparticular, universal access to private piped water, electricity, and separate kitchens with clean cook-ing fuels would save the lives of 10.4, 5.5, and 33.6 children per 1,000 live births, respectively, inrural India. The analysis also shows that girls face a higher mortality risk than boys after the firstmonth of birth (suggesting a significant gender bias in household resource allocation) and that childmortality is higher among those born to mothers with no education. Universal female primary edu-cation would reduce the under-five mortality rate from 99.9 to 77.0 deaths per 1,000 live births.

What does this mean? It means that environmental and human development goals are closelyintertwined. In order to reduce child mortality, for example, developing countries need to mobilizecontributions from multiple sectors—health, of course, but also education, energy, and water—andincorporate cross-cutting perspectives (environment, gender).

Source: van der Klaauw and Wang 2003.

BOX 8.1 Multisectoral interventions to achieve the MDGs: lessons from childmortality in rural India

The U.N. Task Force on Environmental Sustainability is one of 10 task forces making up the U.N.Millennium Project. Its preliminary findings conclude that progress toward environmental sustain-ability has been unsatisfactory, largely because environmental and biodiversity issues have beenpoorly integrated into mainstream thinking and policymaking. The task force places particularemphasis on the importance of biodiversity conservation and ecosystem integrity for environmen-tal sustainability. Although a framework for action is still in progress—the final report is due in June2005—the background paper suggests the main lines of action should be improving environmentalawareness among policymakers and the general public, enhancing national capacity for ecosystemmanagement, and consolidating international markets for global environmental services.

Source: United Nations Millennium Project 2003.

BOX 8.2 The United Nations Task Force on Environmental Sustainability

� In resource-rich countries, such as Indone-sia, governance issues involving commer-cial resources

� In rapidly urbanizing countries and coun-tries in transition, such as Thailand andPoland, pollution management

� In agriculture-dependent countries, suchas Ethiopia, natural resource management

� In arid countries, such as Mali, land andwater management and adaptation to cli-mate change

� In diverse countries, such as Madagascar,biodiversity conservation and tappingglobal finance to help cover the opportu-nity costs of conservation

� In small island states and low-lying coastalcountries, such as the Maldives or Bangla-desh, adaptation to climate change.

The following elements that broadly definethe environmental policy reform agenda in

developing countries should therefore beadapted to income levels and particular coun-try circumstances:

� Strengthening policy and institutionalframeworks for environmental manage-ment

� Getting the prices right for energy, water,and agricultural inputs

� Facilitating access to basic services, such asimproved water supply, sanitation, andclean energy sources.

Notes1. The GEMS/Air program was funded and

operated jointly by the United Nations Environ-ment Programme, the World Health Organization,and the U.S. Environmental Protection Agency.

2. Cavendish 2000. 3. World Health Organization 2002.

P R O M O T I N G E N V I R O N M E N T A L S U S T A I N A B I L I T Y


IIIDeveloped-Country Policies

The overall prospects for economicgrowth and poverty reduction in thedeveloping world depend critically on

the maintenance of strong and balancedgrowth in the developed countries. Macroeco-nomic policies that promote such growth canthus contribute to progress toward the Millen-nium Development Goals and related out-comes in developing countries.1 Good globaleconomic conditions benefit developing coun-tries in several ways, notably by boosting trade,foreign direct investment, and possibly officialdevelopment assistance. Growth in developedcountries can also benefit the rest of the worldthrough its effects on labor migration andrelated phenomena such as flows of remittancesfrom migrating workers.

Although the prospects for world eco-nomic activity appear to be reasonably brightfor the next few years, sustaining a strongglobal economy will require the major coun-tries to address some outstanding issues andimbalances. The challenges to be addressedinclude the orderly moderation of the U.S.current account deficit; the implementationof further structural reforms in several coun-tries, especially in labor markets; and the fis-cal impact of demographic changes buildingup in many developed countries. Disorderlyadjustment in the largest economies couldretard growth or leave global economic con-ditions vulnerable to exogenous shocks.

These challenges come at a time when manydeveloping countries, struggling to recoverfrom the recent slowdown in world trade, arefacing uncertain prospects. In the next fewyears, the terms of trade are generally expectedto disfavor most developing countries—partic-ularly those exporting fuels. The outlook forprivate capital flows to developing countries,meanwhile, has brightened, but only thehigher-rated emerging markets are likely tobenefit from the improving conditions. Theneed for capital, particularly to finance infra-structure, continues to be large in developingcountries as a whole, and can be partly metonly by vigorous growth in remittances.

In addition, growth in middle-income coun-tries, particularly emerging market economies,is highly susceptible to the negative effects offinancial crises, which typically have the great-est impact on the poorest sections of the pop-ulation. Improvements can be made in theinternational financial architecture to enhanceprospects for stronger and more stable capitalflows to developing countries and to reduce thelikelihood, severity, and duration of financialcrises. Rapid progress is being made in the useof collective action clauses (CACs), whichshould make debt workouts less costly andthus may make developing-country bondsmore attractive to investors, but substantialwork remains to improve practices in sover-eign debt restructuring.


Fostering Growth and StabilityMacro-financial Policies

9

Macroeconomic Policies in Developed Countries: Trends and ProspectsIn the United States, the current very accom-modative stance of macroeconomic policy—with a rising budget deficit that is contributingto a record current account imbalance—isexpected to be tightened in the longer term onthe basis of a return of economic growth torates of 3 to 4 percent (table 9.1 and figure 9.1).As economic growth in the United States gath-ers steam, an orderly tightening of monetaryand fiscal policies would help restore balance.

In Europe, the substantial weakening ofbudgetary positions in recent years leaves littlescope for fiscal stimulus. This adds urgency tothe need to improve economic performance,which is being hampered by difficulties anddelays in implementing needed structuralreforms, especially in labor markets and socialsecurity systems. In the coming years, suchreforms could help raise European growth toa sustainable range of 2 to 3 percent.

In Japan, the economic expansion is ex-pected to continue in the near term. To sus-tain growth over the medium term, however,continued efforts will be needed to end defla-tion, to address the remaining weaknesses inthe financial and corporate sectors, and tostabilize public sector debt.

Risks to the global economy. Alongside thisimproving scenario, substantial risks remain.In particular, the world economy could be hitby a worsening of political and security con-ditions, notably in the Middle East. The mostsignificant risk on a global scale, however,continues to be a possible abrupt adjustmentin the U.S. current account position. From1998 to 2002, the current account deficitmore than doubled to $480 billion, and in2003 it rose further to more than $540 bil-lion. Such large external imbalances are diffi-cult to sustain for a long period, and mostobservers now anticipate an adjustment. Dif-ficulties in financing the U.S. external positioncould lead to a sharp decline in the dollar,

P A R T I I I : D E V E L O P E D - C O U N T R Y P O L I C I E S


TABLE 9.1 Macroeconomic indicators for advanced economies, 1993–2008

2003 2004–08Indicator 1993–97 1998–2002 estimated projected

Inflation (average, percent)United States 2.7 2.3 2.3 2.1European Union 2.7 1.8 2.0 1.8Japan 0.7 –0.4 –0.2 0.7Advanced economies 2.5 1.7 1.8 1.8

Current account balance, percent of GDPUnited States –1.5 –3.6 –5.0 –4.6European Union 0.6 0.2 0.4 0.6Japan 2.3 2.6 3.2 3.2Advanced economies 0.2 –0.6 –0.8 –0.7

Fiscal balance, percent of GDPUnited States –2.9 –0.1 –4.9 –3.6European Union –4.8 –0.9 –2.6 –1.8Japan –3.9 –6.8 –8.0 –6.8Advanced economies –3.6 –1.2 –4.1 –3.0

Real GDP growth, in percentUnited States 3.5 3.0 3.1 3.7European Union 1.8 2.5 0.8 2.5Japan 1.7 0.4 2.7 1.8Advanced economies 2.8 2.6 2.1 3.1

Source: IMF World Economic Outlook database and IMF staff calculations.

however, with possible negative repercussionson the relatively highly indebted U.S. house-hold and corporate sectors—for example, ifthe Federal Reserve were to raise interestrates to prop up the currency.

Given the size of the financing needs aris-ing from the U.S. deficit, the moderation of itsexternal imbalance could free up substantialfinancing for other countries, including devel-oping countries. The degree of substitutabil-ity between finance for developing countriesand the United States is no doubt limited, butto the extent that some crowding out is tak-ing place, emerging market economies couldbenefit from lower needs for external financ-ing in the U.S. economy.

Implications for DevelopingCountriesAn improved global outlook presents anopportunity for higher growth in developingcountries. More robust income growth in thedeveloped countries is expected to stimulateactivity in developing countries through trade,especially higher export volumes. Developingcountries suffered from the slowdown inworld trade in 2002, chiefly because of weakdemand in developed countries. Over themedium term, however, the pickup in globalactivity that began in 2003 should translate

into stronger and more sustained exportgrowth for developing countries as a group(table 9.2). This could be helped further ifprogress is made in reducing trade barriersthrough a successful outcome to the DohaRound (see chapter 10). Stronger growth inworld trade is particularly relevant for devel-oping countries, such as those in East Asia,that trade intensively with developed coun-tries. Although the terms of trade for develop-ing countries are expected to fall in 2004–08,the decline should be limited mostly to fuelexporters; nonfuel primary commodity pricesare projected to remain broadly constant.Although oil markets continue to be subject togreat uncertainty, prices are expected to stabi-lize once production fully resumes in Iraq.

Private Capital Flows

Private capital flows can be an importantsource of growth in developing countries, espe-cially if they take the form of foreign directinvestment (FDI), which can raise investment,provide a vehicle for the transfer of knowledgeto the host country, and enhance the humancapital of the host country’s work force. FDI isalso desirable as a source of finance because itis less volatile than debt and because it pro-vides better risk sharing between originatingand recipient countries.

F O S T E R I N G G R O W T H A N D S T A B I L I T Y : M A C R O - F I N A N C I A L P O L I C I E S


FIGURE 9.1 A robust global economy requires orderly resolution of the large external and fiscal imbalancesCurrent account and fiscal balances, United States and all advanced economies, 1993–2003

2 1 0

–1–2–3–4–5–6–7

Perc

ent o

f GD

P

United States

Current account balance

Advancedeconomies

(all)

1998–20021993–97 2003

2 1 0

–1–2–3–4–5–6–7

United States

Fiscal balance

Advancedeconomies

(all)

Source: IMF World Economic Outlook database.

FDI continues to supply the largestamount of external financing to developingcountries. Although FDI drifted lower in2003, it is expected to recover in 2004–05, inline with the global economic recovery.2

While prospects for FDI flows to developingcountries continue to be favorable, their des-tinations remain highly concentrated. Tencountries captured about 69 percent ofNorth-South FDI in 2003 (down from a highof 78 percent in 2000). Nine of the 10 coun-tries were middle-income countries. Theshare of middle-income countries in total FDIto developing countries is about 90 percent.The size of FDI flows relative to GDP is twiceas large in middle-income countries as in low-income countries (figure 9.2). Only a smallportion of FDI—$9 billion out of $135 bil-lion in 2003—goes to Sub-Saharan Africa,which is the region most in need of invest-ment if it is to achieve its development goals.Half the FDI flows to this region werereceived by three oil exporters. The challengeis to raise the capacity of African countries toattract and to make good use of FDI, whichin turn will require improvements in educa-tion, infrastructure, and governance. Once acountry has a strong enough business climateto attract FDI, inflows can fuel a virtuous cir-cle of growth and further investment.3 Theavailability of risk insurance against a broad

range of political risk through private andofficial agencies could aid FDI flows to poorercountries in Africa and elsewhere.4

Although private debt flows5 represent atransfer of savings from developed to devel-oping countries, they can be a mixed blessing.Because private debt flows depend largely oninvestor confidence, they are not a very reli-able source of financing, as the availabilityand cost of financing can change suddenlyand with little warning (figure 9.3).6 In par-ticular, portfolio holdings are relatively easyfor investors to sell on secondary markets,especially for those emerging market countriesin which secondary markets are well devel-oped. Countries that have benefited from sub-stantial portfolio inflows may therefore be hitsuddenly by a sharp increase in the cost ofcapital, or a loss of access, or both.

The challenge is to sustain the cyclicalrecovery in private debt flows over themedium term by maintaining investor confi-dence, while taking steps to lower the vul-nerability of developing countries to sharpreversals in flows and to widen the set ofcountries enjoying access to financing. More-over, discrimination by investors—signs ofwhich have appeared in the wake of therecent experience with defaults and debt re-structurings in emerging markets—can onlyimprove the stability of private financial flows.



TABLE 9.2 Global economic environment and developing countries(annual average percentage changes, except as noted)

2003 2004–08Indicator 1993–97 1998–2002 (estimated) (projected)

World trade 7.9 5.2 4.1 6.5

Developing countriesVolume of exports of goods/services 9.9 6.8 8.6 7.7Terms of trade 0.8 0.2 0.3 –0.4

Fuel exporters 1.7 7.8 2.8 –2.4Nonfuel exporters 0.8 –0.8 –0.2 –0.1

Private capital inflows net (billions of dollars) 234 175 200 —

— Not available. Source: IMF World Economic Outlook database and IMF staff calculations; private capital flows are from World Bank Global Development Financedatabase.



140

120

100

80

60

40

20

US$

bill

ions

3.5

3.0

2.5

2.0

1.5

1.0

0.5

0.0

Perc

ent

Low incomeMiddle income

14

1.1

121

2.2

Sub-SaharanAfrica

Regions Income groups

Europe &Central

Asia


East Asia& Pacific

9

26

37

57

FDI/GDPFDI inflows net

FIGURE 9.2 Low-income countries receive little foreign direct investmentDistribution of FDI to developing countries, by region and income category, 2003

Source: World Bank Global Development Finance database.

US$

bill

ions

350

300

250

200

150

100

50

0

–50

–100

FDI inflowsTotal flows

1991 1993 1995 1997 1999 2001 2003

Debt flowsPortfolio equity flows

FIGURE 9.3 Private capital flows to developing countries are recovering, led by debt flowsNet private capital flows to developing countries, 1990–2003

Source: World Bank Global Development Finance database.

More and better market information, com-bined with greater experience with emergingmarkets, translates into more appropriate pric-ing of investment opportunities and lower riskof contagion.

Developing countries have substantial infra-structure gaps that must be filled if they are toattain the MDGs, as discussed in chapter 6.Although private sector financing is importantto achieving such long-term investment needs,financing for infrastructure has trended down-ward—led by the financial crises of the late1990s and a weakening of the global infra-structure industry. Global capital markets arepotentially well suited to finance large infra-structure requirements, but considerable insti-tutional and regulatory changes in developingcountries will be needed to attract these flows,including a strong institutional framework forthe protection of creditors’ rights. A greateravailability in global markets of political riskinsurance for project financing could assistinfrastructure flows.



TABLE 9.3 Remittances to developing countries, by region, 2001–03(US$ billions)

Percentage increaseRegion 1990 1995 2001 2002 2003 2001–03

East Asia and Pacific 3.0 9.9 13.7 17.0 17.6 28.9Europe and Central Asia 3.2 5.6 10.2 10.3 10.4 1.9Latin America and Caribbean 5.7 12.9 22.9 26.8 29.6 29.3Middle East and North Africa 11.4 10.0 13.2 13.0 13.0 –1.2South Asia 5.6 10.0 13.1 16.9 18.2 38.7Sub-Saharan Africa 1.5 2.7 3.9 4.1 4.1 3.5Total 30.4 51.2 77.1 88.1 93.0 20.7

Source: IMF Balance of Payments Yearbook and World Bank staff estimates.

Workers’ Remittances

In contrast to official aid and private debtflows, workers’ remittances have been a rel-atively stable and growing source of financ-ing for development. Developing countriesreceived an estimated $93 billion in workers’remittances in 2003, an increase of morethan 20 percent from 2001 and more than200 percent from 1990 (table 9.3).7 Follow-ing steady increases during the 1990s, work-ers’ remittances to developing countries nowrank second after FDI as a source of exter-nal finance.8 The especially sharp surge inrecorded remittances since 2001 in partappears to reflect three new developments.Vigorous efforts to curb money launderingand thwart terrorist financing have helpeddivert remittance flows from informal to for-mal channels. Also, increased fears of depor-tation or other legal action prompted manymigrant workers to remit a larger share oftheir savings to their home country. In addi-tion, the growing importance of remittanceshas resulted in better reporting of data inmany developing countries.

Remittances share some characteristics withcapital flows but are also distinguished byimportant differences (box 9.1). The mainsource of remittances continues to be theUnited States, where the rise in remittancescoincided with the economic boom of the1990s and the liberalization of temporarymigration (especially in the technology sector).9

Latin America and the Caribbean continued tobe the top recipient of remittances, absorbingnearly a third of total flows to developing coun-tries.10 Remittances to South Asia rose by morethan one-third in 2001–03; Pakistan accountedfor half of the increase. At about $4 billion,remittances to Sub-Saharan Africa were rela-tively flat. The region receives only a small frac-tion, under 5 percent, of total remittances.

The financial infrastructure supportingremittances can impose high costs on remit-ting funds. These costs can be as large as 20percent of the remitted amount. It is estimatedthat reducing transaction fees by 5 to 10 per-centage points could yield an additional $5billion to $9 billion in annual remittance flowsto developing countries. Policies that couldhave a positive impact on remittance flowsinclude improving competition among moneytransfer banks, increasing migrant workers’access to banking services in source countriesand households’ access to such services inrecipient countries, and removing exchangerestrictions on remittances.

Strengthening the Environment for Emerging Markets

Since the Asian crisis of 1997–98, many stepshave been taken to strengthen the global pol-icy environment and financial architecture tomake financial crises less likely in emergingmarket countries—and to resolve themquickly when they do occur.

Rapid progress has been made in promot-ing the use of CACs in international sovereignbond issuances.11 Following the first Mexicanissue containing CACs in New York in March2003, there has been a clear shift toward theiruse in international sovereign bonds. Belize,Brazil, Guatemala, the Republic of Korea, thePhilippines, South Africa, and Turkey havefollowed Mexico in issuing sovereign bondswith CACs under New York law. Many ofthese issues were heavily oversubscribed andshowed little evidence of carrying a premium

associated with the use of CACs. Collectiveaction clauses also were included in bondsissued by Uruguay in the context of the coun-try’s recent debt exchange. Several developedcountries have taken steps to introduce CACsin their international sovereign bonds.

Discussions have continued on a voluntaryCode of Conduct to encourage uniformadoption of best practices in sovereign debtrestructuring and so speed up resolution ofsovereign debt crises. These discussions bythe international community on the potential



Remittances differ from private capital flows in several ways. Remittances are unilateral transfersthat do not have to be paid back, whereas capital flows create liabilities that must result in outwardflows—of income, interest, or principal. Remittances are sent by emigrants to their families; privatecapital flows are bound by no such ties. Because the motivation for remittances is largely altruis-tic—to help recipients pay for education, medical care, housing, food, or clothing—they tend to bestable over time and may even rise in times of economic difficulty in the recipient country. Otherprivate capital flows, driven by expectations of financial return, tend to be more volatile, rising andfalling “procyclically” with conditions in the recipient economy. Finally, remittances are made byindividuals, whereas capital flows for the most part involve institutional investors, companies, andgovernments.

Some factors affect remittances and capital flows in a similar manner. Both types of flows arepositively affected by the growth cycle in the source country, although capital flows tend to be moresensitive than remittances. Both are similarly affected by the prevailing investment climate (espe-cially the exchange-rate regime) in the recipient country. During 1996–2000, average remittancereceipts as a share of GDP were 0.5 percent in countries with a higher-than-median level of cor-ruption (as indicated by the International Country Risk Guide index) compared with 1.9 percentin lower-than-median corruption countries. Similarly, remittances were nearly twice as high (1.5percent of GDP) in countries with relatively even income distribution (represented by a lower Giniindex) than in other countries (0.9 percent). Countries that were more open (in terms of trade/GDPratio) or more financially developed (in terms of M2/GDP ratio) also received larger remittances.

It is useful also to distinguish remittances from official aid flows. Sometimes it is argued, forexample in comparing the aid efforts of rich countries, that a smaller aid effort is compensated bya larger amount of outward remittance flows that may result from a more liberal immigration pol-icy. A significant part of official development assistance, like remittances, represents unilateral trans-fers to developing countries. But the similarity between aid and remittances ends there. Aid isfinanced by governments using taxpayers’ money. Remittances, on the other hand, are personalflows, paid by migrants from incomes earned in exchange for providing many essential services tothe host nation—legal migrants also pay taxes, and hence, indirectly also contribute to the govern-ment’s aid effort. Aid is essential for meeting the MDGs in many countries. Sub-Saharan Africareceives about five times more net ODA than remittances. Even if border barriers were sharplyreduced, there is no guarantee that (migrants would come from and) remittances would flow to thecountries that most need financial and technical assistance.

Source: Ratha 2003.

BOX 9.1 Differences between remittances and capital flows

benefits of a voluntary code have highlighteda range of views on the central issues. Themain challenge has been to achieve a balancebetween flexibility (to accommodate diversecountry circumstances) and concreteness (toenhance predictability).

International standards and codes alsoappear to be playing an important role instrengthening the functioning of internationalmarkets. The goal of the standards and codesinitiative launched by the IMF and the WorldBank is to encourage countries to adopt goodpractices, identify potential weaknesses inpolicies and institutions, and enhance trans-parency. The initiative covers 12 areas includ-ing data dissemination, fiscal transparency,and monetary and financial transparency. AnIMF and World Bank report finds that majormarket participants appear to be using infor-mation on countries’ observance of standardsin making investment decisions.12

In response to the challenges created byfinancial market developments, the Basel Com-mittee on Banking Supervision proposed a newcapital adequacy framework. It is expected tobe finalized by mid-2004 and implemented bythe end of 2006.13 The Basel II accord is basedon three “mutually reinforcing pillars”:14

� Minimum capital requirements of at least8 percent of risk-weighted assets, with amore sensitive weighting of the riskiness ofdifferent assets

� More stringent supervisory review of capi-tal adequacy—some banks may be requiredto have a capital ratio above the minimum8 percent level

� Public disclosure—to enable stronger mar-ket discipline through enhanced public dis-closure and transparency.

Basel II is designed to enhance the sound-ness of the banking industry by measuring thetrue financial risks in banks’ portfolios and byaligning capital requirements with thoserisks.15 But even as it strengthens the globalbanking industry, the accord could discour-age lending to developing countries. It mayincrease the cost of capital for some borrow-

ers—especially lower-rated borrowers—inpart by undervaluing diversification (and thusoverstating the risk of lending to developingcountries) and by causing bank lending tomove more tightly with the business cycle.

Notes1. Simple regression analysis using data from

1971 to 2000 indicates that a 1 percent change inreal GDP growth in developed countries is associ-ated with a 0.4 percent change in GDP growth indeveloping countries. The strength of the correla-tion, however, is much weaker for primary com-modity exporters (IMF 2001).

2. The downturn in FDI flows to developingcountries in 2003 largely reflects the sharp declinein investment flows in a small number of middle-income countries that suffered a continuing weak-ening in privatization and cross-border merger andacquisition activities. By contrast, FDI to low-income countries remained stable.

3. The impact of FDI on growth has been diffi-cult to verify empirically. Positive effects may kickin only when the absorptive capacity of the hostcountry is above a certain threshold, measured interms of education levels, capital infrastructure, orsimilar indicators (Borensztein, De Gregorio, andLee 1998).

4. The demand for political risk insurance hasbeen rising. Private risk insurers comprise 50–60percent of the market. National export agenciesand the Multilateral Investment Guarantee Agencyaccount for the rest. Most industrialized countrieshave investment guarantee programs, but thereare differences in the amount of such guaranteesprovided.

5. Debt flows display a geographical dispersionsimilar to that of FDI, with the most going to LatinAmerica and East Asia and the least to Sub-Saha-ran Africa.

6. Most of the increase in private debt flows in2003 has gone to countries in Europe and CentralAsia, Latin America and the Caribbean, and EastAsia and Pacific.

7. Official data underestimate the actual size ofremittance flows. Officials in major fund transferagencies argue, based on the size of the funds thatflow through their system, that unrecorded remit-tances may be larger than recorded remittances.

8. World Bank 2004. For data on remittancesand migration flows, see also Adams 2003. In low-income countries, remittances exceed FDI inflows



(net). The geographical movement (temporary orpermanent) of unskilled migrant workers fromdeveloping countries, where unskilled labor isabundant, to developed countries, where produc-tivity is higher, is likely to benefit both sending andreceiving economies. However, immigration poli-cies in developed countries favor highly skilledworkers. The effects of this bias on the human cap-ital endowment in the home country can be signif-icantly negative, particularly if highly skilledworkers never return (Skeldon 1997). The transferof knowledge from the new immigrants back tothe home country and the incentive effects on edu-cational achievements in the home country onlypartially offset the negative effects of this braindrain (Faini 2003).

9. Remittances from the United States exceeded$31 billion in 2002. Those from Saudi Arabia,which is the second largest source of remittances,were about $16 billion.

10. In Mexico, remittances surpassed FDI in thesecond quarter of 2003. In Ecuador, remittancesare second only to oil exports as a source of exter-nal finance. In Sri Lanka, remittances are largerthan tea exports.

11. Collective action clauses are designed tofacilitate coordination among creditors and lowerthe cost of debt restructuring. Early evidence sug-gests that the benefits of CACs in lowering theexpected cost of restructuring outweigh the costs,such as weakening creditor leverage or raising thelikelihood of a default (Eichengreen, Kletzer, andMody 2003). While CACs address the creditorcoordination problem for individual bond issues,they do not solve the problem of cross-issue cred-itor coordination (the so-called aggregation prob-lem) or the problem of creditor coordination forthe existing stock of debt (the so-called transitionproblem) (Chuhan and Sturzenegger 2004).

12. Based on a survey of large, internationallyactive financial institutions, the report found that58 percent of survey respondents used Reports onthe Observance of Standards and Codes (IMF–World Bank 2003).

13. See also the section on the financial sectorin chapter 4.

14. Basel Committee on Banking Supervision2003.

15. Claessens, Underhill, and Zhang 2003;Heid 2003.



Expanding opportunities for interna-tional trade can be a powerful enginefor spurring growth and accelerating

progress toward poverty reduction and theother MDGs. Reducing trade protection willproduce large benefits, not only for develop-ing countries but for developed countries aswell.

While the actions needed to enhance theability of developing countries to use tradeas an instrument of growth and poverty re-duction are primarily their own, developedcountries can do much to help. Estimates ofthe overall trade restrictiveness of policiesaffecting merchandise trade reveal that someOECD countries effectively discriminateagainst low-income developing countries.Correcting trade-distorting agricultural sup-port policies is particularly important, butthe agenda is broader. In manufacturing, tar-iff peaks and escalation that hurt exportsfrom developing countries are a key issue.Lower barriers to services exports, bothcross-border and those requiring the tempo-rary movement of service suppliers, couldalso produce sizable gains. Liberalization byOECD countries not only would have adirect beneficial effect, but the indirect effectmay be even more important: helping gov-ernments pursue desirable domestic reforms.

Putting the Doha Round back on track isa critical priority. The alternatives, bilateral

and regional agreements, will give rise totrade diversion and discrimination and mostlikely will exclude the more difficult areasof reform—sectors such as agriculture andpolicies such as antidumping. There is hugescope for reducing barriers to access to mar-kets for both goods and services. A refo-cused effort centered on market access couldhelp realize the development promise of theDoha Round.

The effort to dismantle barriers to mar-ket access should be complemented by threeprinciples:

� Enhanced transparency of trade policy.OECD countries continue to use trade pol-icy instruments whose effects are difficult toquantify and compare. An example is theuse of specific tariffs (a tax per unit of quan-tity). These often imply very high ad val-orem equivalents that are not transparent toconsumers. Moreover, high specific tariffsimpose a heavier burden on exporters fromdeveloping countries because they producelower-quality (cheaper) products. Data onkey trade-related policies, especially thoserelating to services trade, are limited andincomplete.

� Increased predictability of trade policy.Uncertainty about the conditions of mar-ket access can be reduced through stron-ger disciplines on the use of contingent


Dismantling Barriers to Trade10

protection and simplification of regula-tory requirements.1

� International agreement on targets fortrade policy to help ensure that trade can,to the fullest extent possible, contribute toachieving the MDGs. Such focal pointscould include: (a) complete elimination byhigh-income OECD countries of tariffs onmanufactured products by a target date;(b) complete elimination of agriculturalexport subsidies and complete decouplingof all domestic agricultural subsidies fromproduction, and reduction of agriculturaltariffs to, say, no more than 10 percent, bya target date; and (c) commitments toensure free cross-border trade in servicesdelivered through telecommunicationsnetworks, complemented by actions to lib-eralize the temporary movement of servicesuppliers from developing countries.

A major question confronting OECDcountries is whether the political commitmentcan be mobilized to pursue a meaningful mar-ket access–driven agenda that removes theexisting bias against poor countries. Theanswer depends in part on the willingness ofindustries in OECD countries to devote therequired resources to support deep reforms inagricultural policies and elimination of tariffpeaks and nontariff barriers that are biasedagainst low-income countries. Equally impor-tant will be credible commitments to ensurethat social policies are in place to assist work-ers and communities to adjust to a more com-petitive environment.

Growth, Poverty Reduction, andthe Doha Development AgendaThe attainment of Goal 8 of the MDGs—a global partnership for development—requires increased effort by the developedcountries to open up market access for goodsand services produced in developing coun-tries. This is particularly the case in agricul-ture, textiles and clothing, and labor-intensiveservices—sectors where developing countriestypically have a comparative advantage. Fur-

ther liberalization of trade can generate sub-stantial increases in real incomes, by up to$500 billion or more if account is taken of thelikely dynamic benefits of greater trade.2

Insofar as such trade expansion is realized, ithas the potential to substantially reducepoverty beyond baseline projections—140million fewer poor people in 2015, a declineof 8 percent (figure 10.1). Such income andpoverty reduction gains are not automatic,however. Much depends on complementarypolicies and the investment response tochanged incentives. For Sub-Saharan Africain particular, analysis suggests that trade cando much to help attain the income povertyMDG, but much more than open access tomarkets will be required. The outcome willdepend greatly on improved trade capacity, afunction of the investment climate. Aid has animportant supportive role to play in helpingcountries enhance their trade capacity.

The 2001 WTO Doha Declaration repre-sented the first time that the needs of devel-oping countries were placed at the top of themultilateral trade agenda. In the context ofthe Doha Development Agenda, and subse-quently in Monterrey in March 2002 at theU.N. Financing for Development Conference,high-income countries reiterated their com-mitments to improve market access for agri-cultural and industrial products, especiallyfor developing countries.3

The trade policy environment throughAugust 2003 benefited from the momentumgenerated in the run-up to the WTO’s CancúnMinisterial Conference in September. TheEuropean Union announced a reform of itsCommon Agricultural Policy (CAP) in June,and agreement was reached in August on theimplementation of the public health provisionsof the WTO’s Trade-Related Intellectual Prop-erty Rights (TRIPs) Agreement. However, theCancún Ministerial failed to move the substan-tive agenda forward. Differences over agri-culture, as well as disagreements in otherareas—including multilateral frameworks forinvestment, competition, government procure-ment, and trade facilitation, the so-called Sin-gapore issues—proved impossible to bridge.



Recent proposals by the United States and theEuropean Union to get the Doha Round backon track, which include the suggestion of aban-doning negotiations on investment and compe-tition policies and seek to break the impasse onagriculture, have improved the medium-termprospects for the Round.

Several risks are associated with continu-ing delays in the multilateral talks. The per-ception of failure can strengthen the hand ofprotectionists. High-profile trade disputes,held in check in order not to disturb the mul-tilateral trade talks, could reemerge and dam-age the trading system, hurting smaller andpoorer countries the most. The Cancún out-come may add momentum to strategies forregional and bilateral trade integration. Thiscould prove to be a growing distraction fromthe Doha Round and alter the balance ofinterests in the multilateral talks. A shifttoward greater discrimination in world trade,while possibly beneficial for countries thatobtain preferred access, does not bode wellfor excluded nations. The majority of theworld’s poor live in countries such as China,India, Pakistan, and Indonesia that are notlikely to be part of regional trade agreements

(RTAs) with the major developed countries.Also, RTAs are unlikely to achieve significantprogress in areas such as agriculture that mat-ter most to the poor. It is important that RTAsminimize discrimination against nonmembersand not divert attention away from the multi-lateral WTO process.

Tariffs and Nontariff BarriersAverage most-favored-nation (MFN) tariffsapplied by developed countries are substan-tially lower than those in developing coun-tries.4 In high-income OECD countries theyaverage around 6 percent (table 10.1). On atrade-weighted basis, these tariffs are lower,averaging around 3 percent. Average col-lected tariffs, which are calculated by divid-ing total tariff revenue by the value of totalimports, are even lower. One explanation forthis difference is the impact of widespreadmembership in preferential trade agreementsand various programs offering trade prefer-ences to developing countries. The value ofthis “applied” average tariff is generally lessthan 2 percent. However, this average masksa significant bias against products of export

D I S M A N T L I N G B A R R I E R S T O T R A D E


FIGURE 10.1 Potential income gains from trade reforms are large and can help reduce poverty

400

350

300

250

200

150

100

50

1997

US$

bill

ions

Developingcountries

High-incomecountries

70

60

50

40

30

20

10

0Sub-

SaharanAfrica

SouthAsia

Middle East& NorthAfrica


Europe &Central

Asia

EastAsia

& Pacific

Mill

ions

Dynamic gainsStatic gains $2 per day$1 per day

Increase in real income in 2015 relative to baseline

Reduction in number of poor in 2015 relative to baseline

Source: World Bank staff simulations.

interest to developing countries and substan-tial discrimination between different tradingpartners for similar types of imports. Indeed,a disproportionate share of total revenues iscollected from low-income countries. Forexample, Mongolians and Norwegians bothpaid the United States about $23 million intariffs in 2002. Mongolia exported $143 mil-lion; Norway $5.2 billion—or 40 times asmuch. In effect, Mongolians pay sixteen centsto sell the United States one dollar’s worth ofsweaters and suits, while the Norwegianspaid half a penny to sell the United States onedollar’s worth of gourmet smoked salmon, jetengine parts, and North Sea crude oil.5

A universally accepted definition of a tar-iff peak does not exist. Instead, two criteriaare widely used to indicate the dispersion oftariffs. The first is the number of tariff linesabove a threshold of 15 percent. The secondis the number of tariff lines above a thresholddefined as three times the national average.Developed countries have a low number oftariff lines that exceed 15 percent. Con-versely, given the dispersion of their tariffprofiles, they tend to have a higher number oflines that satisfy the second criterion. The fre-quency of peaks in the tariffs of developedcountries is problematic for developing coun-tries, since the high tariffs are typically onproducts of export interest to them, especiallymore processed products, causing what iscalled tariff escalation (figure 10.2).

