馴于平不不甲 騙計 j Assessi*ng Ai*d A World Bank Policy Research Report Assessing Aid What Works, What Doesn't, and Why Published for the World Bank OXFORD UNIVERSITY PRESS Oxford University Press OXFORD NEW YORK TORONTO DELHI BOMBAY CALCUTTA MADRAS KARACHI KUALA LUMPUR SINGAPORE HONG KONG TOKYO NAIROBI DAR ES SALAAM CAPE TOWN MELBOURNE AUCKLAND and associated companies in BERLIN IBADAN @ 1998 The International Bankfor Reconstruction and Development / The World Bank 1818 H.Street, NW Washingon, D. C 20433 Published by Oxford University Press, Inc. 200 Madison Avenue, New York, NY 10016 Oxford is a registered trademark of Oxford University Press. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press. Manufactured in the United States ofAmerica First printing November 1998 The boundaries, colors, denominations, and other information shown on the maps in this vol- ume do not imply on the part of the World Bank Group any judgment on the legal status of any territory or the endorsement or acceptance ofsuch boundaries. Library of Congress Cataloging-in-Publication (CIP) Data. Assessing aid : what works, what doesn't, and why. p. cm. -- (A World Bank policy research report) "Publi5hed for the World Bank." Includes bibliographical references. ISBN 0-19-521123-5 1. Economic assistance. 2. Economic assistance--Evaluation. I. World Bank. II. Series. HC60 .A836 1998 338.9--ddc2l 98-33286 CIP 9 Text printed on paper that conforms to the American National Standard for Permanence of Paper for Printed Library Materials, Z39.48-1984 Contents Foreword ix The Report Team xi Overview Rethinking the Money and Ideas of Aid 1 The New International Environment 7 New Thinking on Development Strategy 9 Money Matters-In a Good Policy Environment 14 Aid Can Be the Midwife of Good Policies 17 Money Matters-In a Good Institutional Environment 19 Aid Can Be the Midwife of Good Institutions 21 Money, but More Ideas, Too 23 1. Money Matters-In a Good Policy Environment 28 Different Countries, Different Fortunes 29 Why the Divergence? 32 In Countries with Good Management ... 34 Aid Has Not Favored Countries with Good Management 40 Be Focused 44 2. Aid Can Be the Midwife of Good Policies 47 Money-Good or Bad for Reform? 48 Ownership-What Money Cannot Buy 50 Fomenting Reform 53 If Commitment, Money-If Not, Ideas 58 3. Money Matters-In a Good Institutional Environment 60 Fungibility? 62 Public Spending-Quality not Quantity 74 What Should Donors Do? 78 4. Aid Can Be the Midwife of Good Institutions 83 Better Public Provision 84 Public Provision without Public Providers 92 Making Aid Work for Better Public Services 96 V ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY 5. Money, but More Ideas, Too 103 Vietnam: Adjustment without Lending 105 El Salvador, Pakistan, and Brazil: Education's Decentralization and Reform 108 Cameroon: Health Financing and Delivery 111 Africa's Road Maintenance Initiative 113 Conclusion 115 Appendixes 121 Estimating the Effect of Aid on Growth 121 Explaining the Success or Failure of Structural Adjustment Programs 126 Analysis of the Fungibility of Foreign Aid 130 The Impact of Civil Liberties and Democracy on Government Performance 135 Estimating the Impact of Analytical Work 139 Selected Bibliography 143 Boxes 1. Defining Aid 6 2. Measuring Aid 8 3. Defining Sound Management: Good Policies and Institutions 12 4. Functions of the Development Assistance Committee 13 5. Stakeholder Views on Aid Effectiveness 24 1.1 Aid and Heavily Indebted Poor Countries 41 2.1 Reforming Pension Schemes 54 4.1 Aid and Centralization 86 4.2 A Useful Surge in Nongovernmental Organizations 94 4.3 Resettlement in Development 98 5.1 Reforming Aid Agencies in the 1990s 118 Text Figures 1 Financial Flows to Developing Countries 7 2 Official Development Assistance Relative to Gross National Product, .i Major Donors, 1991 and 1997 9 CONTENTS 3 The Gap between Model and Reality in Zambia, 1961-94 10 4 Per Capita GDP Growth in Low-Income Countries with Sound Management 14 5 Project Performance by Policy and Institutional Environment 15 6 Bilateral Aid and Colonial Past 16 7 Elections, Tenure, and Probability of Successful Reform 18 8 Public Investment from One Dollar of Tax Revenue or Official Development Assistance 19 9 Success in Rural Water Supply Projects with Differing Levels of Beneficiary Participation 22 10 Actual Allocation of Aid, 1996, and Optimal Allocation to Reduce Poverty 23 1.1 Infant Mortality in Selected Countries, 1967 and 1994 30 1.2 Economic Management and Growth in Selected Developing Countries 33 1.3 Growth Rate and Income Level 34 1.4 Aid and Growth in Selected Developing Countries, 1970-93 35 1.5 Marginal Impact on Growth of a Percent of GDP in Aid 36 1.6 Economic Growth and Poverty 38 1.7 Decline in Infant Mortality from a Percent of GDP in Aid 39 1.8 Marginal Impact on Private Investment of a Percent of GDP in Aid 40 1.9 Bilateral, Multilateral, and World Bank Aid Per Capita and Income Level 40 1.10 Aid and Population 42 1.11 Allocation of Bilateral Aid, 1970-93 43 1.12 Allocation of Multilateral Aid, 1970-93 44 1.13 Poverty and Policy, 113 Developing Countries, 1996 45 2.1 Zambia: Aid and Policy 50 2.2 Ghana: Aid and Policy 51 2.3 Bolivia: Aid and Policy 52 3.1 Distribution of Aid, by Type and Sector 61 3.2 Full Fungibility 62 3.3 Partial Fungibility 63 3.4 A Dollar's Worth of Aid and Government Expenditure Estimates from Large Cross National Samples Estimates from Individual Countries and Small Samples 65 3.5 A Dollar's Worth of Aid and Government Investment ("Development") Expenditures 67 3.6 Bilateral Aid and Government Consumption 67 3.7 A Dollar's Worth of Aid to Agriculture and Spending on Agriculture 68 3.8 A Dollar's Worth of Aid to Health and Education and Spending on Health and Education 70 3.9 Public Spending on Health as a Percentage of GDP and Health Outcomes 76 vii ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY 4.1 Participation and Project Success 87 4.2 Civil Liberties and the Probability of World Bank Project Failure 88 4.3 Cost and Benefit of an Additional Week of Analytical Work 101 5.1 Vietnam: Aid and Policy 107 5.2 Pakistan: School Enrollment of Five- to Eight-year-old Girls 110 5.3 Cameroon: Change in Health Center Use and Utilization, 1990-91 112 Box Figures 2 Total Aid: OECD Official Development Assistance and Adjusted Official Aid 8 3 Institutions, Policies, and Growth 12 Text Tables 1.1 Real Per Capita Income Relative to the United States 31 2.1 Features of Successful and Failed Adjustment Programs 52 3.1 Donor Financing of Investment Budgets in Two African Countries (percent) 71 3.2 Evaluating the Effect of Projects with Fungibility 72 Appendix Tables A1.1 Estimating the Effect of Aid on Growth 124 A1.2 Countries in the Dataset 125 A2.1 Estimating the Outcome of Adjustment Loans 129 A3.1 Impact of Foreign Aid on Total, Current, and Capital Public Expenditures, 1971-90 133 A3.2 Impact of Sectoral Concessionary Loans on Sectoral Government Expenditure, 1971-90 134 A.4.1 Base Specification for the Nongovernance Determinants of the Economic Rate of Return of Government Projects, 1974-87 137 A4.2 Impact of Civil Liberties Indicators on the Economic Rate of Return of Government Projects, Controlling for Economic and Project Variables 138 A5.1 Impact of Economic and Sector Work on Project Outcome and Economic Rate of Return 141 Viii Foreword FOREIGN AID IS AS MUCH ABOUT KNOWLEDGE AS IT IS ABOUT MONEY. Helping countries and communities generate the knowledge that they need for development is a prime role of assistance. And aid itself is a learning business that continually evolves as lessons of success and failure become clear. AssessingAidis a contribution to this ongoing learning process. It aims to contribute to a larger "rethinking of aid" that the international com- munity is engaged in-a rethinking in two senses. First, with the end of the cold war, there is a group that is "rethinking aid" in the sense of ques- tioning its very existence in a world of integrated capital markets. In response to this trend, we show that there remains a role for financial transfers from rich countries to poor ones. Second, developing and developed nations alike are reconceptualizing the role of assistance in light of a new development paradigm. Effective aid supports institutional development and policy reforms that are at the heart of successful development. Though tremendous progress has been made in the past 50 years, global poverty remains a severe problem. Research into aid, growth, and poverty reduction provides important evidence about making aid more effective. This report summarizes the findings of recent World Bank research on aid effectiveness. The primary research has been published or will soon be published in professional journals. This report aims to disseminate the findings to a broad audi- ence. It should be stressed that there are important aspects of develop- ment cooperation that have not been covered by the research. The report restricts itself to areas where we have new findings. Aid is channeled through a variety of activities that bring a mix of money and ideas. Two key themes emerge in this report. One theme is that effective aid requires the right timing and the second that it requires the right mix of money and ideas. The timing of assistance is crucial in helping countries improve their policies and institutions. When countries reform their economic policies, well-timed assistance can increase the benefits of reform and maintain popular support for them. Similarly, at the local level, when communi- ties organize themselves to improve services, the availability of aid can make the difference between successful innovation and failure. ix ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY On the mix of activities, we find that money has a large impact-but only in low-income countries with sound management. Before countries reform, finance has little impact. Thus, donors need to rely on other instruments to support development in the distorted environments- and the report looks in detail at what is likely to work and what, not. To have a greater effect on global poverty reduction through equitable and sustainable development, foreign assistance requires a three-way part- nership among recipient countries, aid agencies, and donor countries. The recipient countries must be moving towards sound policies and insti- tutions. The active involvement of civil society can help sustain this sound management in developing countries. The good news here is that many poor countries have initiated serious reforms in policies and governance, so that the climate for effective aid is the best that it has been in decades. Development agencies must shift away from total disbursements and the narrow evaluation of the physical implementation of projects to create high impact assistance. Agencies should be evaluated on the extent to which they have allocated their resources, both of finance and knowledge, to stimulate the policy reforms and institutional changes that improve people's lives. The good news here is that both bilateral and multilateral agencies are trans- forming themselves and cooperating together to become more effective. Finally, citizens of donor countries must continue to support aid. The bad news is that just as aid is poised to be its most effective, the volume of aid is declining and is at its lowest level ever. This report is also aimed at persuading developed countries to continue to give aid and to take an active interest in development and development cooperation. More effective development means improvements in the lives of hun- dreds of millions of people: more food on the table, healthier babies, more children in school. These are things worth fighting for-and properly managed, foreign aid can make a big contribution. Joseph E. Stiglitz Senior Vice-President for Development Economics and Chief Economist The World Bank November 1998 x The Report Team T HIS POLICY RESEARCH REPORT WAS WRITTEN BY DAVID DOLLAR and Lant Pritchett of the Development Research Group. It cul- minates a research program on aid effectiveness initiated and supervised by Lyn Squire. Original research as background for this report includes work by the authors and by Craig Burnside, Klaus Deininger, Shanta Devarajan, William Easterly, Deon Filmer, Jonathan Isham, Dani Kaufmann, Elizabeth King, Jennie Litvack, Luis Serven, Lyn Squire, Vinaya Swaroop, and Jakob Svensson. The authors draw extensively on evaluations and studies done in other parts of the World Bank, which are cited throughout the text. Much of the underlying research was presented and discussed at the joint OECF-World Bank symposium, "A New Vision of Development Cooperation for the 21st Century' (Tokyo, September 1997). The authors would like to acknowledge excellent research assistance from Mita Chakraborty, Charles Chang, Giuseppe larossi, and Pablo Zoido-Lobaton and superb administrative support from Emily Khine, Kari Labrie, and Raquel Luz. Lawrence MacDonald provided helpful support throughout the preparation of the report. Bruce Ross-Larson, with Jessica Moore and Sharifah Albukhary, edited the report for publication. The judgments in this policy research report do not necessarily reflect the views of the World Bank's Board of Directors or the governments they represent. xi xii OVERVIEW Rethinking the Money and Ideas of Aid F OREIGN AID HAS AT TIMES BEEN A SPECTACULAR SUCCESS. Foreign aid has at times Botswana and the Republic of Korea in the 1960s, been a spectacular Indonesia in the 1970s, Bolivia and Ghana in the late success-and an 1980s, and Uganda and Vietnam in the 1990s are all exam- unmitigated failure. ples of countries that have gone from crisis to rapid devel- opment. Foreign aid played a significant role in each transformation, contributing ideas about development policy, training for public policymakers, and finance to support reform and an expansion of public services. Foreign aid has also transformed entire sectors. The agricultural innovations, investments, and policies that created the Green Revolution-improving the lives of millions of poor people around the world-were financed, supported, and disseminated through alliances of bilateral and multilateral donors. Internationally funded and coordinated programs have dramatically reduced such diseases as river blindness and vastly expanded immunization against key childhood diseases. Hundreds of millions of people have had their lives touched, if not transformed, by access to schools, clean water, sanitation, electric power, health clinics, roads, and irrigation-all financed by foreign aid. On the flip side, foreign aid has also been, at times, an unmitigated failure. While the former Zaire's Mobuto Sese Seko was reportedly amass- ing one of the world's largest personal fortunes (invested, naturally, out- side his own country), decades of large-scale foreign assistance left not a trace of progress. Zaire (now the Democratic Republic of Congo) is just one of several examples where a steady flow of aid ignored, if not encour- aged, incompetence, corruption, and misguided polices. Consider Tanzania, where donors poured a colossal $2 billion into building roads over 20 years. Did the road network improve? No. For lack of mainte- nance, roads often deteriorated faster than they could be built. 1 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Foreign aid in different times and different places has thus been highly effective, totally ineffective, and everything in between. Perhaps that is to be expected in a complex endeavor that has spanned half a century, with scores of countries as donors, a hundred countries as recipients, tens of Financial aid works in a thousands of specific activities, and nearly $1 trillion in finance. But good policy environment hindsight is valuable only if it produces insight. The checkered history of assistance has already led to improvements in foreign aid, and there is scope for further reform. The pressing question: How can development assistance be most effective at reducing global poverty? The answer is needed urgently. While there has been more progress with poverty reduction in the past 50 years than in any comparable period in human history, poverty remains a dire global problem. More than a billion people live in extreme poverty-on less than $1 a day. Even more lack basic services that people in developed countries take for granted: clean water, sanitation, electricity, schooling. It is ironic-and tragic- that just as economic reform has created the best environment in decades for effective assistance, donors have cut aid back sharply. In 1997 OECD donors gave the smallest share of their GNPs in aid since comparable sta- tistics began in the 1950s-less than one-quarter of 1 percent. It would take roughly a 50 percent increase even to restore aid to its 1991 level. There have been many excellent studies of foreign aid.' But there are three important reasons to revisit this previously charted territory. First, recent shifts in the global economic and political environment-notably the end of the Cold War and the surge in private capital flows to the devel- oping world-have affected the landscape for development assistance in a way that has left many questioning the very existence of aid. Second, there has been a shift in development strategy that requires a new approach to aid as a tactic within the evolving agenda. Third, there is new empirical evidence that provides insights into the puzzle of what is effec- tive aid and what is ineffective aid. This rethinking of aid produces the following findings: m Financial aid works in a good policy environment. Financial assistance leads to faster growth, poverty reduction, and gains in social indica- tors in developing countries with sound economic management. And the effect is large: with sound country management, 1 percent of GDP in assistance translates into a 1 percent decline in poverty and a simi- lar decline in infant mortality. In a weak environment, however, money has much less impact. A $10 billion increase in aid would lift 2 OVERVIEW 25 million people a year out of poverty-but only if it favors coun- tries with sound economic management. By contrast, an across-the- board increase of $10 billion would lift only 7 million people out of their hand-to-mouth existence. " Improvements in economic institutions and policies in the developing The value of development world are the key to a quantum leap in poverty reduction. True, there have projects is to strengthen been sharp improvements in governance and policies in the past institutions and policies decade, but further reform of the same magnitude would lift another so that services can be 60 million people a year out of poverty. When societies desire reform, effectively delivered. foreign aid can provide critical support-in ideas, training, and finance. Efforts to "buy" policy improvements in countries where there is no movement for reform, by contrast, have typically failed. " Effective aid complements private investment. In countries with sound economic management, foreign aid does not replace private initiative. Indeed, aid acts as a magnet and "crowds in" private investment by a ratio of almost $2 to every $1 of aid. In countries committed to reform, aid increases the confidence of the private sector and supports impor- tant public services. In highly distorted environments, aid "crowds out" private investment, which helps explain the small impact of aid in such cases. . The value ofdevelopment projects is to strengthen institutions and policies so that services can be effectively delivered Aid brings a package of knowl- edge and finance. Most aid is delivered as investment projects in par- ticular sectors such as roads, water supply, or education. Project finance, however, often does not increase spending in a sector any more than an untied grant would have-that is, aid finance is typically fungible. Thus, choosing such laudable sectors as primary health or education cannot ensure that money is well used. Aid is financing the entire pub- lic sector, and the overall quality of policies and institutions is the key to securing a large return from this finance. These findings highlight that the most critical contribution of projects is not to increase fund- ing for particular sectors, but to help improve service delivery by strengthening sectoral and local institutions. The knowledge creation supported by aid leads to improvements in particular sectors, whereas the finance part of aid expands public services in general. * An active civil society improves public services. One good idea that many projects have supported in recent years is a participatory approach to ser- vice delivery, often resulting in huge improvements. The best aid projects support initiatives that change the way the public sector does business. 3 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY The top-down, technocratic approach to project design and service deliv- ery has not worked in areas critical for development-rural water supply, primary education, natural resource management, and many more. m Aid can nurture reform in even the most distorted environments-but it Financial assistance requires patience and a focus on ideas, not money. In some of the poor- must be targeted more est countries of the world, the government is not providing effective effectively to low-income policies or services, which is why government-to-government transfers countries with sound have yielded poor results. Still, there are often champions of local or economic management. sectoral reform, and aid at times has been effective supporting these initiatives. This work is staff-intensive and results in little disburse- ment of funds. Successful assistance here aims to help reformers develop and test their ideas. Making aid more effective in reducing poverty requires five policy reforms. First, financial assistance must be targeted more effectively to low-income countries with sound economic management. In a good pol- icy environment financial assistance is a catalyst for faster growth, more rapid gains in social indicators, and higher private investment (chapter 1). In a poor policy environment, however, aid has much less impact. Clearly, poor countries with good policies should receive more financing than equally poor countries with weak economic management. Up until the early 1990s, however, finance has gone in equal amounts to well man- aged countries and to poorly managed ones. Furthermore, much of aid continues to go to middle-income countries that do not need it. It is pos- sible to make aid more effectively targeted to poor countries and to bet- ter management simultaneously. Second, policy-based aid should be provided to nurture policy reform in credible reformers. Experience shows that donor financing with strong conditionality but without strong domestic leadership and political sup- port has generally failed to produce lasting change (chapter 2). Continued flows to governments that pay only lip service to reform have been a major problem. Policy-based financing should go only to countries with a strong track record or where there is a demonstrable basis for optimism (to sup- port, for example, the concrete actions of domestically initiated reform efforts or a government newly chosen on a reform platform). New gov- ernments in post-conflict situations are often good candidates for sup- port. In countries with poor policies and no credible reform movement, assistance should assume the more modest and patient role of dissemi- nating ideas, transmitting experiences of other countries, training future 4 OVERVIEW policymakers and leaders, and stimulating capacity for informed policy debate within civil society. These measures are relatively inexpensive and do not conflict with the proposal that the bulk of finance should go to countries with sound economic management. Third, the mix of aid activities should be tailored to country and sec- Policy-based aid should tor conditions (chapter 3). Even where institutions and policies are weak, be provided to nurture donors have tried to find something useful to finance. Surely it must be policy reform in credible a good thing to finance primary health care or basic education? The evi- reformers. dence, however, is that aid is often fungible, so that what you see is not what you get. In circumstances where similar projects would have been undertaken anyway, donor money for particular projects and sectors does not necessarily "stick"-simply expands the government's budget. Thus even rigorous project selection or reallocation of donor finance to laud- able activities cannot guarantee the effectiveness of aid in a distorted envi- ronment. To measure the effect of their finance, donors must look at overall allocations and, even more important, at the efficacy of public spending. The allocation of expenditures alone does not guarantee success, for the quality of public spending is as important as its quantity. In countries with sound economic management (of both macroeconomic policy and delivery of public services), more aid can be in the form of budget sup- port, which would simplify administration and reduce overhead. In countries with basically sound policies but weak capacity for delivering services, project aid should be a catalyst for improving the efficacy of pub- lic expenditures. Countries without good policies, efficient public ser- vices, or properly allocated expenditures will benefit little from financing, and aid should focus on improvements in all three areas. Fourth, projects need to focus on creating and transmitting knowl- edge and capacity. The key role of development projects should be to support institutional and policy changes that improve public service delivery (chapter 4). Even where money may not stick, the local knowl- edge and institutional capacity created by the catalyst of aid projects can. Where projects are innovative, it is crucial to have objective and rigor- ous evaluation of outcomes and dissemination of new information. Knowledge about what works in service provision-and what does not-is one of the most important outputs of development assistance. In many cases innovative approaches to service delivery will involve greater participation by local communities and decentralization of decisionmaking. 5 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Fifth, aid agencies need to find alternative approaches to helping highly distorted countries, since traditional methods have failed in these cases (chapter 5). Communities and governments are heteroge- neous, and even in the most difficult environments there will be pock- Aid agencies need to find ets of reform. Donors need to be patient and flexible and look for alternative approaches windows of opportunity to nurture these reform efforts. Typically, ideas to helping highly will be more useful than large-scale finance. Donors' ability to work in distorted countries, these environments has been hampered by an "approval and disburse- since traditional methods ment culture" that does not value small-scale, staff-intensive activities. have failed. In the past agencies have too often focused on how much money they disburse and on narrow physical implementation measures of the "suc- cess" of the projects that they finance. It turns out that neither measure tells much about the effectiveness of assistance. The evaluation of devel- opment aid should focus instead on the extent to which financial resources have contributed to sound policy environments. It should focus on the extent to which agencies have used their resources to stim- ulate the policy reforms and institutional changes that lead to better outcomes. These are not easy questions to answer, but independent reviews of development agencies-with participation of developing country policymakers and project beneficiaries-can help establish whether agencies are doing a good job. Box 1 Defining Aid 1\ I iH LiFFEF -- 1 P - 1 1 .-.It i i i%rLd i agencc .:d dtn. r k2o-,izrnm1Ln[ ' duch a, Lhe J(: Jpnr r a -iscjlnc irid .4fiJil d,lJopirwra L 1 :n '0 IntrI[ IMnICaln31 [)ci%elopni.ra or Japan finin,c: Th tIr,i i, - ib;CL , - 11 .cC.nd :rid Lorim- h .er4L.! E io nik i..pir.-rion f andi. MIullaer.il P fAM" IL 1 th l nil h [ hi, [ aici .1mPan c n1uCd b. cFUMICbutA o frorm %%e.dth% ai f Pcrii :2rirE CcI1p.ni.hl. ( lit;la dc%eloprimn L-tiOUILS dild - dfiiiiircrcd K1, ,. ciic. iuch is the 1110a III fllii,:in chac tl...i ft m dcchpJ Unrcd Niuon. L),.-lopircmn Prorrimm-c .ind Lh C*Un l' .. rPsc n arid 1111. l ct.krcUl ri.10 , [* [he WV.rlid hink ( it all t.. .l I ac-'elopniu-1 a.siMance. *ic\ J p!I~~~ rl. ~c orec:.hi, iia!in~ ng1-[in lrr. c ch' 1 third 1-; n1UI11k :rjI iaic, :1.1,c i:,C,m-imcr,'a1 1a-.Ie. - _r M lnid'c itil ' .raii bil i[t r. d lid i' micd- d"iil I -. it iiiu,t bc u,ecd . cIIC Red *.irh oftl,sJl devel.ipmnr 11111E3DfCC 111d i 111 L )d;i nd L r, [CC- trii Lh, i d)rir coun.1 r marin ,j rk 'zcced [ [h1 pih .rc( c--uncric, 1hK. ia 1U- i.L_', h 4-- * n chi cEcd id rcduc .s i tc .,:r Of [uncc ix [h, primary Cu -fthi rudl . bur minin- of that 1,cir IrL%:L b. alboLu I p, renc. and herc iide - i Imirld inc r rc TC l ani fir [r hc ifr; i CaEc cir- -F6 [H- -pre1d "ZrelMllE n 1.iEu iiA blcril aid %,.ould Ci1 de.el.pifnl[ nir r,* miki 11 illcrc cftall. C .0 i'in: ()I r.L.) , un[rIes hcr fc.ih mlp:i itid . in CP dil.d,71 i. i i H1 b iCTl iLi ha bceii uAi r vren .i , i tr,rn uia I i,d in 191 ) , mululuni :rcrai componenrir BilareraImsanui idmir - .iccouncd G..r onl ib-it , ilfth f all aid 6 OVERVIEW The New International Environment FOREIGN AID IS A POST WORLD WAR 11 PHENOMENON. FROM THE start, it had twin objectives, potentially in conflict. The first objec- tive was to promote long-term growth and poverty reduction in After peaking in 1991, developing countries; the underlying motivation of donors was a combi- aid has fallen. nation of altruism and a more self-interested concern that, in the long term, their economic and political security would benefit if poor coun- Figure 1 Financial Flows to tries were growing. The second objective was to promote the short-term Developing Countries political and strategic interests of donors. Aid went to regimes that were Billions of 1995 U.S. dollars political allies of major Western powers. Thus the strategic and develop- mental objectives were potentially, but not necessarily, at odds. Consider flows Bolivia and Zaire. Both received U.S. aid, partly for strategic reasons, yet 250 the outcomes were vastly different. Bolivia used the resources relatively well after reforms of the mid-1980s and over the past decade has stabi- 200 lized and laid a groundwork for success. In the former Zaire, by contrast, Official developmentO it is hard to see any benefit from its years as a major aid recipient. 5l0me t During the 1970s and 1980s foreign aid from OECD countries rose steadily (figure 1).2 In 1991 oOfficial development \ 100 assistance billion (in 1995 prices; see boxes 1 and 2). In the 1990s, however, three events have lowered the absolute and relative importance of foreign aid: so fiscal problems in OECD countries, the end of the Cold War, and the dramatic growth in private capital flows to developing countries. In recent years OECD countries have been struggling to control fis- 1970 1980 1990 1997 cal deficits and contain growth in government spending. Even though Source: Global Development Finance 1998. foreign aid is a tiny fraction of budgets, it has been one of the first items for the ax. All major donors reduced aid relative to their GNPs between 1991 and 1997 (figure 2). The decline was especially sharp in the United States-aid was a mere 0.08 percent of GNP in 1997. Sweden and other Nordic countries have traditionally been generous, giving almost 1 per- cent of GNP. But among large countries, France is the only one that gives more than 0.45 percent. Collectively, OECD countries contributed just 0.22 percent of their GNP in 1997. The end of the Cold War likely influenced some countries' decisions. The strategic importance of aid has ebbed; as a result it risks losing its broad support among donor governments. At the same time, there has been a surge in private capital flows to developing countries. In the 1970s and 1980s official finance-that is, money from bilateral donors and multilateral institutions-represented 7 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Box 2 Measuring Aid 'I OEOf' 0 O II l- -itr . 1 - . L.. T1 1 Box figure 2 Total Aid: OECD Official puhbIhc. int-.mii'.* *'n it mcrnhcr' .,Id r. 11%'..p- Development Assistance and Adjusted Official Aid ino .*aurtrat [tb [ 1 . pri!n -t' r di.hur;Elc ri I C..iCl- 10 1 in i, lha bi u It1 .1 > e PLIcn Billions of 1993 U.S. dollars rant c s nt r1 dtI-rncn ' 1 r ..i,It 3 'i Er OECD AT I.Ltr t tir * c .nr i ..n11 0i:i dlit 2rn.r ,lzmn t ofrcial .iA .tdd tEat tEst 'dire tr pit!: alt,. U to vb 5 development assistance 1 uPiUch. .10'i i 11 [ ndic mi .r1 cil0ssii anc tL t1I0pcd dtlc i oniid 1, ir 1"i ci, ipi.. ncs i, lc ictirz I 1.z . [ rsr*:c p*il; '.. I ii i.Lin ub'st d11 1nc 30 Adjusted diE.1 I Er. dTh 1dlIsrd 'i.1rc Esd t.K Is.1 ci [crh j th, official aid l onE - m iL ItLir1 c u scoid chc rm.. einc, irc ILsd c'rr% and pr- dacc -irl-Il r rz,I1r, in ICI n0rrit f1l ,1s , d i n I u th 1 r.1 u1 MK Ih 0'. .,ppr ).Cihd,-c- n- [ ha,c i-u.h111s. iO 1s. *.tr l li ;Lrcz 1-FTlii. - cl-fi- 0 ild I L t itrTiis. 1 ThI r.h, AdILsed. idJ tllpigur J-. hi'. 1 h Iapcr C Ile in 1. li In ciz jl i an 10 hc '"E(-D m us-11,1. I l 0 o0 I'tL1iI E inctcrninl.i1 irerL11 rEC V11e bizrt I'.'-. ill 1970 1975 1980 1985 1990 l il L.-a ., o rhe d . m .n t s u n .s i n ll k n. . .. . .. r i 1' i.r i ink iuEnlo A d in 3 It] rlin ,th a? pcr.eral iIHe;.r-7 t % .[ r:tr I- III ill I f u lr L -T 0-W ll In(ii:;1 rac..:- .ii. fCr icm 13d i f r h-c Irl :i 1 rc nt about half of all finance going from the developed to the developing world (see figure 1). With private flows expanding to more than $250 billion in 1996, however, official finance is now only a quarter of all finance available to developing countries. Private capital flows are heavily concentrated in a few countries, however-and some flows are volatile. A surge in the late 1970s receded after the onset of the debt crisis in 1982. Another big rush occurred in the mid-1990s, but with the financial crises rocking East Asia in 1997 foreign investment dropped sharply. The flow of private money to the developing world fell by $80 billion between 1996 and 1997. In any event, private flows will continue to go to a small number of (mostly) middle-income countries. In 1996, 26 countries received 95 percent of private investment; the rest went to the other 140 developing countries. 8 OVERVIEW Figure 2 Offical Development Assistance Relative to Gross National Aid is down everywhere. Product, Major Donors, 1991 and 1997 ODA as a percentage of GNP 1.0 1991 1997 0.8 0.6 0.2 Sweden France Canada Germany Japan United Italy United OECD Kingdom States Source World Development Indicators 1998. In a typical low-income country, foreign aid remains far and away the pri- mary source of external finance, amounting to 7-8 percent of GNP. Developments in the 1990s have sharply altered the climate for for- eign aid. The end of the Cold War opened up new possibilities: with aid no longer constrained by those strategic objectives, it should be possible to make aid more efficient at meeting its primary objective of long-term growth and poverty reduction.3 Yet, given budget problems and rising private flows, donors are clearly rethinking the importance and value of foreign assistance. New Thinking on Development Strategy FOREIGN AID, JUST ONE WAY OF PROMOTING DEVELOPMENT, MUST fit within a broad overall strategy. Past domestic and international political conditions and beliefs about development strategy structured organizations, instruments, and implementation of aid. But those beliefs have undergone enormous, and accelerating, change. 