In the Quad countries (Canada, EuropeanUnion, Japan, United States), between 4 and

8 percent of the tariff lines are above 15 per-cent. More than 60 percent of imports subjectto tariff peaks originate in developing coun-tries. While many of these countries benefitfrom tariff preferences, many others do not.

Nontariff barrier (NTB) coverage ratios forhigh-income countries are in the 15 percentrange, similar to that observed for middle-income countries (table 10.2). The use ofNTBs has declined significantly since the late1980s, due in part to WTO agreements torefrain from the use of voluntary exportrestrictions and to gradually phase out aregime of quantitative import restrictionsimposed on textile and apparel products pro-duced in developing countries—the Multi-fiber Arrangement (MFA). Under the WTOAgreement on Textiles and Clothing, quotasare to be phased out progressively over a 10-year period. The next-to-last stage in elimi-nating quotas was implemented on January2002, when a further 15 percent of restric-tive quotas were eliminated. On January 1,2005, all remaining quotas (amounting toabout 80 percent of originally restrictivequotas) are to be eliminated.

Although core NTBs have become lessprevalent in OECD countries, the use of tech-nical product regulations and mandatory stan-dards has been increasing. At least one-third ofall imports into the European Union, Japan,and the United States are subject to mandatoryproduct standards. In most cases the standardsinvolved are applied on a nondiscriminatorybasis, and in principle they also apply to



TABLE 10.1 MFN tariffs, developed countries, 1990 and 2002(percent)

Simple Simple Trade-weighted Applied Appliedaverage average average average tariff, average tariff,

Group or country 1990 2002 2002 1990a 2002a

High-income OECD, of which: 7.0 5.2 2.6 2.4 1.4Canada 8.8 4.2 3.4 2.8 0.9European Union 8.7 4.7 3.1 2.8 1.5Japan 6.3 4.9 2.4 2.4 1.9United States 6.4 5.2 3.1 3.3 1.7

a. Total collected tariff revenues as a share of total value of imports.Source: WTO; IMF; European Commission budget.

domestic production. Thus they are not anal-ogous to core NTBs, although research sug-gests they may have a disproportionate impacton exporters located in developing countries.

Agriculture

The agricultural policies of developed coun-tries affect developing countries by distortingworld prices and restricting access to mar-

kets. There are several methods for measur-ing agricultural support and the distortionsproduced by it. Each has shortcomings andstrengths. The ones most often used are mea-sures calculated by the OECD: the producersupport estimate (PSE) and the total supportestimate (TSE).6 These indicators do not con-sider the economic inefficiencies (distortions)caused by the underlying policies and thusunderestimate their true cost.



Tari

ff r

ates

(per

cent

)

12

10

8

6

4

2

0

IntermediateRaw Processed

United StatesJapan

All products

EuropeanUnion

Canada

12

10

8

6

4

2

0United StatesJapan

Agricultural products

EuropeanUnion

Canada

FIGURE 10.2 Escalating tariff rates discourage development

Source: WTO.

TABLE 10.2 Simple average NTB coverage ratios, 2001(percent)

Share of value of Share of value of Share of tariff lines Share of tariffimports subject to imports subject subject to technical lines subject

Group technical regulations to core NTBs regulations to core NTBs

High-income OECD, of which: 15a 14 12a 14Canada 11 8 2 28European Union 1a 15 1a 23Japan 33 14 29 13United States 34 19 21 14

Country-weighted average:High-income OECD 14 16 9 19

a. These numbers are a lower bound due to underreporting for the EU, reflecting the fact that regulations imposed by individual member statesare not taken into account in the TRAINS database.Source: UNCTAD TRAINS database.

The need for overall measures of agricul-tural support arises because the instrumentsto support agriculture in OECD countriesare manifold and complex. The two majortypes of intervention are market price sup-port—policies that artificially raise domesticprices—and fiscal transfers to farmers. Bor-der protection (such as tariffs) is an exampleof price-supporting policies. Because suchsupport raises prices to both producers andconsumers, it is more trade distorting thanare direct payments to farmers.7

The share of the support that is providedthrough border measures versus direct pay-ments varies substantially across OECD coun-tries and commodities. During 2000–02, theaverage annual TSE in all OECD countrieswas $315 billion, or 47 percent of the value ofOECD’s agricultural production (table 10.3).The peak level of TSE was attained in 1999,when it stood at $356 billion. Compared withthe base period used in the Uruguay Roundnegotiations (1986–88), the TSE has trendedslightly upward in absolute amount, while thePSE has shown little change. The latter aver-aged $235 billion in 2000–02, or 35 percent ofthe value of OECD’s agricultural production.Market price support has fallen from 77 per-cent of the PSE in 1986–88 to 63 percent in2000–02, as many countries have moved to

pay producers directly rather than use bordermeasures. Some, but not all, of this shift indi-cates progress in decoupling payments fromproduction, making support less trade distort-ing. The amount of decoupled payments basedon historical entitlements is still small (only 4.6percent of the total), but in the base period thiscategory of payments was virtually zero.

During the Uruguay Round, disciplines onagricultural trade policy were substantiallystrengthened. A ban was imposed on quanti-tative restrictions, NTBs were converted intotariffs, minimum market access commitmentswere implemented through tariff rate quotas,and export subsidies and an aggregate mea-sure of support (AMS) were made subject to reduction commitments. Despite theseachievements, the PSE and TSE numbersreveal that agricultural protection remainshigh. Moreover, recent policy developmentsin the United States have implied an increasein support to farmers (the 2002 Farm Billincreased total potential payments to farmersby some 70 percent, or almost $75 billionover 10 years). In the European Union, effortsto reform the Common Agricultural Policyare moving toward less trade distortion bydecoupling farm payments from production.However, such decoupling was only partiallyattained, and border protection was not



TABLE 10.3 Agricultural support, 1986–88 and 2000–02

United States European Union Japan OECD

Type of support 1986–88 2000–02 1986–88 2000–02 1986–88 2000–02 1986–88 2000–02

Total support estimate ($ billions) 69 94 110 104 58 60 302 315(% of agricultural production) 44 48 49 46 77 79 52 47of which: Producer support estimate (PSE)

($ billions) 42 47 94 93 49 48 248 235(% of agricultural production) 27 24 42 41 65 63 43 35

Share of PSE from consumers(price support, %) 46 35 86 57 90 90 77 63Share of PSE from taxpayers(fiscal transfers, %) 54 65 14 43 10 10 23 37Per full-time farmer ($) — 19,000 — 16,000 — 23,000 — 11,000

— Not available Source: OECD.

reduced at all (box 10.1). This illustrates theimportance of the WTO and the currentDoha Round in attaining further reductionsin trade-distorting support.

The Uruguay Round Agreement on Agri-culture (URAA) classified domestic supportinto three “boxes”:

� Amber Box. Trade-distorting support mea-sures that are subject to reduction commit-ments, including market price support, aswell as some product- and nonproduct-specific budgetary outlays.

� Blue Box. Payments based on fixed areaand yields, number of heads for livestock,and base level of production. The mecha-nisms are supposed to constrain produc-tion, but the real motivation was a politicalneed to exempt some key European andU.S. policies. The amount of Blue Box sup-

port was included in the baseline AMS, butwas not subject to reduction commitments.

� Green Box. Support measures that are non-or minimally trade distorting and that arenot subject to reduction commitments. Theyinclude general government services (such asresearch, disease control, infrastructure, andfood security stockholding), certain formsof “production-decoupled” income support,structural adjustment assistance, direct pay-ments under environmental programs, andregional assistance programs.

Reduction commitments in the URAA weremade based on the AMS. This corresponds toAmber Box measures, but with exemptions forde minimis support, defined as support notexceeding 5 percent (developed countries) and10 percent (developing countries) of the valueof production of individual products or, in the



On June 26, 2003, European agriculture ministers reached an agreement on reform of the CommonAgricultural Policy, or CAP. This reform constitutes a substantially watered-down version of theinitial proposals. The scope for decoupling subsidies from production is more limited in key sec-tors, cuts in intervention prices are smaller, and the reform will be implemented later than initiallyenvisaged.

The key elements of the agreement are:

� Partial decoupling of subsidies from production will be implemented differently by memberstates. In particular, member states retain the right to link up to 25 and 40 percent of subsidiesto production in the case of cereals and beef, respectively, and are allowed to invoke regionalexceptions if considered necessary to avoid the abandonment of production.

� Progress was made on market policy, although reform is limited. Plans to cut intervention pricesfor cereals have been abandoned, and price cuts in the dairy sector will be phased in over threeto four years.

� Direct payments to large farms will be reduced in order to finance a new rural development pol-icy, broadly in line with the January 2003 proposals.

� Ministers agreed on a mechanism to enforce financial discipline to ensure that the farm budgetceiling set for the period 2007–13 is not overshot.

� The reform will enter into force later than originally proposed. Although most of the reformsare to be implemented in 2005, member states can request a transitional period until 2007.

� Other CAP reforms were not covered and are scheduled for the future. The review of Mediter-ranean Products (olive oil, tobacco, and cotton sectors) started in September 2003. Other sec-tors, in particular sugar and wine, are scheduled for review in 2004.

Source: IMF staff.

BOX 10.1 The EU’s Common Agricultural Policy reform

case of non-product-specific support, thevalue of total agricultural production. Eachindustrial country’s total AMS was supposedto be reduced by 20 percent (and two-thirdsof this amount for developing countries)from a 1986–88 base over a six-year period(1995–2000). Due to the artificially inflatedbaseline levels, countries have in general beenable to exceed their reduction commitments,with the result that most countries in recentyears have been well below their ceiling forAMS. Deep reductions in the ceilings arerequired for the Doha Round to achieve sig-nificant reduction in trade-distorting support.

Trade-distorting agricultural support poli-cies in OECD countries have a significantdetrimental effect on developing countries.For example, OECD protection rates forsugar are frequently above 200 percent, andproducers receive more than double the worldmarket price.8 OECD support to sugar pro-ducers of $6.4 billion a year roughly equalsthe value of developing country exports. U.S.subsidies to cotton growers totaled $3.6 bil-lion in 2001–02, twice the amount of U.S. for-eign aid to Africa. These subsidies depressworld cotton prices by some 10 percent. InWest Africa, where cotton is a critical cashcrop for many small farmers, annual incomelosses for cotton growers are about $250 mil-lion a year.9 In the European Union, producersupport for beef reaches 84 percent of thevalue of domestic production.10

Although the major agricultural exportingeconomies will themselves benefit most fromglobal agricultural reform, the least devel-oped countries also stand to gain signifi-cantly. On average, some 18 percent of thetotal exports of least developed countriescomprise goods that are subsidized in at leastone WTO member country, compared with 3to 4 percent for other countries (table 10.4).However, a similar observation holds forimports—9 percent of all least developedcountry imports involve products that aresubsidized, compared with 3 to 4 percent forother countries. Given that OECD supporttends to depress world prices of the affectedcommodities, reform could entail adjustmentcosts associated with possible increases in theworld price of key food imports. Comple-mentary financial support from the develop-ment community will be needed to assistdeveloping countries that are negativelyaffected by an increase in world prices.

From the perspective of farmers in devel-oping countries, the key issue is the impact ofagricultural support policies on the pricesthey receive (or pay) for their products. Thisdepends primarily on the extent to which out-put is increased and consumption decreasedby policies in other countries. Researchundertaken at the OECD and World Banksuggests that what matters most is marketprice support, that is, market access.11 Theresearch finds that the impact on world pricesof reducing border barriers (tariffs) is likelyto be much larger than the impact of reduc-ing domestic subsidies, by a factor of 5 to10, depending on products and countries.One reason for this difference is that tariffsare often very high for subsidized products,frequently taking the form of nontranspar-ent specific duties. While minimum marketaccess commitments negotiated during theUruguay Round—implemented through tar-iff rate quotas (TRQs)—ensure some access,in many cases the TRQs are small, and theeffect of the tariffs is to support high domes-tic price levels.

This does not imply that domestic supportpolicies should be ignored—far from it.



TABLE 10.4 Trade shares of products affected by agricultural subsidies(1995–98 average, in percent)

Domestic support Export subsidies

Country category Exports Imports Exports Imports

All (143) 3.6 3.7 4.4 4.4Developed countries (23) 3.1 3.3 4.0 3.9Developing countries (90) 4.2 4.2 5.0 5.0Least developed countries (30) 17.8 8.9 16.7 13.1

Note: 1995–98 is the most recent period for which detailed data were reported to the WTO.Numbers in parentheses indicate the number of countries.Source: Hoekman, Ng, and Olarreaga 2004.

Substantial reduction in production-linkedOECD agricultural support policies is impor-tant for developing countries in its own right,but is also critical from a political economyperspective. OECD subsidy reforms are nec-essary to help developing countries pursuetheir own liberalization, an essential step inmaximizing gains. Without action by OECDcountries to decouple support from produc-tion, it is much harder for developing-countrygovernments to pursue and sustain their ownreforms.

Manufactures and Contingent Protection

Trade policies affecting developing-countryexports of nonagricultural products are gen-erally more liberal than policies affecting agri-culture. As noted previously, in the majority ofcases MFN tariffs are low. The exceptions per-tain to certain labor-intensive sectors, such asclothing (apparel), for which import duties aremuch higher—particularly in Canada, Japan,and the United States, where tariff peaks aver-age 16 to 17 percent and significant tariff esca-lation poses an impediment to poor countriestrying to move their products up the valuechain.12 The textiles and clothing sector hasalso been ridden with quantitative restric-tions on trade in recent decades. The 1995WTO Agreement on Textiles and Clothingrequires the abolition of all textile quotas byJanuary 1, 2005. A challenge that will likelyconfront exporters after that date will be thethreat of contingent protection (safeguardsand, especially, antidumping). Antidumpingis a frequently used instrument in developedcountries against labor-intensive manufac-tures exported by developing countries (table10.5), as well as more capital-intensive indus-tries in which developing countries have builta comparative advantage—such as chemicalsand steel. These instruments are also used tomanage imports of agricultural products—examples include imports of catfish andshrimp in the United States—but are primar-ily applied to manufactures.

The existence of the antidumping instru-ment creates substantial uncertainty regard-

ing the conditions of market access facingexporters. Antidumping investigations have achilling effect on imports (they signalimporters to diversify away from targetedsuppliers). This has been of long-standingconcern to East Asian countries in particular,especially China and Japan. China now facesthe highest incidence of investigations and thehighest average level of duties in the UnitedStates (table 10.6).13 A similar conclusionapplies to the other Quad members. ThusChina is also the most often targeted countryin EU antidumping, accounting for some 20percent of all investigations in recent years,with average duties of 40 percent and in spe-cific cases duties above 100 percent.14

Overall Trade Restrictiveness Index

As noted in chapter 4 for developing-countrypolicies, it is not possible to compare the vari-ous trade policy instruments just discussedacross countries. Thus, an overall trade restric-tiveness index (OTRI) was estimated thatincorporates MFN tariffs (both ad valorem andad valorem equivalents of specific tariffs), coreNTBs, and domestic agricultural support.15 Inaddition, in calculating the OTRI vis-à-vis low-income and least developed countries, tradepreferences granted to developing countries



TABLE 10.5 Antidumping investigations by selected OECD members,1995–2002

Against all economies

Number of Initiations per dollar Country/economy antidumping of imports initiating initiations (Index United States = 100)

New Zealand 35 900Australia 142 741Canada 106 171European Union 255 107Japan 2 2United States 279 100

Note: Annual data on a July–June reporting basis.Source: World Bank staff calculations based on notifications to the WTO Antidumping Committee.

(both unilaterally and through major tradeagreements) were included in the estimation.

The simple average OTRI for high-incomeOECD members was 6.3 percent in 2001(table 10.7). Of the Quad economies, theEuropean Union has the highest OTRI, at6.6 percent. As a result of trade preferenceprograms, the OTRI against low-incomecountries is lower than the overall OTRI (3.9percent), and that against least developedcountries is lower still (0.6 percent). Thispattern of lower OTRIs against poorer coun-tries does not apply uniformly. In the case ofCanada, trade barriers discriminate signifi-cantly against low-income countries that donot fall into the least developed category,whereas in the case of the United States, bar-riers are higher both for all low-incomecountries and for least developed countries

than for richer countries. Indeed, the OTRIis highest against least developed countries.Reasons for this include relatively high pro-tection of apparel and other labor-intensive“sensitive” sectors, and the fact that some ofthe countries that face the highest OTRI onexports to the United States do not have sig-nificant preferences for their main exports.Note, however, that U.S. imports from low-income countries, as a share of total importsfrom developing countries, are higher thanfor the European Union.

OTRI estimates for agricultural productsare generally a multiple of those for manufac-tures, reflecting the higher levels of protectionof agriculture (figure 10.3). For high-incomeOECD countries, the OTRI is 25.6 percent foragriculture, compared to 3.6 percent for man-ufacturing. Only Australia and New Zealand



TABLE 10.6 Average tariff imposed in final U.S. antidumping duty determinations(percent)

Developing countries Developed countries

Investigations Total Lower-income Upper income China Japan Other

1979–98 all cases 46 53 30 95 60 311989–98 cases only 58 66 36 116 74 34

Source: Bown, Hoekman, and Ozden 2003.

TABLE 10.7 Overall trade restrictiveness and import shares, high-income OECD and Quad, 2001(percent)

Share of low-income OTRI toward OTRI toward Share of total countries in total low-income least-developed low-income country imports from all

Country OTRI countries countries exports taken developing countries

High-income OECD, 6.3 4.9 4.2 69.6 4.6of which:

Canada 5.6 7.7 0.1 1.2 4.2European Union 6.6 3.9 0.6 21.1 3.2Japan 3.9 1.5 1.1 11.2 10.8United States 4.6 6.3 8.2 20.3 6.4

Source: World Bank staff estimates based on World Integrated Trade Solution and U.N. Comtrade database.

have a lower OTRI in agriculture than inmanufacturing. Norway has the highest agri-culture OTRI among OECD countries (42percent), 25 times greater than its manufac-turing OTRI. Norway is followed by Switzer-land (35 percent) and the European Union(28 percent).16 The high rates for Norwayand Switzerland mainly reflect high ad val-orem equivalents of specific tariffs on agri-cultural imports.17

Figure 10.4 plots, for high-income OECDmembers, the difference between their OTRIsfor trade with low-income countries and theirOTRIs for trade with all countries. A positivenumber indicates that the importing countryimposes higher trade barriers on importsfrom low-income countries than on thosefrom the rest of the world. Several OECDcountries, including the United States, have amore restrictive trade policy vis-à-vis low-income countries relative to their overalltrade policy stance.18

The OTRI estimates are in part a functionof the estimates of the ad valorem equivalents

(AVEs) of the NTBs that are included. Theabove numbers include the AVEs of specifictariffs, core NTBs, and agricultural producersubsidies. They do not include technical prod-uct regulations because of incomplete report-ing of related data by countries. The exclusionof technical regulations in the estimation ofthe AVEs of NTBs is likely to bias downwardthe OTRI against low-income countries. If theexisting, incomplete data on technical regula-tions are included, the OTRI estimates for alltrade do not change much, but those againstlow-income countries increase. In the case ofJapan, the OTRI against low-income coun-tries rises toward the overall level of theOTRI. That is, the effect of any preferenceprograms is eliminated. In the case of theUnited States, the OTRI against lower-incomecountries also rises.

The estimates of the OTRI vis-à-vis low-income countries are also biased downwardby the assumption made in the estimation thatmajor trade preference programs such as theAfrican Growth Opportunity Act (AGOA) of



26

21

16

11

6

1

–4

High-IncomeOECD

NorwaySwitzerlandJapanEuropeanUnion

IcelandCanadaUSANewZealand

Australia

FIGURE 10.3 Protection in agriculture is high—a multiple of that in manufacturingRatio of agriculture OTRI to manufacturing OTRI in high-income OECD countries, 2001


the United States and the European Union’sEverything-but-Arms Initiative (EBA) are fullyeffective (that is, are fully utilized by recipi-ents). Strict rules of origin may substantiallyreduce or even nullify the value of preferencesfor manufactured products. Moreover, part ofthe rents created by preferences, if any, accrueto importers, not the exporters. Taking thesefactors into account will increase the estimatesof the OTRI against low-income countries. Ifit is assumed that utilization rates are only 50percent, that is, only half of imports from leastdeveloped countries enter duty free, the OTRIagainst these countries is roughly the same asthat applying to all trade: around 6 percent,with the United States having the highest bar-riers (12.6 percent). The potential importanceof preferences for least developed countries isillustrated by the fact that without any prefer-ences, they would confront an average OTRIof about 10 percent in high-income OECDcountries—more than two times higher thanthe restrictions currently applied.19

The sensitivity of these results to assump-tions regarding the effectiveness of prefer-ences and to the coverage of NTBs in nationaltrade policy databases points to the need forbetter data collection and reporting. This ismost important for technical regulations,which the foregoing analysis was forced to

ignore for comparability reasons. Recentresearch has shown how important it is forfirms in developing countries to be able to sat-isfy technical product standards.20 A con-certed effort by the relevant internationalagencies (WTO, United Nations Conferencefor Trade and Development, or UNCTAD) toimprove the coverage of data on technicalregulations, ad valorem equivalents of spe-cific tariffs (both bound and applied rates),and (bilateral) utilization of trade preferencesby developing countries is important from aglobal monitoring perspective. The weaknessin both trade policy data and information onservices trade policies and international trans-actions greatly constrains efforts both todetermine the overall trade policy stance andto monitor changes in it over time.

Trade in ServicesData constraints prevent the inclusion of poli-cies affecting trade and investment in servicesin the OTRI. Services are the fastest-growingcomponent of world trade. Developing coun-tries have expanded exports of services nearlyfourfold in the last decade (a faster rate thangoods exports), and increased their share ofthe global marketplace for services to 18 per-cent, up from 14 percent in the early 1990s.



6

4

2

0

–2

–4

–6

Perc

enta

ge p

oint

s

USA

Switzerland

NorwayNew

Zealand

Japan

Iceland

European Union

Canada

Australia

FIGURE 10.4 Is overall trade policy pro development? Mixed pictureDifference between OTRI applied to low-income countries and OTRI for all trade, 2001


In large part this increase reflects growth inbusiness process outsourcing (BPO) services.This activity arises from the outsourcing (andout-location through foreign direct invest-ment) of noncore business processes through-out the value chain of both manufacturingand services industries. Within BPO activities,the more advanced developing countries aremoving from providing only low-end back-office services (such as data entry) to moreintegrated and higher-end service bundles infields such as customer care, human resourcemanagement, and product development. Forexample, Barbados is increasingly prominentin data processing; the Philippines, India, andChina are actively exporting computer soft-ware; and South Africa is exporting telecom-munications services.

Exports of BPO services are an example ofone of four possible ways in which servicescan be traded: cross-border through telecom-munications networks. The other three“modes of supply” involve either the move-ment of the consumer to the location of theservice providers (as with internationaltourism) or the movement of the providers.The latter may be long term and involve for-eign direct investment (“Mode 3” in the ter-minology of the WTO’s General Agreementon Trade in Services, or GATS) or be tempo-rary and involve the movement of individualservice suppliers such as workers (Mode 4).Of these four modes of trading services,Mode 3 (commercial presence through FDI)is currently the largest in world trade,accounting for an estimated $2 trillion, fol-lowed by Mode 1 (cross-border supply, $1trillion), Mode 2 (travel and tourism, $500billion), and Mode 4 ($50 billion).21

The potential gains from successful liberal-ization of trade in services are huge. The U.S.banking industry alone is estimated to havesaved more than $8 billion over the last fouryears, and the cost savings for the world’s top100 financial institutions could be as high as$138 billion annually. Analysis suggests thatthe real income gains associated with procom-petitive reforms that increase the variety andquality of services and lower their costs could

be a multiple of the potential gains resultingfrom liberalization of trade in goods.22

Greater access to developed countries’ mar-kets for the temporary movement of naturalservice providers and commitments to main-tain liberal policies toward cross-border tradethrough telecommunications networks wouldboth be valuable in themselves and assistdeveloping-country governments in pursuingbeneficial domestic reforms. Currently, Mode1 faces few explicit restrictions in OECDcountries. However, the well-publicized cre-ation of jobs in developing countries’ serviceindustries has created concerns that maytranslate into protectionism. Although only asmall fraction of the work force in the devel-oped world is employed in agriculture andtextiles, two areas that have seen significantprotectionist pressure, more than two-thirdsof the work force is employed in the servicessector and many more have services-type jobswithin manufacturing enterprises. The out-location of even a fraction of such jobs couldprovoke powerful protectionist pressures—signs of which are already visible. In severalU.S. states, and more recently at the federallevel, legislation is pending that would restrictoutsourcing on government contracts. InEurope legal norms designed to protect work-ers in outsourcing deals, known as TUPE(Transfer of Undertakings and Protection ofEmployees), as well as recent privacy direc-tives, could have an inhibiting effect on trade.Addressing the question of securing liberalcross-border trade in services should be highon the agenda of the international community.

The same is true for temporary movement ofservice suppliers (Mode 4). Many developingcountries export services that are “embodied”in people. Although much of the associatedmovement of persons is long term, a substan-tial share of cross-border movements involvesservice industry workers who return to theirhome countries at some point. Remittanceflows associated with migration are now amajor source of external funding for develop-ing countries. Remittances received by devel-oping countries are estimated at $93 billionin 2003, up more than 20 percent from 2001



and approximately three times the level of1990. Latin America and the Caribbean, with$30 billion in remittance receipts, accountsfor nearly one-third of the developing-countrytotal.23

There is little doubt that despite the dra-matic developments in technologies for elec-tronic delivery of services (Mode 1), Mode 4will remain important for a range of services.The movement of service-supplying person-nel remains a crucial means of delivery evenfor the Indian software industry. Nearly halfof Indian exports are still supplied throughthe temporary movement of programmers tothe client’s site overseas. Greater freedom forthe temporary movement of service providerswould enable developing countries to providethe labor component of construction, distrib-ution, transport, and a range of other ser-vices. As their populations age and theiraverage levels of training and education rise,developed countries will face an increasingscarcity of moderately and less skilled labor.Given that there is no substitute for humanlabor in some activities (for example, the car-ing occupations, personal services, and arange of professional services), the demandfor Mode 4 is likely to increase over time.

Most countries impose a range of barriersagainst Mode 4–based transactions. Thetemporary movement of service providersinvariably is affected by regulations imposedby multiple authorities, including immigra-tion legislation and labor market policy.Visa formalities are in themselves a signifi-cant obstacle, and the conditions attachedare often used to implement a range ofrestrictions. These include prohibitions andquotas either explicitly or through a varietyof economic needs tests. Wage-parity condi-tions tend to erode the cost advantage of hir-ing foreigners and have the same restrictiveimpact as quotas. Discriminatory treatmentis implemented through a variety of fiscaland regulatory means. Nonrecognition ofprofessional qualifications poses a particularchallenge because of the difficulty in distin-guishing between legitimate and protection-ist denial of recognition.

Temporary movement of service providerspotentially offers a way to realize the gainsfrom trade while averting social and politicalcosts in host countries, as well as avoidingnegative brain drain effects for poor homecountries. Recent research finds that if OECDcountries were to allow temporary access toforeign service providers equal to just anadditional 3 percent of their labor force, thegains to all countries would exceed $150 bil-lion (figure 10.5).24

To date, only limited progress has beenmade in liberalizing international servicestransactions.25 At Cancún, trade in serviceswas not an area of disagreement. The DraftMinisterial Text recognized the progressmade in the negotiations, which so far con-sists of a large number of confidential butreportedly ambitious requests that members(including several developing countries) havemade to each other for greater market access,as well as reportedly disappointing offers ofimproved access submitted by 30 or so mem-bers (including some developing countries).Developing-country engagement in the nego-tiations has been inhibited by these countries’perception that the “assessment” of trade inservices—mandated by the GATS as a condi-tion for the new round of negotiations—hasnot been adequately conducted; that theirdomestic regulatory institutions are ill-equipped to deal with the demands of a liber-alized market; and that there is little prospectof meaningful liberalization in areas in whichthey have an export interest.

For developing countries, what mattersmost in terms of what developed countriescan do on trade in services—as in goods—isimproved market access. This implies thatnegotiating modalities are needed that giveparticular attention to the removal of barriersto trade on those services and modes of sup-ply in which developing countries have anexport interest. Given that the rapid spread ofinformation technology and telecommunica-tions infrastructure is allowing service firmslocated in developing countries to contestmarkets in richer countries, binding commit-ments to impose no restrictions on such trade



would be valuable. Specific commitments toopen markets through liberalization of tem-porary movement of natural service supplierswould also be very valuable, although moredifficult to achieve.

The Singapore Issues and Specialand Differential TreatmentAs tariffs were gradually reduced, govern-ments in high-income countries increasinglysought to cooperate in disciplining the use ofNTBs and trade-distorting domestic policies.They also began to use trade agreements asinstruments for harmonization of domesticpolicies that have an impact on trade. TheAgreement on Trade-Related Aspects of Intel-lectual Property Rights is the major examplein the WTO. At the 1997 WTO Ministerial inSingapore, four new issues were put forwardfor negotiation: trade facilitation, competi-tion policy, investment policy, and trans-parency in government procurement.

Many developing countries have come toquestion the development payoffs associatedwith regulatory harmonization enforcedthrough the WTO.26 The Cancún Ministerialillustrated the difficulty of expanding the cov-erage of the WTO to new regulatory areas.Poor countries expressed concerns that newmultilateral rules might not be in their inter-est, would do little to promote progress onkey issues such as agriculture, and could giverise to additional implementation burdens. Abasic question now confronting the interna-tional development community is how tobuild in flexibility in rules that takes intoaccount substantial differences in prioritiesand capacity across countries. An example ofsuch flexibility is the mid-2003 agreementclarifying that the TRIPS agreements allowlow-income countries without domestic phar-maceutical capacity to license imports ofneeded drugs on a compulsory basis.

A fundamental challenge confronting theWTO is ensuring that its rules support devel-opment. Traditionally, “special and differen-tial treatment” (SDT) has been a means ofincreasing the development relevance of the

trading system, as reflected in preferentialmarket access for developing countries todeveloped-country markets and greater flexi-bility for them to use trade policies than isotherwise permitted by WTO rules. Efforts todate to agree on ways to strengthen and oper-ationalize such provisions, however, havebeen divisive and unsuccessful.

A good understanding of the costs andbenefits of specific proposals and rulesrequires engagement by developing-countrystakeholders—a big challenge in many coun-tries, given prevailing capacity constraints.One way of recognizing these constraintswould be through SDT provisions thatallowed for greater differentiation amongdeveloping countries in determining the reachof WTO rules. The need for differentiation isgreatest for disciplines that require significantcomplementary legal, administrative, andinstitutional investments or capacity to imple-ment, and for those that may give rise to largenet transfers from low-income developingcountries (as under the TRIPS agreement).The basic rationale for differentiation is that



US$

bill

ions

100

80

60

40

20

Developingcountries

Developedcountries

Skilled labor Unskilled labor

FIGURE 10.5 Potential gains from liberalization of services, especiallymigration, are largeWelfare gains from a 3 percent increase in developedcountries’ temporary labor quota

Source: Walmsley and Winters 2002.

certain agreements or rules simply may not beimmediate development priorities or mayrequire that other conditions be satisfiedbefore implementation can be beneficial. Thespecifics of a new approach to SDT requireconsiderable further thought and discussion.What matters is that such a debate be initi-ated. That would help make “developmentrelevance” more than a slogan.

Trade Capacity, TechnologyDevelopment, and Trade-Related AssistanceA major constraint limiting export growth inmany low-income countries is a lack of sup-ply capacity and a high-cost business envi-ronment. Firms in these countries may alsofind it difficult to deal with ever more strin-

gent regulatory requirements, such as healthand safety standards, that apply in exportmarkets. Development assistance can play animportant role in helping to build the capac-ity needed to benefit from increased tradeand better access to markets. Commitmentsfor trade-related technical assistance andcapacity building have been rising and wereequivalent to 4.8 percent of total officialdevelopment assistance in 2001–02. Despitethis expanding “trade-for-aid,” such supportneeds to be substantially strengthened.

Aid can help address trade-related policy,institutional, and investment priorities; facil-itate adjustment to a reduction in trade pref-erences following further nondiscriminatorytrade liberalization; and assist in dealing withthe potential detrimental effects of a signifi-cant increase in world food prices should that



Given the typically small domestic markets in the poorest countries, international trade can play akey role in achieving sustained growth and poverty reduction. The Integrated Framework for Trade-Related Technical Assistance to Least-Developed Countries (IF) is an international initiative com-bining the efforts of the IMF, ITC, UNCTAD, UNDP, World Bank, and WTO, in partnership withbilateral donors and recipient countries, to identify and respond to the trade development needs ofthe LDCs. The main elements of the IF process are: (a) preparation of a comprehensive diagnostictrade integration study, which documents constraints to trade in both the domestic economy andoverseas and develops a plan of action to address the major constraints; (b) integration of the actionplan into the country’s poverty reduction strategy program; and (c) facilitation of interactionbetween the agencies, donors, and recipient governments in the implementation of the recommen-dations through technical assistance, capacity building, and investments.

Although specific conditions vary across countries, a clear message has emerged from the 13diagnostic studies completed as of the end of 2003 (for Burundi, Cambodia, Djibouti, Ethiopia,Guinea, Lesotho, Madagascar, Malawi, Mali, Mauritania, Nepal, Senegal, and Yemen): success inreforming formal trade policies is not sufficient to ensure that trade can play its role in stimulatinggrowth and poverty reduction. A range of barriers at home (including weak institutions, poor gov-ernance, and deficient infrastructure) and abroad (including trade barriers in key markets and insta-bility in neighboring countries) constrain the ability of these countries to contest internationalmarkets. As long as these constraints persist, the positive impact on the poor of the trade liberal-ization that these countries have undertaken will remain muted.