9 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY In the early days of development assistance, government was seen as the positive agent for change. Domestic markets in developing countries were thought to be nonexistent and incapable of growth. International markets were tainted by the association with colonialism, as well as by the If all the aid to Zambia collapse of markets for commodities and credit in the Great Depression had gone into productive of the 1930s. In many developing countries the first flush of indepen- investment, it would be a dence created optimism about new governments as agents of political, rich country today. social, and economic change. Government-to-government aid had a plausible claim as the best way to promote development. Figure 3 The Gap between Model Development economists believed not only that poor countries were and Reality in Zambia, 1961-94 held back primarily by a lack of physical and human capital (which was, and remains, true) but also that domestic poverty and international mar- Thousands of 1995 U.S. dollars ket failures denied developing countries access to investment funds needed 25 for economic growth.4 Calculating countries' growth "requirements" of The per capita income that Zambia would have investment finance or foreign exchange (or both)-and comparing them 20ad if id ad n with what was available-emphasized the size of the gaps to be filled. The investment had gone natural tactical response was to fill the gaps with foreign aid through trans- into growth according I' to the two-gap model fers of investible foreign exchange. If money was the problem, then "mov- 15 ing the money" was an appropriate objective for aid and aid agencies. The contribution of aid could then be measured in dollars. 10 Sadly, experience has long since undermined the rosy optimism of aid- financed, government-led, accumulationist strategies for development. Suppose that development aid only financed investment and investment ambian per really played the crucial role projected by early models. In that case, aid to capita income Zambia should have financed rapid growth that would have pushed per capita income above $20,000, while in reality per capita income stagnated 1961 1970 1980 1990 at around $600 (figure 3). The past 20 years have seen the death of centrally Source: Easterly 1997. planned economies, stagnation in the leading import-substituting models of the 1970s (Mexico and Brazil), and broad economic failure (if not absolute disintegration) of post-independence Africa, which pursued a state- led strategy.5 The past 20 years have also seen waves of spectacular increases in incomes and exports of East Asian economies-first, Korea, Taiwan (China), Hong Kong (now returned to China), and Singapore, followed by Thailand, Malaysia, Indonesia, and, after important economic reforms, China; the emergence of Chile as the most dynamic Latin American econ- omy and the recent recovery of others; and successes in Africa, such as Botswana and Mauritius. The evidence suggests that rapid development is possible, and should be based on markets and on effective states playing an economically important facilitating, but not dominant, role. 10 OVERVIEW So, there have been three phases of development thinking. In the first, market failures were seen as pervasive and complete, and government as the only solution to all ills. In the second there was a brief period when government failure was seen as pervasive and complete, and markets (if not the solution) as the only hope. Today's third view-pragmatic but not Together with the new ideologically satisfying-is that both markets and governments have per- development strategy vasive failures but that these usually are not complete. This emphasizes comes a broader agenda. that government should focus on areas where the problems in the absence of intervention are greatest-but government must have the capacity to improve the situation. "We need to recognize both the limits and strengths of markets, as well as the strengths, and limits, of government interven- tions aimed at correcting market failures" (Stiglitz 1989, p. 202). The development strategy emerging from this view is two-pronged- put in place growth-enhancing, market-oriented policies (stable macro- economic environment, effective law and order, trade liberalization, and so on) and ensure the provision of important public services that cannot be well and equitably supplied by private markets (infrastructure services and education, for instance). Developing countries with sound policies and high-quality public institutions have grown much faster than those wirhout-2.7 percent per capita compared with -0.5 percent per capita (box 3). Put simply, failures in policymaking, institution building, and the provision of public services have been more severe constraints on development than capital markets. Together with the new strategy comes a broader agenda. Early devel- opment practice focused on growth of per capita income. But in reality, developing countries are concerned with broad improvements in the quality of life-higher incomes, yes, but also reduced poverty, advances in literacy and health, and environmentally sustainable development. The new agenda is reflected in the goals set forth by the donor commu- nity, in consultation with developing country partners: " Reducing by one half the proportion of people living in extreme poverty by 2015. m Achieving universal primary education in all countries by 2015. m Making progress toward equality of the sexes and the empowerment of women by eliminating disparities in primary and secondary educa- tion by 2005. * Reducing by two-thirds the mortality rates for infants and children under age 5 and by three-quarters maternal mortality-both by 2015. 11 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Box 3 Defining Sound Management: Good Policies and Institutions .1 N' 1. -. -i: :i. . .il fi . - i ut i[l i in ., 1¯he diii :. .ite. ..mnim p.lick d i rt b.% figur 5 AndI LIb rhir |iii leid rö r.pid dinc. nmet and *n.mitnc. [birce ïc[rzi it h ic been houv n in p. :rr rcdt,inin ff i .attiul. i . un[r. . DXC I *mpl n 1 ,[adle Lu Alter de. uptfl tlUndtriL- t ltri c -Jiri abiiiu[ . d and h.id poiic--~ tritni [hell ru *~.» h t.1 11 El it [be h L u,rplus. i d traJ 'pen- i t '. [ [iri& 11ï c it ~,hr .t rin . ind nu, A, ivtjm_rcJ 1ït u . in,1 rn rA tllïll Ecilur![ i. di¯ til[ hLi[ ni[ imrup.:.%ibIe [' rr~c .ur .~ .iru[ ~. trh .u r pieL , it . uld hc unc . irh high :1 u' F vi: indit ibir. iittl. n 1 rl c.d imbalince, .und i cd'sed trade regit 1Nix in rht 1 . r mniancel An Box figure 3 Institutions, Policies, and Growth .1..I .. c--d O rini mic p.. e 'uid be Uinda Annual percentage growtri in real GDP per capita ei t i:uire ut In.[iU.u[.. iläl gullity n'.o' e, in ic- mienr e( rbc [Lt-e h il br U ru -- 1 I.. t.he quli- 3 illit p01 bb1t1 kuic.ut ni, -nd [he jen +.uniu .ofc*r- rupci,in . [hi er i, of 1oi~'¯- N ha, 'hown. 2 ·ind.-.liest j 1 I u T-! . rh pir in sut.uuOnl quan . 1 s· dl..niru. itu it niuAn d . i 1-tor [ta suppor[g o,.d iiiaid crmenir. [hu liLJb -.-il v. wi or Ibr,iadun rh, lit iuré bé. rit [ht n .ttrtuCinu.iii tnd 1ii,[1[u- titutntl cjri llriiiim a[Ed Iir:.re. F"r e . i ple. eftoir *. itprin.; ¿dicaton itd huih :rc et it.il tor suit -1 ic util J.1 ip n t Atd u;arnrnnt surPPort ti, [uri ul[iural i,- rch id e rention [ntd v ir m u- r. [ hd.inr eff1-rr mo.de an in ,,prr[nc contribii- Institutional a c ii hk . .9% Economic ' quailty l,. vrueräl rir i, rFa the policy - inta,l e r fim1 ru rnt the letutl.I eNp¿ri,ne,, .t * Providing access through the primary health care system to reproduc- tive health services for all women and girls of child-bearing age as soon as possible and no later than 2015. * Implementing national strategies for sustainable development in all countries by 2005 to ensure that losses of environmental resources are reversed both nationally and globally by 2015. 12 OVER V I E W Box 4 Functions of the Development Assistance Committee ir i i .. - i C l ., 1'Il. FHj i IE( 'IY Cr r che l i[ deade and a h:lf th, DAt ha, wid- LIK ci prnc[n X rqanl (iinGmi[te 1 1.\t .) h.s tunc- id nd pu bl1hed a comprrhen l.. 3[ r nnn ipc, f r [i.l'nd 1tbc principal sira[cgi -citin?,, ind p 'li..' and Eti c \id li SC giiid i n, and hekr praui.. e s _ er perftrml n tcr; i i l . Pgan If Ic maifj'r biL, [cr il d n-trs pL p:.it o 'ricn[ it)nla and ''c ri jnu i isuc s in c, t r.l l[1äcrål id prograrïi ar: Inir nclediand aicaei t md marigcmntic uch , cird!iarn.pr hCrIF implcmenrd in a pråiehal i uum IndvåJ rhe; ein iririce. picrian iince.tchniJ sitan.rO- 'ub| L t' C i n..dcrbl d*mdetUl prcisurv fr-n p':lir- Cureient. and cvciar -a ' cli a flmr ILch biic . .l-tnd cvmic !l interelr grnupi 111 bh dunur dtn ni.s i1 dc\ clpirient par[i ipa[on ind gomd ,ounrri, And bilit, ri .ïd amnckrs Cai be -uke[ o jare rnance, ern irnenr, mid genler ,qu uJin. rhe , 1 LC lind d ur"-m,.--Jrirï rin mie, a' Tbc T D-VZQ \\nrkin- 'arl', 'n Aid EtalUit10n muilul-reral d, ,1.pmnt bm·ikt crtnc incenr.,-, brintg r.gerher the head, nt [, ch,aliiiuoii umri ef 6r raf [, be ppro'al-lcucd ru[ht r tiif tesul t- bil reral,.ndmul [Jrta dt m e h-lC s[nO r t sd [be ci r öe x b moRi C'¯lc med rv mu hnv:h n ''.duilen - apacie, in de\ elping counine, n.r'-nal th'' -n [hirit dcclpmnt prIt rhin t., At [h trraegie lå el, the DAC produced in 1'% pin MC''l EM..~[- ~[ar inlpNK ICIcn[ cr[ t in m i ¯ t~ epitt. us h'. ~ r'l .. '.i ".'i' ; "/ '·t don',r iJenrtic re nieryed. DI; ü'. '. u ''s * ' ' hih pr' ide' the ba;- [be DA(- his been the fIIrum - hre [ht maior fr [hr kind ef p.r[nerships m,,[ lhkthK En pr'duce donors ha\ c .:rkrd [a keep Cihr pr'.ram, umed an Je CliprMen progrr' an [he side ''Cthe dc\ elopin du'el:'pmcnl' i E e [•pr [ 'fl t -rdnati*n. and klit '[.. . S[ring ne c nmmi[i n[ r. an fficriix p. -lc- ro rcn- ic ad 1f"ccxmne'- L,en threc'c ars. cih ens irnmvenr and kve de, coi pmen[ prio'te, aimed DAC nirmbei icl lubiec[ [¯ an ec\a Lii ani -ia '. tiid .it pre--por ,..~rh , and or the deonr otde incre sed riuci.d pcrlk.ranmie by he hc r mcmbers o the finacial suppur tor such phlci on a pror im ariJ ha,id ont[dcs b, rhc OkD ritt and budger leel combmed vi[ an emphasan psrnci- e b e , , pcilal liL:n,,t[cd r. miiner5 drav, n pm[ton and cipativ-buildinc i fAilitiring [ht [r ni,- trom the C-ammirtne Ihcsce elopmeni ( eepera- t r ol krw.xlcdcc ind idy i noi' Re.c;.. ndIuding [h, conclusi,n, rcs. hrd h' the D N , L.P.c been pu lished smec !u-. 4r - .rIrrLL'- 1. U.Ji -Tit These goals are elaborated on in Shaping the 21st Century: The Role of Development Cooperation, produced by the Development Assistance Committee (box 4). All this points to a different role for aid. Development assistance is more about supporting good institutions and policies than providing cap- ital. Money is important, of course, but effective aid should bring a pack- age of finance and ideas-and one of the keys is finding the right combination of the two to address different situations and problems. 13 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Money Matters-In a Good Policy Environment O, GIVEN THE NEW PATH IN DEVELOPMENT THINKING, WHAT IS important for long-term growth? Stable macroeconomic environ- Aid has a large effect ments, open trade regimes, and protected property rights, as well as when countries have efficient public bureaucracies that, can deliver education, health, and sound management. other public services. When developing countries have this kind of sound management, financial aid has a big effect on growth and poverty reduc- Figure 4 Per Capita GDP tion, improving social indicators over and above what good management Growth in Low-Income Countries itself induces. Equally, aid has little effect on the development of coun- with Sound Management tries with poor management. Financial aid will have a lasting effect only in a healthy climate for efficient investment and human capital develop- Annual percentage growth ment. Three sources provide the evidence-cross-country studies, research into the success and failure of investment projects financed by the World Bank, and case studies of aid effectiveness. Few cross-country studies have found a robust effect of aid on growth. The picture changes, however, if countries are distinguished according to their economic management. Aid generally has a large effect in good-man- agement environments: 1 percent of GDP in assistance translates to a sus- 2 tained increase in growth of 0.5 percentage points of GDP. Some countries with sound management have received only small amounts of aid and have grown at 2.2 percent per capita. The good-management, high-aid group, I however, grew much faster-3.7 percent per capita (figure 4). There is no such difference for countries with poor management. Those receiving small amounts of aid have grown sluggishly (if at all), as have those receiving large 0 High aid Low aid amounts. Introducing other variables does not change the picture. The effect of aid goes beyond growth. In a country with sound man- Source: Burnside and Dollar i998. agement, 1 percent of GDP in assistance reduces poverty by 1 percent. Aid has a similar effect on infant mortality, but again only if there is good management. A final lesson from cross-country studies is that, with sound country management, aid works in partnership with private capital. Specifically, 1 percent of GDP in aid crowds in another 1.9 percent of GDP in pri- vate investment. Put another way, assistance to well-managed countries increases private sector confidence and supports important public ser- vices. It hardly needs to be said, but in poorly managed countries aid crowds out private investment. More evidence on the relationships among aid, management, and development comes from analysis of the success and failure of public 14 OVERVIEW Figure 5 Project Performance by Policy and Institutional Environment Projects perform better with better policies and Satisfactory projects (percentage of total) better institutions. 80 60 40 1-.g 20 0- .: Economic High policy Low Institutional quality Source: World Bank 1997a. investment.projects financed by the World Bank-in, for example, roads, power, and education. In countries with good macroeconomic manage- ment and efficient public institutions, projects were 86 percent success- ful, with much higher rates of return. In countries with weak policies and institutions, the corresponding figure is a measly 48 percent (figure 5). Case studies of aid effectiveness have come to similar conclusions. Again, look at some earlier comparisons: the important role of aid to Bolivia after it reformed, compared with the largely ineffective aid to Nicaragua, which had poor policies throughout the 1980s; or the highly effective aid to Botswana (one of the countries with the best institutions and policies) in contrast to the many failures in Tanzania. While it is useful as an illustration to draw a sharp distinction between good and bad management, there is in fact a continuum of policy regimes. Many developing countries fall into a gray area between good manage- ment and poor. The key recommendation from these findings is not that finance should go only to well-managed countries. Rather, we recommend that aid be allocated on the basis of poverty and economic management. 15 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Among countries with similar poverty levels but different policy regimes, more finance should go to the countries with better management. The actual allocation of aid has often been influenced by the strategic interests of donors. Although donor behavior differs, total bilateral aid Being a former colony has favored former colonies and political allies more than open economies attracts more aid than or democracies. An undemocratic former colony gets about twice as having good policies. much assistance as a democratic noncolony, and the same is true for a closed former colony compared with an open noncolony (figure 6). As a Figure 6 Bilateral Aid and result, much bilateral aid has gone to countries with poor management. Colonial Past Indeed, countries with poor management have received about the same amount of bilateral aid as countries with good management (after con- trolling for per capita income and population). so Aggregating the flows from different donors obscures the different 40 criteria on which aid is allocated. In the Nordic countries, strategic 30 variables-such as a colonial past or United Nations voting patterns- 20 play almost no part in allocations. Nordic aid is targeted to the poorest countries, favoring open economies and democracies. Multilateral assis- Colonial tance has been more effectively targeted than bilateral assistance to coun- past tries with sound management, though there is still room for Less No colonial democratic More past improvement. Overall, aid is more effective at promoting growth and democratic reducing poverty when it is channeled to poor countries with sound management. A $10 billion increase in aid would lift 25 million people 30 per year out of poverty-if it favors countries with sound management. 25 By contrast, an across-the-board increase would lift only 7 million out 20 of poverty. 15 There are two reasons to be optimistic that foreign aid can be allo- 10 cated more efficiently in the future. First, the end of the Cold War reduces the pressure to provide aid to strategic allies. In a few highly dis- Colonial 0 past torted economies (Myanmar, Nigeria, Zaire) aid declined dramatically Clsed no colonial in the early 1990s. Second, there has been a worldwide trend toward eco- Opn past nomic reform in developing nations. Thus, a growing number of very Source: AlesinaandDollar1998. poor countries have relatively good policies. Ethiopia, India, Uganda, and Vietnam, for instance, have large populations and many poor peo- ple, but have made great strides in economic reform in recent years. Aid targeted to such low-income reformers can have a big effect on growth and poverty reduction. By the same token, large amounts of finance can- not be put to productive use before countries reform. In the highly dis- torted environments, donors need to find other instruments to support development 16 OVERVIEW Aid Can Be the Midwife of Good Policies MORE THAN JUST PROVIDING MONEY, FOREIGN AID MUST promote sound policies and help develop institutions. Recent literature shows that policy reforms that are not difficult tech- Policy reforms depend nically (stabilization and trade liberalization) can add 2-3 percentage mainly on domestic points to developing countries' growth. Since average developing coun- political and social try growth has been only 1 percent per capita, that would be a huge factors. improvement. Much development assistance is aimed at promoting policy reform, and the record is mixed. Such reforms depend mainly on domestic political and social factors, which are not easy for outsiders to influence. But where development assistance to highly distorted regimes has stim- ulated policy reform, the nonfinancial aspect of aid has often been the important factor. Thus it is possible for donors to make headway in pro- moting policy reform in difficult environments without violating the first recommendation-that is, give more money to good policy per- formers. In poor policy environments, ideas are more important than money. Some of the most important ways in which foreign assistance pro- motes policy reform are hard to measure. There has been a worldwide trend toward economic liberalization in the 1990s, and dissemination by development agencies of ideas about good policy has surely had an influ- ence. Donors and foundations have also played an important role in financing the overseas education of policymakers. The Berkeley-trained group that designed Indonesia's reform package in the 1970s is a classic example. Many of Latin America's impressive reforms in the 1980s and 1990s were engineered by politicians and officials with advanced train- ing that was partly financed by aid. Some efforts may pay off only over a long period. In the early 1990s, for example, the World Bank launched a public education campaign in Ukraine to stimulate debate within civil society about economic reform. Ukraine has yet to achieve really good policies, but the support to popu- lar education may yet have a big payoff. It is difficult to measure the pre- cise effects of disseminating knowledge, educating officials, and stimulating popular debate. Case studies suggest, however, that they are often important to a successful reform program. A more straightforward way to promote policy reform through aid is to make financing conditional on the adoption of certain policies-an 17 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY approach long followed by the International Monetary Fund (IMF) and the World Bank. Results have been far from uniform. In many cases pol- icy measures were not carried out, yet loans were disbursed anyway. About a third of the World Bank's adjustment loans fail in that reform objec- New governments are tives are not met: "policy reforms rarely succeed unless the government is more likely to reform. genuinely convinced that the reforms have to be implemented and con- siders the reform program its own" (World Bank 1997a, p. 37). While Figure 7 Elections, Tenure, and results improved in the 1990s, further progress-incorporating the Probability of Successful Reform lessons of the recent East Asian experience-is still needed. A recent study found that the success (or failure) of adjustment loans o0 ocould largely be predicted by a country's underlying institutional and political features, including whether the leader had been democratically elected and how long the government had been in power. A newly elected reform government had a greater probability of success, than an author- 90 itarian government in power for a long time (figure 7). In general, new Elected governments have a high success rate with reform and are good candi- dates for support. This often includes governments that arise in post-con- 80 flict situations. The study also found that variables under the World Bank's control, such as the number of conditions or resources devoted to preparation and supervision, had no significant effect on the probability 70 Non-elected of success or f&ilure of reform. The implication: conditionality is unlikely to bring about lasting reform if there is no strong domestic movement for change. When 60 domestic constituencies are committed to reform, adjustment loans and 0 5 10 15 Years in power foreign aid can help consolidate policy gains in three ways. First, condi- tional loans are a means by which a reform-minded government can pub- Source: Dollar and Svensson 1998. licly commit to policy measures. Second, conditionality sends a signal to the private sector that a reform program is credible, and this should encourage a quicker response from investors. Third, aid spurs growth in a good policy environment. So, if reforms have moved quickly and already improved the policy environment, financial aid will have a big effect, which in turn should help sustain political support for the reforms. This is the conclusion of a review of eight successfully reforming coun- tries in the postwar period (Sachs 1994). Sachs argues that in all cases the government was already committed to reform, but timely foreign assis- tance played an essential role. What aid did was to "help good govern- ments to survive long enough to solve problems" (p. 512). Donors have gradually learned this lesson, and there is evidence that they have become more selective in providing policy-based aid: in the 1990-95 period, the 18 OVERVIEW success rate ofWorld Bank adjustment loans climbed to 76 percent, com- pared with 67 percent before 1990 (World Bank 1997a). Thus, in supporting policy reform, the mix and timing of ideas and finance are crucial for effective assistance. In countries without reform movements, donors can try to nurture them through analytical work, A dollar of aid increases training, and technical assistance. Such nonfinancial assistance remains public investment by important as reform movements develop and consolidate reform plans. exactly the same amount The timing of large-scale finance must also be exact. If it comes too early, as any dollar of the emergence of a coherent program can be undermined, and the finance government revenue. is unlikely to have much effect on growth. If it comes too late, it is an opportunity missed-to increase the effect of reform on growth and to Figure 8 Public Investment from help consolidate the policy regime. In deciding timing, donors should One Dollar of Tax Revenue or pay attention to the progress of reform, as well as to leading indicators of Official Development Assistance successful reform (a newly elected government, for instance). Conditionality still has a role-to allow government to commit to reform Ii 30 and to signal the seriousness of reform-but to be effective in this it must focus on a small number of truly important measures. 25 Money Matters-In a Good Institutional 20 Environment 15 HE COMPOSITION AND EFFICIENCY OF PUBLIC EXPENDITURES 1o affect both growth and poverty reduction. While public provi- sion of some services is crucial, public involvement in others is 5 at best neutral-and at worst distorts the economy and directs attention and resources from more pressing public sector responsibilities. Although Tax revenue ODA most foreign aid has financed specific investment projects, what you see is not necessarily what you get. In many cases development financing is Source: Feyzioglu, Swaroop, and Zhu 1998. fungible. For instance, although most aid is targeted to finance invest- ment costs, estimates suggest that the net effect of a dollar of aid is to increase public investment by only 29 cents-exactly the amount by which any dollar of government revenue would have raised investment (figure 8). Similarly, an aid dollar used to finance projects in education tends to increase government spending in all sectors to the same extent as a dollar of government revenue from any source. If aid resources are fungible, the effects of "money aid" and "ideas aid" need to be assessed separately. Fungibility means that a government can 19 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY use increased resources as it chooses-to increase spending, fund tax cuts, or reduce the fiscal deficit. The effectiveness of finance depends on the quality of all public investments and expenditures, not simply on aid- financed sectors and projects. This finding has important implications Countries with public for the evaluation and management of aid. Agencies often hone in on the sectors that provide success rate of individual projects as one measure of their effectiveness. effective, high-quality At first glance, this appears to be a focus on "quality." But it can lead to services are prime distorted incentives, depending on the criteria for judging success. Since candidates for large money is often fungible, the return to any particular project financed by amounts of financial aid does not reveal the true effect of assistance. Moreover, if agencies are support. evaluated mainly on the success rate of projects (defined narrowly, with- out accounting for spillover benefits), managers will avoid risky, innova- tive projects in favor of things that are known to work. With fungibility, the impact of aid is not the same as the impact of the aid-financed pro- ject. The return on the finance depends on the overall effectiveness of public expenditures. In addition, there is the important question of how the project was done differently with donor support, compared with how it would have been done without such support. Once fungibility is recognized, choosing the level of support and the instrument for providing assistance requires a view on the overall quality of public sector management. But how to judge the composition and effi- ciency of public spending? Governments need to choose expenditures carefully, with an eye toward the net impact of additional spending on growth and poverty reduction. And that depends on what the private sec- tor would supply without government. Too many developing country budgets have been devoted to activities that have no growth potential and no effect on poverty: wasteful and inefficient public enterprises, middle- class subsidies for fuel, electricity, and more, and spending that benefits mainly the rich, such as credit subsidies and free universities. Moreover, the efficiency of government spending is at least as important as its com- position. Governments should be judged not on how much they spend but on how much they accomplish. Countries with public sectors that provide effective, high-quality ser- vices are prime candidates for large amounts of financial support, and it makes sense to provide some of this as general financing of the budget. This approach recognizes the reality of fungibility and economizes on the cost (both to donors and recipients) of delivering aid. Studies of foreign aid have long pointed to the problem ofpoor coordination among donors and the burden that this imposes on developing countries. Delivering 20 OVERVIEW more finance as general support to the budget would alleviate this problem. As noted, many low-income countries fall into a gray area between good and poor management. Because implementing macroeconomic and trade reforms is technically easier than strengthening institutions (such Donors should be more as the civil service and the rule of law), these countries will often have rel- willing to cut back atively good macroeconomic policies but inefficient service delivery. financing to countries Thus there will have to be more support for building institutions and with persistently low- implementing reforms in different sectors-more ideas, less money. A quality public sectors. greater share of financing should come through projects whose value added is measured by the degree to which design and implementation helped improve performance in those sectors. Finally, donors should be more willing to cut back financing to coun- tries with persistently low-quality public sectors. While it is difficult to withdraw support, attempts to work around a poor-quality public sector are unlikely to produce anything worthwhile or lasting. In these cases donor support should be geared less to financing and more to activities that in the long run may lay the groundwork for institutional and policy reform-again, back to ideas, with only enough finance to put the new ideas to work. The general point is that different country environments require different instruments of support. Aid Can Be the Midwife of Good Institutions ANY IMPORTANT PUBLIC SERVICES-SUCH AS MUCH OF BASIC infrastructure-are difficult (or impossible) to allocate through markets. But the economic characteristics that make these services candidates for government involvement also create problems in designing institutions and incentives that would make the public sec- tor efficient. One major development challenge for 2000 and beyond is getting governments to do well the things that governments must do. Well-designed aid can support effective public institutions and good governance by helping with the experimentation, learning, dissemina- tion, and implementation of new ideas on service provision. Where there is demand for change, aid can make a big contribution, often through projects. As noted, the money element of a project does not necessarily stick with a particular sector because government revenues are fungible. 21 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Thus the main rationale of projects must be to support reform of sector institutions and policies and demonstrate new ways of achieving devel- opment results. Various historical, legal, political, and economic conditions make each Beneficiary country's situation unique-and often require the generation of new participation can knowledge. Aid can help governments by bringing to bear lessons from quintuple project their own, or international, experience, and by creating conditions that success. help experimentation and, more important, evaluation. Aid projects can assist countries to improve the performance of entire sectors. Figure 9 Success in Rural Water Unfortunately, aid has sometimes been part of the problem. Past aid Supply Projects with Differing has gone almost exclusively to (or through) central governments and has Levels of Beneficiary Participation affected how public services are delivered. Even though some local gov- ernments can provide services more effectively, developing countries tend Successful projects (percentage of total) to have unnecessarily centralized service provision, to which aid has con- 70 tributed. The traditional design and management of aid packages have also reduced the participation of local communities in the design and 60 implementation of development projects. Donor responses to weak institutions have in the past often been inef- so fective. Faced with low implementation capacity and pressure to "move the money," aid agencies have a long history of attempting to "cocoon" their projects using free-standing technical assistance, independent pro- 3o ject implementation units, and foreign experts-rather than trying to improve the institutional environment for service provision. In many 20 cases these efforts have been as much, or more, of a failure as efforts to induce macroeconomic policy reform. They have neither improved ser- 1o vices in the short run nor led to institutional changes in the long run. Aid can help improve institutions and policies at the sector level in Low High many ways: participation participation Source: Isham, Narayan, and Pritchett 1995. m Analytical work of donor agencies, such as the World Bank, has been shown to increase the probability of success of public projects. In one study an additional $10,000 in analytical work returned $80,000 in stronger benefits from projects. " In a sample of donor-financed rural water supply projects, the success rate was 68 percent for projects with high beneficiary participation, but only 12 percent for projects with low participation (figure 9). More important, donor efforts to promote beneficiary participation helped the adoption of this approach to service provision. " Comparing project success across countries shows the importance of 22 the institutional environment for public investments. In general, more OVERVIEW civil liberties (freedom of the press, freedom to assemble, and so on) lead to more project successes. Effective provision of community ser- vices requires that citizens' voices be heard. m About 40 percent of World Bank projects have had a substantial effect on institutional development-that is, they have helped change the Poor countries with good way the public sector does business. policies should get more aid than ones with The findings suggest that the key role of aid projects as a way of pro- mediocre policies-but in viding development assistance is not to move money but to support the fact they get less. creation of an effective public sector. Aid agencies can provide ideas about how to improve services and finance innovative approaches. Learning Figure 10 Actual Allocation of from these innovations generates knowledge about what works and what Aid, 1996, and Optimal Allocation does not. This view of development projects has important implications, to Reduce Poverty not only for the choice and evaluation of projects, but for the way aid Aid (percentage of GDP) agencies are designed and evaluated. In particular, the evaluation of pro- jects from the donors' point of view should focus on whether they had a 4 Optimal allocation to reduce poverty positive impact on the institutions and policies of the sector. Actual 3 allocation Money, but More Ideas, Too 2 D EVELOPING COUNTRIES ARE TO A LARGE EXTENT MASTERS OF their own fate. Domestic economic management matters more than foreign financial aid. Economies that lag are held back - more by policy gaps and institutions gaps than by a financing gap. Aid as money has a large impact only once countries have made substantial . progress with reform of policies and institutions. Poor countries with a Mediocre Good policy Poor policy Mdor odplc good policies should get more aid than ones with mediocre policies-but in fact they get less (figure 10). Source: Collier and Dollar 1998. Foreign aid has concentrated too much on the transfer of capital with (often) scant attention to the institutional and policy environment into which resources were flowing. This approach resulted from misunder- standings about development-overemphasizing finance at the expense of policies and institutions-and from external and internal pressures on aid institutions. Disbursements (of loans and grants) were easily calcu- lated and tended to become a critical output measure for development institutions. Agencies saw themselves as being primarily in the business of dishing out money, so it is not surprising that much went into poorly managed economies-with little result. 