Improving domestic market integration is essential if the potential benefits of trade are to beextended to the poorer segments of the population. The studies of Malawi and Nepal show thatreduction of domestic transportation costs would bring strong welfare gains for farming house-holds. Among these households, the poorest would benefit disproportionately because transporta-tion costs make up a larger percentage of their expenditures. Case studies in Ethiopia and Guinea

BOX 10.2 Lessons from Integrated Framework diagnostic trade integration studies

materialize. At the September 2003 WTOMinisterial meeting, the IMF and the WorldBank announced their readiness to help coun-tries benefit from better market access oppor-tunities and to undertake trade-relatedreforms and complementary actions. TheIMF will provide financial support in the con-text of new or existing IMF-supported pro-grams to countries that may face a netnegative impact on their balance of paymentsfrom trade-related adjustments in other coun-tries. The World Bank has been enhancing itslending and technical assistance portfolio inthe trade area and strengthening learning pro-grams and partnerships. New lending fortrade capacity building doubled from $132million in 1998–2000 to $267 million in2001–03. Through the Integrated Frame-work for Trade-Related Technical Assistance,

a program supported jointly by several bi-lateral donors and multilateral agencies,progress is being made in integrating tradepolicy more closely into country developmentstrategies and addressing the related “behind-the-border” agenda (box 10.2).

Facilitation of technology transfer alsowould help enhance the capacities of devel-oping countries to benefit from integrationinto the world economy. Developed countriescould promote technology transfer in severalways. Examples include helping to buildcapacity in intellectual property rights andtechnical regulations and standards, estab-lishing public and public-private researchfacilities, facilitating technology-related ser-vices trade, and supporting training programsin the functioning of modern technology mar-kets. Many governments in high-income



reveal that many of the poor would be left behind if there were overall trade reform but no improve-ments in markets within the country.

Transport costs are a major component of total costs for exporters. The studies show how a mul-titude of negative factors can contribute to extremely high transportation costs. These factors includedeficient infrastructure, poor conditions of the vehicle fleet, lack of security, numerous formal andinformal checkpoints, absence or nonenforcement of transit agreements, and lack of competition. Inpoor countries, especially those that are landlocked, the conjunction of these factors poses dauntingchallenges to domestic producers, seriously undermining the competitiveness of their exports. In Malitransport costs represent on average 30 percent of the value of traded goods (this ratio stands at some18 percent for other Western Africa countries; in developed countries, it is typically around 5 per-cent). In Malawi, transport infrastructure is reasonably well developed, but high prices due to lackof competition in transport services add nearly 50 percent to sugar production costs.

Another major bottleneck clearly identified in the studies is inefficiency in customs clearance.Some countries have managed to reduce clearance times—for example, from an average of three tofour weeks in Senegal to less than one week—but they still have a way to go to match moreadvanced developing countries, where customs clearance is usually achieved within a matter ofhours. Customs procedures generally suffer from a lack of transparency and the prevalence of dis-cretionary measures.

Although all the countries studied are eligible for trade preferences in developed-country mar-kets, the impact of these schemes has been very uneven. For example, while Lesotho has signifi-cantly expanded its garment exports under the U.S. African Growth and Opportunity Act, othercountries, such as Ethiopia and some West African countries, have largely failed to do so. EU pref-erences on sugar and tobacco for exporters in Malawi have been substantial, but the wider eco-nomic impact is less apparent. For most countries trade preferences have done little to mitigate theimpact of the high transportation and transaction costs that they face compared with other tradingcountries and to stimulate trade volumes and diversification into a wider range of products.

Source: World Bank, Integrated Framework diagnostic trade integration studies.

countries offer incentives to firms to locate inor provide technologies to lower-incomeareas within their own countries. One optionis for governments to offer similar incentivesto firms transferring technologies to low-income countries. They could also offer thesame incentives for research and developmentperformed in poor countries as for R&Dconducted at home.

Notes1. A simple rule that would move in this direc-

tion would be to forbid any WTO member fromimposing tariffs at the tariff-line level that are higherthan three times its average tariff. Predictabilitycould also be enhanced through full binding by allWTO members of their tariff lines. For example, ofthe 41 African WTO members, only 9 have boundall their manufacturing tariff lines, while 15 havebound less than 10 percent of their tariffs.

2. These estimates correspond to the pro-poorscenario analyzed in chapter 1 of Global EconomicProspects 2004 (World Bank 2003). They assumethe elimination of agricultural export subsidies anddomestic support, a tariff ceiling of 10 percent foragricultural products and 5 percent for manufac-turing in OECD countries, and a tariff ceiling of 15percent for agricultural products and 10 percent formanufacturing in developing countries.

3. WTO 2001, paragraphs 13 and 16.4. Countries can apply a number of different

tariffs on imports, depending on the country oforigin of those imports. The MFN tariff is thehighest tariff rate that is applicable to all imports,although, as is becoming increasingly common,this rate can be reduced on imports originatingfrom certain countries. Such preferential rates canbe either reciprocal or nonreciprocal in nature.Applied tariffs, whether they are preferential orMFN, also need to be distinguished from boundtariffs, which are negotiated ceilings on the valueof a tariff. These ceilings, or bindings, are notifiedto the WTO and are applicable to the MFN tariffof a WTO Member.

5. WTO 2002.6. The PSE incorporates the effect of trade poli-

cies by measuring the difference between the worldprice of a product at the border of the country andthe domestic producer price. This difference is mul-

tiplied by domestic production to derive the “mar-ket price support” (MPS) for each product. Directpayments to producers—classified according to thebasis on which the payments are made—are addedto the MPS to derive the PSE. To derive the TSE,an estimate of expenditures on General ServicesSupport (spending on collective services, such asresearch, agricultural education, inspection, infra-structure, marketing, and public stockholding) isadded to the aggregate PSE, as are governmenttransfers to consumers (such as food stamps).

7. Thus, for example, a 10 percent increase inproducer price support (a form of direct payment)has only the production-increasing effect, while a10 percent tariff raises the producer price to thesame degree, which has the same production effectbut also reduces consumption. The tariff has a“double-barrel” impact on imports. See, for exam-ple, Anderson (2003); Messerlin (2002); andHoekman, Ng, and Olarreaga (2004) for adetailed discussion.

8. Mitchell 2004.9. Baffes 2003.10. Messerlin 2002.11. Beghin, Roland-Holst, and van der Mens-

brugghe 2002; World Bank 2003; Hoekman, Ng,and Olarreaga 2004.

12. The European Union has no tariffs above15 percent for textiles and clothing. In the UnitedStates and Canada, most tariff peaks affect indus-trial products (over 85 percent), whereas in the EUand Japan most peaks affect agricultural products(91 and 77 percent, respectively). See Hoekman,Ng, and Olarreaga (2002).

13. Bown, Hoekman, and Ozden (2003) notethat the number of cases against developing coun-tries is much higher than their share in U.S.imports. The average duty on developed countries(excluding Japan) was 31 percent, compared with53 percent for developing countries. Similar con-clusions hold for other major users of antidumping.

14. Liu and Van den Bussche 2003.15. OTRIs are computed bilaterally for every

importing country and then aggregated acrosspartner countries. See chapter 4 for a descriptionof the methodology.

16. Note that the economic significance forlow-income countries of very high trade protectionin Norway or Switzerland can be smaller than thatof more moderate levels of protection in muchlarger markets, such as the European Union.



17. Agriculture is more highly protected in bothdeveloped and developing countries. However, rel-ative to manufactures, agricultural protection ishigher in richer countries.

18. The bilateral OTRI estimates indicate thattrade policies in the Quad are most restrictivetoward lower-middle-income countries with GDPper capita in the $1,500 to $5,000 range.

19. Preference utilization by developing coun-tries is generally quite low. According to a WTOdocument (WT/COMTD/LDC/W/31, 29th Sep-tember 2003), the GSP (Generalized System of Pref-erences) utilization rates for preferences granted byCanada, the European Union, Japan, and theUnited States are 61, 31, 46, and 67 percent, respec-tively. In the case of Canada, duty-free access forLDCs was implemented as of January 1, 2003.

20. Otsuki, Wilson, and Sewadeh 2001; Wilsonand Abiola 2003.

21. These numbers are very rough estimatesbased on balance of payments statistics and the

very incomplete information reported on the ser-vices sales by affiliates of foreign-owned firms inhost countries; see Karsenty 2002. Statistics ontrade in services are very weak and do not allow astraightforward mapping between recorded for-eign exchange flows and the four modes of supply.

22. Stern (2002) surveys the literature.23. World Bank 2004.24. This conclusion is conditional on the model

used; relative orders of magnitude may differ usingan alternative global model. See Walmsley andWinters 2002.

25. Schiff and Winters 2003; Mattoo 2003.26. There has been a substantial amount of

research on the economics of the Singapore issues.This research generally concludes that these issuesare important from a development perspective—all relate to the investment climate—but expressesdoubts regarding the appropriateness of dealingwith them in the WTO. See, for example, WorldBank (2003).



While trade policy reform, as dis-cussed in chapter 10, would ex-pand opportunities for growth in

developing countries, adequate foreign aidwould be important in enhancing their capac-ities to exploit those opportunities. That isespecially so for low-income countries, forwhich aid is the largest source of externalfinancing. The Monterrey Consensus recog-nized the need for substantial additionalexternal resources if low-income countriesare to achieve the MDGs, and donors havecommitted themselves to a 7 percent annualincrease in real terms up to 2006. If thesecommitments are realized, official develop-ment assistance would rise by about $18.5billion by 2006 from the 2002 level of about$58 billion. While encouraging, the indicatedincrease in aid is well short of what is neededand can be productively used by developingcountries. Moreover, a much higher propor-tion of aid will need to be provided in theform of cash, and in flexible ways, so that itcan be deployed in accordance with countrypriorities to finance the costs of meeting theMDGs.

Equally important is the need to raise thequality of aid. It is widely accepted that aid ismost effective in countries with sound policiesand institutions. As reviewed in part II of thisreport, developing countries in general areimproving their policies and institutions and

creating conditions that are more conduciveto the effective use of resources—domesticand foreign. Recent evidence supports a grow-ing trend among donors toward aid selectiv-ity, with more aid allocated to countries withbetter policies. Yet there is considerable vari-ation in the selectivity focus of donors, reflect-ing their multiple objectives. The largestdonors are not particularly selective, and thescope for improving the allocation of aid issubstantial.

There is also much scope for increasing theeffectiveness of aid by better aligning aid withrecipients’ development priorities—as mani-fested in country-owned and -led povertyreduction strategies and other developmentframeworks—and by harmonizing donorpolicies and practices. The Rome High-LevelForum of early 2003 has given a welcomeimpetus to the alignment and harmonizationagenda. Consistent follow-through on thatagenda will be important

The Heavily Indebted Poor Countries(HIPC) Initiative has contributed to loweringthe debt burden of these countries and freeingdomestic resources for growth and povertyreduction. Despite these achievements, main-taining debt sustainability is becoming an in-creasingly important issue for these and someother low-income countries, given their needfor large increases in external financing tomeet the MDGs. In some countries, financing


Providing More and Better Aid11

the MDGs with new concessional borrowingcould lead to an unsustainably large debtburden. Additional aid will need to have anappropriate mix of grant financing and newcredits.

Aid Flows: Recent Trends,Prospects, and RequirementsTrends in Aid Volume: A Mixed Picture

After falling substantially in the second halfof the 1990s, aid volumes rose in 2002. NetODA flows, as estimated by the OECD’sDevelopment Assistance Committee, rosefrom $52.3 billion in 2001 to $58.3 billion in2002. The ratio of ODA to donors’ GNI,which fell from 0.34 percent in the early1990s to 0.22 percent in 2001, rose to 0.23percent. Preliminary OECD DAC estimatesindicate that net ODA rose to $68.5 billion in2003; the increase in real terms was muchsmaller—to $60.5 billion (at 2002 prices andexchange rates).1 Aid volumes are set to risefurther as DAC members start delivering on

their Monterrey commitments. If the com-mitments made by donors at or since theMonterrey Conference of March 2002 arerealized, total ODA would increase by about$18.5 billion over the 2002 level (32 percentin real terms), that is, from $58 billion to $77billion (figure 11.1).2 That would raise theODA-to-GNI ratio to 0.29 percent in 2006.The aid effort and new commitments varywidely across donors (table 11.1).

The increase in development assistance isencouraging, but there is concern that a largepart of it may not be directed to financing theincremental costs of meeting the MDGs. Ofthe roughly $6 billion nominal increase in netODA by DAC donors in 2002 (the increase inreal terms was about $4 billion), debt reliefaccounted for $2.9 billion, technical coopera-tion $1.9 billion, and emergency and disasterrelief and food aid $0.7 billion (figure 11.2).Moreover, the increase in bilateral net ODAwas concentrated in a few countries: Serbiaand Montenegro ($1.3 billion), Mozambique($0.9 billion), and Côte d’Ivoire ($0.7 billion),with the additional aid mostly reflecting debt



US$

bill

ions

120

110

100

90

80

70

60

50

40200620042002200019981996199419921990

$18.6 billionMonterreycommitments

Sept. 2003 DevelopmentCommittee Paper> $30 billion

Additional $50 billion ≈ 0.35 percent of donors’ projected GNI in thelatter half of the 2000s, roughly the same as in the early 1990s.

FIGURE 11.1 Aid is rising but is well short of what is neededNet ODA from DAC donors

Note: The Development Committee paper estimated, on a conservative basis, that developing countries could effectively absorb at least $30 bil-lion in additional aid annually (Development Committee 2003b). Source: OECD 2004a; Development Committee 2003b.

forgiveness; and in frontline states in the waron terror, such as Afghanistan ($0.7 billion).Looking ahead, there is some concern thatadditional aid flows could be significantlyinfluenced by donors’ strategic agendas, suchas the war on terrorism and conflict andreconstruction in Afghanistan and Iraq.3 Itwill be important to ensure that such strategicobjectives do not crowd out development aid.

Non-DAC donors’ ODA, which has aver-aged around 2 percent of DAC ODA inrecent years, was also up in 2002. At $3.2billion, these flows were at their highest levelin more than 10 years. The increase was ledby Arab aid, especially development assis-tance from Saudi Arabia, which rose by morethan 400 percent (mainly humanitarianassistance to neighboring countries and anincrease in development lending). In the1990s, Arab aid fell as oil prices fell, but ithas rebounded in tandem with the recentincrease in oil prices. The favorable trend innon-DAC aid appears to reflect some of thesame factors motivating DAC donors—arenewed focus on development spurred bythe MDGs and the Monterrey Compact, aswell as strategic considerations.

The rise in ODA has occurred in an envi-ronment of lower overall flows to developingcountries. Official flows to these countrieshave declined sharply, mostly due to a fall innet nonconcessional lending from multilateral

P R O V I D I N G M O R E A N D B E T T E R A I D


TABLE 11.1 Actual ODA and post-Monterrey commitments (DAC donors), 2002 and 2006

Net ODA ODA/GNI Net ODA ODA/GNI Real change in ODA in 2006 in 2002 in 2002 In 2006 in 2006 compared with 2002

Donor US$ billions Percent 2002 US$ billions Percent 2002 US$ billions Percent

Five largest donors (by aid effort)Denmark 1.64 0.96 1.53 0.83 –0.11 –7Norway 1.70 0.89 2.07 1.00 0.37 22Sweden 1.99 0.83 2.25 0.87 0.26 13Netherlands 3.34 0.81 3.57 0.80 0.23 7Luxembourg 0.15 0.77 0.21 1.00 0.06 41

Five largest donors (by amount)United States 13.29 0.13 19.54 0.17 6.23 47Japan 9.28 0.23 10.50 0.26 1.22 13France 5.49 0.38 7.38 0.47 1.89 34Germany 5.32 0.27 7.10 0.33 1.78 33United Kingdom 4.92 0.31 6.91 0.40 1.98 40

Total (DAC Members) 58.27 0.23 76.85 0.29 18.58 32Total (EU Members) 29.95 0.35 39.63 0.42 9.68 32

Note: Data for 2006 are projections based on announced commitments.Source: OECD 2004a.

6

5

4

3

2

1Debt relief:$2.9 billion

Technical cooperation:$1.9 billion

Emergency anddisaster reliefand food aid:

$0.7 billion

Other bilateral: $0.3 billion

Multilateral:$0.2 billion

Total increase (2001–02) = $6 billion

FIGURE 11.2 The increase in ODA in 2002 was concentrated in special-purpose grantsBreakdown of the $6 billion nominal increase in net ODAby DAC donors in 2002

Source: OECD DAC database.

sources. Less nonconcessional borrowing fromthe IMF, repayments on past crisis financingpackages, and substantial prepayments ofloans to the World Bank are behind thisdecline. Concessional lending by multilateralshas also edged lower. As reviewed in chapter 9,private flows have begun to recover, but for-eign direct investment to developing countrieshas declined.

At about $9 billion, private grants by non-governmental organizations (using their ownresources) are becoming an important sourceof finance for ODA recipients.4 Private grantshave expanded rapidly, experiencing a near70 percent increase during 1997–2002, andnow account for 0.03 percent of DAC mem-bers’ GNI. The growth in private donor fundsreflects the success that NGOs have had ininfluencing a broad array of developmentissues, from reducing debt burdens of thepoorest countries to environmental, health,and governance issues.5

Aid Commitments Lag Countries’Absorptive Capacity and Needs

While the international community generallyagrees on the need for substantially moreresources if the MDGs are to be realized, thereis a concern that countries may not be able toeffectively absorb large increases in resourcetransfers. One reason is that large inflows ofaid could complicate macroeconomic man-agement because of the well-known “Dutchdisease” phenomenon.6 Another reason is thataid flows could undermine government insti-tutions when these institutions are weak. Incountries with weak governance, the elite cancapture resource transfers and keep the bene-fits from flowing to the poor. Large amountsof aid could also overwhelm local capacity touse funds well.7 Complex donor proceduresand practices may add to the problems ofabsorptive capacity.

While the ability of poor countries toproductively absorb increases in externalresources is crucial to aid effectiveness, it isimportant to note that ODA increases in line

with post-Monterrey commitments (see table11.1) would leave the size of ODA relative torecipients’ GNI below the level of the early1990s.8 Recent and ongoing reforms in manydeveloping countries have improved the set-ting for effective use of aid.9 A recent WorldBank study found that countries with rela-tively good policies and institutions couldeffectively utilize a substantial increase inaid to boost progress toward reaching theMDGs.10 It estimated that a doubling offlows could be productively used in thelarge Asian low-income countries, includingBangladesh, India, Indonesia, Pakistan, andVietnam. These countries have prospects ofcontinued good policies, a large gap withrespect to MDG targets, and relatively lowaid dependence. Sub-Saharan African coun-tries with relatively good policies could onaverage effectively absorb increases in aid onthe order of 60 percent. Of course, differencesacross countries are substantial. The studyconcluded that, as a conservative estimate, aninitial increment of at least $30 billion annu-ally could be effectively used by recipientcountries. As countries improve their policiesand governance over time and upgrade theircapacities, the amount of aid that can be usedeffectively would rise into the range of $50billion plus per year, the amount that mostestimates suggest is likely to be necessary tosupport adequate progress toward the MDGs(box 11.1).

Modalities for Mobilizing Additional Aid

Apart from efforts to increase the amount ofaid, the time frame of the MDGs—to 2015—has encouraged consideration of ways to“frontload” future aid commitments. Themost fully developed of such approaches isthe United Kingdom’s proposal for an Inter-national Finance Facility (IFF). As a way tofrontload significant aid flows toward achiev-ing the MDGs, the IFF appears technicallyfeasible—legally binding donor pledges couldbe used to back AAA-rated market borrow-ings by a treasury vehicle. Key requirements





The costs of achieving the MDGs are hard to estimate. There are several reasons for this. For one thing, putting aprice tag on achieving these goals requires distinguishing between average and marginal cost. In education, for exam-ple, the marginal cost of enrolling a child could be higher than the average cost because children who are not in schoolmight be harder to induce to attend school, or they may be in more scattered populations. For another thing, progresson one goal contributes to progress on other goals. Thus, safe drinking water and good sanitation promote betterhealth. The interdependence of goals implies that costing each goal separately could result in double counting. Thereare multiple determinants for each goal, and they cut across many sectors. Measuring the multisectoral dimensionsis difficult. The task of measuring the costs of attaining the MDGs is also complicated by the fact that the effective-ness of additional expenditure depends on appropriate changes in policies and institutions.

Studies on costing the MDGs fall into two broad categories: global costing exercises and country-level estimates.Global elasticities and average costs guide the former, while the latter use country-level information to scale up tothe global level. Neither type of study effectively incorporates the multisectoral dimensions of MDGs.

Global costing exercises. The two well-known global studies are a report prepared for the United Nations by theHigh-Level Panel on Financing for Development led by Ernesto Zedillo (United Nations 2001) and a study at theWorld Bank by Devarajan, Miller, and Swanson (2002). Both studies assume that developing countries implementnecessary reforms. The U.N. report estimated that roughly $50 billion a year in additional ODA would be requiredto achieve the MDGs. The cost of halving the proportion of people living in extreme poverty was estimated at $20billion more in aid; and an additional $30 billion in ODA was estimated for the service goals. The report cautionedabout the difficulties of estimation and noted that a firmer, more comprehensive estimate would need to be based onindividual country estimates. Devarajan, Miller, and Swanson used two different approaches. One approach esti-mated the MDG resource needs by calculating the economic growth rate, and in turn the investment, required toachieve the goals. The second approach separately estimated the costs of achieving the individual goals. Bothapproaches yield comparable estimates—$40 billion to $60 billion a year in additional ODA. Another World Bankstudy (Development Committee 2003a) estimated that $30 billion in additional aid would be required to achieveselected targets within the service delivery MDGs. The study noted that this estimate did not include certain costs,notably that of complementary infrastructure.

Country-level costing. A World Bank study prepared for the September 2003 Development Committee meeting(Development Committee 2003b) estimated how much additional aid recipient countries could productively use toaccelerate progress toward the MDGs. The approach linked aid requirements to countries’ initial conditions, thesoundness of their policies, and their capacity to make effective use of additional resources. The study analyzed 18low-income countries with relatively good policies. Reflecting the different country characteristics, estimates of theadditional aid that countries could effectively utilize vary greatly, from a range of 20 to 100 percent or more. Basedon this analysis, the study concluded that developing countries could effectively absorb at least $30 billion in addi-tional aid annually—a conservative estimate that does not fully take into account all costs, especially the costs ofinfrastructure investments needed to accelerate growth and improve service delivery and the costs of global publicgoods (such as combating infectious diseases and addressing environmental concerns).

A study by the United Nations Millennium Project (2004) took a different approach. It asked what it would costto achieve the MDGs, with the cost estimate relatively unconstrained by absorptive capacity limitations of recipientcountries. The study identified the full range of inputs and actions considered necessary to reach the goals, includingcosting detailed interventions in 14 sectors or categories. Preliminary results for five countries—Bangladesh, Cam-bodia, Ghana, Tanzania, and Uganda—for 2005–15 show that these countries would need substantial increases inexternal development assistance, with their requirements estimated at $44 to $57 per capita per year, which trans-lates into a range of $7.5 billion annually for Bangladesh to about $1 billion annually for Cambodia. Scaling up fromthese country-level results, the study concluded the total ODA would need to rise substantially from current levels,but the total requirements would be less than the ODA target of 0.7 percent of rich countries’ national income.

BOX 11.1 Estimating the cost of the MDGs

for the IFF to be feasible in practice wouldinclude breadth and strength of political back-ing for the obligations being undertaken, andresolution of issues on the treatment of pledgesfrom a fiscal viewpoint by national authorities.Consideration could also be given to alterna-tive structures based on the same principles.

Recent discussions of official financing fordevelopment have broadened the range ofmodalities under consideration beyond tradi-tional aid. As a result, there is a need for moresystematic analysis and comparison of a menuof approaches that might provide additional-ity, greater efficiency of aid, or both. Thewhole spectrum of concessionality needs to beexamined from three points of view. First isthe case for extending concessionality beyondits traditional country-defined boundaries tofocus on MDG investments within countriesnot usually eligible for concessional lending.Second is the use of tailored concessionalityto make more flexible financing available topoor countries, particularly when they facedebt distress or exogenous shocks. Third iswhether added concessionality is an appropri-ate means to address global and regionalexternalities such as public health. Work isunder way on this menu of issues, from dif-ferent perspectives, within the World Bankand the IMF. As this work progresses over thecoming months, more defined options andproposals will emerge.11

The Changing Composition of AidRising Share of Technical Cooperation

In addition to adequacy of amount, aid needsto be provided in forms that are responsive torecipients’ needs. A significant feature ofchange in the composition of aid has been therising share of technical cooperation andother special-purpose assistance. Technicalcooperation now accounts for nearly a thirdof gross bilateral ODA. Technical coopera-tion is of course useful, as it transfers much-needed skills and knowledge. Often, however,it is associated with high-cost expatriate con-

sultants and with absorbing local expertise inprojects without building sustainable localcapacity. Although technical assistance is notincluded in official statistics on aid tying, itcan essentially be considered as a form of tiedaid. Consequently, estimates of progress onaid untying should be interpreted cautiously.According to DAC estimates, the proportionof untied aid has increased from 60 to 80 per-cent in the past two decades (for those coun-tries that report). These numbers excludetechnical cooperation, administrative costs,and ODA by members that do not report thetying status of their aid.12

Tying is costly for recipients. A recent studyof several projects in Ghana found that 11 to25 percent of resources could have been savedif aid had been untied.13 An earlier study foundthe costs to be in a similar range: 15 to 30 per-cent.14 With DAC members agreeing in 2001to untie all financial aid to the least developedcountries, more progress on untying aid islikely. Efforts should also be made to evaluatethe effectiveness of technical cooperation.

Need for More Flexible Forms of Aid

More aid needs to be provided in forms thatcan finance the incremental costs of achiev-ing the MDGs. Of the total ODA, assistanceprovided through bilateral channels accountsfor about 70 percent. Of the total bilateralassistance, currently only about 30 percent isavailable for project and program expendi-tures in recipient countries.15 The rest is allo-cated to special-purpose grants such astechnical cooperation, debt relief, emergencyand disaster relief, food aid, and cost of aidadministration. The proportion of bilateralODA net of special-purpose grants has beendeclining steadily. The current proportion of30 percent compares with a high of about 60percent in the early 1980s (figure 11.3). Thistrend needs to be reversed. A much higherproportion of additional aid will need to beprovided in more fungible resources—cashtransfers—that countries can deploy in linewith their priorities to finance the costs ofmeeting the MDGs.



Aid delivery needs to be more flexible, withaid provided in forms that reflect recipients’needs and circumstances.16 As countries’ pub-lic financial management systems improve,more aid can be provided through budget sup-port or sectorwide programs. Depending onrecipients’ circumstances, part of the assistanceprovided through these means should be avail-able to finance recurrent costs—much of theincrease in costs associated with the improve-ment of education and health programs thatare crucial to progress toward the MDGs isrecurrent. Such flexible provision of aid wouldbe appropriate in countries with sound budgetframeworks, for example, Uganda and BurkinaFaso. Where budget frameworks are at an earlystage of reform, as for example in Madagascar,financing for recurrent expenditures could beprovided through investment projects.

Quality and Effectiveness of AidThe achievement of the MDGs requires bothmore and better aid. Issues relating to the qual-ity and effectiveness of aid are receiving increas-ing attention. There is broad recognition thatby complementing improved policies in recipi-ent countries, better donor policies and prac-tices can contribute significantly to enhancingthe impact of aid. The agenda spans several keydimensions of the quality of aid: selectivity(allocation of aid across countries), alignment(consistency with recipients’ development pri-orities), and harmonization (streamlining andcoordination of donor procedures and prac-tices and harmonizing them with those of therecipients). Related aspects include the reduc-tion of aid volatility and fragmentation.

Aid Effectiveness and Selectivity

Aid has a greater impact on growth if it is tar-geted to countries with better policies andinstitutions (figure 11.4). The impact dependson the interaction between aid and policy.With bad policies, aid has little effect ongrowth; if anything, the relationship may benegative. With good policies, the effect of aidis positive and is stronger if aid is given in suf-

ficiently large amounts. There is now generalacceptance that policies and institutions mat-ter for aid effectiveness, which is reflected ina trend toward increased selectivity in theallocation of aid.17



70

60

50

40

30

20

Perc

ent o

f tot

al b

ilate

ral O

DA

1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002

FIGURE 11.3 The proportion of aid provided in cash and more flexibleforms should be rising, not fallingTotal bilateral ODA minus special-purpose grants

Source: OECD DAC database.

FIGURE 11.4 Institutions and policies matter for aid effectivenessThe relationship between the quantity of aid, the qualityof policies, and GDP growth

1.5

1

0.5

0

–0.5

–1

–1.5

GDP

grow

th (p

erce

nt)

Medium

Volume of aid

Low High

Bad policy

Medium policyGood policy

Source: Dollar and Levin 2004.

Recent research finds a significant rela-tionship between donor assistance and thequality of recipients’ policies.18 About three-fourths of the donors exhibit a positive rela-tionship between their aid allocations and thesoundness of recipients’ policies and institu-tions. Moreover, donors today are in generalmore selective than a decade ago. Selectivityhas also increased with respect to poverty,with more aid being allocated to poorer coun-tries, given the quality of policies. However,variation among donors is wide. On average,multilateral assistance is much more selectivethan bilateral assistance. Among bilateraldonors, the largest ones (in terms of theabsolute amount of aid provided) are lessselective (box 11.2). This means that the typ-ical donor is more selective than the typicalaid dollar. Actions being taken by some ofthese donors, such as the establishment by theUnited States of the Millennium ChallengeAccount (MCA), are expected to increase theselectivity of their aid. The MCA aims to

improve aid effectiveness by tying increasedassistance to performance.

The more selective donors also providemuch more aid per donor country’s citizen(figure 11.5). The top providers of assistanceper donor country citizen to low-incomecountries with better policies and institutionsduring 1999–2002 were Denmark (providing$78 per Danish citizen a year on average dur-ing this period), Luxembourg ($53), Norway($48), and the Netherlands ($39).19 Corre-sponding figures for some of the largestdonors in absolute size are quite low (forexample, $4.5 for the United States), with theresult that aid provided to low-income coun-tries with good policies and institutions perdonor country citizen averaged a relativelymodest $10.3 for DAC members as a wholeduring this period.

Even as donors move to target aid better,they need to take account of the special needsof the conflict-affected and other low-incomecountries under stress (LICUS). The challengeis to balance issues of weak policies and insti-tutions with the need to maintain criticalengagement. Appropriately timed and directedaid, sensitive to local efforts to rebuild and toinstitutional capacity constraints, can play avery useful role. There is no set pattern of aidinterventions targeted to LICUS countries.They can include meeting humanitarian needs,improving basic service delivery, and prevent-ing or responding to conflict. There is also widevariation among the LICUS in their capacity toabsorb aid: for example, postconflict countriesmay be better able to absorb larger amounts ofaid effectively than are other LICUS.20 Bosnia,Rwanda, and Timor Leste have demonstratedthe potential for aid-supported recovery andprogress toward the MDGs. The new DACinitiative on difficult partnerships should helpto shape more effective ways of providing aidto help countries under stress. 21

Recent research shows that well-timed aidcan also be quite productive followingadverse exogenous shocks, helping to limitthe diversion of development resources intoshort-term relief efforts.22 Low-income coun-tries are particularly vulnerable to natural



80

70

60

50

40

30

20

10

0

Coun

trie

s w

ith

stro

ng p

olic

ies/

inst

itut

ions

0 10 20

Denmark

Luxembourg

Netherlands

Norway

Sweden

France

Japan

30Countries with weak policies/institutions

40 50 60 70 80

GermanyGermanyDAC averageDAC average

United Kingdom

United States

FIGURE 11.5 More selective donors provide more aid per capita tocountries with stronger policies and institutionsAid per bilateral donor country’s capita to low-incomecountries, average for 1999–2002 ($ per capita)

Source: Dollar and Levin 2004.


Dollar and Levin (2004) develop two indexes to measure the selectivity of aid. A policy selectivityindex measures the elasticity of aid with respect to the quality of recipients’ policies and institutions(controlling for recipients’ per capita income and population). A poverty selectivity index measuresthe elasticity of aid with respect to the recipients’ per capita income level (controlling for their pol-icy and institutional quality and population). An overall index is computed as an average of thetwo—average of the policy elasticity and the (negative of) the poverty elasticity—to measure howselective donors are in terms of focusing assistance on low-income countries with better policies andinstitutions.

The results show that in 2002, a 100 percent increase in the quality of recipients’ policies wason average associated with 176 percent more aid. The relationship is much stronger for multilateralassistance (elasticity of 2.57) than for bilateral assistance (elasticity of 0.63). However, policy elas-ticity varies considerably among bilateral donors. The Nordic countries, the Netherlands, and theUnited Kingdom show high policy selectivity (elasticity of 2.5 or higher). Broadly similar results areobtained for poverty elasticity.

Some of the largest donors in absolute amount are less selective. France and the United Statesdo not appear to have been particularly selective along either the policy or poverty dimension. Japanappears to be selective on policy but not on poverty, reducing the overall selectivity of its aid.

The authors test the robustness of their analysis to the choice of the measure for recipients’ poli-cies and institutions. The results reported in the table use the World Bank CPIA as the measure. Usingtwo other measures of institutional quality that are available for a large number of countries—theInternational Country Risk Guide rule-of-law index and the Freedom House democracy index—theauthors find that the rankings of aid selectivity are quite similar. Thus, whichever measure of the qual-ity of policies and institutions is used, the same donors appear to be more or less selective. Theauthors also test the sensitivity of their analysis to different definitions of aid flows—ODA includingloans and debt relief and grants only—and find that they do not significantly alter the results.

It should be noted that the selectivity index captures only one, albeit important, dimension of thequality and effectiveness of aid—its allocation across countries. Other dimensions of the quality of aiddiscussed elsewhere in this chapter include the form and composition of aid, its predictability, andalignment and harmonization. Moreover, the policy selectivity index needs to be interpreted cautiously.Some aid can be effective in situations where policies and institutions are still weak, such as in certainpost-conflict and post-shock situations. This is an area where more research would be useful.

BOX 11.2 Measuring aid selectivity

There is substantial scope for improving the allocation of aidAid selectivity index, 2002

Indicator Policy selectivity Poverty selectivity Overall selectivity

Total aid 1.76* –0.49* 1.12Bilateral aid 0.63 –0.38* 0.50Multilateral aid 2.57* –0.83* 1.70Five largest donors (by amount)

United States 0.66 –0.76* 0.71Japan 1.90 0.01 0.94France –0.07 –0.28 0.10Germany 2.06* –0.47* 1.27United Kingdom 3.66* –1.06* 2.36

Good practice examplesDenmark (bilaterals) 4.77* –1.11* 2.94IDA (multilaterals) 4.23* –4.20* 4.22

* Elasticity different from zero at 10 percent significance level.Note: The last column is an average of the first two. Emergency and disaster relief are excluded from these calculations.Source: Dollar and Levin 2004.

disasters, terms-of-trade shocks, and otheradverse shocks, and those with inadequatecushions of external reserves or fiscalresources are likely to suffer especially severeeffects on growth and poverty. The main pol-icy challenges here are to accelerate the inter-national response to such shocks, increase theamount of assistance, and make more use ofgrants rather than lending for this purpose.23

About a third of all ODA goes to middle-income countries. Many of these countrieseither have met or are on track to meet sev-eral of the MDGs well before 2015. The bulkof the resources needed to achieve the devel-opment goals in these countries will have tocome from domestic sources, and their exter-nal financing needs will be met primarilyfrom private or nonconcessional flows—espe-cially in the upper-middle-income countries.Aid, however, can play an important role inthe middle-income countries: as a catalyst forreform; as a reinforcer of domestic efforts totackle large pockets of poverty (middle-income countries remain home to 280 millionpeople who live on less than $1 a day and 870million who live on less than $2 a day); andas a provider of countercyclical support to

reduce vulnerability to financial shocks andhelp deal with their consequences. Over time,as income levels rise further in these countriesand their domestic resource availabilities andaccess to private capital markets expand,their use of ODA should decline.