23 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY So, what can the international community usefully do in highly dis- torted environments? Obviously there is no simple answer to this ques- tion. Almost nothing positive has happened in Burma (Myanmar) or Nigeria in the past three decades, so it would be hubris to pretend that there is some magic solution to the problems of these and similar coun- tries. Nevertheless, there are examples of successful assistance that has improved the lives of people in such countries. The final chapter of the report presents four case studies of effective aid under difficult conditions: adjustment without adjustment lending in Vietnam; support to education decentralization in El Salvador, Pakistan, and Brazil; a health financing innovation in Cameroon; and the Box 5 Stakeholder Views on Aid Effectiveness di . -. d ".'lh a idc n i lh lder; r. rFI . Lt .r v12 ot [c rlk.J.i. de. b[,pim r nd dt je.. epd ·:_ rk_ l fithouzh [here ej:a t . I t P'i t% tTt.t inc-, ItåhlsJ r e -u ~tiIii itLciLiiL in .i-, c . [li- L run!: c.. . : .- u:. t * ,. t.di.'i -hih ihere s brod ar.nt un oh t.t Ti,r hinor,ukt a-n o co n eiprn,oe.»CuL.nr Jip£n,OECF J. i*. '.t'.'t~ Fvit _Yrv .. 7.; ä t . I pI. ,t, ljt r i '..,. J t t luc At Che Ph- ppinus 24 OVERVIEW Africa road maintenance initiative. Running through these cases of suc- cess in the difficult environments are four themes: a Finding a champion. Countries, governments, and communities are heterogeneous. While it is fair to characterize Burma overall as "poorly managed," there are likely to be reform-minded elements in the com- For reform to take root munity and even in the government. If aid can find and support these requires a demonstration reformers, it can have a big impact. that it actually works. n Having a long-term vision ofsystemic change. Successful reformers have a vision of how things could be different in 10 years-different both in outcomes (more kids going to school, graduating, getting jobs) and in process (community involvement in schools, broad public support for reform policies). a Supporting knowledge creation. While reformers typically have a long- term vision, they often need to develop the details of reform through innovation and evaluation. Furthermore, for reform to take root requires a demonstration that it actually works. Financing and evalu- ating innovations is a key role of development assistance. m Engaging civil society. In the highly distorted environments, the gov- ernment is failing to provide supportive policies and effective services. This is why government-to-government financial transfers produce poor results. Effective aid in this case often involves supporting civil society either to pressure the government to change or to take service provision directly into its own hands. The foregoing points concern characteristics of promising environ- ments for change and how donors can promote those characteristics. The lessons of successful assistance also provide guidance about donor behav- ior. Aid is more effective when: * Agencies focus on long-term reform. It is noteworthy that success in dif- ficult environments typically involves intensive staff input from donors and small disbursements of money. It also goes "beyond pro- jects" to support systemic reforms. In the difficult environments, effec- tive assistance is more about ideas than it is about money or projects. Money may be one input, as in the examples in Chapter 5. Projects can also be a useful input. But the focus of assistance in these cases was on supporting people with new ideas or looking for new ideas. m Donors work in partnership rather than in competition. Studies of aid have long pointed to the proliferation of donors and lack of coordi- 25 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY nation among them. Well-managed countries force coordination on donors, but in the weak environments they often run amok. It is hard to explain this behavior, except that different donors like to "plant their flags" on something tangible. In the cases of successful assis- In cases of successful tance, we tend to find strong partnership among donors with a focus assistance donors on larger transformations, not on individual projects and flags. focus on larger transformations, not As the focus of aid expands to money and ideas, management and eval- on individual projects uation become trickier. Agencies are trying to support policy reform and and flags. institutional development at both the macroeconomic and sectoral levels. The privileged units of account should be the country or sector program. Individual projects will be important in supporting sectoral development, but they should be evaluated primarily on their contribution, as building blocks and testing grounds, to the reform of sectoral institutions and poli- cies. "Evaluating the Bank's effectiveness can no longer focus only on the project, it must also measure the impact of the full range of bank activi- ties at the sector and country levels. This calls for new evaluation instru- ments and modalities . . ." (World Bank 1997a, p. 1). Donor agencies have learned from their success and failures. In the 1990s all of the major agencies have instituted reforms aimed at strength- ening the focus on results on the ground. Most agencies have also devel- oped country assistance strategies so that individual activities now must fit into a larger game plan for policy reform and institutional change. In the same vein, the focus of evaluation has risen above the level of the pro- ject to overall country program reviews. These essentially ask: Have agen- cies used their resources to stimulate institutional and policy changes that have led to improved services and better quality of life? This is not an easy question to answer. But it is the right one to ask. The Overseas Development Council recently sponsored reviews of the overall impact of aid in eight African countries, conducted jointly by scholars from the developing countries in question and from major donor countries- which seems to be a fruitful approach. Similarly, the shareholders of the IMF recently turned to a group of outside scholars to assess the Fund's support to low-income reformers. And the World Bank has initiated country assistance reviews with input from a wide range of stakeholders. Called for are independent reviews of development agencies with strong input from developing countries and focusing on two questions: Has the bulk of financing gone to sound institutional and policy envi- ronments? And have agencies contributed to policy reform and institu- 26 OVERVIEW tional change? Evaluating the right things should feed back into the man- agement and incentives within agencies. With better management and evaluation, development agencies should become: m More selective-putting more money into economies with sound There is reason to be management. optimistic that reform of a More knowledge-based-using resources to support new approaches to aid agencies will service delivery, expanding knowledge about what works, and dissemi- succeed. nating this information as a core business. m Better coordinated-results-oriented agencies should worry less about planting their flags on particular projects and more about how commu- nities, governments, and donors working together can improve services. m More self-critical-agencies should be asking themselves continually: Why do we do what we do? And what is the impact? With a better understanding of development and aid effectiveness and with the end of Cold War strategic pressures, there is reason to be opti- mistic that reform of aid agencies will succeed. Notes 1. Even a partial list would include books by those out- mote peaceful conflict resolution. Furthermore, humani- side aid agencies (Cassen 1986, updated in 1994; Riddell tarian aid, which accounts for less than 10 percent of devel- 1996; Mosley 1987; Krueger, Michalopoulos, and Ruttan opment assistance and focuses on short term mitigation 1989), various institutional evaluations, and large bibli- efforts, is a separate issue not addressed in this report. ographies on specialized topics, such as aid and macroeco- nomics (White 1992), policy-based lending (Killick 4. The oft-made suggestion that the early consensus 1991), aid and nongovernmental organizations (Riddell, was on physical investment and that the new consensus is Bebbington, and Peck 1995), and aid and post-conflict a shift to human investment is belied by the literature and reconstruction (World Bank 1998b). by reality. Gunnar Myrdal's classic Asian Drama, written in the 1950s, signaled the intellectual shift to roughly equal 2. Because of data availability, the focus of this study is emphasis on physical and human capital assistance from OECD countries to the developing world. It does not cover South-South aid (for example, from 5. It would be wrong to be overly critical of earlier Kuwait or Saudi Arabia) or the Soviet Union's extensive views or those who held them, as economics is an inexact foreign aid. science and reading the evidence is difficult. A 50-year-old development economist making policy in 1960 would 3. The report focuses on international efforts to sup- have witnessed in his or her lifetime: the Great Depression, port long-term growth and poverty reduction in the devel- the success of World War II planning in ending unem- oping world. There are other important types of ployment and raising wartime production, the spectacular international cooperation not covered by the study, such rise of the Soviet Union under central planning, and the as efforts to reduce cross-border narcotics traffic or to pro- catch-up of Japan to the European powers. 27 CHAPTER 1 Money Matters-In a Good Policy Environment Many of the measures ANY DEVELOPING COUNTRIES HAVE RECEIVED that promote long-term large amounts of foreign aid over long periods. growth also help reduce But has aid reduced poverty and infant mortal- poverty. ity or increased per capita income and private investment? Would more effective aid raise incomes and significantly lower poverty in the future? And can aid's effect on poverty be improved by reallocating it from some countries to others? All aid is ultimately aimed at promoting growth and reducing poverty, whatever the immediate objectives. In an ideal world, $1 million of aid to one country would have the same marginal effect on growth and poverty reduction as channeling the same amount to another country. In the real world, this does not happen. Some countries benefit more than others. Why? Aid effectiveness largely depends on the institutions and policies of recipient countries. Where economic management is sound, aid leads to higher private investment, more rapid growth, lower infant mortality, and a faster decline in poverty. Where economic management is poor, aid has little effect on development. The next section of this chapter briefly reviews the divergent growth experiences of developing countries. Growth helps reduce poverty and improve social indicators, which is why the relationship between aid and growth is important. Then some stylized facts about growth, institutions, and policies are examined, and the question asked: why should aid affect growth? After that, we look at the relationship between aid and growth, aid and poverty reduction, and aid and private investment. Then another question: how has aid-both bilateral and multilateral-been allocated? Finally, recommendations are made on how aid can be made more effective. 28 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT Different Countries, Different Fortunes IN 1966 ETHIOPIA, INDIA, AND THAILAND WERE ALL POOR COUNTRIES. In each society a majority of people were living on less than $1 a day (in 1985 prices). Compared with many other developing countries, "Rates of growth of real per capita India did well, doubling real per capita income between 1966 and 1990, GNP are diverse, even over sustained and reducing poverty to 53 percent of the population. But over the same periods. For 1960-80 we observe, for period Thailand did very well, tripling per capita income and cutting example: India, 1.4percentperyear; poverty to just 2 percent. This difference is what Robert Lucas means by Egypt, 3.4 percent; South Korea, 70 the "staggering consequences for human welfare" of different rates of percent....Between the 60sand70s, growth and development. The contrast between Thailand and Ethiopia Indonesia 's growth increased from is even starker. In 1966 Thailand had about four times Ethiopia's per 3.9 to 75. capita income. By 1990 Ethiopians had seen no income growth, and "I do not see how one can look at Thailand was more than 10 times wealthier. figures like these without seeing them Growth matters not for its own sake, but because it raises living stan- as representing possibilities. Is there dards. Of every 1,000 babies born in 1967 in Thailand, 84 did not sur- some action a government of India vive the first year of life. By 1994 that figure had been cut by nearly could take that would lead the Indian two-thirds. India and Ethiopia also made progress, but not as much. In economytogrowlikeIndonesias?Ifso, India infant mortality fell by a half over this period. In Ethiopia it fell by what, exactly? The consequences for only 27 percent. Other social indicators tell a similar story. By 1990 human welfare involved in questions Thailand had achieved universal primary education for girls, while the like these are simply staggering: Once figure for India was 91 percent and for Ethiopia, 19 percent. one starts to think about them, it is Growth does not eliminate poverty and deprivation, but per capita hard to think about anything else. " incomes and social indicators tend to improve (or deteriorate) together: Robert E. Lucas, Jr. life expectancy, school enrollment, infant mortality, and child malnutri- "On the Mechanics of tion are all closely related to per capita income. Although there have been Economic Development" great improvements in living standards globally, the huge differences p. 4-5 across countries show the importance of development. Take infant mor- tality: in nearly every developing country it fell between 1967 and 1994. The extent of the decline, however, varied widely (figure 1.1). Economically stagnant countries such as Zaire (now Democratic Republic of Congo), Ethiopia, Niger, and Zambia saw only modest falls, while such faster-growing countries as Botswana, Chile, Lesotho, Mauritius, and Tunisia saw improvements of 70 percent or more. Rising incomes of the poor expand their capacity to improve their health, education, and living standards. Moreover, many of the measures that promote long-term growth-basic education or an open trade regime, for example-also help reduce poverty. But economic growth does not solve all ills: environmental degradation, crime and violence, 29 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY From the same starting Figure 1.1 Infant Mortality in Selected Countries, 1967 and 1994 points in 1967, infant Infant mortality mortality fell to 33 per (per 1,000 live births) thousand in Botswana in 250 1994 while remaining *1967 S1994 virtually unchanged at Mali 110 in Zambia. 200 Niger Ethiopia * iTurkey 150 Bangladesh India Lesotho Tunisia Congo Peru i Brazil cuador loo0 Zmia TanjChi le Zamb Thailand Tanzania { so Kenya Inoeiu IdeiHondurasc Botswana Mauritius 0 Source: World Development Indicators 1998. sex discrimination, social exclusion, and the like may or may not improve with income. When people improve their status from low to middle income, new challenges emerge for society. Consider, for exam- ple, the choking haze-from fires set to clear plantations-that enveloped Southeast Asia in the summer of 1997. Or the many devel- oping country cities that have severe problems with traffic congestion and air pollution. Still for low-income countries, generating self-sustaining growth is a crucial first step in development. If Ethiopia's total output were divided evenly among households, everyone would be poor. Growth (an increase in per capita income) is clearly required to solve poverty and its problems: malnutrition, poor health, and lack of education and basic social services. Overall, developing countries have made tremendous progress: more babies are surviving, more food is on the table, more children are in school, and there are fewer deaths from easily preventable diseases. There has been more global improvement in life expectancy in the past 40 years than in the previous 4,000 (World Bank 1993). The revolu- 30 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT tionary expansion of education can be seen across generations: 62 per- cent of Chinese in their sixties received no schooling as children, com- pared with just 5 percent of young adults. In Indonesia 80 percent of 60-year-olds had no schooling, compared with just 15 percent of young adults. But in all dimensions, progress has been uneven geographically. The greatest improvement has been in fast-growing economies. Since World War II, developing countries have fallen into three categories: fast grow- ers, slow and stalled growers, and nongrowers and reversals (Pritchett 1998). A small but significant group-Botswana, Indonesia, the Republic of Korea, and Thailand-has grown much faster than indus- trial countries (table 1.1). The gap between these countries and the United States (the leader in income) has narrowed impressively. Countries can, and do, move in (and out) of the group of fast grow- ers. Many were sure that Korea was a basket case before growth took off in the early 1960s. Indonesia in 1966 was plagued by slow growth, high Table 1.1 Real Per Capita Income inflation, and social unrest. Before the boom in 1974, Mauritius's GDP Relative to the United States per capita was lower than in 1960. Other countries have joined the high- * ,.f growth set more recently: China since the late 1970s, and Chile since .S.ole 1984. Countries also fall out of the fast-grower set. Brazil's growth was 1v66 9!0 4.2 percent from 1965 to 1980 but then -0.2 percent from 1980 to 1992. And Indonesia's growth will certainly slow as a result of its current crisis. L an - I Slow growers, with growth rates similar to those of industrial coun- I ri d n tries, include Bangladesh, India, Malawi, and Pakistan. They have not n-I.c_; been falling behind, but they are not catching up either. In 1990 per capita income in the United States was 18 times India's, about the same as in 1966. But since per capita income roughly doubled for each coun- ndo .n try, the absolute gap is now much wider. Slow-growing countries have seen some improvement in poverty and social indicators, but not nearly " as much as the fast growers. For a significant number of people in the world, there has been no per . 