Donor Fragmentation

Bilateral aid is highly fragmented, with possi-ble negative implications for aid quality. Theargument against aid fragmentation is that ifeach donor has only a small share in a recip-ient country’s total aid, then the donor’s stakemight be more narrowly focused on the suc-cess of the donor’s own projects and less onthe country’s overall development. Anotherpossible consequence is that if a project’sfixed costs are high and there are returns toscale, then too many individual projects mayimpair efficiency. Fragmentation may alsoimpose high transaction costs on recipients,as more time is taken up by donor require-ments. Donors such as Canada, Finland, theNetherlands, and the United States (throughthe Millennium Challenge Account) areproposing to reverse the trend toward dis-persion by concentrating more aid in fewerrecipient countries.

A donor fragmentation index constructedfrom the shares of individual donors inannual aid disbursements shows that frag-mentation has been rising (figure 11.6).24 Theindex ranges from 0 to 100, with higher val-ues implying more fragmentation. The indexis 0 if one donor accounts for all of a recipi-ent’s aid. The degree of fragmentation rises asthe number of donors rises, as their aid sharesbecome more equal, or both. The issue ofdonor fragmentation is of course more seri-ous in some countries than others. For Tan-zania, for example, the fragmentation indexin 2002 was a high 93 (compared to an aver-age of 67). Egypt, in contrast, has an indexvalue of 62. The degree of fragmentation ismoderately correlated with the size of ODAin recipients’ GNI, suggesting that greater aiddependency is associated with a higher degreeof fragmentation.



75

70

65

60

55

502002199819941990198619821978

FIGURE 11.6 Aid fragmentation is highDonor fragmentation index, 1975–2002

Source: OECD DAC database and World Bank staff estimates.

Volatility in Aid Flows

Volatility in aid flows can put much stress onthe ability of recipient countries to plan anduse resources well and thereby hurt aid effec-tiveness. In many countries, aid tends to bemore volatile than fiscal revenues.25 This isespecially a problem for countries with highaid dependency; these countries tend to expe-rience more volatility and possess the leastoptions for coping with it (figure 11.7).26

Faced with uncertain aid flows, poor coun-tries have, in principle, three options. Theycan try to apply a flexible fiscal framework inwhich taxes and expenditures are adjusted inresponse to aid receipts. They can try tosmooth out aid fluctuations by allowing thelevel of international reserves to fluctuate.And they can use domestic nonmonetaryfinancing to fill the gaps in aid levels. All threeoptions present serious difficulties for suchcountries. Aid shortfalls are mostly absorbedby reductions in spending, and sometimes byincreases in taxes.27

As countries build a track record of policyperformance, their efforts should be supportedwith timely, more predictable, and longer-termaid commitments. Progress toward the MDGsis necessarily a long-term process and requiressustained reform of policies and institutions.Commitment to sustained reforms, especiallyin countries with a higher dependence on aid, isdifficult without a reasonable degree of assur-ance about the aid flows and the circumstancesunder which they are likely to materialize.

Alignment and Harmonization

Alongside improvements in the allocation ofaid across countries, two other factors arekey to enhancing development effectiveness:(a) alignment of aid with partner countrydevelopment strategy and priorities, and (b)harmonization of donor policies and prac-tices around strengthened partner countrysystems. Improving alignment and harmo-nization is now an active item on the agendaof the international community, reflecting thedevelopment lessons of the past decade that

underscore the importance of country own-ership and leadership of the developmentprogram for its effectiveness. There is alsorecognition that the wide variety of donorrequirements and processes for deliveringaid in the past were generating high andunproductive transaction costs for partnercountries and drawing down their limitedinstitutional capacities.

Donors are beginning to anchor their sup-port in country-owned and -led strategies. Inlow-income countries, the Poverty ReductionStrategy Paper (PRSP) provides the centralframework for strategic alignment of inter-national aid with national development pri-orities and for achieving better coherenceand coordination in donor support activities.Finding effective ways of aligning aid with thePRSP and harmonizing donor policies andpractices around it constitute the core of thealignment and harmonization agenda in low-income countries.

The strengthened focus on, and high-levelpolitical commitment to, the alignment and



FIGURE 11.7 Aid flows are typically more volatile than fiscal revenuesin aid-dependent countries

100

80

60

40

20

0

Perc

enta

ge d

istr

ibut

ion

of c

ount

ries

High aid-dependency

15

85

Low aid-dependency

38

62

Aid < fiscal revenues

Relative volatility

Aid > fiscal revenues

Note: High aid-dependency refers to countries with an aid-to-revenue ratio larger than 50percent; low aid-dependency refers to countries with an aid-to-revenue ratio between 10 and50 percent.Source: Bulir and Hamann 2003.

harmonization agenda was evident at theRome High-Level Forum in early 2003. TheForum imparted a strong impetus to thisreform agenda that is now being actively fol-lowed up. A new institutional architecturehas emerged to take this work forward.

A new institutional architecture. The follow-up work to Rome is being coordinated by theOECD DAC Working Party on Aid Effective-ness and Donor Practices established in May2003. The work, carried out in active collab-oration with bilateral and multilateral donorsand partner countries, aims to develop pro-posals for action in several areas important toaid effectiveness. These include alignmentand harmonization, public financial manage-ment, procurement, managing for develop-ment results, and untying of aid.28

Task teams have been set up for each ofthese areas. The work on alignment and har-monization is focused on facilitating theimplementation of the Rome agenda throughdissemination and exchange of experience;tracking progress on more effective aid deliv-ery; and improving existing mechanisms forexercising peer pressure (including using thepeer review mechanism to monitor broaderissues of policy coherence). Work in the areaof public financial management is focused ondeveloping a performance measurementframework for public financial management;increasing the predictability of aid; betterintegrating aid flows in recipient country bud-gets; aligning budget support with PRSPprocesses; and developing an accountingstandard for aid. The work on procurementaims at mainstreaming good procurementpractices and supporting capacity building inpartner countries. The work on results aimsto articulate a core set of principles of goodpractice in managing for development results,drawing on the emerging practices in differ-ent institutions. And the work on untying aidsupports the 2001 DAC agreement to movetoward untying aid to all least developedcountries and to explore how to broadentheir recommendation.

The Working Party plans to report onprogress on its work to the DAC’s Senior-Level Meeting in late 2004 and to the SecondHigh-Level Forum scheduled to be held inMarch 2005 in Paris.

Monitoring progress on alignment and harmo-nization. To enable systematic monitoring ofprogress on the alignment and harmonizationagenda, a preliminary set of indicators has beendrawn up that is now being field-tested. Theindicators cover the key dimensions of owner-ship, alignment, streamlining, and improve-ment of practices (table 11.2). A survey will becarried out in 14 developing countries to cap-ture progress on alignment and harmonizationat the country level. Data collected from thisexercise will be used to report to the SecondHigh-Level Forum in 2005.

Emphasis on country-level implementation.In the effort to improve alignment and har-monization, emphasis is being placed oncountry-level implementation, to test initiativesand assess their effectiveness. This process isincreasingly being led by recipient countries,with the support of local-level donors. A num-ber of countries have stepped forward as par-ticipants in country-level efforts to advance andmainstream harmonization.29 The experienceof Vietnam, one of the frontier harmonizationcountries, offers some concrete examples ofimplementation progress (box 11.3).30

Partner countries and donors are also in-creasingly looking to harmonize and align atthe sector level, through the application ofsectorwide approaches, or SWAps. In SWApsdonors support government programs basedon agreed sector strategies. Wherever possible,they use the country’s systems for procurement,financial management, environmental assess-ment, and monitoring and reporting, and pooltheir funds to provide harmonized support. Inthe health sector, for example, about 20 coun-tries currently receive support through SWAps.

In sum, an encouraging consensus hasemerged in the international community onthe need to better align and harmonize aid,





TABLE 11.2 Framework of indicators on progress on harmonization and alignment (provisional)Rome Commitments on Harmonization and Alignment

Partner countries coordinate development assistance

1. Percentage of partner countries where partners and donorshave agreed on an agenda for greater harmonization.

2. Percentage of partner countries that lead localcoordination processes, including organizing and chairingconsultative groups or roundtables.

3. Comment: The Joint Venture on Public FinancialManagement is currently elaborating a framework on publicfinancial management. A decision will be made on whetherthis might provide the basis for an appropriate indicator.

4. Percentage of partner countries where partners and donorshave agreed to an action plan and have committedresources to build institutional capacity to managedevelopment assistance.

1. Partner countries assume leadership roles in thecoordination of development assistance (*3)

2. Partner countries, when necessary, reform their systemsand procedures and adopt international principles andgood practices (*3)

3. Donors increasingly support partner countries’ capacity tomanage development assistance effectively (*3)

4. Development assistance is increasingly delivered inaccordance with partner countries’ priorities (*5.1)

5. Donors rely increasingly on partner country systems andprocedures (*5)

5. Percentage of donors within a country whose countryassistance strategies are consistent with PRSs or equivalentnational frameworks.

6. Percentage of donors within a country whose developmentassistance is aligned with budget cycle.

7. Percentage of donors within partner country relying on (a) supreme audit institution to audit donor funds and (b) partner countries’ procurement systems.

6. Donors implement common arrangements for planning,managing, and delivering aid (*6)

7. Donors reduce missions, reviews, and reports whereappropriate (*5.2; *5.7)

8. Donors are transparent about their activities (*6)

8. Percentage of donors within a country that rely on acommon conditionality framework consistent with PRSs orequivalent national frameworks.

9. Percentage of SWAps where funding, reporting, andmonitoring are guided by a common framework.

10. Number of agreements on delegated cooperation.11. Number of donor missions per partner country, of which

joint donor missions.12. Number of financial reports per country prepared by the

Ministry of Finance and sector ministries solely for donors.13. Percentage of donors within a partner country that disclose

information on (a) planned aid flows and (b) countryanalytic work.

9. Donors review key policies and procedures to support harmonization (*5.2)

10. Global and regional programs increasingly promote and support harmonized approaches (*5.7; *5.9)

14. Comment: These dimensions will be captured in the peerreview reports and reports to the DAC SLM Level Forum.

Note: Number in parentheses indicates number of the paragraph that refers to the specific commitment in the Rome Declaration. Paragraph 7 of the Rome Declarationon Harmonization called for tracking and refining indicators of progress on harmonization. This list of commitments and the accompanying list of indicators of effectivepartnerships address that undertaking.Source: OECD 2003.

2. A

lignm

ent

4. P

ract

ices

3. S

trea

mlin

ing

1. O

wne

rshi

p

Policies, procedures, practices, and incentives foster harmonization

Donors streamline aid delivery

Donors align aid with partner country priorities and systems

and the main elements of the agenda havecrystallized in the past year. What is needednow is consistent implementation of theseefforts and concerted monitoring to keeptrack of progress and ensure learning andcross-fertilization.

Progress on the HIPC InitiativeFor the heavily indebted poor countries, debtrelief is crucial to create the fiscal space formuch-needed increases in spending to pro-

mote growth and reduce poverty. Overall,substantial progress has been made in theimplementation of the HIPC Initiative. As ofApril 2004, 27 HIPCs—or more than two-thirds of the 38 countries that potentiallyqualify for assistance under the initiative—have reached the decision point and arereceiving debt relief. In net present value(NPV) terms, these 27 countries account for85 percent of the total expected relief for the34 HIPCs for which data are available. Thir-teen HIPCs have also reached the completion



In Vietnam, a pioneer of harmonization, and a European Union harmonization pilot, donors havebeen supporting the implementation of the government’s harmonization priorities, with progress inseveral areas.

At the strategic level. The Vietnamese government has taken the lead to develop a clear harmo-nization program and utilized its poverty reduction strategy for ODA planning and to ensure sec-tor policy coherence.

Budget alignment. The Poverty Reduction Support Credit (PRSC) and other budget support willbe aligned on an interim basis starting in 2004 and regularized in 2005.

Joint analytic work. The Asian Development Bank and the World Bank jointly undertook pro-curement and financial management diagnostics. The IMF, UNDP, Denmark, the Netherlands, theUnited Kingdom, and the World Bank jointly conducted a public expenditure review.

Harmonized financial management. The AsDB, Agence Française de Developpement (AFD),Japan, Kreditanstalt fur Wiederaufbau (KfW), and the World Bank have agreed on a commonreporting format for selected ongoing and new projects; common assessments of implementingagency financial management arrangements; common content, format, and frequency of reportsrequired by the donors; and common auditor acceptability criteria, with a shared list of acceptablefirms, common audit terms of reference, and an agreed follow-up approach to audit findings includ-ing sanctions.

Harmonized procurement. Agreement has been reached among the AsDB, AFD, Japan, KfW,and the World Bank on common bidding documents, common upper-limit thresholds for Interna-tional Competitive Bidding/National Competitive Bidding, common assessment of thresholds forprior/post reviews, and eligibility of dependent state-owned enterprises.

Environmental assessment. Harmonization efforts among the AsDB, AFD, Japan, KfW, and theWorld Bank are moving toward agreement on environmental assessment (EA) coverage, consulta-tion through sharing of EA documentation, mitigation measures, management plans, and EA dis-closure and timing.

Joint portfolio reviews. This is an ongoing collaborative effort by AsDB, Japan Bank for Inter-national Cooperation, and the World Bank, which has more recently been joined by AFD and KfW.

Capacity building. The Like-Minded Donor Group, constituting seven to nine bilateral donors,has teamed up with the government and other donors in a $10 million multidonor trust fund tostrengthen government capacity related to procurement, project management, financial manage-ment, and budget support.

Source: World Bank, Operations Policy and Country Services Vice Presidency.

BOX 11.3 Vietnam’s comprehensive government-led harmonization program

point—when creditors provide the full debtrelief committed at the decision point on anirrevocable basis.

The process of reaching the completionpoint has generally taken longer than earlierenvisaged. The delay is attributable to thechallenges of achieving and maintainingmacroeconomic stability and preparing andimplementing poverty reduction strategies.Most of the countries in this situation,however, either have adopted the necessarymacroeconomic policy changes or are makingefforts to do so. Similarly, while the preparationof PRSPs has taken longer than expected, mostcountries that are in the interim period havefinalized their strategies and are not expected tobe constrained by the requirement for one-yearsatisfactory implementation. Progress has alsobeen made in most cases on social and struc-tural completion point triggers.

Challenges remain for countries that havenot yet reached the decision point. Domesticconflict, unsettled postconflict situations, andprotracted arrears continue to impede coun-tries from reaching the decision point. Theissue thus arises as to whether a number ofcountries could be left behind when the HIPCInitiative expires. The initiative is open to alleligible countries that establish a performance

track record leading to the decision pointbefore the end of 2004, when the sunsetclause of the HIPC Initiative will take effect.Staffs of the Bank and Fund will presentoptions for extension of the sunset clause totheir respective Boards by September 2004.

HIPC relief is projected to substantiallylower debt stocks and debt service ratios formost HIPCs that have reached the decisionpoint. NPV debt stocks in the 27 HIPCs thatreached the decision point by July 2003 areprojected to decline by about two-thirds oncethey reach their respective completion points.HIPCs in the interim period have benefitedfrom Paris Club debt relief as well as fromrelief from several multilateral creditors underthe HIPC Initiative. The ratio of debt serviceto exports for the 27 decision-point countriesdeclined from 16.9 percent in 1998 to an esti-mated 9.8 percent in 2003 and is projected tofall to 7.9 percent by 2006 (table 11.3).31

Poverty-reducing expenditures have risenin the countries that have reached the decisionpoint. For these 27 countries, such expendi-tures in 2003 were estimated to be about threeand a half times the amount of debt-servicepayments.32 Annual debt service is projectedto be about 24 percent lower during 2001–06than it was in 1998–99, a reduction averaging



TABLE 11.3 Debt service and poverty-reducing expenditure by the 27 HIPCs that have reached the decision point(in US$ millions, unless otherwise indicated)

Actual Estimate Projected

Expenditure 1998 1999 2000 2001 2002 2003 2004 2005 2006

Debt service 3,696 3,179 3,131 2,451 2,419 2,729 2,711 2,534 2,794Debt service/exports (%)a 16.9 14.5 13.9 10.4 10.0 9.8 8.7 7.6 7.9Debt service/government revenue (%)a 25.2 21.8 21.8 16.5 14.9 14.6 12.8 10.9 11.0Debt service/GDP (%)a 3.9 3.4 3.4 2.5 2.4 2.4 2.2 1.9 1.9Poverty-reducing expenditure 6,067 5,934 6,712 7,581 9,104 10,832 11,521 12,465Poverty-reducing expenditure/

government revenue (%)a 40.9 41.4 45.2 46.8 48.6 51.0 49.7 49.0Poverty-reducing expenditure/GDP (%)a 6.4 6.4 7.0 7.5 8.0 8.7 8.6 8.7

Note: Debt service figures for 1998 and 1999 reflect debt relief already provided to Bolivia, Guyana, Mozambique, and Uganda under the original framework. Debt servicefigures do not include the impact of the “topping off” granted to Ethiopia and Niger in April 2004. Data on poverty-reducing expenditures are not available for all coun-tries, particularly for 2003–06. In aggregate, the last available period was used for future years, thus understating the likely level of poverty-reducing spending. a. Weighted averages.Source: HIPC country documents and IMF staff estimates.

about $830 million a year. Poverty-reducingspending, meanwhile, increased from about$6.1 billion in 1999 to $9.1 billion in 2003and is projected to increase to about $12.5 bil-lion in 2006 (see table 11.3).33

Debt relief to HIPCs provided by tradi-tional mechanisms, through the HIPC Ini-tiative, and on a voluntary basis togethergenerally reduces the ratio of HIPC net pre-sent value debt to exports to the averagelevel of other low-income countries. Contin-ued measures are needed by HIPCs and bycreditors to ensure that debt sustainability ismaintained after the completion point, justas similar measures are needed for otherlow-income countries. Bank and Fund staffhave therefore proposed a debt sustainabil-ity framework for low-income countries toaddress this issue. The Executive Boardshave endorsed the key elements of theframework, and work is continuing towardmaking the framework fully operational.34

This framework is intended to provideguidance on several issues related to thefinancing strategies for low-income coun-tries, including the range of indicators forassessing debt sustainability, the role of poli-cies in determining appropriate debt thresh-olds, the importance of including domesticdebt in such assessments, and the appropri-ate mix of grants and new credits. Theseissues are becoming increasingly importantin light of the need for large increases inexternal financing to meet the MDGs. Insome countries, financing the MDGs withnew concessional borrowing could lead toan unsustainably large debt burden. In April2004, Ethiopia and Niger were grantedadditional “topping up” assistance underthe HIPC Initiative, on an exceptional basison account of exogenous factors that hadadversely affected their debt sustainability. Itneeds to be stressed, however, that debt sus-tainability is not only a resource-flow issue.It also depends on increasing growth, diver-sifying exports, increasing access to globalmarkets, and adequately mitigating theeffects of exogenous shocks.

Although creditor participation has im-proved under the Initiative, some non–ParisClub bilateral and commercial creditors havepersisted in not committing to provide HIPCrelief. Most of the costs attributable to bilat-eral creditors continue to be borne by mem-bers of the Paris Club. Commercial creditorsrepresent less than 5 percent of the NPV costof relief, in part because of measures by theDebt Reduction Facility for IDA-only coun-tries, which has reduced the stock of com-mercial debt in HIPCs. Moral suasion,however, remains the principal measure forencouraging participation and discouraginglitigation by remaining commercial creditors.With respect to multilateral debt, 23 of the 31multilateral creditors have indicated theirintent to participate in the Initiative, repre-senting more than 99 percent of the total debtrelief required.

A key premise of the HIPC Initiative is thatdebt relief should be additional to otherforms of external financing assistance. Animportant issue is whether countries receivingdebt relief are also receiving additionalresources or whether debt relief crowds outother aid flows. Merely observing the size offlows does not provide conclusive evidence ofadditionality, because the amount of aid thesecountries would have received without theHIPC Initiative is not known. There are alsosubstantial difficulties in measurementbecause different donors account for debtrelief in different ways. Debt relief is some-times explicit, such as through grants for debtrelief, or implicit, such as through debt ser-vice reductions.

The September 2003 HIPC Status of Imple-mentation Report found, based on a review ofbalance of payments data, that gross and netflows of official external resources to the 27countries that have reached the decision pointincreased substantially during 1999–2002.35

Net flows (that is, the difference between grossresource inflows and debt service payments)increased from $4.5 billion in 1999, beforeprovision of debt relief under the enhancedinitiative in 2000, to $7.3 billion in 2002



once the enhanced initiative got under way.Although it is difficult to determine whatresource flows would have been withoutHIPC in 1999–2002, the data do suggest thatdebt relief has been additional to other formsof external financing. The September reportnoted, however, that official external financ-ing flows to the 27 decision-point countriesdeclined substantially in the mid-1990s (asthey did to other low-income countries).

Notes1. OECD 2004b.2. OECD 2004a. Amounts are at 2002 prices

and exchange rates.3. At the October 2003 International Donors

Conference for the Reconstruction of Iraq, donorspledged more than $33 billion in assistance—loans, grants, and export credits—over four years.There is uncertainty as to when these pledges willbe realized and whether over time they will repre-sent an increase in total aid, as most of the initialannouncements indicate, or a reallocation of aidfrom other countries.

4. The contribution of NGOs to development isdifficult to quantify in dollars, for a number of rea-sons. First, many of their staff work on a voluntarybasis and so the value of their labor is never quan-tified in money terms. Second, even when they arepaid, as for example volunteers working as teach-ers or community workers in developing countries,the value of this compensation may be considerablyless than the market value of their services. Third,many of the activities undertaken by NGOs are notstrictly developmental in aim, but may promotecultural or recreational activities.

5. While NGOs have been successful in raisingtheir own funds and in channeling official flows(about 7 percent of ODA supports NGO activity),they do face the challenge of continuing to beeffective as they grow in size.

6. A recent study by Prati, Sahay, and Tressel(2003) suggests that the adverse trade effect may notbe very large, and even a doubling of aid flows from10 to 20 percent of recipients’ GDP is estimated tolead to a real appreciation of only 6 percent.

7. Knack and Rahman 2004.8. With a 7 percent annual increase in real

terms for ODA and projected medium-term realgrowth for low-income countries of 4.8 percent a

year, the ODA-to-GNI ratio would rise fromaround 2.7 percent to more than 3 percent. Aiddependence varies widely across low-income coun-tries: for example, aid in 2001 was 0.4 percent ofincome in India and 28 percent in Mozambique.

9. See Goldin, Rogers, and Stern (2002) for areview of evidence on the increased productivity ofaid.

10. Development Committee 2003b. Theresults are based on an analysis of 18 low-incomecountries with relatively good policies.

11. Development Committee 2004.12. DAC members’ reporting of tying status is

uneven. For example, the United States does notreport tying data. Tying statistics cover only about40 percent of total bilateral ODA.

13. Aryeetey, Osei, and Quartey 2003.14. Jepma 1991.15. About 30 percent of total ODA takes the

form of contributions to multilateral institutions. Ofthe total aid provided in turn by the multilateralinstitutions, close to 30 percent also is in the form ofspecial-purpose grants (mainly from U.N. agencies).

16. Also see Development Committee 2003b.17. Dollar and Levin 2004. Figure 11.4 draws

on this paper and plots the unexplained (residual)component of growth, aid, and policies. The paperalso refers to studies that find different results onthe relationship between the quality of policies andinstitutions and the effectiveness of aid and pro-vides an assessment.

18. Dollar and Levin 2004. Other recent stud-ies of aid selectivity include Birdsall, Claessens,and Diwan (2003) and Roodman (2003).

19. For this calculation, low-income recipientcountries are divided into those with strong poli-cies and institutions and those with weak policiesand institutions, where the latter is the bottom halfof low-income countries on the CPIA ranking.This binary distribution has its limitations ofcourse as it does not capture the relative positionof recipients in terms of the CPIA score within thetwo categories. In the figure, the selectivity of agiven donor is indicated by the angle of the lineconnecting the dot point to the origin relative tothe 45-degree line.

20. Collier and Hoeffler 2002.21. Development Committee 2003b; OECD

2004a.22. Collier and Dehn 2001.23. In this regard, the IMF is examining

whether responses to shock-induced financing



needs can be made more consistent in the contextof programs supported by the Poverty Reductionand Growth Facility (PRGF), and whether theterms of the Fund’s instruments targeted to shockfinancing can be made more appropriate for low-income countries.

24. The methodology for computing the donorfragmentation index is from Knack and Rahman2004.

25. Bulir and Hamann 2001, 2003.26. Bulir and Lane (2002) find that aid flows on

average are procyclical, so that countries receivemore aid when economic activity is stronger.

27. Gemmell and McGillivray 1998.28. See OECD 2004a for a detailed discussion

of the work of the Working Party.29. Sixteen countries or country groups stepped

forward before or in Rome to expand or main-stream country-level efforts to streamline donorprocedures and practices: Bangladesh, Bolivia,

Cambodia, Ethiopia, Honduras, Jamaica, Kenya,Kyrgyz Republic, Morocco, Nicaragua, Niger, thePacific Island group, the Philippines, Senegal, Viet-nam, and Zambia.

30. Other examples of harmonization activitiesat the country level are listed in the Country Imple-mentation Tracking Tool, at http://www.aidhar-monization.org.

31. See IMF–World Bank 2004b.32. The definition of poverty-reducing expen-

ditures varies across countries. Commonlyincluded are primary education, basic health ser-vices, and rural development.

33. Country authorities are putting in placepublic expenditure management systems thatwould ensure the efficiency of poverty-reducingexpenditures. See IMF–World Bank 2003b.

34. IMF–World Bank 2004a. Also see IMF2004.

35. IMF–World Bank 2003a.



Developed countries need to step upaction in support of key global publicgoods and areas for global collective

action related to the millennium goals. Giventheir weight in the global system, developedcountries bear a special responsibility withrespect to the global public goods agenda.Developing countries, of course, also need toplay their part in this effort.

Six areas surface in most discussions ofglobal public goods: health, environment,financial stability, peace, trade, and knowl-edge.1 Of course not all issues in these areasare of a public good nature. Only those thattranscend national boundaries and are eitherregional or global in nature are relevant in thiscontext. So the relevant actions include thoseaimed at preventing adverse health and envi-ronmental spillovers, preserving internationalfinancial stability and peace, and promotingthe gains from international trade and thespread of knowledge. Together, these actionsconstitute an important contribution to theattainment and sustainability of development.

Some global public goods have been deliv-ered more successfully than others, reflectingthe special circumstances of each good andthe mechanism chosen for its delivery.Ranked from relative success to relative fail-ure, the following examples are often men-tioned: aviation safety, postal systems, theInternet, eradication of smallpox, protection

of the ozone layer, prevention of globalwarming, safety of the oceans (including pol-lution control), and sustainable use of naturalresources such as fisheries.

Effective delivery of global public goodsraises important institutional and financingissues. Institutional approaches could take theform of using existing institutions to assumeresponsibility for a specific global public goodprogram, usually with a governance frame-work that ensures involvement and oversightby donors and other affected countries. Exam-ples are the Consultative Group on Inter-national Agricultural Research (CGIAR) andthe River Blindness Control Program withinthe World Bank, and peacekeeping programsof the United Nations. At the other extreme,some programs have involved setting up newinstitutional delivery mechanisms. An exam-ple is the Global Fund to Fight AIDS, Tuber-culosis, and Malaria (GFATM). Some criticshave voiced concern about the proliferationof such funds and institutions. Between thesepoles lie some innovative arrangements, suchas the Global Environment Facility (GEF),which has an independent governance frame-work but relies fully on existing institutionsfor implementation. Networks and coali-tions are increasingly common, particularlyin the areas of standard setting for commu-nication, commerce, and trade. Importantlessons can be learned from networks with a


Fulfilling Responsibilities for Global Public Goods

12

long and successful experience, such as thosefor civil aviation, traffic safety, and postalservices.

Appropriate models for financing globalpublic goods have received considerablescrutiny recently. In addition to questions ofexternalities, free riders, and scarcity of funds,the need to finance complementary programsat the national level must be taken intoaccount. Official development assistance is alarge source of public funding for global pub-lic goods. Many multilateral and bilateral pro-grams directly finance global public goods orchannel resources to trust funds and special-ized institutions. Each year approximately $5billion in official development assistance isdevoted to global public goods.2 In recentyears, private sources—foundations, compa-nies, individuals—have become important infinancing global public goods, usually in part-nership with official sources. CGIAR and theRiver Blindness Control Program were earlyexamples of mixed private-public funding. TheGlobal Alliance for Vaccines and Immuniza-tion (GAVI), GFATM, and the Prototype Car-bon Fund are prominent recent examples.Some of these partnerships are designed to cre-ate markets by providing up-front support andincentives for private companies to invest in

the creation and delivery of key global publicgoods. Global public goods dealing with activ-ities that fall predominantly in the commercialsphere—financial sector, telecommunications,air transportation—are financed largely on aprivate basis through appropriate industrycooperative arrangements and fees.

Global public goods recently have receivedmuch greater attention, a reflection of thetrend toward globalization and the associatedshrinking of the world. One manifestation ofthe increased attention is the InternationalTask Force on Global Public Goods, estab-lished last year by France and Sweden to lookinto recent work and initiatives on the topicand to identify priorities for action (box12.1). Another manifestation is the increasedattention to global public goods and relatedprograms at multilateral institutions, as setout in chapter 13.

Economic theory suggests that global pub-lic goods typically will be undersupplied, ascountries fail to adequately take into accountthe implications of their policies and actionsfor people living in other countries;3 hence,the importance of monitoring country per-formance in meeting commitments underinternational agreements and in signing ontosuch agreements in the first place.



The mandate of the International Task Force on Global Public Goods, chaired by Ernesto Zedillo(former president of Mexico) and Tidjane Thiam (former Minister of Planning, Development, andCoordination, Côte d’Ivoire), is to encourage the provision of international public goods, globaland regional, that are of critical importance for eliminating poverty and achieving the MDGs.

The task force is to systematically assess and clarify the notion of global and regional publicgoods, to identify those public goods that contribute most directly to reducing poverty and meet-ing the other MDGs, and to make recommendations to policymakers and other stakeholders onhow to provide and finance those goods. The task force is also to propose responsibility for follow-up, including monitoring of effectiveness and results.

Meeting for the first time in September 2003 at Yale University in New Haven, Connecticut, thetask force agreed on areas for further work, to be discussed at its next meeting in Istanbul in earlyMarch. The task force’s final report is due in 2005. Before then, a wide consultation is planned, topromote broad understanding of and consensus around the report’s recommendations for action.

Note: For more information on the task force, visit http://www.gpgtaskforce.org/bazment.aspx.

BOX 12.1 International Task Force on Global Public Goods

For developed countries, it is worthwhileto organize the assessment of their policiesand actions with respect to global publicgoods into three blocks: their participation ininternational agreements and institutions thatcommits them to act in a certain way; theirsupport for capacity building and other pro-grams (bilateral, multilateral, philanthropic)that help developing countries take criticalactions; and their voluntary actions (if any)taken solo or in networks with like-mindedpartners.

This framework is used below to analyzecertain issues relating to environmental sus-tainability—a key global public good. Futureglobal monitoring reports will address otherglobal public goods as well. In the meantime,financial stability, trade (including someissues relating to knowledge and technologytransfers), and health are treated in the con-text of other chapters of this report.

Developed-Country Impact on the Global Commons:Limited ProgressDeveloped countries shoulder the predomi-nant responsibility for preserving the globalenvironmental commons and helping devel-oping countries pursue environmentally sus-tainable development, both because of their

larger contributions to global environmentaldegradation and their greater financial andtechnological resources for prevention andmitigation.

Policies aimed at improving environmen-tal performance in and by the developedcountries can be divided between those thatwill improve local environmental conditionsand those aimed at reducing negative im-pacts on the global commons. An analysis ofdeveloped-country policies affecting theirlocal environment is beyond the scope of thisreport, but the experience of the OECD withenvironmental performance reviews offersvaluable lessons for developing countries(box 12.2).

The environmental performance of devel-oped countries in relation to the state of theglobal commons is mixed. The ratification ofinternational environmental conventions—with some notable exceptions—indicates ahigh degree of commitment to global environ-mental issues (box 12.3). The process of globalenvironmental governance has witnessed someresounding successes, most prominently theMontreal Protocol, an international agreementto protect the ozone layer. The success of theprotocol was facilitated by the existence ofcost-effective alternatives to ozone-depletingsubstances, the availability of adequate fund-ing for developing countries,4 and the fact that

F U L F I L L I N G R E S P O N S I B I L I T I E S F O R G L O B A L P U B L I C G O O D S


Since 1991, the OECD has carried out environmental performance reviews (EPRs) of its membercountries. Each EPR report is peer reviewed and approved by the OECD Working Party on Envi-ronmental Performance. Progress is assessed by establishing baseline conditions and examining pol-icy commitments, institutional arrangements, and routine capabilities for carrying out nationalevaluations. Along with the traditional benefits of country analytic work, the peer review processpromotes environmental improvement and continuous dialogue through benchmarking (the shar-ing of information on the policies, approaches, and experiences of the reviewed countries) and peerpressure.

The EPR process can be successfully replicated in developing countries. Indeed, since 1993 theUnited Nations Economic Commission for Europe (UNECE) has been carrying out EPRs on coun-tries in Europe and Central Asia that are not members of the OECD.

Source: OECD and UNECE Web sites; World Bank staff.

BOX 12.2 Progress toward environmental sustainability through performancereview and peer pressure

inaction would exact high costs in the nearfuture. But most global environmental prob-lems do not share those characteristics, result-ing in slower progress to resolve them.

Sectoral policies in developed countriesmake or break progress toward global envi-ronmental sustainability. The maintenance ofhigh subsidies to fisheries and the failure ofenergy policies to forcefully reduce green-house emissions are just two vivid illustra-tions of the lack of rapid overall progress.More than a quarter of the world’s fisheriesare overexploited or depleted, as overfishingby local communities is aggravated by fishingfleets from developed countries. Global sub-sidies for fishing, meanwhile, conservativelyestimated at $10 billion to 15 billion a year(about a quarter of the annual $56 billiontrade in fish), are driving unrestricted andhighly advanced fish harvesting.

The MDGs call for a reduction of green-house gas emissions. The 1997 Kyoto Proto-col places most of the burden for reducinggreenhouse gas emissions on rich countries,

home to just 16 percent of the world’s popu-lation but the source of more than half suchemissions. Kyoto calls on developed countriesas a group to reduce carbon dioxide emis-sions by at least 5 percent of 1990 levels by2008–12. Overall, progress to date has beendisappointing (figure 12.1). Within the Euro-pean Union only two countries, Sweden andthe United Kingdom, are on track to meettheir targets. The overall EU target is to cutgreenhouse gas emissions by 8 percent before2010, but with existing policies only a 0.5percent cut will be achieved. The UnitedStates, which produces 25 percent of globalgreenhouse gas emissions, has explicitlydeclined to ratify the protocol and con-tributes growing emissions.

Aid for the Environment: Long StagnantOfficial development finance for environ-mental concerns has followed the stagnantpath of overall development financing in



The last 20 years have seen the birth of six major international environmental agreements, the rat-ification of which reflects a solid consensus on the importance of multilateral efforts to tackle globalenvironmental problems. Between 150 and 190 countries have ratified the Montreal Protocol to theVienna Convention for the Protection of the Ozone Layer (1987), the Basel Convention on theTransboundary Movement of Hazardous Wastes and Their Disposal (1989), the Convention onBiological Diversity (1992), the Framework Convention on Climate Change (1992), and the Con-vention to Combat Desertification (1994). Of the 151 signatories to the most recent agreement, theStockholm Convention on Persistent Organic Pollutants (2001), 41 have already ratified it. (It willenter into force with the 50th ratification.)

A notable exception to these positive results is the Kyoto Protocol to the Framework Conven-tion on Climate Change (1997). To enter into force, the protocol must be ratified by 55 parties thattogether account for at least 55 percent of global greenhouse gas emissions, but the 120 parties thathave ratified represent only 44.2 percent of emissions. Six so-called Annex I parties (developed andtransition economies) have not yet ratified the protocol: Australia, Croatia, Liechtenstein, Monaco,Russia, and the United States (though Russia has recently indicated that it will). The United Stateshas announced that it will not ratify the Kyoto protocol. It has not ratified the Basel or the Bio-diversity conventions.

Source: Convention Web sites; World Bank staff.

BOX 12.3 International environmental agreements: toward global cooperation,with some notable exceptions

1990–2000. A slight increase in official bilat-eral aid for environmental purposes followedthe 1992 Rio Convention, but commitmentspeaked in 1996 and the increase was reversedby the end of the decade. A slightly differentpattern emerges in multilateral commitments.After declining sharply in 1993, with total aidcommitments, they rebounded in 1994 onlyto resume a declining trend. Commitments toenvironmental concerns averaged 3 percentof total bilateral aid and 5 percent of totalmultilateral aid, bringing total environmentalaid to $2 billion per year—far from the com-mitments made at Rio (table 12.1).

Commitments to environmental funds setup to confront global issues are no exception.The Global Environment Facility was estab-lished in 1991 to help developing countriesfund projects and programs that protect theglobal environment, initially in the fields of bio-diversity, climate change, ozone depletion, andinternational waters. Since the 1994 restruc-turing, new funding for the GEF has remainedstagnant. The first replenishment, covering1994–98, saw pledges of $2.02 billion. Newcommitments were $1.99 billion for 1998–02and $2.2 billion for 2002–06. Despite expan-sion to cover new environmental issues, such asland degradation and persistent organic pollu-tants, GEF funding has actually declined byalmost 10 percent as a share of the combinedGDP of the 38 contributing nations.

Developed Countries and GlobalEnvironmental Sustainability:Widely Varying PerformanceThe performance of developed countries onglobal environmental sustainability is far fromhomogeneous. MDG 7 aims to “reverse theloss of environmental resources.” Table 12.2assesses developed country efforts to this endon three dimensions: depletion of the globalcommons through carbon dioxide (CO2) emis-sions (an MDG indicator); provision of globalpublic goods through participation in multilat-eral environmental agreements; and assistance,through environmental aid, to poor countries

in their quest to halt and reverse environmentaldegradation. Sweden and Switzerland emergeas exemplary global citizens, whereas Australiaand the United States tail the group of devel-oped countries (table 12.2).

The contributions of developed countriesto carbon dioxide emissions and climatechange vary widely. In per capita terms, Aus-tralia, Luxembourg, and the United Statesemit more than three times the amount ofSweden and Switzerland. Nor is the participa-tion of developed countries in environmentalagreements homogeneous. While 15 countrieshave ratified the Kyoto Protocol, the BeijingAmendment, and the Biodiversity Conven-tion, Australia and the United States haveeach ratified just one. And developed coun-tries commit remarkably different amounts ofaid for environmental sustainability. Den-mark, the Netherlands, Norway, Sweden, andSwitzerland commit around 0.02 percent oftheir GDP to environmental aid. Ireland andItaly give less than 0.001 percent—more thana 20-fold difference.



All high-income

countries

Other high-income

countries

UnitedStates

JapanEuropeanUnion

14

12

10

8

6

4

2

CO2

emis

sion

s in

bill

ions

of

met

ric

tons

20001990

FIGURE 12.1 MDGs and Kyoto Protocol call for reduction of greenhouseemissions, but results tell a different storyCarbon dioxide emissions, 1990 and 2000

Note: Carbon dioxide emissions are the most important component of greenhouse gas emissions.Source: World Bank, World Development Indicators.



TABLE 12.1 Aid for the environment, 1990–2000(US$ millions)

Aid 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000

Bilateral 1,033 652 921 737 806 1,083 1,822 1,526 1,012 1,394 984Multilateral 657 604 1,683 368 1,942 822 744 812 1,439 590 962Total 1,690 1,256 2,604 1,105 2,748 1,905 2,566 2,338 2,451 1,984 1,946Total (% of bilateral

donors’ combined GDP) 0.010 0.007 0.014 0.006 0.013 0.008 0.011 0.011 0.011 0.008 0.008

Note: The table uses the World Bank definition of environmental aid, as applied in Pagiola and others (2002). Alternative definitions generate different levels of aid forenvironment but similar trends.Source: Adapted from Pagiola and others 2002.

Table 12.2 Performance of developed countries on global environmental sustainability

Depletion of the Ratification of multilateral Aid forglobal commons environmental agreementsa environment

Carbon dioxide emissions Beijing Amendment Biodiversity Bilateral aid for Country (metric tons per capita, 2000) Kyoto Protocol to Montreal Protocol Convention environment (% of GDP)b,c

Australia 18.0 No No Yes 0.0054Austria 7.5 Yes No Yes 0.0078Belgium 10.0 Yes Yes Yes 0.0012Canada 14.2 Yes Yes Yes 0.0031Denmark 8.4 Yes Yes Yes 0.0193Finland 10.3 Yes Yes Yes 0.0060France 6.2 Yes Yes Yes 0.0025Germany 9.6 Yes Yes Yes 0.0070Greece 8.5 Yes No Yes ··Iceland 7.7 Yes No Yes ··Ireland 11.1 Yes No Yes 0.0007Italy 7.4 Yes No Yes 0.0004Japan 9.3 Yes Yes Yes 0.0070Luxembourg 19.4 Yes Yes Yes ··Netherlands 8.7 Yes Yes Yes 0.0222New Zealand 8.4 Yes Yes Yes ··Norway 11.1 Yes Yes Yes 0.0230Portugal 6.0 Yes No Yes 0.0003Spain 7.0 Yes Yes Yes 0.0041Sweden 5.3 Yes Yes Yes 0.0145Switzerland 5.4 Yes Yes Yes 0.0216United Kingdom 9.7 Yes Yes Yes 0.0064United States 19.8 No Yes No 0.0021

·· Negligible or not available.a. “Yes” indicates ratification or equivalent (acceptance, accession, approval).b. Calculated using the World Bank definition of environmental aid; see Pagiola and others (2002) for details.c. Average of 1998, 1999, and 2000; except Ireland (2000) and Portugal (1999 and 2000).Source: World Bank, World Development Indicators; convention Web sites; World Bank staff.

Priorities for ActionThe case for greater efforts by developedcountries in supporting and financing globalpublic environmental goods is a strong one.Rich nations bear a special responsibility forthe environmental commons. Conservation,particularly the protection of biodiversity, isanother important responsibility. Thatresponsibility is partly a question of ability topay. But it is also a question of willingness topay—many citizens of developed countriesare highly motivated to conserve naturewhere it is most at risk in developing coun-tries. Tapping that willingness to finance con-servation in poor nations will be essential inpreserving resources at risk.

Without diminishing the importance ofother needed policy changes, such as reducingfisheries subsidies, priorities for action in richcountries include:

� Increasing aid for environmental sustain-ability to help developing countries establishadequate frameworks for environmentalmanagement. This includes financing biodi-versity conservation through mechanismssuch as the GEF and actions aimed atimproving natural resource managementand reducing the burden of environment-related diseases in developing countries.

� Controlling greenhouse gas emissions indeveloped countries.

� Buying offsets or reductions in greenhousegas emissions from developing countries.

� Assisting developing countries to adaptto climate change and increased climaticvariability.

Notes1. Ferroni and Mody 2002; Kaul, le Goulven,

and Mendoza 2003. See also Development Com-mittee 2000, 2001 and African Development Bank2002.

2. World Bank 2001.3. This conclusion follows from the growing lit-

erature on global public goods, whose originsmost observers trace back to the seminal work byPaul Samuelson (1954) on the pure theory of pub-lic goods. That work formalized earlier thinkingon the provision of public goods and for manyyears was used as the starting point for discussionsof local public goods and public finance. With themore recent focus on global public goods, it hasbeen used to understand those issues as well.Under the theory, a pure public good has twoattributes: “nonrivalry” (that is, my consumptionof the good in no way reduces your ability to con-sume it) and “nonexcludability” (that is, no one inthe immediate vicinity can be prevented from con-suming the good). Also under the theory, purepublic goods are chronically undersuppliedbecause those who pay for them do not reap thefull gains they generate.

4. Since 1994 some $150 million a year hasbeen made available through the multilateral fundassociated with the protocol.



IVRole of International Financial Institutions

How well are the international finan-cial institutions contributing to theachievement of the MDGs and related

development outcomes? This chapter takes amodest first step toward answering that ques-tion, focusing on the International MonetaryFund, the World Bank, and the major regionaldevelopment banks—the African Develop-ment Bank, the Asian Development Bank, theEuropean Bank for Reconstruction and Devel-opment, and the Inter-American DevelopmentBank—and drawing primarily on the self- andindependent evaluations carried out by theinternational financial institutions themselves.Future efforts will extend and deepen theframework developed herein, including bybringing in the findings of external evaluationsand surveys more explicitly,1 and by broaden-ing the framework to include the issue ofdeveloping-country voice in institutional gov-ernance, where the different ownership struc-tures of the various IFIs can provide a usefulplatform for comparative analysis.

In this context, the chapter finds evidencethat the IFIs are generally becoming morecountry focused, more collaborative, andmore results oriented, while respecting theirdistinctive mandates and modalities for pro-viding support for country, regional, andglobal programs. But it is inconclusive on thecritical questions of comparative perfor-mance and whether the whole of the IFI con-

tribution is larger (or smaller) than the sum ofthe IFI parts. Going forward, greater avail-ability and comparability of independent andself-evaluation data from all IFIs should helpimprove comparative monitoring exercisessuch as this one and, in turn, reporting to thetaxpaying public in all countries. The jointwork program on results endorsed by themultilateral development banks at theRoundtable on Managing for DevelopmentResults held in Marrakech in February 2004should provide a vehicle for progress on thiskey issue.2

The chapter is structured as follows. It firstsummarizes the framework used for consider-ing the contribution of the IFIs individuallyand in the aggregate. It then applies the frame-work in turn to the IMF, the World Bank, andthe major regional development banks. Next itlooks at the trends in systemic coherence andwhether the contribution of the IFIs is larger orsmaller than the sum of the individual contri-butions. The last section draws out conclu-sions and implications for action.

Framework for Monitoring IFI Contribution IFIs do not achieve development outcomesdirectly through their individual actions,but they can contribute to those outcomes.For IFIs and other agencies, success depends


Monitoring the IFIs’ Contribution13

on effectively deploying assistance in high-quality ways in countries and programs thatdeliver, and on influencing country policiesand programs—often at a distance. Thiscomplicates the job of assessing the IFI con-tribution to the achievement of the MDGsand related outcomes, but no more than itcomplicates the job of designing IFI pro-grams to maximize that impact. An under-lying theme of the chapter is the need forcoherence between ex ante program appraisaland ex post evaluation. Assessing an institu-tion’s contribution after the fact is complex,but it is rendered more so if it is not antici-pated before the fact, and incorporated intooperational design.

What are the channels of influence throughwhich the IFIs affect the attainment of theMDGs and related development goals, andhow can they be measured? Simply countingthe number of projects or volume of lendingfor MDGs is clearly not the answer, althoughnumbers on IFI inputs may constitute part ofthe evidentiary picture. Rather, it is the IFIs’catalytic role that matters the most in theachievement of development outcomes—both by supporting policy and institutionaldevelopment and by directing that support tothose countries and sectors where the payoffto growth and poverty reduction is likely to

be the largest. Significant IFI impact maycome just as easily through a partnership ona global program, the policy dialogue on apiece of analytic work, or support tostrengthen country capacity to design a fullyowned reform program, as through a tradi-tional lending operation. In such a setting, themeasurement challenge is to discern the IFIinfluence on the global and regional contextand the country policies and actions in thecritical areas identified in the preceding chap-ters. Meeting this challenge is best pursuedthrough a thoughtful and transparent results-chain analysis that links final outcomes to thespecific inputs (including the level and distri-bution of IFI financing across countries) andoutputs needed to produce them, and thatcan be evaluated upon program completion.

The framework for considering the IFIcontribution is structured around fourthemes: country programs, global programs,partnership, and results. These themes areused to examine the IFIs individually andthen collectively (box 13.1). The first twothemes reflect the fact that increasingly theIFIs have two major product lines that are rel-evant to the achievement of the MDGs andrelated outcomes: their country programsand their global programs. Of course, the bal-ance between the two—as well as the corre-

P A R T I V : R O L E O F I N T E R N A T I O N A L F I N A N C I A L I N S T I T U T I O N S


Country programs� Alignment with countries’ priorities for poverty reduction and other MDGs� Relevance and selectivity in program design

Global programs� Support for capacity building for regional and global public goods � Anchor in international system for global public goods

Partnership� Harmonization of institutional policies and practices � Coordination of country support programs

Quality and results� Results orientation, including monitoring, evaluation, and reporting systems� Findings on quality and results of strategies and operations

BOX 13.1 Framework for assessing IFI contributions

sponding instruments of support—variesacross the individual IFIs, in line with theirdistinctive mandates and comparative advan-tages, as elaborated in the sections below. Forexample, the IMF’s country program focus ison macroeconomic and fiscal sustainability,which is essential for sustainable growth andpoverty reduction. Its global program focus ison international financial stability, in linewith its role as the institutional anchor of theinternational financial system. The WorldBank’s country program focus is broader,covering the full range of social and structuralissues important for the achievement ofpoverty reduction and the MDGs. Its globalfocus is on partnership and capacity-buildingprograms designed to benefit developingcountries. Focusing on countries in theirrespective regions, the regional banks alsohave broad development mandates. Theexception is the European Bank for Recon-struction and Development, whose mandateis more narrowly focused on transition. Allthe regional banks have substantial programsfor regional public goods.

Partnership has been singled out in theanalysis for special attention, given its centralimportance to the IFIs’ joint and individualeffectiveness in helping countries achievetheir development goals. The lessons of expe-rience (and common sense) suggest thatwhere many donors are involved in a countryor global program, it is essential that theyharmonize their rules and reporting require-ments to avoid overtaxing scarce countrycapacity. In addition, where there is a PovertyReduction Strategy Paper, donors should syn-chronize the timing of their own countrybusiness strategies with it, so that the countrycan know the resources it is likely to haveavailable when it is doing its programmingand budgeting. But maximizing the gainsfrom partnership requires going beyond har-monization and synchronization. It alsorequires strategic alignment, with agenciesbeing coordinated and strategically selectivein their support programs in line with coun-try priorities and their distinctive compara-tive advantage vis-à-vis other agencies. Such

action will also help to ensure that the wholeof IFI and donor support adds up to morethan the sum of the individual parts.

Finally, the critical bottom line is results.Here the chapter zeroes in on two topics: (a)the systems the IFIs have in place for assess-ing their results, and (b) the emerging pictureof what those systems suggest about the IFIcontribution. As the main tool for demon-strating results, evaluation is fundamental tothis work—both self-evaluation undertakenby the management of programs and inde-pendent evaluations undertaken on behalf ofgovernance bodies that oversee management,such as the board of directors.3

Activities of Individual IFIsThis section looks in turn at the activities ofthe IMF, the World Bank, and the regionaldevelopment banks, using the organizingstructure outlined above.

International Monetary Fund

The IMF contributes to the achievement of theMDGs through several important, if indirect,channels primarily related to the macroeco-nomic policy environment. These channelsinclude the Fund’s work on the global eco-nomic and financial system and industrialcountry surveillance; its financing to membercountries; and its policy advice and technicalassistance to developing countries. Given thenature of these contributions and the fact thatthey draw on many aspects of the IMF’s work,there is no simple yardstick for assessing effec-tiveness. Information from a variety of sourcesmust thus be brought to bear in making suchan assessment.

C O U N T R Y P R O G R A M S

A key responsibility of the IMF is to providefinancing to member countries so that theycan address external imbalances withoutresorting to measures destructive of nationalor international prosperity. This financing isprovided on the strength of a program ofpolicies designed to address the underlying

M O N I T O R I N G T H E I F I S ’ C O N T R I B U T I O N


imbalances that necessitated the financing.For low-income countries facing temporaryshocks and for middle-income countries fac-ing capital account crises, the IMF’s financingalleviates the burden of immediate adjust-ment that could otherwise set back thosecountries’ efforts to achieve the MDGs.

Beyond the direct benefits of its financing,the Fund provides advice to its member coun-tries, in the context of financial arrangements(where its advice is linked to program condi-tionality) and of its regular surveillance of allmember countries. From the perspective ofachieving the MDGs and related develop-ment objectives, the IMF’s role in providingadvice is particularly important. The primaryorientation of the IMF’s advice is toward theachievement and maintenance of macroeco-nomic stability over the medium term, givenits importance for growth and poverty reduc-tion. Moreover, country experience is repletewith episodes in which macroeconomic dis-turbances have derailed progress towardgrowth and prosperity. In emerging-marketcountries, crisis prevention and managementare a central focus of the IMF’s work, whichtakes stock of lessons from crises in Asia,Latin America, and elsewhere.4

A major challenge for the IMF’s policyadvice is to ensure not just that the advice iscorrect—itself a major area of controversy—but that members are helped to implementappropriate policies. This challenge involvesissues of capacity and ownership. Capacityissues should be taken into account indesigning programs. Capacity can also bestrengthened through technical assistance(TA), which the IMF provides in a range ofareas related to its responsibilities, includingfiscal issues, the monetary and financial sec-tor, and statistics. The IMF recently steppedup its provision of TA through the establish-ment of two African technical assistance cen-ters (AFRITACs). There are also plans toestablish a center in the Middle East. Themost recent review of the IMF’s TA, under-taken in 2004, focused on strengtheningeffectiveness and management with a view toaddressing the institution-building needs of

members, particularly low-income countries.It found that, while clear progress has beenmade, attention still needs to be given tothese areas and to developing a clear strate-gic direction for TA in the future.5

Country ownership is an element that hasbeen examined in connection with the IMF’sconditionality review during 2000–02.6 Thefocus on ownership reflects concerns that inmany cases IMF-supported programs may notbe adequately aligned with member countries’own priorities and that this gap may bereflected in weak implementation. Researchsuggests that program implementation dependsmainly on domestic factors rather than theefforts of IMF staff. The conditionality guide-lines approved in September 2002 stress thatownership is an essential foundation of anIMF-supported program. The implementationof these guidelines will be reviewed periodi-cally, beginning in mid-2004.7

In low-income countries, the PRSP is avital tool for building ownership and orient-ing policies toward the MDGs. The PRSPprovides a framework in which a country canformulate its strategy for poverty reductionon the basis of a domestic consultativeprocess. It also therefore provides a basis fordonors and IFIs to align the conditions fortheir financial support with the country’s ownstrategy. A review of the PRSP process under-taken jointly by the World Bank and the IMFin 2002 found that the PRSP was widely seenas useful, but that the consultative processesused to prepare the PRSP needed to bestrengthened.8 A key issue is the alignment ofthe PRSP and a program supported by aPoverty Reducation and Growth Facility(PRGF) financial arrangement. The formershould provide a road map to the MDGs,while the latter must be based on realisticprojections of macroeconomic developmentsand external financing, even if those projec-tions fall short of what is needed. The mainchallenge is to use the tension between thecountry’s needs on the one hand and its avail-able resources and capacity on the other toelicit more action from donors to provide thefinancing and from the country authorities to



undertake the reforms needed to bring realityinto line with aspirations.9 As a related issue,extensive evidence suggests that the growthprojections underlying IMF-supported pro-grams have tended to be biased toward over-optimism. This points to a need to strengthenthe analytical framework in which IMF-supported programs are designed.

The IMF’s financial arrangements havecome to play a longer-term role in low-incomecountries, typically providing financing onconcessional terms through successive PRGFarrangements. A long-term engagement withlow-income countries helps to ensure that theFund’s financing and policy advice are appro-priately directed at helping these countriesachieve their development goals. The nature ofthe Fund’s engagement, however, may changeas the country’s situation changes. Some low-income countries have episodic financingneeds associated primarily with temporarymacroeconomic imbalances to be addressedthrough policy corrections; providing suchfinancing is a natural role for the Fund. In con-trast, for others, financing is required on amore continuous basis to facilitate the institu-tional reforms and investments in human cap-ital and infrastructure needed to achievedevelopment goals.10 In the latter context, theFund has tended to remain engaged and tocontinue to provide financing, but typically asa relatively small part of overall financingflows from other international financial insti-tutions and donors. In such cases, a financialarrangement is, to a considerable extent, avehicle for policy advice and monitoring, andthe IMF’s involvement is a signal of the coun-try’s macroeconomic management to otherproviders of official financing. Associated withthis signaling role is the risk that the IMF couldbe drawn into prolonged program relation-ships that may not be justified on the basis offinancing need and the quality of policies—pointing to the need to consider alternative,nonprogram, modes of IMF engagement. Thepolicy on assessment letters enables the IMF toprovide a more authoritative assessment ofpolicies, either in or outside the context of aprogram relationship.

G L O B A L P R O G R A M S

Within the international system, the IMF istasked with promoting a stable and openglobal economic and financial environment,which is essential for the achievement of theMDGs and related development outcomes.The Fund contributes to such an environmentfirst of all through the surveillance of system-ically important countries. The Fund’s ArticleIV consultations with industrial countries area vehicle for promoting appropriate policies,such as curbing domestic imbalances thatmay pose risks for the global economy. Mul-tilateral surveillance, in the context of theWorld Economic Outlook and the GlobalFinancial Stability Report, can highlight bothglobal macroeconomic and financial risksand urge changes in policies by the majorcountries. In these and other contexts, theFund is an advocate for increased foreign aidto achieve the MDGs and for increased mar-ket access for developing countries’ exports.The effectiveness of the IMF’s surveillancewill next be reviewed later in 2004.11

P A R T N E R S H I P

In carrying out its mission and mandate in con-tributing to the achievement of the MDGs, theIMF works with partner agencies, especiallythe World Bank but also with other multilateraland bilateral providers of aid and financing.

The IMF’s engagement in low-income coun-tries puts it in a position to facilitate and possi-bly to catalyze other financing. In many cases,the approval or successful review of an IMFarrangement is itself a signal to other interna-tional institutions and donors that policies aresufficiently sound that it is prudent to providetheir financing. The IMF recently took steps toaddress concerns that these signals were not asclear as they should be. Notably, it clarified theuse of assessment letters to convey the institu-tion’s views for both on- and off-track pro-grams. The IMF also plays a more direct role infacilitating other official financing through itsparticipation in donor meetings.

It is increasingly recognized that macro-economic stability and growth depend heav-ily on structural and institutional factors.12



Providing advice on many of these issues isprimarily the responsibility of the World Bankand other development agencies. Coordina-tion with the World Bank is therefore critical.Due in part to concerns that weaknesses incoordination had contributed to an expansionof IMF policy conditionality, new guidelinesfor Bank-Fund coordination in programdesign and conditionality were approved inMay 2002.13 The most recent review of Bank-Fund collaboration in relation to these guide-lines indicates that collaboration appears tobe improving, but that there is room for fur-ther improvement. At the same time, collabo-ration has some natural limits, due to thedifferent mandates and time frames of the twoinstitutions and their different organizationalstructures. The review highlights efforts thathave been made to strengthen collaboration inparticular areas, such as in the work on low-income countries and in public expendituremanagement issues.14

Q U A L I T Y A N D R E S U L T S

The IMF has long had internal systems forensuring the quality of its advice and pro-grams, in the form of management controlsand oversight by the Executive Board. Also,periodic reviews by staff have taken stock ofemerging results and lessons learned, feedingthis information back into the design of newprograms and instructions to staff. Notably,the review of conditionality that will be under-taken in 2004 will examine the objectives andoutcomes of Fund-supported programs andthe design of the policy framework, in addi-tion to reviewing the application of the 2002conditionality guidelines. These reviews havebeen supplemented by independent externalreviews on key topics, such as the externalevaluation of the Enhanced Structural Adjust-ment Facility (ESAF) in 1998.15

The creation in 2002 of the IndependentEvaluation Office (IEO)—which reportsdirectly to the IMF’s Executive Board, notthrough management—complements theseefforts.16 To date, three reports have beenprepared, discussed by the Board, and thenpublished. The first report examined the pro-

longed use of Fund resources, documentingthe number of member countries with longperiods of program engagement and examin-ing case studies.17 The report raised concernsthat long-term engagement could underminethe revolving nature of the IMF’s financialresources for balance of payments adjust-ment; that in some cases, prolonged engage-ment reflects persistent weaknesses inprogram design and implementation; that theuse of an IMF financial arrangement as a sig-nal to donors may distort decisions on IMFfinancing; and that the presence of the IMFmay inhibit the development of domesticdecisionmaking processes. In discussing thisreport, the Executive Board noted that inmany instances—particularly in low-incomecountries—prolonged program engagementmay play a constructive role in tackling deepstructural problems whose solution requiresprotracted effort. The Board initiated proce-dures for ex post assessments of memberswith longer-term program engagement andfor semiannual factual reports on the inci-dence of such engagements. The report alsogave added impetus to work assessing pro-gram design and strengthening surveillance inprogram countries.

The IEO’s second report examined the roleof the IMF in three recent capital accountcrises: Indonesia in 1997–98, Korea in1997–98, and Brazil in 1998–99.18 It notedthat IMF surveillance was more successful inidentifying macroeconomic vulnerabilitiesthan in recognizing and analyzing the risksarising from financial sector and corporatebalance-sheet weaknesses. Insufficient candorand transparency also limited the effective-ness of surveillance on policies. With regardto the design and implementation of IMF-supported programs in response to thesecrises, the report noted that macroeconomicoutcomes turned out to be very different fromprogram projections and that in a number ofrespects the policies incorporated in the pro-grams could, with the benefit of hindsight,have been improved. The report noted that,in response to these crises, reforms have beenundertaken to strengthen the IMF’s capacity



to anticipate and help avert crises as well asto better manage the crises that do occur.

The third IEO report examined the experi-ence with fiscal adjustment in IMF-supportedprograms.19 This report found that the Funddoes not follow a one-size-fits-all approach inits programs, as claimed by some critics. Butit found that Fund staff could improve thepresentation in staff reports of the rationalefor their fiscal policy advice. The report alsomade recommendations for anchoring fiscaladjustment and reforms in a medium-termframework, both in the context of IMF-supported programs and surveillance. Stepsare now under way to implement a numberof these recommendations.

During 2004 the IEO will undertakeassessments of the PRSP process and thePRGF (jointly with the World Bank’s Opera-tions Evaluation Department), the IMF’stechnical assistance activities, and the IMF’srole in Argentina.

World Bank

The World Bank contributes to the achieve-ment of the MDGs and related developmentoutcomes in a number of ways. Its most tan-gible contribution is the support it provides todevelopment programs in low- and middle-income countries. Closely related is the sup-port the Bank provides for global andregional programs, where considerable ex-pansion has taken place in recent years. TheBank’s country and global program work iscarried out with developing- and developed-country partners—as well as with partners inother international institutions—and isincreasingly focused on results.

Against this background, this sectionlooks at the evidence on what the Bank isdoing in its country and global programs,how it behaves as a partner, and by whatyardsticks it measures its contribution todevelopment results. The analysis draws onthe extensive assessment and evaluationmaterial available on World Bank activities,both self-assessments carried out by centralunits within the Bank and independent eval-

uations carried out by the Bank’s OperationsEvaluation Department (OED). These docu-ments are prepared periodically, providing aseries of snapshots on the Bank’s contributionand institutional performance in criticalareas. For the most part, they show an insti-tution whose performance has been improv-ing in recent years, but with a number ofspecific areas still needing further work.

This picture is broadly consistent withrecent survey evidence. In the global poll ofworld opinion leaders in late 2002 to early2003 that the Bank commissioned,20 mostrespondents reported a generally positiveview of the Bank, with many saying that theBank has been doing a better job in the pastfew years in a number of areas, especiallypoverty reduction. But a number of criticismsalso were expressed, especially with respect tohow respondents saw the Bank’s effectivenessin fighting corruption, the social impact of thepolicy reforms it supports, and the arroganceand bureaucracy with which it is perceived tooperate. Similarly, in the survey of countryauthorities carried out for the recent reviewof Bank-Fund collaboration, reform pro-grams supported by the Bank were reportedto be largely or fully owned by the country in88 percent of cases, and a similar majoritypointed to Bank participation widening own-ership of the authorities’ development strat-egy. 21 But 11 percent of the respondentsreported only partial ownership.


The World Bank’s approach to country pro-grams recognizes that one size does not fit all,and that the Bank’s support must be tailoredto country conditions and grounded innational strategies setting out countries’development vision, objectives, and priori-ties.22 As the Bank’s central operational doc-ument for each country program, theCountry Assistance Strategy (CAS) is a keyplace to start in assessing the Bank’s contri-bution. The CAS summarizes the diagnosis ofcountry conditions, the lessons learned fromprevious external support, and the forwardprogram. It is prepared by the country team



in cooperation with the authorities and part-ners, approved by senior management, anddiscussed by the Executive Board. In present-ing the Bank’s prospective program, the CASis meant to explain what is proposed as wellas why, in terms of the particular value addedand additionality that the Bank can bring tothe table. For most low-income countries, thePRSP is the vehicle setting out the nationalstrategy that the Bank helps countries to for-mulate and that serves as the foundation forthe CAS. For middle-income countries, thereis not an agreed format like the PRSP for set-ting out the national strategy, and the Bankrelies on a wide variety of country-specificvehicles as a basis for the CAS.

Alignment with Country Priorities. In prac-tice, how well do Bank CASs align withcountry priorities? According to a series ofretrospectives prepared by Bank manage-ment and OED reviews, most CASs do pro-vide fairly full coverage of the country’sdevelopment program, describing the gov-ernment’s development objectives and thekey elements of the development strategy—highlighting strengths and weaknesses.23 Themost recent CAS retrospective judged thetreatment of the national development pro-gram and priorities to be substantial in 86percent of the 39 CASs and CAS ProgressReports in its review cohort. This representsa significant increase from the levels reportedin the previous retrospectives (60 percent inthe first and 80 percent in the second). How-ever, it must be noted that the CASs in thereview cohort were discussed by the Boardbefore the end of 2000. As the period sincethen has seen enhanced focus on countryownership and national strategies, throughthe PRSP in the case of low-income countriesand country-specific processes in the case ofmiddle-income countries, these trends arelikely to have intensified. Informal reviews ofmore recent CASs do point to a further step-up in country focus, and the expectation isthat this finding will be validated in the nextCAS retrospective.

In part, this expectation is based on paral-lel developments with respect to PRSPs, whichhad been prepared for 37 countries as of theend of March 2004. Only two CASs in the ret-rospective review cohort were based on fullPRSPs (Burkina Faso and Uganda), and thetreatment of the national strategy was strongin both. Also, the CASs for Albania, Guyana,Mauritania, Vietnam, and Yemen, which wereprepared on the basis of full PRSPs (albeitafter the closure of the cohort for the last ret-rospective), were similarly strong, with analy-sis clearly linked to the PRSP, providing a firmbasis for consideration of the CAS diagnosticand programming options. CASs also willbenefit from ongoing improvements in thePRSPs themselves. A recent Bank-Fund staffreview of PRSPs reports evidence of progressas more recent PRSPs build on the experienceof earlier ones, and countries more advancedin the process gain implementation experi-ence.24 But the review also identified a num-ber of challenges that affect the PRSP’susefulness in underpinning the CAS. Theseinclude the importance of (a) bringing in theMDGs and articulating expected outcomesmore explicitly; (b) striking the right balancebetween ambition and realism in setting PRSPtargets; (c) identifying reliable indicators ofprogress and outcomes; and (d) ensuringappropriate prioritization across PRSP tar-gets. Going forward, Bank management hassuggested—and clients appear receptive to theproposal—that it is important for countries tostrengthen the results focus of PRSPs andhighlight the links with the longer-term MDGtargets and the associated policy and institu-tional reforms and domestic and foreignfinancing requirements.