2 capita growth in recent decades, or even (especially recently) substantial I decline. Many of these people live in Africa: for instance, C6te d'Ivoire, i a Ethiopia, and Zambia. Non-Africans include Haitians, Nicaraguans, and Iranians. In relative terms, these countries are falling further behind the 1 , rest of the world. In 1966 per capita income in the United States was 20 times Zaire's; now it is 50 times. 31 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Why the Divergence? H OW TO EXPLAIN THESE WIDELY DIFFERING GROWTH PERFOR- mances? It was once thought that the key factor holding back There is plenty of poor countries was a lack of savings and foreign exchange for evidence that good investment. Part of the initial rationale for foreign aid was to help coun- macroeconomic tries overcome a "savings gap" to finance necessary investment and a management provides a "foreign exchange gap" so that imported machinery could be the corner- fertile environment for stone of that investment. Development agencies worked with a "two gap" growth. model that made imports and investment in physical capital the driving force of growth. The role of aid in promoting growth was clear, since aid can help fill both gaps. The slow stagnation and sudden collapse of the Soviet system of cen- tral planning, which was the intellectual father of "gap thinking" and development planning, made it clear that investment alone cannot guar- antee growth. In the 1990s the focus of theoretical and empirical work on growth has gone deeper. Emphasis has shifted from investment to incen- tives. That is, from capital to the underlying institutions and policies that promote growth by encouraging efficient investment, by supporting human capital development, and by facilitating technological advance. What are the conclusions of this new growth literature? A stable macroeconomic climate is crucial. High inflation is bad for investment and growth (Fischer 1993). Similarly, large fiscal deficits hold back growth (Easterly and Rebelo 1993). An outward orientation and a reasonable environment for international engagement are essential: most trade liberalizations accelerate growth (Sachs and Warner 1995).1 Fiscal, monetary, and trade policies show whether a country is well managed at the macroeconomic level, and there is plenty of evidence that good macroeconomic management provides a fertile environment for growth. Good institutions and economic management are also needed at the microeconomic level. The strength of private property rights and the rule of law and the quality of the civil service affect long-term growth (Knack and Keefer 1995). Similarly, corruption in the public bureaucracy is bad for growth (Mauro 1995). For their studies of aid and growth, Burnside and Dollar (1997, 1998) compiled a dataset covering 56 aid-receiving developing countries and averaging growth, aid, and other variables over four-year periods (start- ing with 1970-73 and ending with 1990-93). They created an economic management index based on the Sachs-Warner measure of openness, the 32 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT Figure 1.2 Economic Management and Growth in Selected Developing Countries with better Countries policies can grow faster, and reform can get them Annual percentage growth in GDP per capita there-as in Ghana and 8 Botswana D Bolivia. 1986-89 6 El Salvador Indonesia 4 1990-93 1982-85 Thailand Bolivia 1974-77 2 1990-93 OGhana Ghana India 1986-89 0 1986-89 1990-93 Tanzania -2 1982-85 Bolivia _ Zaire Zambia 1982-85 Peru 1986-89 1986-89 1986-89 -e Nicaragua E11986-89 -8 Poor Mediocre Good Economic management index Source: Burnside and Dollar 1998. budget surplus, the inflation rate, and a measure of institutional quality. Obviously there is more to economic management, but these serve as a useful, and measurable, proxy.2 The relationship between per capita growth and the economic management index reconfirms that sound management-at both the macroeconomic level and institutional lev- els-is important for growth (figure 1.2). This finding suggests that poor countries have been held back not by a financing gap, but by an "institutions gap" and a "policy gap." If they can overcome them, they can begin to grow successfully. Botswana and Thailand are examples of good management-Tanzania and Zambia of poor. The difference in management between, say, Thailand and Tanzania may have been worth about 4 percentage points of growth. (Note that while Indonesia had solid macroeconomic policy, it scored badly on the measure of institutional quality.) Countries can reposition themselves over time. Bolivia and Ghana had poor management in the early 1980s, but became good-management countries in the 1990s. The macroeconomic policies included in this index can be changed quickly if society and government want to reform. Other institutional aspects of good management, such as the rule of law, take longer to improve (North 1990). 33 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Growth rates are also much more volatile in developing countries than in industrial countries (Easterly and others 1993). Developing countries may grow at 5 percent per capita in one four-year period, and then at -3 percent in the next. Empirical studies of growth can explain about half Poor countries have the of the variation in growth rates among developing countries. potential for rapid growth. In Countries with Good Management ... Figure 1.3 Growth Rate and Income Level T HERE ARE POTENTIAL ADVANTAGES TO BACKWARDNESS, BUT in GDP per capita these can be realized only with good policies (Sachs and Warner 3.1 1995). The Burnside-Dollar sample can be divided into good Good and poor management groups, to show the relationship between income 3.0 management group at the beginning of a period and subsequent growth (figure 1.3). Among 2.5 the good-management group, middle-income countries grow faster than 2.0 high-income ones. The superstars-the East Asian tigers (before the 1997 crisis) or Chile-are found in the good-management, middle- 1.s income group, which grew much faster than high-income countries. By 1.0 Poor and large, good-management middle-income countries attract a lot of management foreign direct investment, learn technologies from more advanced 0.5 group economies, and have investment with a high return to capital. They are a catching up with the industrialized world, but as they move closer -0.5 growth tends to slow, as there are fewer opportunities to learn, and the return to capital diminishes. In contrast, with poor management there -1.0 Low Lower-middle Middle is no advantage to backwardness: such countries grow slowly at all income income income income levels. Countries that cut themselves off from globalization Source: Burnside and Dollar l998. (through restrictive trade practices and unstable macroeconomic poli- cies) are the biggest losers. When low-income countries, such as Mali, put good policies in place, they perform better, but they still do not match the results for the mid- die-income countries. There are several reasons for this. These countries may have other characteristics that hurt growth-being landlocked, for instance. But it is also possible that their ability to save and invest is ham- pered by poverty itself, even when good policies are in place. If interna- tional capital markets were perfect, private money would perhaps flow to seriously reforming countries. But imperfections mean that private investors cannot accurately assess the situation and may hold back, under- mining what would otherwise be a successful reform program. 34 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT This suggests a role for foreign aid in sparking developing country growth. Financial aid to poor countries that have good policies should have a high return. It can help them accelerate to rapid growth. On the flip side, there is no reason to suppose that pouring aid into poor-man- agement countries will accelerate growth. Development assistance can also encourage policy reforms (how it can do so is the subject of the next chapter). Serious reform can add 2-3 per- centage points to growth. This would make an enormous difference in developing countries that have had little or no growth in per capita income over the past 25 years. ... Aid Spurs Growth The simple relationship between aid and per capita growth in devel- oping countries is weak, if it exists at all (figure 1.4).3 Some countries get a great deal of assistance and grow slowly (Zambia, for instance) while others also get a lot and grow quickly (Botswana, Ghana). From this simple starting point, there are two problems in analyzing the effect of aid on growth. First, other factors that affect growth must be Figure 1.4 Aid and Growth in Selected Developing Countries, 1970-93 Some countries received Annual percentage growth in GDP per capita a lot of assistance, and 1.o incomes fell-some countries received little 8 0 9aid, and incomes rose. 6 P 4 2 o a -6 0 3 6 9 12 15 Aid as a percentage of GDP Source: Burnside and Dollar 1998. 35 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY considered. Second, in some cases aid may be deliberately given to coun- tries that are growing poorly. Consider a country hit by a typhoon that destroys the rice crop-thus reducing growth. This calamity may induce a temporary increase in aid. Thus, while the simple relationship might With good management appear to be a negative association between aid and growth, it would be an additional 1 percent a mistake to interpret this as evidence that aid reduces growth. of GDP in aid increases A recent study addressed both of these problems and still found no growth by 0.5 relationship between aid and growth (Boone 1994). Even when a range percentage point. of institutional, political, and policy variables are added to the equation the result is still the same-no relationship. Figure 1.5 Marginal Impact on What changes the picture completely, however, is separating good- Growth of a Percent of GDP in Aid management from poor-management countries. The proposition that aid has no effect where incentives are weak was tested by Burnside and Percentage points of economic growth Dollar (1997). For countries with poor management it is true: whatever 0.5 the amount of aid, growth was minuscule, or even negative. That some 0.4 countries have received large amounts of aid for decades yet shown no growth has given aid a bad reputation. 0.3 Taking the good-management group and dividing them into high-aid 0.2 and low-aid groups produces striking results. The good-management, low-aid group grew at 2.2 percent per capita, but the good-management, 0.1 high-aid group grew almost twice as fast-at 3.7 percent per capita. 0.0 Including other variables (institutional, political, and policy) shows that -0.1 aid has a large, positive effect on growth in good-management countries .. (appendix 1). Removing middle-income countries, which receive little -0.2 aid, from the sample makes the effect of aid even stronger. With good -0.3 management, an additional 1 percent of GDP in aid increases growth by -0.4 0.5 percentage points-a rate of return of roughly 40 percent, if depre- Good Average Poor clation is 10 percent or so a year (figure 1.5). Countries with sound man- Economic management agement that have received a lot of aid and done well include Bolivia, El Source: Appendix 1. Salvador, Ghana, Honduras, and Mali in the 1990s. There are diminishing returns to aid so that-even with good man- agement-the benefit will decline as aid is increased. However, given the amount of assistance that well-managed countries actually receive, there would still be a strong positive return to increasing aid to these countries. In Ghana in the 1990-93 period, for example, a doubling of aid would have been necessary to reach the point at which the marginal return to further finance would have been zero. Moreover, if a country is receiving even the average amount of aid (about 2 percent of real purchasing power parity GDP), a 1 point increase 36 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT in its management index leads to an increase in growth of 1.3 percentage points. With twice that amount of aid the effect of good management is even stronger-growth is 1.9 percentage points higher. Other researchers have reached similar conclusions. In a recent study comparing aid effectiveness in Bolivia, Costa Rica, and Nicaragua, Policies have a critical Cecilia Lopez, Colombia's Minister of Planning, found that aid did not influence on the have much effect on growth in Nicaragua because of the country's highly effectiveness of aid, for distorted policies-a large fiscal deficit and high inflation. Aid was largely the same reasons that effective in Bolivia, which successfully completed a reform program in they affect economic the 1980s. Lopez's conclusion from the three case studies: growth. n Foreign aid in itself is neutral with respect to development, for its pos- itive or negative effects depend on government policies. Effects on eco- nomic development will tend to be positive when aid is used to build up capital or to finance public investment that contributes to the prof- itability of private capital, or for human capital development. m Fiscal policy should generate government's current savings, so that both domestic and foreign resources finance public investment. If this does not happen, foreign resources may end up financing the govern- ment's current expenses and not investment projects, as happened in Nicaragua. m The relationship between aid's positive effects and good domestic poli- cies always holds, even during adjustment (Lopez 1997). A study of seven African countries by researchers from donor and recipient countries (under the auspices of the Overseas Development Council) pointed to good economic policies as a key precondition for successful assistance: "Clearly, the large influx of aid has not fostered rapid economic growth in most African countries and has been only partly success- ful in poverty alleviation. How can one reconcile this lack of suc- cess with the field-level evidence just cited and reach a clear verdict on aid's economic impact? Unfortunately, a number of factors pre- vent the establishment of a clear link between aid and the overall performance of the economy. First, government economic policies themselves can prevent aid from having the impact it otherwise would have. Policies have a critical influence on the effectiveness of aid, for the same reasons that they affect economic growth; 37 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY economic policies that engender balance of payments difficulties, high budget deficits, or high rates of inflation are likely to foster a climate of economic uncertainty, which dampens the private sec- tor's response to the public investment represented by aid. A clas- In declining economies sic example of such a phenomenon is an aid-funded agricultural an average drop of 7 extension project that does not achieve its objectives of output or percent in per capita productivity growth, because macroeconomic policies have resulted income led to an in an overvalued exchange rate, which makes the farmer's output increase in poverty of 19 noncompetitive. Similarly, policies that create extensive market dis- percent. tortions are likely to undermine the quality of scarce resource use, resulting in socially suboptimal outcomes. Taking up the example Figure 1.6 Economic Growth and of the agricultural extension project, farmer response is muted Poverty because the government's agricultural policies provide farmers with Percent disincentives to adopt yield-enhancing new technologies. These are 20 not fortuitous examples: by the mid-1980s, most experts were con- vinced that donor and government attempts to expand and mod- ernize agriculture in many African states had been largely undone by macro and sectoral policies that were exacerbating the impact of 10 Aer2ge the decline in world prices for agricultural commodities. They cred- per c.ipita Ipie omdte.Te inoe-ua ited the prevailing policy environment with the superior perfor- 5 growth mance of the agricultural sector in Kenya, for example, and the poor performance in Tanzania and Zambia" (van de Walle and Johnston a 1996, p. 36-37). -s Aeiage ... Aid Reduces Poverty change -o in poverty The main aim of aid is to reduce poverty. Poverty reduction in devel- Expanding Declining cutisireae to apagowh economies economies oping countries is closely related to per capita income growth. Development economists used to worry that the benefits of growth would Source: Ravallion and Chen 1997. be undone by increases in income inequality as poor countries developed. Recent evidence has shown conclusively, however, that this is not so. One study examined trends in the distribution of income for 45 coun- tries for which high-quality household income or expenditure data are available over time (Li, Squire, and Zou 1998). In 29 countries there has been no trend in either direction. The remaining 16 countries are divided: 8 have shown rising inequality, 8 declining. If income distribution does not change much over time, the gains in per capita income will affect dif- ferent segments of society to about the same degree. Thus, in countries with rapid growth, incomes of the poorest will rise rapidly and the inci- 38 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT dence of poverty will decline. But in countries with no per capita growth and a stable income distribution, there will be no poverty reduction. Another study examined recent per capita growth and poverty reduc- tion in 67 countries for which household data were available. It found that every country with increasing per capita household income saw With good management poverty decline, and every country with declining per capita income saw an extra 1 percent of poverty increase. In the expanding economies, per capita income grew 4 GDP in aid leads to a