These findings are broadly in line withOED’s conclusions, which are drawn from itsCountry Assistance Evaluations (CAEs) andother evaluations, such as the multidonorreview of the Comprehensive DevelopmentFramework (CDF)25 and the ongoing PRSPassessment that OED is conducting in part-nership with the IMF’s IEO. OED has consis-tently stressed the importance of more realistic



targets in PRSPs—cautioning that manycountries and regions will not achieve theMDGs by 2015—as well as greater attentionto monitoring and evaluation, which it notedis particularly important given the differenttime frames for PRSPs (typically 3 years) andthe MDGs (typically 10–15 years).26 Relat-edly, OED has urged the Bank to identify theobjectives and targets of its CASs withgreater specificity and to implement internalchanges to foster the design and implemen-tation of the multisectoral strategies andoperations that will be essential for achievingthe MDGs. Also highly relevant for thePRSP-CAS nexus, OED finds evidence of thePRSP process contributing to the develop-ment of country ownership, especially whena highly inclusive participatory process isconducted through the country’s normalpolitical processes and institutions; OED fur-ther points to the need to lessen the tensionbetween the PRSP’s role as a process forbuilding domestic consensus and ownershipand the role that it plays with regard to debtrelief and access to aid resources.27

The CAS model applies equally to themiddle-income countries (MICs), for whichBank management has just completed amajor review of its program and strategy.28

A key finding of this review is that despitemany examples of successful Bank engage-ment in individual MICs—for example,Brazil, China, Mexico, and Turkey—trendsin Bank lending are not in line with theobjective of scaling up support to MICs,given the vast numbers of poor people livingthere. In large part, the disconnect reflects asecular decline in Bank infrastructure lendingto MICs, as well as reduced lending to coun-tries that have gained access to financialmarkets. However, some other multilateraldevelopment banks have maintained or evenincreased their lending over the same timeperiod, suggesting that internal Bank factorsrelated to the cost of doing business may bepartly to blame. Going forward, the Bank isadopting a back-to-basics approach, cuttingred tape by relying more on MICs’ stronger

policy development and systems for fiduciaryand environmental safeguards, proactivelyengaging in value-adding operations in infra-structure and service delivery, and promotingthe use of IBRD risk-management instru-ments—all grounded in high-quality eco-nomic and sector work (ESW). The approachis designed to reinvigorate the Bank’s engage-ment with this critical set of clients and helpthem promote sustainable, equitable, job-creating growth; raise living standards; andreduce poverty.

Relevance and Strategic Selectivity. Strategicselectivity involves systematically examiningthe tradeoffs among possible Bank Groupactivities, assessing their relative impact, andestablishing priorities while taking resourceconstraints into account. It requires looking atthe potential magnitude of impact, the likeli-hood of successful country action—including,importantly, through policy reforms—and thepossible availability of alternative sources ofsupport as a way to assess the expected value-added of the Bank’s contribution. Getting thispart of the CAS right is essential for maximiz-ing the Bank’s impact. It goes to the heart ofensuring that the Bank is doing the rightthings—in addition to doing them right.Ensuring the relevance of Bank support is akey objective of the Bank’s agenda on manag-ing for results and is the driving force in thedevelopment of the results-based CAS, cur-rently being piloted (box 13.2).

The last CAS retrospective found animprovement in selectivity, with more than70 percent of CASs rated satisfactory or bet-ter, compared with only 50 percent in the pre-vious retrospective. However, when theretrospective raised the bar on selectivity toinclude discussion of tradeoffs and the ratio-nale for instrument choice, fewer than 60 per-cent of the cohort CASs were ratedsatisfactory or better. Clearly, as Bank man-agement has stressed, there is ample scope forimprovement in this important area, espe-cially along three dimensions: (a) the specificchannels through which the proposed Bank



program was expected to work—in otherwords “the results chain”; (b) the analysis ofthe Bank’s comparative advantage vis-à-vis itspartners to explain the CAS’s strategic selec-tivity; and (c) the specific implications of thelessons learned from past country and Bankimplementation experience for the design ofthe strategy.

Strategic relevance and selectivity are alsocentral themes of OED’s CAEs and othercountry program evaluations, which havebeen carried out in more than 60 countriessince the first such evaluations, for theArgentina and Ghana programs, in fiscal1995.29 As OED notes: “While each projectproposed in the CAS may individually beconsistent with CAS objectives, it is notalways clear that the summation of Banklending in the CAS is the best way to achieve

CAS objectives.”30 In several country evalua-tions, OED has rated overall country pro-gram outcomes as unsatisfactory even thoughthe large majority of the individual projectoutcomes earned a satisfactory rating. Inother words, the Bank did things right at theproject level but did not necessarily do theright things to achieve stated CAS objectives.


A critical part of the World Bank effort tosupport country development is the analyticand advocacy work it does in the globalarena, especially with respect to the policiesand actions of developed countries on trade,aid, and debt relief, given their importance inachieving the MDGs, as discussed in earlierchapters. In addition, as noted there, theBank is monitoring international scaling-up



The results-based Country Assistance Strategy is designed to improve the strategic relevance andselectivity of Bank country programs and provide greater support for strengthening country capac-ity to manage for results, thereby increasing the Bank’s contribution to country outcomes. The con-cept of a results-based CAS was first elaborated in 2002 as a central element of the Bank’s actionplan on managing for results.

The results-based methodology involves a change in mindset and approach to formulating theCAS—from starting with programming inputs and then analyzing their likely impacts, to startingwith desired outcomes and then identifying what inputs and actions (by the Bank and others) areneeded to achieve them. Supporting and complementing this change in mindset, the results-basedCAS introduces a framework for articulating expected outcomes and identifying indicators fortracking implementation progress and evaluating outcomes at program completion. Strengthenedmonitoring and evaluation at the CAS level is critical to success, including introduction of a CASCompletion Report, or self-assessment, of progress under the previous CAS and review of this reportby OED.

Interim guidelines on results-based CASs have been issued,a and a pilot phase of preparingresults-based CASs is under way. Five results-based CASs were presented to the Board in 2003—Brazil, Cameroon, Mozambique, Sri Lanka, and Ukraine. More pilots are being completed in fiscal2004, and other country teams are beginning to apply the results-based methodology to CAS design.Further work includes implementing regional plans to support country teams in preparing results-based CASs, evaluating the pilot phase for results-based CASs in late fiscal 2004, issuing a revisedoperational policy and a good-practice note for results-based CASs, and mainstreaming the results-based CAS in fiscal 2005.

a. See (http://opcs/CAS/cs-g.html).Source: World Bank 2004d.

BOX 13.2 Results-based CAS

efforts in four priority areas for the servicedelivery MDGs—Education for All, HIV/AIDS, health, and water and sanitation. It isalso directly providing support in these andother areas through its country and globalprograms. One example is its support for theGlobal Fund to Fight AIDS, Tuberculosis,and Malaria, for which the Bank is trustee.

Bank support for global programs beganthree decades ago, with the establishment ofthe Consultative Group on InternationalAgricultural Research (CGIAR), for whichthe Bank is a convener and donor to the sys-tem, as well as a lender to developing coun-tries for complementary activities.31 Newglobal programs were gradually added overtime, with a major step-up in global partner-ships and associated program support activi-ties commencing in the late 1990s, reflectingthe rapid pace of globalization, the sharplyincreased attention to global policy issues inthe development community, and the Bank’sincreased partnership orientation. In Septem-ber 2000 the Development Committee en-dorsed the Bank’s priorities in supportingglobal public goods, focusing on five areas—public health, protection of global commons,financial stability, trade, and knowledge.32

Global programs are now reflected inBank corporate strategy papers and opera-tional activities, with about 50 programs(managed by either the Bank or externalrecipients) receiving grants from the Devel-opment Grant Facility. In these programs, theBank is working in capacity-building andsupport programs with countries, to helpthem meet their requirements under interna-tional agreements, and in partnership pro-grams focused on the delivery of global andregional public goods, including by providingseed money for new such programs.33

The assessment framework for global pro-grams—both within the Bank and in otheragencies—is at a much earlier stage of develop-ment than it is for country programs, reflectingthe more recent vintage of most global pro-grams. Within the Bank, the self-assessmentframework is still being developed—as is the exante appraisal framework—while on indepen-

dent evaluation, OED has recently completedits Phase I review of the Bank’s global pro-grams and is in the final stages of the analysisof Phase II. As a general matter, OED con-cluded that the Bank has played a useful rolein these programs by providing a platform forlearning, advocacy, and collaborative action toaddress key global challenges. With notableexceptions, including large and high-profileprograms such as the Global EnvironmentFacility (GEF) and CGIAR, for the most partOED found the programs to be underman-aged, especially relative to country programs,with too little attention to formal appraisaland evaluation criteria, too little policy con-tent, and unclear accountabilities.34

Building on the work of a high-level inter-nal review team, Bank management has gen-erally endorsed OED’s recommendations andis paying increased attention to strategicfocus, country alignment (with any imple-mentation at the country level included in theCAS program), developing country voice,business planning and resource management,and risk management and quality assurance inthe Bank’s global programs.35 Its enhancedefforts in these areas aim to build on a seriesof measures adopted in recent years toimprove the governance and oversight of theBank’s global programs and partnershipsthrough new processes for screening proposedprograms and ensuring their strategic focus,better systems for tracking implementation,and enhanced attention to independent andself-evaluation on program completion.


The World Bank’s policy is to operate jointlywith partners when addressing major devel-opment issues. This policy applies equally toBank country and global program activities,broadly as follows. First, the Bank worksclosely with the IMF, the MDBs, the UnitedNations and U.N. agencies, OECD-DAC,the European Union and bilateral donors,WTO, and other partners in its country andglobal work. Second, as a global develop-ment institution with broad coverage acrosscountries and issues, the Bank often plays a



strategic role in providing what smaller andmore specialized agencies cannot because oftheir size or narrower mandate—a role thatunderpins the way strategic selectivity mani-fests itself in Bank CASs, as discussed earlier,and that shapes the particular Bank contri-bution to global and regional partnershipsand programs. Third, in carrying out its part-nership activities, the Bank wears differenthats as befits the occasion—leader, follower,adviser, helper, and so on. It need not anddoes not play the lead role in every instance.This point warrants emphasis both internallyin guiding staff behavior and externally,especially in light of the findings of theBank’s global poll with respect to percep-tions of institutional and staff arrogance.36

Underpinning the Bank’s partnership pol-icy is its extensive work on harmonization—designed to better align its processes andprocedures with those of clients and withthose of other agencies. In turn, such align-ment reduces the transaction costs of devel-opment assistance. To this end, the Bank isintensifying its collaboration with interestedclients and partners to extend and deepen theincreasing number of country-level harmo-nization activities. Many of these were begunbefore the Rome Harmonization Forum inFebruary 2003 and several have been initi-ated over the past year.37 The Bank is cur-rently pursuing the harmonization agenda ina number of countries, playing either leader-ship or supportive roles as the occasion andcircumstances warrant. These countriesinclude the seven associated with the RomeDeclaration, as well as a growing number ofothers in which harmonization programs andactivities have just started or are being broad-ened or deepened.38 This work, of course,relates closely to the Bank’s ongoing supportfor country-led partnerships for the imple-mentation of national development strate-gies, whether in the context of PRSPs forlow-income countries or of country-specificvehicles in the case of middle-income coun-tries. And, for all countries, Bank manage-ment is using internal review processes(including the Operations Committee) for

CASs and lending operations, to proactivelyidentify opportunities for further country-level harmonization. The World Bank is alsoplaying leadership and supportive roles in thecontext of the DAC Working Party on AidEffectiveness, which was established after theRome forum to support and facilitate harmo-nization efforts.39

Supporting and complementing these coun-try-level efforts, the Bank has taken severalimportant steps in recent years to modernizethe fiduciary framework governing its lendingoperations. These steps have helped to set thestage for Bank participation in harmonizationwith country and partner systems. New finan-cial management guidelines allow borrowersto submit project reports based on their ownfinancial reporting systems.40 New fiduciaryprocesses permit the Bank to participate inpooled financing arrangements in sectorwideapproaches (commonly known as SWAps)characterized by common arrangements forfinancial reporting, auditing, procurement,and disbursement. The Bank’s audit policieshave been aligned with international auditingstandards and good practice, allowing foradaptations of audit scope to assessed projectrisk.41 A new loan administration platform isbeing designed to facilitate harmonization ofdisbursement procedures with country proce-dures and those of other lenders. Finally, majorinvestments have been made in the Bank’sfinancial management diagnostic work, espe-cially on the country financial accountabilityassessment, to upgrade quality, work jointlywith MDB and other partners, and inform thecountry policy dialogue on financial manage-ment systems. These investments complementjoint work with the IMF, the European Com-mission, and bilateral donors on the PublicExpenditure and Financial Accountability(PEFA) initiative and the OECD-DAC harmo-nization process to help clients strengthen theirpublic financial management systems.

The Bank has taken complementary stepson procurement procedures to improve thescope for harmonization with country sys-tems and with MDB (and other) partners andto upgrade their capacity building and learn-



ing content. To these ends, the Bank hasintroduced higher prior-review thresholdswhen client capacity warrants and clarifiedwhen local procurement laws and practices—including for e-procurement—can be used inBank-financed (and cofinanced) projects. Aswith financial management, the Bank has alsotaken steps to transform the country pro-curement assessment into a diagnostic toolfor the policy dialogue, with country pro-curement assessments now carried out jointlywith clients and MDB and other partners.The Bank is also working with the OECD-DAC on an initiative for helping countriesstrengthen their procurement systems, whichalso will lead to stronger in-country systemsaround which donors can harmonize theirsupport. Finally, based on the master docu-ments produced and agreed by the MDB Pro-curement Harmonization Group, the Bankhas issued new documents on the procure-ment of civil works and goods. In November2003, the Board approved adjustments to theBank’s Procurement Guidelines that reflectedthese and other changes.


The bottom-line measures of the Bank’s con-tribution are the quality and results of itsoperational products and services. This sec-tion looks at Bank systems for measuring andmonitoring quality and results and then sum-marizes what those systems suggest about theBank’s contribution to development out-comes and its institutional performance.

Monitoring Systems. Historically, the Bankmeasured its operational performance pri-marily by lending commitments—both dol-lars lent and projects approved. Building onthe findings of and the follow-up to theWapenhans Task Force Report,42 the qualitydimension was added in 1996 as a second pri-mary indicator of operational performance.Adding the quality of lending and analyticand advisory services as an indicator servedto focus Bank management attention on thatdimension, and, after the investment of muchtime, resources, and commitment— including

the creation of the Quality Assurance Group(QAG), with a major program and the fullsupport of top management—the Bank’s per-formance on quality improved substantially.Adding results as a third indicator was a log-ical next step, taken in 2002 to furtherimprove the Bank’s effectiveness by subject-ing this important measure to more system-atic management scrutiny. In the context ofthis decision, the Bank adopted an ambitiousplan to better measure, monitor, and managefor results; the Development Committeeendorsed this plan in September 2002.43

Since then, there has been significantprogress in designing and piloting the neces-sary changes in Bank systems to implementthe results agenda. Central to the agenda areeffective monitoring and evaluation systems.This is true not only for investment lending,for which there is a long tradition of monitor-ing and evaluation, although implementationperformance needs to be improved, but alsofor the CAS, adjustment lending, and non-lending services, for which monitoring andevaluation is a more recent development. Inall these areas, work is under way to see howbest to apply, adapt, and improve existingapproaches, building on the lessons learnedfrom recent monitoring and evaluation pilotsand OED reviews. As noted earlier (see box13.2), country teams are piloting results-basedCASs that identify country outcomes (fromthe PRSP or other national strategy) to whichthe Bank will contribute, along with interme-diate indicators linked to the particular prod-ucts and services that the Bank will provide.With respect to the Bank’s lending and non-lending products and services, operationalpolicies and processes are being reviewed withthe aim of expediting implementation and theachievement of results, while documentationrequirements are being reviewed with the aimof increased transparency in the reporting ofresults objectives and achievements.

Consistent with the above, OED’s mostrecent assessment of Bank systems concludedthat the framework for self-evaluation byBank management, and independent evalua-tion by OED, is strongest at the project



level.44 At the country level, OED found thatthe framework for independent evaluation,comprising OED CAEs, was well established,and it welcomed the recent introduction ofCAS Completion Reports, which it saw asresponding to its long-standing recommen-dation and closing an important gap in self-evaluation at the country level. But OED alsostressed the importance of the remainingweaknesses in independent evaluation of sec-tor strategies and global programs and ofgaps in self-evaluation of sector strategies,nonlending operations, trust funds, andknowledge initiatives, while also acknowl-edging that new initiatives are under way onself-evaluation of global programs. On thelatter, the first annual Sector Strategy Imple-mentation Update will provide a basis forimproved monitoring and self-assessment ofsector strategy implementation, includingthrough the development of strengthenedmonitoring indicators.

Meanwhile, in response to a request fromthe IDA Deputies, a two-tiered Results Mea-surement System has been developed to mea-sure progress on selected country outcomesacross IDA countries and to measure IDA’sperformance at the institutional level incontributing to development outcomes (box13.3).45 The results-based CAS framework isexpected to provide the necessary inputs in themedium term, but in the meantime, Bank man-agement expects to use IDA portfolio indica-tors derived from OED ratings for projectoutcomes and QAG quality assessments,which are leading indicators of success inachieving CAS outcomes. Although OED doesnot routinely conduct impact evaluations,some increase in such activities may be war-ranted, in view of their relevance to the resultsagenda, and to the overall framework forassessing the Bank’s contribution. For exam-ple, a recent impact evaluation of the Bank’ssupport to education in Ghana found—on thebasis of a careful analysis of the results chainand examination of data collected specificallyfor the study—a very substantial positiveimpact on education outcomes.46

Effectiveness and Results. Looking beyondthe systems to the actual outcomes of theBank’s assistance, what does the evidenceshow? Here the real issue is how to demon-strate Bank impact. OED considers that theBank’s country strategies have been on thewhole fairly successful, based on what it seesas relatively high CAE outcome ratings—with 65 percent of all years assessed and 75percent of the post-1998 years rated moder-ately satisfactory or better. Once the results-based CAS has been mainstreamed, theself-assessment framework for such judg-ments also would be available. CASs wouldbe capable of being evaluated to start with,providing a basis for judging and comparingscores in achieving CAS outcomes. But inview of the considerable lead time required tohave a large enough cohort for meaningfulassessment, the leading indicators of CASimpact will need to be relied on until fiscal2006 or fiscal 2007.

Overall, these indicators—OED and QAGscores—point to consistently improving port-folio quality from fiscal 1997 up to fiscal 2002,as the major steps that management was tak-ing were being reflected in a step-up in quality.OED evaluations of projects exiting the port-folio each year confirm the positive trend. Fis-cal 2002 marks the third consecutive year ofproject performance exceeding the Bank’sStrategic Compact target of 75 percent satis-factory or better outcomes.47 By number ofprojects, the score was 79 percent; weighted byvalue, it was 85 percent. More dramatic is thecontinued upward climb in sustainability rat-ings to similarly high levels from much lowerstarting points. However, for fiscal 2003 exits,early OED ratings point to a decline in out-come scores. Pending the completion of OED’sanalysis of the full cohort of projects com-pleted in fiscal 2003, QAG has launched a spe-cial analysis of possible issues that may need tobe addressed by management in ensuring thecontinued improvement in portfolio manage-ment and quality.

Both QAG and OED have emphasized therole of economic and sector work in the Bank’s





effectiveness. QAG findings point to theimportance of ESW, and its increasing quality,as leading indicators of positive outcomes inBank operations and country programs.Assessments by QAG confirm the continuingimprovement in the quality of the Bank’s coun-try analytic work, reflecting the increasedattention to it by senior Bank management inrecent years. OED’s findings are that in coun-tries where recent outcomes of Bank assistancewere evaluated as satisfactory, high quality, rel-evant, and timely, ESW generally made a sub-stantial contribution. Where outcomes werenot satisfactory, deficiencies of ESW in one ormore of these dimensions were a contributingfactor. CAEs report favorable outcomes whenhigh-quality ESW was timed to precede Bankoperations and country programs, and unfa-vorable outcomes when ESW was not timed inthis way. Even when ESW was timely and ofhigh quality, the relevance of Bank strategieswas reduced when its findings were not used orused only selectively in programs and lendingoperations. ESW was found to be particularlyimportant for first-time or renewed borrowersand for stop-go reformers.

Regional Development Banks

This section looks at the role and contribu-tion of the four major regional developmentbanks—African Development Bank, AsianDevelopment Bank, European Bank forReconstruction and Development, and Inter-American Development Bank—also using theorganizational structure of box 13.1. 48


Notwithstanding differences in mandates andinstruments, each of the regional banks isfocused on country issues. All have countrystrategy papers and are increasingly involvedin the PRSP process, in line with the agreedMDB/IMF Protocol on Collaboration on thePRSP. The regional banks are increasinglyfocused on the MDGs, with the EBRDfocused on transition impact. Their strategiesin turn drive their country lending and non-

lending programs. Their independent evalua-tion departments assess their strategies onprogram completion, thus affording oppor-tunities to test and validate results and takestock of the lessons learned.

Work is under way in each of the fourregional banks to adapt and improve existingapproaches to country strategies, in line withthe results agenda. The AfDB has aligned itscountry strategy papers more closely withPRSPs, deepening the analysis, based onpoverty diagnostics, sectoral priorities, andoutcome indicators articulated in country-owned PRSPs, and spelling out in the countrystrategy paper how the proposed AfDB lend-ing and nonlending activities will contributeto poverty reduction. In the AsDB, there alsohas been progress in linking country strategypapers to PRSPs, and work is under way todesign a results-based country strategy paper.In the EBRD, individual country strategiesare results based as they are evaluated againstthe performance of the country portfolio ofprojects and against the sector reform andtransition challenges tracked by the Transi-tion Report and sector strategies. The IDBintroduced new guidelines for its countrystrategies last year, requiring, inter alia: anexplicit focus on results, with a “strategymatrix” articulating the link between thecountry’s own development objectives andstrategy and the development objectives andstrategies of all donors, including the IDB;identification of the proposed IDB-assistedinterventions together with performance indi-cators and targets of expected results; andreflection of the findings of the country pro-gram evaluations prepared by the IDB’s inde-pendent evaluation office.


The regional banks are involved in global andregional programs covering financial stability,trade, environment, post-conflict assistance,and knowledge, with all but EBRD alsoinvolved in the control of infectious diseases.49

In many cases the banks are focused onregional public goods (RPGs), or on regional

aspects of global public goods, looking to theWorld Bank on the global aspects. They alsoare involved in helping their regional clientsbuild country capacity to meet requirementsunder global agreements. For the AfDB, criti-cal issues are post-conflict assistance andhealth, especially in the face of the HIV/AIDSepidemic. For the AsDB, key issues are theenvironment, health, and knowledge, with a

particular focus on those issues where thereare spillover effects within the region, orwithin the AsDB’s subregional coverage. Forthe EBRD, nuclear safety is an area of specialfocus, where the Bank has the internationallead in supporting transition countries in thedecommissioning of capacity, along with theresolution of other environmental liabilities ofthe earlier era. Another is financial stability,



The proposed IDA14 results-measurement system is designed to reflect the priorities and processesof national poverty reduction strategies, be linked to the MDG framework, show aggregated resultsacross IDA countries, and assess IDA’s contribution to development results. It measures results ontwo levels:

Aggregate country outcomes. The first tier includes 17 possible indicators to measure theprogress of IDA-eligible countries on core development outcomes (see table). The majority of theseindicators were chosen to be consistent with country priorities articulated in national poverty reduc-tion strategies, aligned with MDG indicators, and relevant to IDA’s mandate and activities in bor-rowing countries. Indeed, most of the indicators are considered in PRSPs, either as specific targetsor as subjects for discussion in the text. Ten are MDG indicators. The others are complementary,relating to growth for poverty reduction, and reflecting IDA’s support for the economic growth, pri-vate sector development, and public sector management that are necessary to reduce poverty.

IDA’s contribution to country outcomes. The second tier involves introducing a stronger focuson results and a self-assessment system in World Bank Country Assistance Strategies in IDA-eligi-ble countries, and assessing the quality and outcomes of projects in the IDA portfolio, drawing ondata from the Operations Evaluation Department and the Quality Assurance Group. The follow-ing indicators have been put forward to monitor progress in this tier: the number of countries thatuse a results-based CAS; CAS final outcome ratings as validated by OED through the CAS Com-pletion Report review; project outcome ratings as validated by OED through the ImplementationCompletion Report review; and quality-at-entry indicators for IDA projects as assessed by QAG.

In April 2003 IDA Deputies reviewed this architecture and found it to be a sound basis for mov-ing forward. At their first meeting of the IDA14 replenishment in February 2004, Deputies reem-phasized the importance of measuring development effectiveness and results at the country leveland identified this area as a key theme for further discussion and work during the replenishment.Over the coming months, IDA Deputies will face important decisions about the monitoring ofaggregate country outcomes within the results-measurement system. They will need to reach con-sensus on a set of country outcome indicators, each with different merits. The 17 indicators pro-posed for consideration are those that—at this time—best meet the three criteria of relevance to keydevelopment outcomes, sensitivity to policy actions, and measurability in a sufficient number ofIDA countries. However, to varying degrees, the ability to monitor these indicators on a regularbasis—and the quality of the resulting information—are dependent on expanded coverage,increased periodicity, and standardization of questions within household surveys and other data-gathering mechanisms. The existence of reliable and relevant indicators to measure developmentprogress will also depend on the inclusion of the issue of statistics and evidence-based policymak-ing in the policy dialogue with countries. The implications for the international community aretwofold. First, a greater financial and technical commitment will be needed to strengthen statisticalcapacity and monitoring and evaluation systems in low-income countries and to reinforce inter-national reporting systems. Second, expectations must remain realistic for improvements across a

BOX 13.3 Proposed IDA14 results-measurement system

especially the adoption of the standards andcodes underpinning market economies. TheIDB has five priority areas in the provision ofregional and global public goods—financialsector assessments, regional integration, curb-ing of infectious diseases, promotion of envi-ronmental services, and support for researchin agriculture and regional policy dialogue. Ithas prepared a new policy framework for its

support for RPGs, including a financing facil-ity geared to providing grant financing forwhat it calls “early stage RPGs,” where dia-logue among countries is needed; “later stageRPGs,” where larger institutional resources tomanage the emerging program are needed;and the initial stages of “club RPGs,” whichwill likely be financially self-sustaining oncethey are up and running.50



Outcome indicators under consideration for IDA14 results-measurement system

PRSPs that PRSPs Availability Agency include covering in WDI Typical responsible

indicator subject database frequency for data Indicator % % % of reporting compilation

1. Proportion of population below 13 100 72 Every 3–5 years World Bank$1/day poverty line

2. Prevalence of underweight children 44 63 97 3 years UNICEF, WHOunder five years of age

3. Under-five mortality 72 97 100 3 years UNICEF, WHO4. Proportion of year-old children 9 72 100 Annual UNICEF, WHO

immunized against measles5. HIV prevalence rate of pregnant 3 69 91 Only 1999 UNAIDS, UNICEF

women 15–246. Proportion of births attended by 59 72 84 3–5 years UNICEF, WHO

skilled health personnel7. Ratio of girls to boys in primary, 69 78 94 Annual UNESCO

secondary, and tertiary education8. Primary school completion rate 31 100 100 Annual UNESCO9. Proportion of population with 94 94 100 3 years UNICEF, WHO

sustainable access to an improved water source

10. Fixed lines and mobile telephones 25 34 97 Annual ITU, World Bankper 1,000 inhabitants

11. Formal cost of business registration 3 13 72 Annual World Bank12. Time for business registration 6 13 75 Annual World Bank13. Public expenditure management 0 100 — TBD World Bank14. Agricultural value added 22 66 94 Annual UNSD, World Bank15. GDP per capita 41 100 94 Annual UNSD, World Bank16. Access of rural population to an 16 44 — TBD World Bank

all-season road17. Household electrification rate 34 59 — 3 years World Bank

— Not available.Note: TBD = to be determined.

range of indicators in countries with limited capacity. Thus, it is especially important to reach con-sensus on a small number of indicators that countries identify as highly relevant for managing theirdevelopment processes.

Source: World Bank, Concessional Finance and Global Partnerships Vice Presidency.


For the regional banks, the partnership topichas two interrelated aspects. One, at the insti-tutional level, is their role and participationin the harmonization effort; the other, at thecountry level, is their performance as part-ners. Reflecting their growing interest in both,the regional development banks are activeparticipants in the DAC Working Party onAid Effectiveness.

Working in partnership with the WorldBank and others, the regional developmentbanks are continuing their work on the unfin-ished institutional harmonization agenda andthematic areas of common concern. Centralto the agenda is the focus on the harmoniza-tion of results reporting emerging from theMarrakech Roundtable. As noted there,developing countries have made significantprogress on country ownership, strategicvision, and donor alignment through thepoverty reduction strategy process, but manycountries continue to struggle with the diver-sity of donor reporting requirements, espe-cially with respect to project monitoring andevaluation—hence the urgency of action.Legal documentation and disbursement pro-cedures are examples of issues that haverecently become subjects of discussion andwill be taken forward. Meanwhile, existingworking groups have continued their workon institutional harmonization in tandemwith their country-level work. The FinancialManagement Group, for example, has com-pleted good-practice papers and continues towork with relevant professional bodies on aninternational accounting standard for devel-opment operations. In addition to its work oncountry-level capacity building, the Procure-ment Group has completed harmonized bid-ding documents for prequalification for civilworks and civil works contracts and forselection of consultants. The EnvironmentalGroup has completed a common frameworkfor environmental impact assessments, whilefocusing most of their efforts on country-levelwork. And the Evaluation Group has fol-lowed up on good-practice papers on publicand private sector evaluations, through a

benchmarking exercise, and put joint countryevaluations on its agenda. The MDB GenderGroup, operating in cooperation with U.N.and OECD-DAC gender groups, has recentlyco-sponsored a workshop on Gender and theMDGs.

For country-level activities, the partner-ship issue is typically less about partnershipsbetween individual regional banks, becausethere is little overlap among them. Rather theissue is more often about partnerships withthe World Bank, IMF, U.N. agencies, andbilateral agencies, with which the regionalbanks are linking up, together with govern-ments interested in harmonization efforts atthe country level. To a large extent, the issuehere is one of country strategy and the selec-tion of operational support vehicles takinginto account what other partners are provid-ing. But it may also involve harmonization ina specific lending context, especially wherenew sectorwide approaches are challengingprevious methods of support for countrydevelopment and co-financing arrangements.Concrete cases, such as the Bangladeshexample (box 13.4), put the focus on wherethe really tough work on harmonizationlies—reconciling differences in institutionalguidelines for procurement and financialmanagement within a particular countrycontext in a real-time situation.


The recent Marrakech Roundtable, spon-sored by the MDBs in collaboration with theOECD-DAC, provided an important oppor-tunity for the banks to take stock of wherethey are in their application of results-basedapproaches and the progress made since thefirst Roundtable in June 2002.51

Systems and Processes. The regional develop-ment banks have all been actively pursuingthe results agenda in the context of theircountry assistance strategies, as discussedabove, and in their business processes forassuring the quality and development effec-tiveness of their operations. To this end, allthe banks are giving greater attention to the



quality and impact of their operationsthrough enhanced monitoring and evaluationand portfolio management. In previous years,the AfDB had taken a number of measures toimprove project quality at entry, supervision,monitoring and evaluation, and portfoliomanagement. These measures are now beingextended and deepened in line with the resultsagenda, where the focus is on “higher level”interventions and evaluations at the country,sectoral, and regional levels. AsDB has signif-icantly improved its portfolio managementand monitoring of project implementation tomake it more results focused. Project docu-mentation throughout the project cycle hasbeen refocused on development objectives,with lessons learned highlighted in projectcompletion and audit reports. Special atten-tion is also being paid to the quality and ana-

lytical content of AsDB’s economic, technical,and sector work, as a basis for the policy dia-logue with clients. Last year, the EBRD intro-duced a new Transition Impact MonitoringSystem, with periodic monitoring and report-ing on project performance. The new systemis designed to facilitate a portfolio approachto project monitoring. Progress reporting isdone both on a project and on a portfoliobasis to allow a comparison of the transitionimpact potential of the portfolio against othersources of project risks and returns, and othersector and country dimensions. The IDB hastaken a number of important steps to upgradeproject quality and management. Building ona pilot quality-at-entry assessment, improve-ments are being made to project design andevaluation, including the introduction of astructured self-evaluation system for use



The Bangladesh Second Primary Education Development Project is a concrete example of how theharmonization and alignment agenda can be implemented at the country level to improve aid effec-tiveness and support the MDGs. Investment in primary education is a key to reducing poverty,enhancing opportunity, and improving the quality of life. Bangladesh has made progress in primaryeducation in the last two decades. Nonetheless, significant problems remain: in the quality dividebetween the rich and poor, and in access of the poor, especially of girls. Donors have responded torequests to assist—indeed so much so that 11 donors were funding 27 separate projects in primaryeducation, all with their own separate donor-specific operating requirements. To try to achievegreater coordination, the government made a bold move. Supported by the local donor community,the government adopted and endorsed a Macro-Plan, and a sector strategy, which outlines the objec-tives for primary education over the next six years and sets out the policy framework and imple-mentation plan for improving education quality, enhancing access to schooling, and amelioratingoverall management and oversight of primary education. Rather than a series of projects, the gov-ernment worked with donors, led by the Asian Development Bank, to develop the Second PrimaryEducation Development Project as a coherent program (a sectorwide approach, or SWAp) in sup-port of the strategy, which covers all public primary schools in the country.

While the program approach may stretch government capacity, the government and develop-ment partners have achieved a satisfactory congruence of goals and aspirations through the col-laborative and participatory process, including agreement on results that can be monitored andquantified. The multidonor-supported program approach aims to reduce transaction costs for thegovernment. Key aspects of the SWAp are to replace 27 different project structures with 1 overar-ching structure, with cofinanced funds from the AsDB, IDA, and six other partners disbursedthrough a pooled account, on which AsDB reports to other donors, and with 90 percent of pro-curement carried out by the government under a newly enacted law.

Source: World Bank, Operations Policy and Country Services Vice Presidency.

BOX 13.4 Sectorwide approach to primary education development

during project supervision, where reporting isbeing strengthened, and also on project com-pletion, where the coverage by the indepen-dent evaluation unit is being expanded.

The regional banks also are using corpo-rate reporting for informing shareholders andstakeholders about progress on the resultsand quality agenda, recognizing the criticalrole such reporting plays in sustaining man-agement attention to the agenda and in turninstitutional follow-through. The AfDBreports on results through three main chan-nels. First, much operational reporting onquality, results, and evaluation takes placethrough the Board Committee on Develop-ment Effectiveness. Second, AfDB is pilotingthe Results Measurement Framework, whichit provides to the African Development FundDeputies, as part of the latter’s oversightfunction. Finally, AfDB is beginning to imple-ment a Corporate Balanced Scorecard inreporting to its Executive Board on strategicplanning and budgeting.

The AsDB is reporting through the dia-logue on results with the recently createdBoard Committee on Development Effective-ness and with the Asian Development FundDeputies, including in the context of thedesign and implementation of the AsianDevelopment Fund Results MeasurementSystem. In addition to project-based resultsmanagement, the EBRD uses results-basedmanagement for its own institutional perfor-mance and budgeting, with a focus on transi-tion impact and financial indicators in itsinstitutional scorecard, which it shares withits Executive Board. Both the transitionimpact and financial performance objectivesare reviewed annually by the Board in thecontext of EBRD’s Medium-Term StrategyUpdate and Strategic Portfolio Review; thesereviews inform the EBRD’s annual dialoguewith its Board on the budgeting process. TheIDB is also strengthening its corporate report-ing, including to its Board Committee on Pol-icy and Evaluation. At the 2004 AnnualMeetings in Lima, Peru, the Committee of theBoard of Governors discussed progress onIDB’s development effectiveness reform

efforts to date, and a Medium-Term ActionPlan for enhancing IDB’s development effec-tiveness and results orientation is currentlyunder preparation.

Effectiveness and Results. The independentevaluation departments of the four regionalbanks fill broadly similar roles within theirorganizations, contributing to institutionallearning and accountability. To some extentthey also provide a basis for comparing perfor-mance across the institutions—although thereare limits to comparability, as the measuresused in the different banks are not exactly thesame. Going forward, harmonization of evalu-ation criteria, both for individual operationsand for aggregate institutional scores, would beworthwhile, building on the earlier effort by theEvaluation Cooperation Group to assess thedifferences across the MDBs and to produce agood-practice standard.52

In the meantime, the following profiles areillustrative of the situation across the regionaldevelopment banks, based on publicly avail-able material. The AfDB reports in its annualreport that in general the latest annual evalu-ation review found that the overall perfor-mance rating for outcomes was satisfactory orbetter for more than two-thirds of projects;the percentage scores for institutional devel-opment and sustainability were appreciablylower.53 According to AsDB’s annual evalua-tion reports, 40 percent of completed projectsand 100 percent of programs are selected forevaluation each year. In 2001, of the 17 proj-ects and 3 programs evaluated, none werefound to be unsuccessful, with 14 percenthighly successful, 41 percent successful, and45 percent partly successful. In 2002, of the27 projects and 6 programs, none were unsuc-cessful, while 12 percent were highly success-ful and 18 percent only partly successful.

In the EBRD, which provides extensiveand transparent information on its evalua-tion activities and findings, the focus is ontransition impact, covering privatization,competition, and corporate governance, andon overall performance, which includesfinancial and environmental performance in



addition to transition impact. 54 On overallperformance, 52 percent of operations eval-uated during 1996–2002 were given success-ful or highly successful ratings, while 74percent were rated successful or highly suc-cessful on transition impact (the differenceswere largely attributable to low scores onfinancial performance). For IDB, 67 percentof operations reviewed in 2001 were foundto have adequately defined output indicators,according to a report of the independentevaluation office, a finding that contributedto the adoption of some of the reforms men-tioned earlier.55

Systemic IssuesThere is evidence that the Bank, the Fund, andthe MDBs are working better together, as wellas with other partners, including the U.N.agencies, WTO, OECD-DAC, and EU andbilateral donors. This brings benefits to clientsand partners alike, through the improved effi-ciency and effectiveness of the IFI system. Butfurther progress is possible, which calls forcontinued attention to the issue.

Bank-Fund Collaboration

There has been progress on Bank-Fund col-laboration in operational work in recentyears, reflecting major investments by bothinstitutions toward that end, involving staff,management, and the Executive Boards of thetwo institutions. 56

The PRSP process provides a basis forcoherent and consistent work of the Bankand the Fund, fully aligned with the country’sdevelopment strategy. The Joint Staff Assess-ments (JSAs), which evaluate the soundnessof the PRSPs, have proved instrumental inenabling staffs to develop common views.Reinforcing upstream engagement and coor-dination of the two institutions, using in par-ticular the opportunity offered by the JSAprocess, would further efforts to promotesynergies, better delineate responsibilities insupport of the PRSPs, and reduce gaps andoverlaps.

In middle-income countries, there is noexplicit framework for country-led coordina-tion as in the case of countries with PRSPs.The wide variety of development needsamong MICs and difficult-to-predict shifts incountry needs for balance-of-payments sup-port lead to differences in the timing andnature of Bank and Fund support and in thecontent and scope of conditionality, makingthe implementation of collaboration morechallenging.57 Nevertheless, the principles foreffective collaboration are similar, includingearly consultation on program design andconditionality, and division of responsibilitiesbased on respective mandates and compara-tive advantage. Meanwhile, a key priority formany MICs, especially those with access tofinancial markets, relates to joint Bank-Fundwork on the prevention and resolution ofcrises and the strengthening of member coun-tries’ financial and fiduciary systems.

Staff coordination is supported by twoinstitutional coordination mechanisms: theJoint Implementation Committee (JIC) forcooperation on HIPC/PRSP countries, and theFinancial Sector Liaison Committee (FSLC)for cooperation in financial sector work. TheJIC, established in 2000, has provided a use-ful framework for institutional coordinationon the work on low-income countries. Goingforward, the JIC’s role is being expanded toanchor monitoring of progress on overallBank-Fund collaboration in an institutionalframework. The mandate of the JIC is toaddress cross-cutting issues on Bank-Fundcollaboration; monitor progress on imple-mentation of the framework of collaborationon country programs and conditionality; and,when needed, provide an additional instru-ment to help country teams in the two institu-tions to reach agreement on priorities, thusensuring coherence of policy advice and pro-gram design. This covers issues that arise inboth the low- and middle-income countrycontexts. As such, the JIC will provide the in-stitutional framework for monitoring progresson overall Bank-Fund collaboration.

The FSLC, established in 1998 for cooper-ation on financial sector work, continues to



be useful as a forum through which staffexchange information, coordinate work pro-grams, undertake joint missions, and provideconsistent policy advice to country authori-ties, including for financial sector conditionsin lending operations. These staff interactionsare supported and complemented by broad-based and well-developed mechanisms ofinstitutional coordination. Regular meetingsbetween the Managing Director of the Fundand the President of the Bank, as well asbetween the Managing Directors of the Bankand the Deputy Managing Directors of theFund, provide the foundations for a regulardialogue at the most senior level of both insti-tutions. Area Department management in theFund and Regional Vice-Presidencies in theBank are also in close and regular contact;and so are respective central units on key pol-icy issues of mutual interest, with a growingnumber of joint activities.

Finally, the Executive Boards of both insti-tutions are engaged in monitoring Bank-Fund collaboration, with periodic reports onprogress in key areas. Meanwhile, regularand transparent reporting in board docu-ments of the views of each institution onreform priorities, program conditionality,and progress in implementation of the agreedprogram constitutes a crucial element forensuring consistency of views, transparency,and staff accountability.58 One mechanism,introduced in 2002,59 for improved collabo-ration and communication with the respec-tive Executive Boards, is the enhancedannexes on IMF and World Bank relations inprogram documents. Aside from keeping theBoards abreast of developments, theseannexes are intended to help ensure up-stream engagement between the staff of thetwo institutions, delineate the division ofresponsibilities, and provide an assessmentof the country’s reform efforts.

Multilateral Development Banks

Among the MDBs, cooperation also has con-tinued to grow. As with the evolution ofBank-Fund collaboration, the progress

reflects the response to the signals emanatingfrom top management and from sharehold-ers. The presidents of the MDBs meet at leasttwice a year to review substantive and strate-gic issues, including collaboration amongtheir respective agencies. They also haveissued joint statements on a number of com-mon themes, and cosponsored importantinternational events, the Marrakech Round-table on Managing for Development Resultsbeing only the most recent. At their last meet-ing, the presidents also agreed to strengthenMDB cooperation on capacity building andinfrastructure.

Operational vice presidents and their man-agement teams from the World Bank and therespective regional development banks alsomeet regularly. These meetings include peri-odic consultations on Memoranda of Under-standing (MoU), which have become strategictools for setting out the division of labor inspecific countries and sectors. These MoUs arenow reviewed and updated regularly. Areview of the MoU between the AfDB and theWorld Bank was completed in July 2002. InDecember 2003 a meeting was held in whichan agreement was reached on a correspondingaction plan for country, sector, and regionalthematic cooperation under the MoU. Areview of the MoU between the AsDB and theWorld Bank is scheduled for later this calen-dar year. The EBRD and the World Bank arecosignatories to three country- and issues-specific MoUs, each of which is reviewed onits own time line. The MoU between the IDBand the World Bank is currently under review.

The quality of MDB cooperation in theircountry work is the critical test for determin-ing the benefits this partnership can bring tothe clients. Notwithstanding occasional stafftensions on individual operational issues,there is evidence of an improved relationshipin operations, and better strategic and the-matic coherence is gradually showing up inspecific operations. All MDBs are committedto supporting countries in preparing PRSPs,basing their assistance strategy on the PRSPsor other such country-owned strategies, andcoordinating the strategy with the other



MDB(s) also involved with the country.Cooperation in support of the PRSP processis progressing, but there is recurring frictionover the different roles of the World Bank(and IMF) and the regional banks, not least inrelation to the preparation of the Joint StaffAssessments. After an uncertain start, coor-dination of World Bank and regional devel-opment bank country strategies is nowbecoming more accepted, with recent goodexamples from Honduras and Nicaragua.Discussions have taken place on closer coor-dination or even joint country strategies inCambodia, Papua New Guinea, Mongolia,Sri Lanka, the Pacific Islands, and Uganda.The viability and relevance of such coordina-tion, particularly in larger middle-incomecountries, are not uniformly accepted andwill continue to be discussed. That said, gov-ernment leadership is accepted as the norm toaim for in MDB support for country devel-opment in both low- and middle-incomecountries. Meanwhile, MDBs have continuedto strengthen cooperation (with each otherand other partners) on individual operationalproducts, at times doing such work jointlyand in other cases dividing the work betweenthem. Joint portfolio reviews, public expen-diture reviews, fiduciary assessments, andprocurement assessments are examples of thistype of cooperation, as are activities in sup-port of PRSP preparation. Similarly, there is agrowing number of examples of joint or coor-dinated lending or grant operations, in edu-cation, health, infrastructure, environment,and other areas.

IFIs and Beyond

Until now, this chapter has focused on IFIactivities, both individually and collectively.But also of interest to the Development Com-mittee and others is the question of how theseactivities fit into the evolving landscape forcountry and global development, includingthe activities of other multilateral agencies. Acompanion paper has been prepared on thattopic,60 building on the analysis developed incollaboration with staff in partner agencies

last year and discussed with World BankExecutive Directors.61

That paper sets out a three-tiered architec-ture, governing the coherence, coordination,and cooperation among the various agencies,around which the detailing and assessment ofspecific actions will be structured.

� The first tier—where development actuallytakes place, and the MDGs will be met, ornot—is at the country level. Here the con-sensus both in rhetoric and increasingly inreality is the centrality of national owner-ship, national policies, national systems,and national leadership for successfuldevelopment. This principle is at the coreof the CDF, the PRSP process, the WorldBank’s strategy for assisting middle-incomecountries, the United Nations DevelopmentAssistance Framework, the MDBs’ countrystrategies and country strategy papers, andthe various strategy documents of bilateraland other donor agencies. It was central tothe discussions at Monterrey, Rome, andMarrakech, and as indicated in the paper,there has been much progress across agen-cies—the IFIs, U.N. agencies, the EU, andbilaterals—taking it forward.

� In the second tier are all the multilateraland bilateral support agencies, chargedwith helping countries achieve their devel-opment outcomes. Clearly these agenciesmust coordinate with each other in theircountry work. Increasingly they are doingso, under the leadership of the countryitself, with each agency supporting coun-try priorities according to their respectivemandates and comparative advantage. Butagency headquarters also need to coordi-nate to establish the scope (and policies)for their country representatives to sup-port national activities; shape the strategicdirections and future staffing profiles andskills mix of their respective agencies toensure that, taken as a whole, the interna-tional capacity needed to support countrydevelopment is in place; and share infor-mation and data for monitoring, to informdeveloping countries of how they are



doing vis-à-vis comparators and to tellglobal taxpayers how the development sys-tem is performing. An example of suchcooperation is the inputs to the currentGlobal Monitoring Report provided byOECD-DAC, U.N., and WTO staff and theinputs provided by Bank and Fund staff tothe annual reports of the Secretary Generalto the General Assembly on the implemen-tation of the Millennium Declaration.

� The third tier is in the realm of ideas andobjectives, with the convergence of inter-national thinking of recent years on theends and means of development—cappedby the MDGs and the Monterrey Consen-sus. As discussed in earlier chapters, thereis broad agreement on the MDGs andrelated outcomes as the goals of develop-ment and on the responsibilities of devel-oping and developed countries in terms ofpolicies and actions for achieving them,with the focus increasingly on the imple-mentation of those policies and actionsand on specific measures for monitoringprogress. In turn, this growing consensusand coherence on objectives have openedthe way for greater specialization in agen-cies, grounded in their respective man-dates, and complemented by increasedcooperation with partner agencies to fillgaps in international support for countrydevelopment.

Conclusions Where does this leave us? First, the evidencepresented above points to progress in indi-vidual IFIs on transparency, country focus,and results orientation. However, thereclearly is no room for complacency; a keychallenge will be to extend and deepen theprogress that has been made so far. Importantareas for further improvement include:

� For the IMF, the priority is to continue torefine its role in assisting low-incomecountries to confront the macroeconomicchallenges of achieving sustained high lev-els of growth and poverty reduction. To

that end the Fund is adapting its instru-ments of financial and technical support tothe needs of its low-income members, withparticular attention to how such supportcan be used to catalyze other donor assis-tance, deal with post-conflict situations,assist members in responding to exoge-nous shocks, and establish institutions thatwill enable low-income countries to gainincreasing access to private sector financ-ing. The Fund’s work agenda also aims atstrengthening the design of Fund-supportedeconomic programs in low-income coun-tries, while enhancing alignment with thePRSP. A third element of the Fund’s ongo-ing work, together with the World Bank, isto develop an effective and flexible frame-work for assessing debt sustainability inlow-income countries.

� The World Bank’s country support priori-ties are to continue to work with partneragencies to support country efforts todeepen the PRSP process as a basis for thedesign of its assistance strategies in low-income countries; to adapt its approaches,instruments, and institutional processes tothe evolving needs of middle-income coun-tries; and to complete the major agendathe Bank has set out on managing forresults, harmonization, and simplification.Supporting and complementing the deep-ening of country-led approaches in theBank’s assistance strategies is the strength-ening of its analytic, knowledge, andadvocacy work. A key priority for Bank-supported global and sectoral programs isthe implementation of an effective frame-work for appraisal, monitoring, and eval-uation that is every bit as strong as theframework for country programs.

� The regional development banks also havelarge agendas before them—in their coun-try programs and their support for regionalpublic goods. All need to complete theirongoing reforms associated with the resultsagenda, as they set out at the MarrakechRoundtable. In addition, like the WorldBank, greater efforts are needed on theoverall governance and accountability



framework for their regional and sectoralprograms.

Second, looking across the IFIs, there alsohas been progress—both institutionally and inday-to-day work at the country level. Bank-Fund collaboration and coordination amongthe MDBs are smoother and more productivethan they were as recently as five years ago. Intandem with the increase in partnership andcoordination, there is a healthy trend towardgreater specialization in line with institutionalcomparative advantage. This reverses thetrend of the early 1990s when overlaps inagencies’ capacities increased, as the consen-sus on the comprehensiveness of the develop-ment paradigm was beginning to grow. Butthe “gains from trade” between and amongIFIs have not all been harvested yet. Opportu-nities include increased selectivity of agencyprograms in line with comparative advantage,harmonization of agency practices aroundnational poverty reduction strategies and sys-tems, and joint evaluations of their support.

Finally, it is none too early to begin think-ing beyond the above agenda to the nextphase of IFI reform, focused on dynamiccomparative advantage. Small steps in thatdirection include proactive encouragementof cross-IFI secondments, the broadening ofthe professional networks from individualinstitutional networks into IFI networks, andthe explicit consideration of the IFI dimen-sion in sector strategy papers. Beyond thesemeasures, it will be necessary for the IFIs totackle strategic issues more directly and totake advantage of the enhanced partnershipsacross IFIs, so that they can organize andstaff themselves effectively to support coun-try, regional, and global development.

Notes1. See, for example, Scott 2004 and World

Bank 2003d. 2. Joint Marrakech Memorandum 2004.3. OED 2003b. 4. Ghosh and others 2002; Collyns and Kincaid

2003.

5. IMF 2004b.6. IMF 2001; Boughton and Mourmouras

2004. 7. IMF 2002b. 8. IMF 2002c; IMF–World Bank 2002b. 9. IMF 2003a. The forthcoming PRSP Progress

Report will examine a related set of questions.10. IMF 2003e. 11. For the most recent review, see IMF 2003b. 12. IMF 2003f, chapter III.13. IMF-World Bank 2002c. 14. IMF-World Bank 2004.15. IMF 1998. 16. For external assessments, see IMF 2004a,

pp. 41-51.17. IMF 2002a. 18. IMF 2003c. 19. IMF 2003d. 20. World Bank 2003d.21. IMF–World Bank 2004. 22. Development Committee 2000b. 23. The latest retrospective report, issued in

March 2003, examined the 28 CASs and 11 CASProgress Reports discussed by the Executive Boardin fiscal 2000 and the first half of fiscal 2001. Itassessed the progress made in improving the qual-ity of the different dimensions of the CAS since theprevious retrospective, issued in May 2000. SeeWorld Bank 2000, 2003b.

24. Development Committee 2003. 25. OED 2003e. 26. OED 2003a. 27. OED 2004a.28. World Bank 2004b.29. OED 2003c. 30. OED 2003a. 31. OED 2003d. 32. Development Committee 2000a.33. World Bank 2003c. 34. OED 2002. 35. World Bank 2003f.36. World Bank 2003d. 37. World Bank 2003e. 38. Including Bangladesh, Bolivia, Brazil, Cam-

bodia, Colombia, Dominican Republic, Egypt,Ethiopia, Fiji, Ghana, Honduras, India, Jamaica,Kenya, Kyrgyz Republic, Mexico, Mongolia,Morocco, Mozambique, Nicaragua, Niger, PacificIslands, Rwanda, Senegal, Serbia, Sri Lanka, Tajik-istan, Tanzania, Uganda, Vietnam, and Zambia.

39. See chapter 11.40. This flexibility is being implemented in all

new Bank-financed projects; Bank staff are working



with country authorities and other donors to agreeon reporting formats.

41. The associated guidelines for staff andannual financial reporting and auditing for Bank-financed activities were issued in June 2003. Thenew policy applies to all projects appraised fromJuly 1, 2003.

42. Wapenhans Task Force 1992.43. Development Committee 2002. 44. OED 2003b.45. IDA 2003. 46. OED 2004b. 47. OED 2003a.48. This section is based on the progress reports

the regional development banks prepared for Mar-rakech, as well as other sources as specifically noted.See www.managingfordevelopmentresults.org.

49. MDB Joint Liaison Group 2002.50. IDB 2004. 51. World Bank 2004c.

52. MDB Evaluation Cooperation Group 2002. 53. AfDB 2002, p. 56.54. EBRD 2003.55. IDB 2003.56. This section is based on IMF–World Bank

2004.57. The considerations affecting Bank-Fund

collaboration in MICs are discussed in IMF–WorldBank 2001.

58. As indicated earlier in the chapter, in addi-tion to reporting on country matters, Bank andFund staff have been reporting jointly to theirExecutive Boards on a wide range of thematic andpolicy issues of joint relevance to the two institu-tions, such as public expenditure management,trade, FSAPs, and ROSCs.

59. IMF–World Bank 2002a. 60. World Bank 2004a. 61. World Bank 2003a.



Chapter 1 Monitoring Framework

Asrey, Steven. 1996. “Water, Waste and Well-Being: A Multicountry Study.” American Jour-nal of Epidemiology 143(6): 606–23.

Barro, Robert. 2001. “Human Capital andGrowth.” American Economic Review, Papersand Proceedings 91(2): 12–17.

Behrman, Jere, Andrew D. Foster, Mark RichardRosenzweig, and Prem Vashishtha. 1999.“Women’s Schooling, Home Teaching, andEconomic Growth.” Journal of Political Econ-omy 107(August): 682–714.

Berg, Andrew, and Anne Krueger. 2003. “Trade,Growth, and Poverty: A Selective Survey.” IMFWorking Paper WP/03/30. Washington, D.C.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2002. Development CommitteeCommuniqué. Washington, D.C. September22.

———. 2003a. Achieving the MDGs and RelatedOutcomes: A Framework for Monitoring Poli-cies and Actions. DC2003-0003. Washington,D.C. March 26.

———. 2003b. “Background Paper.” DC2003-0003/Add.1. Washington, D.C. March 28.

———. 2003c. “Development Committee Com-muniqué.” Washington, D.C. April 13.

———.2003d. “Poverty Reduction StrategyPapers: Progress in Implementation.” DC2003-0011. Washington, D.C.


References

The word processed describes informally reproduced works that may not be commonly available.

———. 2003e. “Supporting Sound Policies with Adequate and Appropriate Financing.”DC2003-0016. Washington, D.C. September13.

———. 2003f. “Global Monitoring of Policiesand Actions for Achieving the MDGs andRelated Outcomes: Implementation Report.”DC2003-0013. Washington, D.C. September15.

———. 2003g. “Development Committee Com-muniqué.” Dubai. September 22.

Filmer, Deon, and Lant Pritchett. 1999. “The Impactof Public Spending on Health: Does Money Mat-ter?” Social Science and Medicine 49.

Glewwe, Paul. 1999. “Why Does Mother’sSchooling Raise Child Health in DevelopingCountries: Evidence from Morocco.” Journalof Human Resources 34(1): 124–59.

Glewwe, Paul, and E. King. 2001. “The Impact ofEarly Childhood Nutritional Status on Cogni-tive Development.” World Bank EconomicReview 15(1): 81–113.

Goldin, Ian, Halsey Rogers, and Nicholas Stern.2002. The Role and Effectiveness of Develop-ment Assistance: Lessons from World BankExperience. Washington, D.C.: World Bank.

Harrison, Makiko, Eric Swanson, and Jeni Klug-man. 2003. “Targets and Indicators for MDGsand PRSPs: What Countries Have Chosen toMonitor.” World Bank. Processed.

IMF (International Monetary Fund). 2001. “HowDo Fluctuations in the G-7 Countries AffectDeveloping Countries?” World Economic Out-look (October).

Klasen, Stephan. 2002. “Low Schooling for Girls,Slower Growth for All? Cross-Country Evi-dence on the Effect of Gender Inequality inEducation on Economic Development.” WorldBank Economic Review 16: 345–73.

Miguel, Edward, and Michael Kremer. 2001.“Worms: Education and Health Externalities inKenya.” Working Paper 8481. National Bureauof Economic Research, Cambridge, Mass.

Pritchett, Lant, and Lawrence H. Summers. 1996.“Wealthier Is Healthier.” Journal of HumanResources 31(4): 841–68.

Schultz, Paul. 1987. “School Expenditures andEnrollments, 1960–1980: The Effects ofIncome, Prices and Population Growth.” InGale Johnson and Ronald Lee, eds., PopulationGrowth and Economic Development. Madi-son: University of Wisconsin Press.

Udry, Christopher. 1996. “Gender, AgriculturalProduction, and the Theory of the Household.”Journal of Political Economy 104(5): 1010–46.

United Nations. 2000. “Millennium Declaration.”A/RES/55/2. New York, September 18. www.un.org/documents/ga/res/55/a55r002.pdf.

———. 2001. “Road Map towards the Imple-mentation of the United Nations MillenniumDeclaration: Report of the Secretary General.”A/56/326. New York, September 6.

———. 2002a. Outcome of the InternationalConference on Financing for Development,Monterrey Consensus. A/57/344. New York.

———. 2002b. Report of the World Summit onSustainable Development. Johannesburg.

———. 2003. “Implementation of the UnitedNations Millennium Declaration: Report of theSecretary General.” A/58/323. New York. Sep-tember 2.

Wagstaff, Adam. 2003. “Intersectoral Synergiesand the Health MDGs: Preliminary Cross-Country Findings, Corroboration, and PolicySimulations.” World Bank, DECRG/HDNHE,Washington, D.C. Processed.

WHO (World Health Organization). 2001.Macroeconomics and Health: Investing inHealth for Economic Development. Report ofthe Commission on Macroeconomics andHealth. Geneva.

———. 2002. World Health Report 2002.Geneva.

World Bank. 2003. Global Economic Prospects2003. Washington, D.C.

WTO (World Trade Organization). 2001. “Minis-terial Declaration, Adopted on 14 November,2001, Doha.” WT/MIN (01)DEC/1. Geneva.

R E F E R E N C E S


Chapter 2 MDG Prospects: Reasonsfor Optimism, Grave Concerns

Adams, Richard. 2002. “Economic Growth,Inequality and Poverty: Findings from a NewData Set.” Policy Research Paper 2972. WorldBank, Washington, D.C.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2003. “Supporting Sound Policieswith Adequate and Appropriate Financing.”DC2003-0016. Washington, D.C. September13.

Dollar, David, and Aart Kraay. 2002. “Growth IsGood for the Poor.” Journal of EconomicGrowth 7(3): 195–225.


OECD (Organisation for Economic Co-operationand Development). 2004. Development Co-operation Report 2003. Paris.

Preston, Samuel H. 1980. “Causes and Conse-quences of Mortality in Less Developed Coun-tries during the Twentieth Century.” In RichardEasterlin, ed., Population and EconomicChange in Developing Countries. Cambridge,Mass.: National Bureau of Economic Research.

Ravallion, Martin. 1995. “Growth and Poverty:Evidence for Developing Countries in the1990s.” Economic Letters 48(3–4): 411–17.

———. 2001. “Growth, Inequality and Poverty:Looking Beyond Averages.” Policy ResearchWorking Paper 2558. World Bank, Develop-ment Research Group, Poverty and HumanResources, Washington, D.C.

Rokx, Claudia, Rae Galloway, and Lynn Brown.2002. “Prospects for Improving Nutrition inEastern Europe and Central Asia.” Health,Nutrition, and Population series. World Bank,Human Development Network, Washington,D.C.

UNESCAP/UNDP (United Nations Economic andSocial Commission for Asia and Pacific/UnitedNations Development Programme). 2003.“Promoting the Millennium DevelopmentGoals in Asia and the Pacific.” Bangkok.

UNESCO (United Nations Educational, Scientific,and Cultural Organization). 2003. Gender andEducation for All—The Leap to Equality.Paris.

UNICEF (United Nations Children’s Fund). 2002.“End of Decade Databases: Child Mortality.”www.childinfo.org/cmr/revis/db2.htm.

United Nations. 2003. Implementation of theUnited Nations Millennium Declaration:Report of the Secretary General. A/58/323.New York. September.

United Nations Millennium Project Task Force onEducation and Gender Equality. 2004.“Interim Report: From Promises to Action.Recommendations for Gender Equality and theEmpowerment of Women” (draft).www.unmillenniumproject.org/documents/tf3genderinterim.pdf. February.

World Bank. 2001a. Engendering Development:Through Gender Equality in Rights, Resources,and Voice. New York: Oxford University Press.

———. 2001b. Global Economic Prospects 2002.Washington, D.C.

———. 2002. A Case for Aid: Building a Consen-sus for Development Assistance. Washington,D.C.

———. 2003a. “East Asia and Pacific RegionalUpdate.” http://www.worldbank.org/eapup-date.

———. 2003b. “East Asia Update: From CyclicalRecovery to Long Run Growth.” www.devel-opmentgoals.org.

———. 2003c. Global Economic Prospects 2004.Washington, D.C.

———. 2003d. World Development Report 2004:Making Services Work for Poor People. Wash-ington, D.C.

———. 2004. The Millennium DevelopmentGoals for Health: Rising to the Challenges.Washington, D.C.

Chapter 4 Improving EnablingClimate for Growth—Economic and Financial Policies

Alesina, Alberto, Silvia Ardagna, Guiseppe Nico-letti, and Fabio Schiantarelli. 2003. “Regula-tion and Investment.” Working Paper 9560.National Bureau of Economic Research, Cam-bridge, Mass.

Anderson, James, and Peter Neary. 2003. “TheMercantilist Index of Trade Policy.” Inter-national Economic Review 44(2): 627–49.

Aziz, Jahangir, Francesco Caramazza, and RanilSalgado. 2000. “Currency Crises: In Search ofCommon Elements.” IMF Working PaperWP/00/67. International Monetary Fund,Washington, D.C.

R E F E R E N C E S


Barro, Robert J. 1997. “Determinants of Eco-nomic Growth: A Cross-Country EmpiricalStudy.” Lionel Rogers Lectures. Cambridge,Mass.: MIT Press.

Batra, Geeta. 2003. “Investment Climate Mea-surement: Pitfalls and Possibilities.” DiscussionPaper. World Bank, Investment ClimateDepartment, Washington, D.C.

Berg, Andrew, and Anne Krueger. 2003. “Trade,Growth, and Poverty: A Selective Survey.” IMFWorking Paper WP/03/30. International Mon-etary Fund, Washington, D.C.

Dollar, David, Mary Hallward-Dreimeier, and T.Mengistae. 2003. “Investment Climate andFirm Performance in Developing Countries.”World Bank, Washington, D.C. Processed.

Dollar, David, Mary Hallward-Dreimeier, AnqingShi, Scott Wallsten, Shuilin Wang, and LixinColin Xu. 2003. “Improving the InvestmentClimate in China.” Working Paper 26876.World Bank, Development Research Group,Investment Climate, Washington, D.C.

Fallon, Peter, and Robert Lucas. 2002. “TheImpact of Financial Crises on Labor Markets,Household Incomes and Poverty: A Review ofEvidence.” World Bank Research Observer17(1): 21–45.

Hallward-Dreimerier, Mary, Scott Wallsten, andLixin Colin Xu. 2003. “The Investment Cli-mate and the Firm: Firm-Level Evidence fromChina.” Working Paper 3003. World Bank,Development Research Group, Investment Cli-mate, Washington, D.C.

IMF (International Monetary Fund). 2003. WorldEconomic Outlook. Washington, D.C. Octo-ber.

Kee, Hiau Looi, Alessandro Nicita, and MarceloOlarreaga. 2004. “Calculation of the OverallTrade Restrictiveness Index.” World Bank,Development Research Group, Washington,D.C. Processed.

Maskus, Keith, and John S. Wilson, eds. 2002.“Quantifying the Impact of Technical Barriersto Trade: Can It be Done?” University ofMichigan Press.

Mattoo, Aaditya, Randeep Rathindran, andArvind Subramaniam. 2001. “Measuring Ser-vices Trade Liberalization and Its Impact onEconomic Growth: An Illustration.” PolicyResearch Working Paper 2655. World Bank,Development Research Group, Washington,D.C.

Scarpetta, Stefano, Philip Hemmings, ThierryTressel, and Jaejoo Woo. 2002. “The Role of

Policy and Institutions for Productivity andFirm Dynamics: Evidence from Micro andIndustry Data.” OECD Economics DepartmentWorking Paper 329. Organisation for Eco-nomic Co-operation and Development, Paris.April.

Standard and Poor’s. 2003. Global Stock MarketsFactbook. New York.

Wilson, John S. 2002. “Liberalizing Trade in Agri-culture: Developing Countries in Asia and thePost-Doha Agenda.” Policy Research WorkingPaper 2804. World Bank, Policy ResearchDepartment, Washington, D.C.

World Bank. 2003. Doing Business in 2004.Washington, D.C.

Yaron, Jacob, McDonald Benjamin, and GerdaPiprek. 1997. Rural Finance: Issues, Design,and Best Practices. Washington, D.C.: WorldBank.

Chapter 5 Upgrading Public SectorGovernance

Beck, Thorsten, George Clarke, Alberto Groff,Philip Keefer, and Patrick Walsh. 2000. “NewTools and New Tests in Comparative PoliticalEconomy: The Database on Political Institu-tions.” Policy Research Working Paper 2283.World Bank, Washington, D.C.

Bruno, Michael, Martin Ravallion, and LynSquire. 1998. “Equity and Growth in Develop-ing Countries: Old and New Perspectives onthe Policy Issues.” In V. Tanzi and K. Chu, eds.,Income Distribution and High-QualityGrowth. Cambridge, Mass.: MIT Press.

Dollar, David, and Aart Kraay. 2002. “Growth IsGood for the Poor.” Journal of EconomicGrowth 7(3): 195–225.

Evans, Peter B., and James E. Rauch. 1999.“Bureaucracy and Growth: A Cross-NationalAnalysis of the Effects of ‘Weberian’ StateStructures on Economic Growth.” AmericanSociological Review 64: 748–65.

IMF (International Monetary Fund). 2003.“Growth and Institutions.” World EconomicOutlook (chapter 3). Washington, D.C. April.

IMF–World Bank. 2002. “Actions to Strengthenthe Tracking of Poverty-Reducing PublicSpending in Heavily Indebted Poor Countries(HIPCs).” Washington, D.C. March.

———. 2003. “Bank/Fund Collaboration on Pub-lic Expenditure Issues.” Washington, D.C. Feb-ruary 14.

R E F E R E N C E S


Kaufmann, Daniel, Aart Kraay, and MassimoMastruzzi. 2003. “Governance Matters III:Governance Indicators for 1996–2002.” PolicyResearch Working Paper 3106. World Bank,Washington, D.C.

Kaufmann, Daniel, Aart Kraay, and Pablo Zoido-Lobaton. 1999. “Governance Matters.” PolicyResearch Working Paper 2196. World Bank,Washington, D.C.

Knack, Stephen, and Philip Keefer. 1995. “Institu-tions and Economic Performance: Cross-Coun-try Tests Using Alternative InstitutionalMeasures.” Economics and Politics 7 (Novem-ber): 207–27.

Knack, Stephen, Mark Kugler, and Nick Manning.2003. “Second Generation Governance Indica-tors.” International Review of AdministrativeSciences 69(3): 345–64.

Lambsdorff, Johann Graf. 2003. “FrameworkDocument 2003.” Background Paper to the2003 Corruption Perceptions Index.http://www.transparency.org/cpi/2003/dnld/framework.pdf.

Mauro, Paolo. 1995. “Corruption and Growth.”Quarterly Journal of Economics 110:681–712.

NEPAD (New Partnership for Africa’s Develop-ment). 2003a. “African Peer Review Mecha-nism.” Sixth Summit of the NEPAD Heads ofState and Government Implementation Commit-tee, Base Document. http:/www.touchtech.biz/nepad/files/documents/49.pdf.

———. 2003b. “Objectives, Standards, Criteriaand Indicators for the African Peer ReviewMechanism.” http:/www.touchtech.biz/nepad/files/documents/110.pdf.