percent, and poverty declined 5 percent. In declining economies an aver- decline in infant age drop of 7 percent in per capita income led to an increase in poverty mortality of 0.9 percent. of 19 percent (figure 1.6). Poverty data are not available for all countries and time periods in the Figure 1.7 Decline in Infant Burnside and Dollar study, so direct examination ofthe effect ofaid is impos- Mortality from a Percent of sible. But according to Bruno, Ravallion, and Squire (1998), on average a 1 GDP in Aid percent increase in per capita income in developing countries reduces Decline in infant poverty by 2 percent. Put another way, in countries with sound management mortality (percent) an extra 1 percent ofreal GDP in aid results in a0.5 percentage point increase 1.0 in growth and hence a 1 percent decline in poverty. In countries with poor management the expected effect of aid on poverty reduction is far less. Aid can affect well-being in many ways. Consider changes in infant 0.8 mortality, an important social indicator for which data are widely avail- able. Burnside and Dollar found that aid helps reduce infant mortality if 0.6 a country has good management-an extra 1 percent of GDP in aid leads to a decline in infant mortality of 0.9 percent (figure 1.7). In contrast, if 0.4 a country has poor management, there is no marginal impact from another 1 percent of GDP in aid. 0.21 ... Aid "Crowds in" Private Investment 0 The impact of aid on growth in good-management countries is high. Good Average Poor This may be because once good management is in place, there are high- Economic management productivity public investments to be financed by aid. If an economy is Source: Burnside and Dollar 1998. growing rapidly, the return on investment in road rehabilitation, for example, is high. If the economy is stagnant, the return is lower or nonex- istent. Similarly, economic growth makes it easier to get children into school and have a high return from this investment in human capital. In a sluggish economy the incentives to send children to school are weak. Thus, with good management in place, aid helps government make more of these high-return expenditures. Evidence at the microeconomic level supports this finding. Public investments financed by the World Bank, 39 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Figure 1.8 Marginal Impact on for instance, have more success in a good institutional and policy envi- Private Investment of a Percent ronment: 86 percent, compared with only 48 percent in a weak institu- of GDP in Aid tional and policy environment (see overview figure 5). There is another possible explanation for the relationship between aid Private investment as a percentage of GDP and growth-that when a poor country puts good policies in place, domes- 2.0 tic and foreign private investors are uncertain about the sustainability of the reform. If there is fear of reversal, investors will wait. And even with good 1.5 management, low-income countries are hamstrung by other impediments, including weak infrastructure. Foreign aid to reforming governments may 1.0 improve the environment for private investment-both by creating confi- dence in the reform program and helping to ease infrastructure bottlenecks. 0.5 In a poor-management country foreign aid may crowd out private investment: it may encourage the public sector to make commercial 0 investments that would otherwise be undertaken by private investors. In a good-management country the effects of foreign aid would likely be -0.5 od magnified by crowding in private investment. Good Average Poor There is evidence to support these notions (Dollar and Easterly 1998). Economic management One percent of GDP in assistance increases private investment an extra Source: Dollar and Easterly 1998. 1.9 percent of GDP in good-management countries (figure 1.8). Thus the combination of good management and foreign aid is welcomed by the pri- vate sector, and this helps to explain the strong effect aid has on growth in such an environment. In a poor-management country, however, 1 percent and World Bank Aid Per Capita of GDP in aid is estimated to reduce private investment by 0.5 percent of and Income Level GDP, which may explain why the net effect of such aid is small. Aid per capita These findings for the effect of aid on growth and private investment (log scale in dollars) shed some light on the problems of the Heavily Indebted Poor Countries and how they can be helped (box 1.1). $7 Multilateral Aid Has Not Favored Countries with Good Management $1 World Bank ARGETING ASSISTANCE TO POOR COUNTRIES WITH SOUND institutions and policies will make the most of scarce aid $0.14- T resources to encourage investment, spur growth, and reduce 15CI 400 1,100 3,000 Real GDP per capita poverty. But between 1970 and 1993 aid allocations-by bilateral and (log scale) multilateral donors-were dominated by politics-both the interna- Source: Calculated from Burnside-Dollar dataset. tional politics of the Cold War and the internal politics of aid agencies. 40 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT Foreign aid is targeted to poor countries-but imperfectly. Bilateral aid per capita received by developing countries has a weak, negative rela- tionship with their per capita income (figure 1.9). Generally, a doubling of per capita income leads to a 33 percent reduction in aid. But the behav- ior of bilateral donors varies widely (Ehrenpreis 1997). Half of all Swedish aid has gone to the poorest 12 percent of countries (weighted by popu- lation); of all bilateral aid, only 20 percent went to these countries. Another study also found that Canadian, Dutch, and Nordic assistance is sharply targeted to poor countries (Alesina and Dollar 1998). The rela- tionship between aid per capita and income is stronger for multilateral aid than for bilateral and stronger still for the World Bank's International Development Association (IDA) facility, which is part of multilateral aid. For IDA, a doubling of per capita income has been associated with a 90 percent reduction in aid per capita. Another important factor in aid allocation is population. Countries that have small populations get more assistance per capita or more rela- Box 1.1 Aid and Heavily Indebted Poor Countries .- . i - - )*lI ai' . *il .I L -* - *,i rh, -- h-Acc conri -i.. :E i. . vp ii i J.d .i.. I. L. 11ch .ii, rr.l th11 bC J L i c Li frl i -r cr.,:c ri c Ce i n ri *. L., c.E 'runi in aid ill -caiiric 'n' er- t 1nr L. u [cIL iLJ4 ijtiicAfl[! 'c h.'ni.'..i 1i 'LL ..rid B±d',. L1W ib& . and o.rhcr jon.-.r% 1!1-, It c ". ni I' proi . [ I P1.11 r i, [ 11 . r* tr- 1.11 n . 6 Ii 1 1[ Uc 10011, [11% . ' c I Ir al-iIL( . - - hi -LirLc I [ ncif-.r - flifti.c i ltl L frc I did dciir.i hiLh hIl 'LAld'c!. hc' Iu iiri - mvr rfl ; iL [ ur, .- l cc - ' M I ILchI 11rt .LIF-c: A id h.% c. ii rh I., K r,. i nip.l Ir i.jli *Lm I c.. . Aii idd i.l nil i cl [L,L1.i111rL hi p.***cla 1 . IrIIIi1 LhC p ,1 i. [ Tli[T. r.. Iib ['Ih uno p iLblC l c. 1' n 1d L i -l r IkM L-0 - iri-_.nic Li''TEi , ILL il t , -rall. II IiC 1, uiliceffi inn icII[ I un mn- hii.d'.Irll . . p crc1ulpill *=ri.\ [1 [l h1 I3 dL- r Ii -1 I * .'cui I ci'rlihe [1'c Llm.inc [ile clif. vrv Ibc r rti i Lip Iar-L tri.. olit, . o r m.i j-'-rncj pi Likc-. [h.t couicrn lucd .. - l c r dLI.f Ykc,c [ 11. In i', i- .r- 100 put r** r[. * i-c a11iLld. Ic, Inr rki Cir,r pIcL r Idtli a t-lric I iid LL tIIl. d I r! .[ I !'TI i ilk Ci *t.,i I rtirr i aid .ill ha- i rc-eiCr in[ ,t ir, i ;. .J-Il in" i,-cflcm n[ i rh, I )'.) Iii .i IlId 1EtCd [""IT Lh.uilllfl c p tA. IliiECti t I; . *i i [o ' LiLLL Il.iiiL ifnn c,Ll [ 1cc1i i c.r r.d,!iii -n,11 IrI I f IIF 'L . %cL bLIICIIii t .r ilt iic:- r[ ii L At Er ciAu LiIii i - dic,: LIbr .[ i tL , I. nud ' he h.i.ld IIL ri)l I-a ll, in ebrcd ,ibr iiUir11 rt iu]ilic i i ri1k hic k- ..IL 0 P.. c i il IaICm 1C . prpiil ii*I. rzIr- rta'rd .4 p-K rc i II i .1[licrL- i hc-Il L L - .rd l c.p c It LC L ,.c ICL [IL' [raim l [ fdi tin i n 1 fli - rc IIi--1.L.fii3 ,Ii.. ilk. 1 JE..c deif . .cil111 ftt'Lt 1 iiiLLri ilr k, [* " hi- Ui1.ith I[ Id-iricl( p. 'Ic 41 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Figure 1.10 Aid and Population Aid as a percentage of GDP (log scale) 55 0 7 0 00 00 01 10 00 500 o~00~0 0~ 4%0 0* 000 0 0.40 00~0 00 e t* 000 0 0 0 0* 0~ 0 00 0 1 10 100 500 Population (millions, log scale) Source: Calculated from Burnside-Dollar claraser. tive to GDP (figure 1.10). This makes sense for some assistance. For instance, it costs about the same to provide technical assistance to the cen- tral bank of India as to the central bank of Lao PDR. But if aid is financ- ing public services, such as education or road construction, there perhaps. should not be such discrimination against large countries. More than half the world's poor live in India and China: in 1990-93 they received $2 and $1 per capita in aid. Yet small countries often receive $50 per capita or more. This discrimination against large countries is one reason the rela- tionship between aid allocation and income is not strong. A second factor that undermines this relationship is that aid allocation often depends on the political or strategic interests of donors-for exam- ple, U.S. aid in the Middle East or European aid to former colonies. Burnside and Dollar (1997) found that political factors help explain the allocation of bilateral aid, but that multilateral aid is not strongly influ- enced. These factors are not important for only a handful of bilateral donors-Canada, Denmark, Finland, Netherlands, Norway, and Sweden (Alesina and Dollar 1998). The dual objectives-pursuing strategic goals and rewarding good policies-work against each other. This can be seen by looking at the cor- 42 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT relation between bilateral aid and management (figure 1. 11). So, how have bilateral donors treated two countries with the same income and population but different management? Good and bad management regimes have received roughly the same average assistance. Being a for- mer colony of a major donor is more valuable in attracting bilateral assis- tance than having good management (Alesina and Dollar 1998). The allocation of multilateral aid has depended on income, popula- tion, and good management. Political and strategic considerations were not significant. Thus, for multilateral aid, lower-middle income coun- tries with good management received 30 percent more than the typical poor-management country with the same income and population (figure 1.12). For very low-income countries the difference between good man- agement and poor management has been minor. Since most official development assistance is bilateral, the alloca- tion of all aid together (multilateral and bilateral) shows little rela- tionship with the quality of country management. Cold War aid driven by strategic considerations may have accomplished its political goals, but aid that went to countries with poor management did little to reduce poverty. Figure 1.11 Allocation of Bilateral Aid, 1970-93 Aid as a percentage of GDP 21 ~ ~ 0_ Low income Go income Imanagement Lower-middle Poor income management Real GDP per capita Source: Calculated from the Burnside-Dollar dataset. 43 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Be Focused C LEARLY, DEVELOPMENT ASSISTANCE NEEDS TO BE MORE CONCEN- trated on where it can be most effective in reducing poverty. It At the moment plenty of needs to take more account of the environment in which it is countries combine mass placed. Finance is most effective in reducing poverty in those countries poverty with good that have both mass poverty and a good policy and institutional envi- policies and institutions. ronment. Conversely, it is less effective in countries that already have rel- atively little poverty, or have poor policies and weak institutions. But is there much scope for a more focused approach? Perhaps mass poverty is so strongly linked with poor policies and weak institutions that we just don't find the combination of mass poverty and good policies in which aid is highly effective? Not so. At the moment plenty of countries combine mass poverty with good policies and institutions. Why? Because of the wave of policy reform and institution building that has swept through poor countries during the 1990s. Figure 1.13 classifies 113 developing countries based on their inci- dence of poverty and the quality of their policies in 1996. What creates a great opportunity for foreign aid is the upper-right- hand quadrant. Thirty-two countries have poverty rates above 50 per- Figure 1.12 Allocation of Multilateral Aid, 1970-93 Aid as a percentage of GDP 3 2 - . management Poor Low management Real GDP per capita Source: Calculated from Burnside-Dollar data set. 44 MONEY MATTERS-IN A GOOD POLICY ENVIRONMENT cent, and also have above average policies. In such conditions aid is highly effective in reducing poverty. Some of the countries in this quadrant are among the poorest in the world: for example, Ethiopia, Uganda, Mali, and India. This is the High Impact quadrant for aid. Conversely, in the bottom-left-hand quadrant aid is radically less effec- Why? Because of the rive. Sixteen countries already have a low incidence of poverty, as well as wave of policy reform below average policies. There are fewer poor people to help, and aid does and institution building less for them because it is handicapped by the poor policies, that has swept through In the remaining two quadrants aid is less effective than in the High poor countries during Impact quadrant. the 1990s. In the upper-left-hand quadrant there are 32 countries with poverty rates above 50 percent. There are millions of poor people in these economies in need of help, but unfortunately the policy envi- ronments are not good enough for aid to have much effect. The pri- ority for the world community in these countries is to help in the domestic political and social process of policy change: that is, in con- tributing knowledge rather than big finance. Of course, some finan- cial flows provide opportunities for dialogue and knowledge transfers. But aid to these economies has to be justified more for its indirect contribution to policy change than for its direct effect on poverty reduction. Figure 1.13 Poverty and Policy, 113 Developing Countries, 1996 Poor policy Good policy High poverty High poverty II I Poor policy Good policy Low poverty Low poverty III IV Source: Collier and Dollar 1998. 45 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY In the lower-right-hand quadrant there are 31 countries that have above-average policies. Aid works in these environments, but there is less work for it to do, since there is already a low incidence of poverty. So, there is certainly scope for a focused approach to aid: financial assis- An across-the-board tance to High Impact countries is far more effective than in countries in $10 billion increase in other quadrants. But is there scope for aid to be more focused than it is at aid would lift 7 million present? Surely, donors are already taking into account poverty and poli- people out of poverty, cies in determining where to allocate finance-aren't they? while a targeted increase Actually, not very much. To see this, suppose that the world commu- could lift 25 million out nity raised an extra $10 billion and allocated it proportionately to exist- of poverty. ing aid allocations. So allocated, this extra aid could raise 7 million people out of poverty. Now suppose instead that the $10 billion were allocated to the High Impact quadrant. The extra impact would be dramatic: 25 million people would be raised out of poverty (Collier and Dollar 1998). Being more focused can thus spectacularly increase the effectiveness of aid in reducing poverty. Nearly four times as many people could be lifted out of poverty for a given amount of aid. And of course, if aid became much more effective, there would be more of it. Rapid poverty reduction is the major global challenge: demonstrated effectiveness can be expected to produce greater support for aid. Notes 1. All three of these studies consider whether good poli- 1991). In low-income countries, this measure of GDP is cies bring about growth or growth spurs good policies- often three times higher than GDP at domestic prices and conclude that policy reform leads to more growth. because prices of nontraded goods tend to be low in poor countries. In the developing world the average aid receipt 2. Collier and Dollar 1998 show that the finding that relative to PPP GDP is about 2 percent. The same amount aid has more impact in a good management environment would be 6-7 percent of GDP in domestic prices. holds in the 1990s using a much broader measure of insti- tutions and policies, including issues such as corruption, 4. The measure of policy used to construct this figure is governance, equity, and safety nets. For research purposes, the Country Policy and Institutional Assessment of the this broader measure has the disadvantage that it is not World Bank. It has more components than the index in the available back through time. Burnside-Dollar study, including such areas as social sector policies and safety nets. It has the advantage of covering a 3. For the growth analysis, aid is measured relative to large number of countries and being up-to-date. But it is GDP at purchasing power parity (Summers and Heston not available as a consistent measure going back in time. 46 CHAPTER 2 Aid Can Be the Midwife of Good Policies OUND MANAGEMENT THAT PRODUCES MACROECONOMIC Between the mid-1980s stability, openness, rule of law, and absence of corruption and mid-1990s there leads to growth and poverty reduction. It also creates the were huge advances in right environment for aid to reduce poverty. While a $10 bil- the quality of policies in lion increase in aid could lift 25 million people out of poverty developing countries. each year, an even bigger assault on global poverty will require further institution building and policy reform. Between the mid- 1980s and mid-1990s there were huge advances in the quality of man- agement in developing countries. A further improvement of the same magnitude could add a full percentage point to developing country growth and lift another 60 million people a year out of the poverty