OECD (Organisation for Economic Co-operationand Development). 1994. Taxation and Invest-ment Flows: An Exchange of Experiencesbetween the OECD and the Dynamic AsianEconomies. Paris.

Pritchett, Lant, and Lawrence H. Summers. 1996.“Wealthier Is Healthier.” Journal of HumanResources 31(4): 841–68.

Tanzi, Vito, and Howell Zee. 2000. “Tax Policyfor Emerging Markets: Developing Countries.”National Tax Journal 53(2): 299–322.

UNECA (U.N. Economic Commission for Africa).Forthcoming. African Governance Report.Addis Ababa.

WHO (World Health Organization). 2002.“Mobilization of Domestic Resources forHealth: Report of the Working Group 3 of the

Commission on Macroeconomics and Health.”Geneva. August.

World Bank. 2003. “Lessons on GovernanceReform from the 1990s. A Concept Note.”World Bank, Public Sector Governance Group,Washington, D.C. November. Processed.

Chapter 6 Strengthening Infrastructure

Angel, Shlomo. 2000. Housing Policy Matters: AGlobal Analysis. New York: Oxford UniversityPress.

APHRC (Africa Population and Health ResearchCenter). 2002. Population and Health Dynam-ics in Nairobi’s Informal Settlements. Nairobi.

Bacon, Robert. 1999. “A Scorecard for EnergyReform in Developing Countries.” Public Pol-icy for the Private Sector. Note 175. WorldBank. April.

Caldéron, Cesar, William Easterly, and LuisServén. 2003. “The Output Cost of LatinAmerica’s Infrastructure Gap.” In WilliamEasterly and Luis Servén, eds., The Limits ofStabilization: Infrastructure, Public Deficits,and Growth in Latin America. Palo Alto,Calif.: Stanford University Press.

Camdessus, Michel. 2003. “Financing Water forAll.” In James Winpenny, ed., Report of theWorld Panel on Financing Water Infrastruc-ture. Marseilles: World Water Council.

Datt, Gaurav, and Martin Ravallion. 1998. “WhyHave Some Indian States Done Better than Oth-ers at Reducing Poverty?” Economica 65(257):17–38.

De la Fuente, Angel, and Antonio Estache. 2004.“A Survey of the Evidence on the GrowthEffects of Infrastructure.” World Bank, Infra-structure Vice Presidency, Washington, D.C.Processed.

Deininger, Klaus, and John Okidi. 2003. “Growthand Poverty Reduction in Uganda, 1999–2000:Panel Data Evidence.” Development PolicyReview 21(4): 481–509.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2003. “Water Supply and Sanita-tion and the MDGs.” DC2003-004/Add.3.Washington, D.C. April.

DFID (U.K. Department for International Devel-opment). 2002. “Making Connections: Infra-structure for Poverty Reduction.” January.Processed.

R E F E R E N C E S


http://62.189.42.51/DFIDstage/FOI/dc/7mar02_making_connections.pdf.

Easterly, William, and Luis Servén, eds. 2003. TheLimits of Stabilization: Infrastructure, PublicDeficits, and Growth in Latin America. PaloAlto, Calif.: Stanford University Press.

Estache, Antonio, Vivien Foster, and QuentinWodon. 2002. “Accounting for Poverty inInfrastructure Reform—Learning from LatinAmerica’s Experience.” World Bank InstituteStudies in Development Series, Washington,D.C.

Fan, Shenggen, Linxiu Zhang, and Xiaobo Zhang.2002. “Growth, Inequality and Poverty inChina: The Role of Public Investments.”Research Report 123. International Food Pol-icy Research Institute, Washington, D.C.

Fay, Marianne, and Tito Yepes. 2003. “Investingin Infrastructure: What Is Needed from 2000 to2010?” Policy Research Paper 3102. WorldBank, Infrastructure Vice Presidency, Washing-ton, D.C. Processed.

Foster, Vivien, and Jean-Philippe Tre. 2000. “Mea-suring the Impact of Energy Interventions onthe Poor: An Illustration from Guatemala.” InInfrastructure for Development: Private Solu-tions and the Poor. Conference volume forPrivate Provision of Infrastructure AdvisoryFacility (PPIAF). London: Department for Inter-national Development (U.K.) and World Bank.

Global Water Partnership. 2000. Towards WaterSecurity: A Framework for Action. Stockholm.

Jalan, Jyotsna, and Martin Ravallion. 2001.“Does Piped Water Reduce Diarrhea for Chil-dren in Rural India?” Policy Research WorkingPaper 2664. World Bank, DevelopmentResearch Group, Washington, D.C. Processed.

Jayasuriya, Ruwan, and Quentin Wodon, eds.2003. “Efficiency in Reaching the MillenniumDevelopment Goals.” Working Paper 9. WorldBank, Washington, D.C.

JMP (Joint Monitoring Programme for Water andSanitation of WHO-UNICEF). 2000. Data for2000. http://www.wssinfo.org.

Kingdom, William, and Nicola Tynan. 2002. “AWater Scorecard: Setting Performance Targetsfor Water Utilities.” Viewpoint 242, PrivateSector and Infrastructure Network, WorldBank, Washington, D.C.

Leipziger, Danny, Marianne Fay, and Tito Yepes.2003. “Achieving the Millennium Develop-ment Goals: The Role of Infrastructure.” Pol-icy Research Working Paper 3163. World Bank,Washington, D.C. Processed.

Qiang, Christine Zhing-Wen. 2004. “RegulatoryDevelopments in ICT.” World Bank CITDepartment, INF Vice Presidency, Washing-ton, D.C. Processed.

Ramirez, Maria Teresa, and Hadi Esfahani. 2000.“Infrastructure and Economic Growth.” Bankof Colombia, Bogotá. Processed.

Roberts, Peter. 2003. “Transport Indicators.”World Bank TWU Department, Washington,D.C. Processed.

Roja, F. K. 2003. “The Penalties of Inefficiency inInfrastructure.” Review of Development Eco-nomics 7(1): 127–37.

United Nations Millennium Project Task Forceon Water and Sanitation. 2003. “Achievingthe MDG for Water and Sanitation: What WillIt Take?” Interim Full Report. New York.December.

Winpenny, James, ed. 2003. Report of the WorldPanel on Financing Water Infrastructure. Mar-seilles: World Water Council.

World Bank. 1994. World Development Report.Washington, D.C.

———. 2000. World Business Environment Survey.http://info.worldbank.org/governance/wbes/.

———. 2002. “Water Sector Reform Scorecard: AGlobal Overview of Urban Water Supply andSanitation Sector Reform.” Washington, D.C.Processed.

World Economic Forum. 2003. Global Competi-tiveness Report, 2002–03. Cologny/Geneva.

Chapter 7 Accelerating HumanDevelopment

Blank, Lorraine, Margaret Grosh, and ChristineWeigand. 2003. “Patterns in Social ProtectionExpenditures.” World Bank, HDNSP, Wash-ington, D.C.

Bruns, Barbara, Alain Mingat, and RamahatraRakotomalala. 2003. Achieving Universal Pri-mary Education by 2015: A Chance for EveryChild. Washington, D.C.: World Bank.

Dehn, Jan, Ritva Reinikka, and Jakob Svensson.2003. “Survey Tools for Assessing Performancein Service Delivery.” In Francois Bourguignonand Luiz Pereira da Silva, eds., The Impact ofEconomic Policies on Poverty and Income Dis-tribution: Evaluation Techniques and Tools.New York: World Bank and Oxford UniversityPress.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the World

R E F E R E N C E S


Bank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2004. “Education for All (EFA)–FastTrack Initiative: Progress Report.” DC2004-0002/1, Washington, D.C. March 26.

Di Tella, Raphael, and William Savedoff. 2000.Diagnosis Corruption: Fraud in Latin Amer-ica’s Public Hospitals. Latin America ResearchNetwork. Washington, D.C.: Inter-AmericanDevelopment Bank.

Dreze, Jean, and Haris Gazdar. 1996. “UttarPradesh: The Burden of Inertia.” In Jean Drezeand Amartya Sen, eds., Indian Development:Selected Regional Perspectives. New Delhi:Oxford University Press.

Filmer, Deon, and Lant Pritchett. 1999. “TheImpact of Public Spending on Health: DoesMoney Matter?” Social Science and Medicine49: 1309–23.

Kattan. 2003. “User Fees in Primary Education.”World Bank, Washington, D.C. January.Processed.

Khandker, Shahidur, Mark Pitt, and NobuhikoFuwa. 2003. “Subsidy to Promote Girls’ Sec-ondary Education: The Female Stipend Pro-gram in Bangladesh.” World Bank,Washington, D.C. March. Processed.

Lewis, Maureen. 2000. “Who Is Paying for HealthCare in Europe and Central Asia?” EasternEurope and Central Asia Region Monograph.World Bank, Washington, D.C.

Reinikka, Ritva, and Jakob Svensson. 2004.“Local Capture: Evidence from a Central Gov-ernment Transfer Program in Uganda.” TheQuarterly Journal of Economics 119(2).

Rubio, Gloria, and Kalanidhi Subbarao. 2003.“Impact Evaluation in Bank Projects: A Com-parison of Fiscal 1998 and 1999.” World Bank,Poverty Reduction and Economic Management(PREM) Network, Washington, D.C.

Schneider, Pia, A. K. Nandakumar, DenisPorignon, Manjiri Bhawalkar, DamasceneButera, and Courtney Barnett. 2000. “RwandaNational Health Accounts 1998.” TechnicalReport 53. Partnerships for Health ReformProject. Bethesda, MD: Abt Associates Inc.

UNAIDS (Joint U.N. Programme on HIV/AIDS).2003. “Progress Report on the GlobalResponse to the HIV/AIDS Epidemic, 2003.”Geneva.

UNDP (United Nations Development Pro-gramme). 2003. Human Development Report.New York.

USAID (United States Agency for InternationalDevelopment). Various years. Demographicand Health Survey. Washington, D.C.

World Bank. 2003a. Restoring Fiscal Disciplinefor Poverty Reduction in Peru. World Bank,Washington, D.C.

———. 2003b. World Development Indicators2003. Washington, D.C.

———. 2003c. World Development Report 2004:Making Services Work for Poor People. Wash-ington D.C.

———. 2004. The Millennium DevelopmentGoals for Health: Rising to the Challenges.Washington, D.C.

Chapter 8 Promoting EnvironmentalSustainability

Cavendish, William. 2000. “Empirical Regulari-ties in the Poverty-Environment Relationship ofRural Households: Evidence from Zimbabwe.”World Development 28(11): 1979–2003.

United Nations Millennium Project. 2003. “Back-ground Paper of the Millennium Task Force onEnvironmental Sustainability.” New York.Processed.

van der Klaauw, Bas, and Limin Wang. 2003.“Child Mortality in Rural India: Determinantsand Policy Implications.” World Bank, Wash-ington, D.C. Processed.

WHO (World Health Organization). 2002. WorldHealth Report 2002. Geneva.

Chapter 9 Fostering Growth andStability: Macro-financial Policies

Adams, Richard, Jr. 2003. “International Migra-tion, Remittances, and the Brain Drain: AStudy of 24 Labor-Exporting Countries.” Pol-icy Research Working Paper 3069. WorldBank, Washington, D.C.

Basel Committee on Banking Supervision. 2003.“The New Basel Accord, Consultative Docu-ment.” Bank for International Settlements,Berne.

Borensztein, Eduardo, José De Gregorio, andJong-Wha Lee. 1998. “How Does ForeignInvestment Affect Growth?” Journal of Inter-national Economics 45(June): 115–35.

Chuhan, Punam, and Federico Sturzenegger. 2004.“Default Episodes in the 1980s and 1990s:What Have We Learned?” In Joshua Aizenmanand Brian Pinto, eds., Managing Volatility and

R E F E R E N C E S


Crisis: A Practioner’s Guide. World Bank,Washington, D.C. http:/www.wbweb.world-bank.org/prem/prmep/economicpolicy/mv/handbook.htm.

Claessens, Stijn, Geoffrey R. D. Underhill, andXiaoke Zhang. 2003. “Basel II CapitalRequirements and Developing Countries: APolitical Economy Perspective.” Paper pre-sented at the workshop on Quantifying theImpact of Rich Countries’ Policies on PoorCountries, October 23–24, organized by theCenter for Global Development and the GlobalDevelopment Network. World Bank, Washing-ton, D.C.

Eichengreen, Barry, Kenneth Kletzer, and AshokaMody. 2003. “Crisis Resolution: Next Steps.”IMF Working Paper WP/03/196. InternationalMonetary Fund, Washington, D.C.

Faini, Ricardo. 2003. “Is the Brain Drain anUnmitigated Blessing?” Discussion Paper2003/64. World Institute for Development Eco-nomics Research, Helsinki. September.

Heid, Frank. 2003. “Is Regulatory Capital Pro-cyclical? A Macroeconomic Assessment ofBasel II.” Paper presented at Basel CommitteeBanca d’Italia Workshop, March 20–23. Rome.

IMF (International Monetary Fund). 2001. WorldEconomic Outlook. Washington, D.C. October.

IMF–World Bank. 2003. “International Stan-dards: Strengthening Surveillance, DomesticInstitutions, and International Markets.”Washington, D.C.

Ratha, Dilip. 2003. “Enhancing Workers’ Remit-tances to Developing Countries.” World Bank,Washington, D.C.

Skeldon, Ronald. 1997. Migration and Develop-ment. Harlow, England: Addison Wesley Long-man.

World Bank. 2004. Global Development Finance2004. Washington, D.C.

Chapter 10 Dismantling Barriers to Trade

Anderson, Kym. 2003. “How Can AgriculturalTrade Reform Reduce Poverty?” BackgroundPaper prepared for the United Nations Millen-nium Project Task Force on Trade. New York.Processed.

Baffes, John. 2003. “Cotton Market Policies:Issues and Facts.” Policy Research WorkingPaper 3218. World Bank. Washington, D.C.Processed.

Beghin, John, David Roland-Holst, andDominique van der Mensbrugghe. 2002.“How Will Agricultural Trade Reforms inHigh-Income Countries Affect the TradingRelationships of Developing Countries?”World Bank, Washington, D.C. Processed.

Bown, Chad, Bernard Hoekman, and CaglarOzden. 2003. “Developing Countries and U.S.Antidumping: The Path from Initial Filing toWTO Dispute Settlement.” World TradeReview 2(3): 349–71.

Hoekman, Bernold, Francis Ng, and MarceloOlarreaga. 2002. “Eliminating Excessive Tar-iffs on Exports of Least Developed Countries.”World Bank Economic Review 16(1): 1–21.

———. 2004. “Reducing Agricultural Tariffs ver-sus Domestic Support: What’s More Importantfor Developing Countries?” World Bank Eco-nomic Review 18(1).

Karsenty, Guy. 2002. “Trends in Services Tradeunder GATS: Recent Developments.” Presenta-tion to the WTO Symposium on Assessment ofTrade Services, March 14–15. Geneva.

Liu, Xiang, and Hylke Van den Bussche 2003. “EUAntidumping Cases against China.” Licos Dis-cussion Papers 11902. Catholic University,Leuven, Belgium.

Mattoo, Aaditya. 2003. “Services in a Develop-ment Round.” World Bank, Washington, D.C.Processed.

Messerlin, Patrick. 2002. “Agriculture in the DohaAgenda.” Policy Research Working Paper3009. World Bank, Washington, D.C.

Mitchell, Donald. 2004. “Sugar Policies: Oppor-tunities for Change.” Policy Research WorkingPaper 3222. World Bank, Washington, D.C.

Otsuki, Tsunehiro, John S. Wilson, and MirvatSewadeh. 2001. “Saving Two in a Billion:Quantifying the Trade Effect of European FoodSafety Standards on African Exports.” FoodPolicy 26(5): 495–514.

Schiff, Maurice, and L. Alan Winters. 2003.Regionalism and Development. New York:Oxford University Press and World Bank.

Stern, Robert. 2002. “Quantifying Barriers toTrade in Services.” In Bernard Hoekman, PhilipEnglish, and Aaditya Mattoo, eds., Develop-ment, Trade and the WTO: A Handbook.Washington, D.C.: World Bank.

Walmsley, Terri, and L. Alan Winters. 2002.“Relaxing Restrictions on Temporary Move-ment of Natural Persons: A Simulation Analy-sis.” University of Sussex, Economics

R E F E R E N C E S


Department. Brighton, U.K. Processed.Wilson, John S., and Victor Abiola, eds. 2003.

Standards & Global Trade: A Voice for Africa.Washington, D.C.: World Bank.

World Bank. 2003. Global Economic Prospectsand the Developing Countries 2004: Realizingthe Development Promise of the Doha Agenda.Washington, D.C.

———. 2004. Global Development Finance,2004. Washington, D.C.

WTO (World Trade Organization). 2001. “Minis-terial Declaration, Adopted on 14 November,2001, Doha.” WT/MIN (01)DEC/1. Geneva.

———. 2002. Speech by the Director General,April 12, Montreux.

Chapter 11 Providing More andBetter Aid

Aryeetey, Ernest, Barfour Osei, and Peter Quartey.2003. “Does Tying Aid Make It More Costly?A Ghanaian Case Study.” Paper presented atthe Workshop on Quantifying the Impact ofRich Countries’ Policies on Poor Countries,Center for Global Development and the GlobalDevelopment Network, Washington, D.C.October 23–24.

Birdsall, Nancy, Stijn Claessens, and Ishac Diwan.2003. “Policy Selectivity Forgone: Debt andDonor Behavior in Africa.” The World BankEconomic Review 17(3): 409–35.

Bulir, Ales, and A. Javier Hamann. 2001. “HowVolatile and Unpredictable Are Aid Flows, andWhat Are the Policy Implications?” IMF Work-ing Paper WP/01/167. International MonetaryFund, Washington, D.C.

———. 2003. “Aid Volatility: An EmpiricalAssessment.” IMF Staff Papers 50(1): 64–89.

Bulir, Ales, and Timothy Lane. 2002. “Aid and Fis-cal Management.” IMF Working PaperWP/02/112. International Monetary Fund,Washington, D.C.

Burnside, Craig, and David Dollar. 2000. “Aid,Policies, and Growth.” American EconomicReview 90(4): 847–68.

———. 2004. “Aid, Policies, and Growth: Revis-iting the Evidence.” Policy Research Paper 0-2834. World Bank, Washington, D.C.

Collier, Paul, and Jan Dehn. 2001. “Aid, Shocksand Growth.” Policy Research Working Paper2688. World Bank, Washington, D.C.

Collier, Paul, and A. Hoeffler. 2002. “Aid, Policyand Growth in Post-Conflict Countries.” Policy

Research Working Paper 2902. World Bank,Washington, D.C.

Devarajan, Shantayanan, Margaret J. Miller, andEric V. Swanson. 2002. “Goals for Develop-ment: History, Prospects and Costs.” PolicyResearch Working Paper 2819. World Bank,Human Development Network, Washington,D.C.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2003a. “Progress Report and Crit-ical Next Steps in Scaling Up: Education for All,Health, HIV/AIDS, Water and Sanitation—Synthesis Paper.” DC2003-0004. Washington,D.C.

———. 2003b. “Supporting Sound Policies withAdequate and Appropriate Financing.”DC2003-0016.Washington, D.C. September.

———. 2004. “Financing Modalities toward theMillennium Development Goals: ProgressNote.” DC2004-003. Washington, D.C. April12.

Dollar, David, and Victoria Levin. 2004. “TheIncreasing Selectivity of Aid, 1984–2002.” Pol-icy Research Working Paper 3299. WorldBank, Washington, D.C.

Gemmell, Norman, and Mark McGillivray. 1998.“Aid and Tax Instability and the Budget Con-straint in Developing Countries.” CREDITResearch Paper 98/1. University of Notting-ham, Centre for Research in Economic Devel-opment and International Trade, U.K.


IMF (International Monetary Fund). 2004. “IMFDiscusses Operational Framework for DebtSustainability in Low-Income Countries.” Pub-lic Information Notice 04/34. Washington,D.C. April 5. www.imf.org/external/np/sec/pn/2004/pn0434.htm.

IMF–World Bank. 2003a. “Heavily Indebted PoorCountries (HIPC) Initiative: Status of Imple-mentation.” IDA/SecM2003-0477/1. Washing-ton, D.C. September.

———. 2003b. “Update on Implementation ofAction Plans to Strengthen Capacity of HIPCsto Track Poverty-Reducing Spending.”IDA/R2003-0043. Washington, D.C. March.

R E F E R E N C E S


www.imf.org/external/np/hipc/doc.htm.www.worldbank.org/publicsector/pe/HIPCUpdate.pdf.

———. 2004a. “Debt Sustainability in Low-Income Countries—Proposal for an Opera-tional Framework and Policy Implications.”Washington, D.C. February 3. www.imf.org/external/np/pdr/sustain/2004/020304.pdf.

———. 2004b. “Heavily Indebted Poor Countries(HIPC) Initiative: Statistical Update.” IDA/SecM2004-0184. Washington, D.C. April.

Jepma, Catrinus J. 1991. “The Tying of Aid.”Organisation for Economic Co-operation andDevelopment, Development Centre Studies,Paris.

Knack, Stephen, and Aminur Rahman. 2004.“Donor Fragmentation and Bureaucratic Qual-ity in Aid Recipients.” Policy Research Work-ing Paper 3186. World Bank, Washington, D.C.

OECD (Organisation for Economic Co-operationand Development). 2003. “OECD Question-naire, Indicators of Progress in Harmonizationand Alignment, Annex A, Explanatory Note”(first draft for discussion). Paris. December.

———. 2004a. Development Cooperation Report2003. Paris.

———. 2004b. “ODA Statistics for 2003 andODA Outlook.” DCD/DAC(2004/22. Paris.

Prati, Alessandro, Ratna Sahay, and Thierry Tres-sel. 2003. “Is There a Case for Sterilizing For-eign Aid Flows?” IMF Seminar Series 2003-88.International Monetary Fund, Washington,D.C. June.

Roodman, David. 2003. “An Index of Donor AidPerformance.” Center for Global Development.Washington, D.C.

United Nations. 2001. “Report of the High-LevelPanel on Financing for Development.” Presentedto the General Assembly, June 28, New York.

United Nations Millennium Project. 2004. “Mil-lennium Development Goals Needs Assess-ment.” New York.

World Bank. 2002. A Case for Aid: Building aConsensus for Development Assistance. Wash-ington, D.C.

Chapter 12 Fulfilling Responsibilitiesfor Global Public Goods

African Development Bank. 2002. “MDB Supportfor Global Public Goods Provisions: ProgressReport.” Addis Ababa, Ethiopia.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2000. “Poverty Reduction andGlobal Public Goods: Issues for the World Bankin Supporting Global Collective Action.”World Bank Report to the Development Com-mittee. Washington, D.C.

———. 2001. “Poverty Reduction and GlobalPublic Goods: Progress Report.” World BankReport to the Development Committee. Wash-ington, D.C.

Ferroni, Marco, and Ashoka Mody. 2002. Inter-national Public Goods: Incentives, Measure-ment, and Financing. Boston, Mass.: KluwerAcademic Publishers; Washington, D.C.: WorldBank.

Kaul, Inge, Pedro le Goulven, and Ronald Men-doza. 2003. “Providing Global Public Goods:Managing Globalization.” New York: OxfordUniversity Press.

Pagiola Stefano, Roberto Martin-Hurtado, PriyaShyamsundar, Muthukumara Mani, and Patri-cio Silva. 2002. “Generating Public SectorResources to Finance Sustainable Develop-ment.” Technical Paper 538. World Bank, Envi-ronment Department, Washington, D.C.

Samuelson, Paul. 1954. “The Pure Theory of Pub-lic Expenditure.” Review of Economics andStatistics 49(2): 257–81.

World Bank. 2001. Global Development Finance2001: Buildings Conditions for EffectiveDevelopment Finance. Washington, D.C.

Chapter 13 Monitoring the IFIs’Contribution

AfDB (African Development Bank). 2002. AnnualReport. Addis Ababa, Ethiopia.

Boughton, James, and Alex Mourmouras. 2004.“Whose Programme Is It? Policy Ownershipwith Conditional Lending.” In David Vines andChristopher Gilbert, eds., The IMF and Its Crit-ics: Reform of Global Financial Architecture.Cambridge, U.K.: Cambridge University Press.

Burnside, Craig, and David Dollar. 2000. “Aid,Policies, and Growth.” American EconomicReview 90(4): 847–68.

Collyns, Charles, and Russell Kincaid, eds. 2003.Managing Financial Crises—Recent Experi-ence and Lessons for Latin America. IMF

R E F E R E N C E S


Occasional Paper 217. International MonetaryFund, Washington, D.C.

Development Committee (Joint Ministerial Com-mittee of the Boards of Governors of the WorldBank and the International Monetary Fund onthe Transfer of Real Resources to DevelopingCountries). 2000a. “Poverty Reduction andGlobal Public Goods: Issues for the World Bankin Supporting Global Collective Action.”DC2000-16. Washington, D.C. September.

———. 2000b. “Supporting Country Develop-ment: World Bank Role and Instruments inLow- and Middle-Income Countries.”DC2000-19. Washington, D.C.

———. 2002. “Better Measuring, Monitoring,and Managing for Development Results.”DC2002-0018. Washington, D.C. September.

———. 2003. “Poverty Reduction StrategyPapers: Progress in Implementation.” DC2003-0011. Washington, D.C. September.

EBRD (European Bank for Reconstruction andDevelopment). 2003. Annual Evaluation Over-view Report. http://www.ebrd.org/projects/eval/showcase/main.htm.

Ghosh, Atish, Timothy D. Lane, Ales Bilir, M.Schulze-Ghattos, and Alexandros T. Mour-mouas. 2002. “IMF-Supported Programs inCapital Account Crises.” IMF OccasionalPaper 210. International Monetary Fund,Washington, D.C.

IDA (International Development Association).2003. “IDA Results Measurement System:Update Note.” IDA/R2003-0189. Washington,D.C. October.

IDB (Inter-American Development Bank). 2003.Analysis of Project Evaluability Year 2001. RE-275. Washington, D.C. January.

———. 2004. Regional Public Goods: TheOpportunity for an Expanded Bank Role.Washington, D.C.

IMF (International Monetary Fund). 1998. Exter-nal Evaluation of the ESAF—Report by aGroup of Independent Experts. Washington,D.C.

———. 2001. “Strengthening Country Ownershipof Fund-Supported Programs.” Washington,D.C. December.http://www.imf.org/external/np/pdr/cond/2001/eng/strength/120501.htm.

———. 2002a. “Evaluation of the Prolonged Useof Fund Resources.” SM/02/87. Washington,D.C. September.

———. 2002b. “Guidelines on Conditionality.”IMF Executive Board Decision 12864-(02/102). Washington, D.C. September.

———. 2002c. “Review of the Poverty Reductionand Growth Facility.” SM/02/51. Washington,D.C. February.

———. 2003a. “Alignment of the PRGF and thePRSP.” SM/03/94. Washington, D.C.

———. 2003b. “Enhancing the Effectiveness ofSurveillance: Operational Responses, theAgenda Ahead, and Next Steps.” Public Infor-mation Notice 03/50. Washington, D.C. April10.

———. 2003c. “Evaluation of the Role of theFund in Recent Capital Account Crises—Report by the Independent Evaluation Office.”SM/03/171. Washington, D.C. May.

———. 2003d. “Independent Evaluation Office—Evaluation of Fiscal Adjustment in IMF-Sup-ported Programs.” SM/03/291. Washington,D.C. August.

———. 2003e “Role of the Fund in Low-IncomeMember Countries over the Medium Term—Issues Paper for Discussion.” SM/03/257.Washington, D.C. July.

———. 2003f. World Economic Outlook (April).Washington, D.C.

———. 2004a. Finance and Development. Wash-ington, D.C. March (articles by Kenen, Lissak-ers, Milleron, and Welch).

———. 2004b. “Technical Assistance EvaluationProgram—Findings of Evaluations andUpdated Program.” March 1.http://www.imf.org/external/np/ta/2004/eng/030104.htm.

IMF–World Bank. 2001. “Strengthening WorldBank-IMF Collaboration on Country Programsand Conditionality.” SecM2001-0461/1 andSM/01/219, Supplement 1, Revision 1. Wash-ington, D.C. August.

———. 2002a. “Operationalizing Bank-FundCollaboration in Country Programs and Con-ditionality.” Staff Guidance Note. Washington,D.C.

———. 2002b. “Review of the Poverty ReductionStrategy Paper Approach.” SM/02/53. Wash-ington, D.C. February.

———. 2002c. “Strengthening IMF–World BankCoordination on Country Programs and Con-ditionality—Progress Report.” World BankDocument SecM2002-443 and IMF DocumentSM/02/271. Washington, D.C. August.

R E F E R E N C E S


———. 2004. “Strengthening IMF–World BankCollaboration on Country Programs and Con-ditionality—Progress Report.” SecM2004-0072/1. Washington, D.C. March 26.

Joint Marrakech Memorandum. 2004. “Promot-ing a Harmonized Approach to Managing forDevelopment Results: Core Principles; andAction Plan on Managing for DevelopmentResults.” Presented to the Second Roundtablefor Managing Results, February 4–5, spon-sored by the World Bank and four regionaldevelopment banks.

MDB (Multilateral Development Banks) Evalua-tion Cooperation Group. 2002. “Good Prac-tice Standard on MDB-Supported Public SectorEvaluations.” June.

MDB Joint Liaison Group. 2002. “The Role of theMDBs in the Provision of Global PublicGoods.” Washington, D.C.

OED (Operations Evaluation Department, WorldBank). 2002. “The World Bank’s Approach toGlobal Programs: An Independent Evalua-tion.” Washington, D.C.

———. 2003a. “2002 Annual Review of Devel-opment Effectiveness—Achieving DevelopmentOutcomes: The Millennium Challenge.” Wash-ington, D.C.

———. 2003b. “2003 Annual Report on Opera-tions Evaluation.” Report 25861. Washington,D.C.

———. 2003c. “Approach Paper: Country Assis-tance Strategy Evaluation Retrospective.”CODE2003-0071. Washington, D.C. Novem-ber.

———. 2003d. “The CGIAR at 31: An Indepen-dent Meta-Evaluation of the ConsultativeGroup on International AgriculturalResearch.” Washington, D.C.

———. 2003e. “Multi-party Evaluation of theComprehensive Development Framework.”CODE2003-0034. Washington, D.C. May.

———. 2004a. “2003 Annual Review of Devel-opment Effectiveness—The Effectiveness ofBank Support for Policy Reform.” Report28290. Washington, D.C.

———. 2004b. “Books, Buildings, and LearningOutcomes: An Impact Evaluation of WorldBank Support to Basic Education in Ghana.”Report 28779. Washington, D.C.

Scott, Alison. 2004. “Assessing Multilateral Effec-tiveness.” U.K. Department for InternationalDevelopment, London. February.

Wapenhans Task Force. 1992. “Report of thePortfolio Management Performance TaskForce.” World Bank, Washington, D.C.

World Bank. 1998. “Country Assistance Strate-gies: Retrospective and Outlook.” SecM98-242. Washington, D.C. March.

———. 2000. “Country Assistance Strategies:Retrospective and Implications.” R99-228/2.Washington, D.C. May.

———. 2003a. “Coherence, Coordination, andCooperation among Multilateral Organiza-tions.” SecM2003-0112. March, andSecM2003-0380. August. Washington, D.C.

———. 2003b. “Country Assistance Strategies:Retrospective and Future Directions.” Wash-ington, D.C. March.

———. 2003c. “Delivering Global Public GoodsLocally: Lessons Learned and SuccessfulApproaches.” Development Grant FacilityTechnical Note. Washington, D.C.

———. 2003d. “The Global Poll: MultinationalSurvey of Opinion Leaders 2002.” Prepared for

R E F E R E N C E S


the World Bank by Princeton Survey ResearchAssociates. May 2003. www.worldbank.org.

———. 2003e. “Harmonization Follow-up:Global Architecture and World Bank Activi-ties.” SecM2003-0387/1. Washington, D.C.November.

———. 2003f. “Update on Management ofGlobal Programs and Partnerships.”SecM2003-0095. Washington, D.C. March.

———. 2004a. “Coherence, Coordination, andCooperation among Multilateral Organiza-tions: Progress and Challenges.” SecM2004-0152. Washington, D.C. April.

———. 2004b. “Enhancing World Bank Supportto Middle-Income Countries.” SecM2004-0071. February.

———. 2004c. “Managing for DevelopmentResults in the Multilateral DevelopmentBanks.” Prepared with inputs from MDBs.Washington, D.C. February.

———. 2004d. “Managing for DevelopmentResults in the World Bank.” Washington, D.C.

THE WORLD BANK ™xHSKIMBy358597zv":;:>:&:*ISBN 0-8213-5859-6

The turn of the century was marked by significant and promising events forworld development. The Millennium Development Declaration—signed by189 countries in September 2000—led to the adoption of the MillenniumDevelopment Goals (MDGs), which set clear targets for reducing poverty andother sources of human deprivation and promoting sustainable development.

A meeting of world leaders in Monterrey, Mexico, in March 2002 established a shared understanding of the broad development strategy and policies needed to achieve the MDGs. The Monterrey Consensus ushered in a new compact between developing and developed countries that stressed their mutual responsibilities in the quest for the development goals.

With broad agreement on development goals and the means to achieve them, the key task now is implementation—to translate vision into action. Is implementation happening? Whatprogress has been made? Are the various parties delivering on their parts of the compact? Whatare the priorities in the agenda? Global Monitoring Report 2004 addresses these questions.

Publication of this book launches a series of annual reports that will assess how the world isdoing in implementing the policies and actions for achieving the MDGs and related

outcomes. Produced jointly by the World Bank and the International MonetaryFund, in collaboration with partner agencies, this report provides a uniquely

comprehensive review of the global development policy agenda. It covers the whole spectrum of policies and institutions that form part of the

development effort. And it covers all key actors in this effort—developing and developed countries as well as international agencies.

The report is organized into four sections: framework, developing-country policies, developed-country policies, and the role of international financial

institutions. Part I describes the analytic framework for monitoring progress in reaching the MDGs and assesses the prospects for attaining those goals. That

assessment overarches the subsequent three sections, detailing the urgent need to scale up policies and actions. Part II evaluates policies in developing countries, focusing on low-

income countries, where the challenge to achieve the development goals is greatest, but also covering issues important to middle-income countries. Part III assesses policies of developedcountries and suggests actions required if they are to fulfill their commitments in support of the development goals. Part IV looks at how multilateral agencies are playing their parts, bothindividually and collectively, in supporting countries’ progress toward the development goals.

Global Monitoring Report 2004 is essential reading for development practitioners and thosewith an interest in international affairs and global development.

Documents

Global Monitoring Report 2004