abyss. Policy reform and capacity building are the keys. Ideas and money- together-have the potential to do far more than finance alone. In this chapter we review the relationship between aid and policy reform at the macroeconomic and sectoral levels. Has the amount of finance that countries receive affected their policies? Some conservative critiques of aid hold that finance will lead to poor policies. What is the evidence? We also examine policy-based (or conditional) lending and the extent to which it has helped improve economic policies in developing countries. Then we focus on less tangible mechanisms through which aid can affect policy-dissemination of ideas, education of future leaders, stimulation of policy debate within civil society. Arguably, this is one of the most important roles for aid. It is hard, however, to make generalizations about aid and policy. First, it is not true that countries that receive large amounts of aid have bad poli- cies; there is surprisingly little relationship between the amount of aid and policies. Second, policy-based lending has a mixed record. Many societies have initiated serious reform programs, and adjustment lending has been 47 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY successful in supporting those reforms and helping lock in policy gains. At the same time there is a long legacy of failed adjustment lending where there was no strong domestic constituency for reform. Foreign aid cannot take the lead in promoting reform if there is little local movement in that Conditional lending is direction. Finally, and somewhat speculatively, case studies suggest that aid worthwhile where can play an important role in stimulating reform over the long haul. reforms have serious Foreign aid has helped finance the overseas education of many leaders, an domestic support. investment that may not have much effect on policies for 20 years or more. There are also cases in which foreign assistance has stimulated policy debates within civil society, but it is difficult to measure the impact. Conditional lending is worthwhile where reforms have serious domes- tic support. As with aid in general, however, donors have not been suffi- ciently selective with policy-based lending. In countries with poor policies and no movement toward reform, it is worth thinking about how to cre- ate reformers and popular movements for reform, but it will not be easy. Adjustment lending has not been a useful tool in this respect, and may have been counterproductive. The role of aid in difficult environments is to educate the next generation of leaders, disseminate information about policy, and stimulate public debate where possible. Money-Good or Bad for Reform? SOCIAL SCIENTISTS HAVE LONG DEBATED THE ROLE OF OFFICIAL AID in supporting reform. If policy reforms have short-term costs affect- ing particular segments of the population, foreign aid can help get those reforms off the ground. Stabilization typically requires fiscal adjust- ments that lead to higher taxes or fewer services for some groups. Trade liberalization hurts firms and workers in previously protected industries. State enterprise reform and privatization often lead to transitional unem- ployment. If a government wants to implement these reforms, foreign aid can help with adjustment costs. One recent study analyzed eight major postwar economic reformers: Bolivia, Chile, Germany, Israel, Mexico, Poland, and Turkey (Sachs 1994). In each case aid was a crucial contributor, though all governments were committed to reform before large-scale aid arrived. The study con- cluded that aid helps "good governments to survive long enough to solve problems" (p. 512). Moreover, foreign aid increases the effect of policy 48 AID CAN BE THE MIDWIFE OF GOOD POLICIES reform on growth, thus raising benefits relative to costs. This should increase the likelihood that reform will be sustained. Another study, however, points out that "aid can also help bad gov- ernments to survive" (Rodrik 1996). "For debating purposes, one can cite at least as many cases as Sachs does to demonstrate an association between Aid supported plentiful aid and delayed reform .... One of the pieces of conventional governments with bad wisdom about the Korean and Taiwanese reforms of the 1960s is that policies to about the these reforms took place in large measure because U.S. aid, which had same extent that it been plentiful during the 1950s, was coming to an end" (p. 31). supported reforming Burnside and Dollar examined the relationship between aid and their governments. index of macroeconomic and trade policies for the 56 countries in their sample. They showed first that policies can largely be explained by under- lying country characteristics, such as the rule of law, ethnic splits (which are associated with poor policies), or political instability (also linked to poor policies). When aid is added to this analysis, it has no effect on the policy index. This does not refute Sachs's view that aid contributed to reform in some cases-instead, it shows only that aid supported govern- ments with bad policies to about the same extent that it supported reforming governments. While there is not much relationship between the amount of aid that countries get and the level of their policies, it is still possible that aid supports changes in policies. If donors were good at anticipating "turn- ing points" and at increasing their aid just before reform, we would observe aid flowing to poor policy regimes, but the flows would be fol- lowed by reform. Alesina and Dollar (1998) investigate this possibility. In a sample of 60 countries, they identify 87 episodes in which there is a surge in aid (a large change relative to what the country had been receiving). In only six of the 87 episodes was the surge followed by sig- nificant reform. In 92 cases in which there was a large decline in aid, 16 were followed by reform. Thus, reform is more likely to be preceded by a decline in aid than an increase in aid. But the main thing that emerges from this work is that there is little relationship between changes in aid and policy reform. It is not generally the case that donors have successfully anticipated "turning points" and increased their assis- tance in advance of reform. The varied relationship between aid and policy can be seen in indi- vidual countries. Zambia, for example, could support the view that aid enables governments to delay reforms. Policies in Zambia were poor and getting poorer throughout 1970-93, yet the amount of aid that it 49 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY received rose continuously-reaching 11 percent of real GDP by the early 1990s (figure 2.1). For every Zambia, however, there is a Ghana. Ghana received little aid when it had bad policies-but has received strong donor support In Zambia, policies got since it reformed (figure 2.2). Case studies of Ghana generally find that worse while aid foreign assistance helped consolidate a good reform program. In increased . . . Burnside and Dollar's 56 country sample, these different experiences cancel out: aid and policy are virtually uncorrelated. When other vari- Figure 2.1 Zambia: Aid and Policy ables known to affect policy are introduced, there is still no relationship between aid and policy. Aid as a percentage Policy index of GDP So, there is no simple relationship between the amount of aid that Mediocre 12 countries receive and the quality of their policies. Aid may still have con- Policy Aid tributed to policy reform in some cases, however, either through the lever- 0 age of conditionality or through the dissemination of ideas. &10 8 Ownership-What Money Cannot Buy #6 IF AID IS AS LIKELY TO DELAY REFORM AS ENCOURAGE IT, WHY NOT make assistance conditional on policy reform? After all, financial sup- Poor I port from the International Monetary Fund and structural adjust- 2 ment lending from the World Bank are designed to disburse only as 1966-69 1990-93 reform measures are carried out. Although these flows are only a fraction of official finance, other donors pay attention to structural adjustment Source: Dollar and Easterly 1998.. programs when making decisions about aid allocations. For various reasons, however, conditionality may fail to generate per- manent improvements in policy. First, it is inherently difficult to moni- tor. Take, for example, a seemingly simple condition: that the fiscal deficit not exceed a certain level. The fiscal deficit is influenced not just by gov- ernment policy, but by shocks outside government control. So a country may miss an agreed target because of a shock; in fact, doing so may be desirable because a target that is good policy in one environment becomes poor policy in another. Whether or not a policy target has been met requires some subjective judgment. The subjectivity becomes more acute as reforms become more complex institutionally. The second problem with conditionality is that it has force only dur- ing the life of the adjustment program. A government in financial diffi- culty may agree to reforms and carry them out to obtain conditional 50 AID CAN BE THE MIDWIFE OF GOOD POLICIES resources. If there is no strong commitment to these reforms, they can - and likely will-be reversed at the end of the program. The third and probably most serious problem concerns incentives within donor agencies. Disbursing funds is one of the important ratio- nales for these agencies. Since monitoring policy reform requires some . . . but in Ghana, aid subjective judgment, donors will likely find that governments are mak- rose in lockstep with ing a good effort-whether they are or not-and disburse their funds. better policies. The Economist describes this kind of donor behavior as follows: Figure 2.2 Ghana: Aid and Policy "Over the past few years Kenya has performed a curious mating rit- Aid as a percentage ual with its aid donors. The steps are: one, Kenya wins its yearly Policy index of GDP pledges of foreign aid. Two, the government begins to misbehave, Good 4 backtracking on economic reform and behaving in an authoritar- Aid ian manner. Three, a new meeting of donor countries looms with exasperated foreign governments preparing their sharp rebukes. Four, Kenya pulls a placatory rabbit out of the hat. Five, the donors are mollified and the aid is pledged. The whole dance then starts 2 2 again" (August 19, 1995, p. 37). Policy There is a mountain of literature on structural adjustment lending and \ its effect on policies (for example, Mosley 1987, Mosley and others 1995, and Thomas 1991). All the studies conclude with skepticism about the ability ofconditionality to promote reform in countries where there is no Poor 0 strong local movement in that direction. One concludes that, in Africa, 1966-69 1990-93 structural adjustment lending from the World Bank affected the policies Source: Dollar and Easterly 1998. of recipients "a little, but not as much as the Bank hoped" (Mosley and others 1995). The main problem was that the World Bank had strong incentives to disburse funds-and thus was inclined to see a good effort even where there was none. In fact, in the authors' sample of adjustment loans, only 53 percent of loan conditions were met. Even so, almost all adjustment loans were disbursed. Zambia, for example, received 18 adjustment loans between 1966-69 and 1990-93 while its policies got worse (see figure 2.1). In Kenya the World Bank provided aid to support identical agricultural policy reforms five separate times. Each time reforms were either not implemented or later reversed. Yet all adjustment loans were disbursed. The lesson? A conditional loan is no guarantee that reforms will be carried out-or last once they are. Yet, adjustment lending from the IMF and the World Bank has sup- ported many successful reform programs. Bolivia is a good example of a 51 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY country where adjustment lending provided finance to a determined reforming government, with assistance increasing in lock-step with policy reforms (Lopez 1997; figure 2.3). Much of this increase in finance came through adjustment loans. Another review of policy-based lending con- In Bolivia, adjustment cluded: "It seems clear that the lending cum conditionality process works lending provided finance well only when local polities have decided, largely on their own, possibly to a determined with outside technical help, to address their reform needs, effect certain reforming government. policy changes sequentially, and approach the international community for financial help in getting there" (Ranis 1995, p. 10). Figure 2.3 Bolivia: Aid and Policy In its own internal reviews the World Bank has come to the same con- Aid as a percentage clusion-"ownership," or strong domestic support of reforms, is essen- Policy index of GDP tial for adjustment lending to succeed. Before 1990 about a third of Good Aid 4 adjustment loans failed to achieve expected reforms, and the lack of bor- Policy / rower ownership or commitment was a key factor in the failures (Branson and Jayarajah 1995). In a sizable sample of World Bank adjustment loans (105 cases where reforms were successfully carried out, and 55 where they were not), a ' recent study found several political and institutional features associated with successful reform programs (Dollar and Svensson 1997). Most important, 52 percent of governments that implemented successful reforms were democratically elected, whereas only 29 percent of govern- ments overseeing failed programs were democratically elected. Governments that had been in power for a long time were less likely to Poor - 0 implement reform successfully. Furthermore, political instability was 1966-69 1990-93 highly correlated with failure (table 2.1). Political and economic variables Source: Dollar and Easterly 1998. successfully predicted the outcomes of 75 percent of the adjustment loans (appendix 2). Table 2.1 Features of Successful and Failed Adjustment Programs Faded 5 2 1 n I i nd. r,:, I~ ~ c-n,1-1101111' '' L I _h; 52 AID CAN BE THE MIDWIFE OF GOOD POLICIES The study also examined factors under the World Bank's control, including: size of the adjustment loan, number of conditions, resources used to prepare the loan, and resources devoted to analytical work in the four years prior to the loan. It found that these "Bank effort" variables are similar for successful and failed adjustment programs (see table 2.1). In the past the World When all the variables are tossed into the analysis, it becomes clear that Bank did not sufficiently successful reform depends primarily on a countrys institutional and take into account that political characteristics. success or failure of In the past the World Bank did not sufficiently take into account that reform depends largely success or failure of reform depends largely on a countrys own effort. A on a country's own case in point: Zambia. In the 1980s the Bank approved four structural effort. adjustment loans totaling $212 million, and they were almost fully dis- bursed. After completion, the World Bank rated three of the four loans as failures: that is, reforms supported by the loans were not satisfactorily implemented. The Dollar-Svensson results suggest that this was largely predictable. Conditions in Zambia at the time were not conducive to reform. The government, not democratically elected and in power for a long time, was likely a nonreformer. It may have been worth taking a chance on the first adjustment loan, but (admittedly with 20/20 hind- sight) a succession of policy-based loans for Zambia was not the best use of resources. Donors have gradually learned this lesson and become more selective about providing policy-based assistance. In 1990-95, the suc- cess rate of World Bank adjustment loans increased to 76 percent (World Bank 1997a). These studies suggest that policy reform depends largely on countries' institutional and political features. Borrower ownership of reform is increasingly recognized as a prerequisite for success. But once a serious reform program is started, financial assistance can help consolidate it. Fomenting Reform EMPIRICAL LITERATURE HAS SHOWN THAT REFORMS WITHIN reach for low-income countries could add 2-3 percentage points to their growth. Given that long-term growth has been zero to 1 percent per capita for most of these countries, that would be a huge improvement. For many of the poorest countries-those that are falling further behind in relative terms-initiating serious policy reform is 53 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY perhaps the most important issue on their agenda. At the same time, it is not easy to generate reform, and conditional lending into weak insti- tutional and policy environments has failed. So, can donors promote policy reform in the highly distorted environments of, for example, Myanmar, Nigeria, or the Democratic Republic of Congo? Or do they simply throw up their arms and walk away? Clearly, to desert difficult countries is unacceptable. As poverty is reduced in developing countries with sound policies, countries with highly distorted policies will account for an ever-increasing share of the poor. What is the answer? There is evidence that donors can make a dif- ference without large-scale financing. Intangible and low-cost efforts can promote policy reform over the long term-by disseminating develop- ment ideas, training the next generation of leaders, and stimulating pol- icy debate in civil society. Development agencies have played an important role in disseminat- ing information about what is good policy, both at the macroeconomic level and in individual sectors such as education, health, and pension reform (box 2.1). But good policy is not something subjectively decided by the World Bank; lessons emerge from the experiences of developing countries. Good management is, objectively, what has increased growth and reduced poverty in those countries. Some key development policies (macroeconomic stability and open- ness) have become known as the "Washington consensus." This is an ironic designation because in the 1970s agencies such as the World Bank were not promoting these policies. In fact, the Bank was taken Box 2.1 Reforming Pension Schemes 11:. t-1 . I q..<*'.- I ir. -. l.. hcIr '.I-aE- , unr.- \r,cnrin., 1 hin . HLungar:. 11'. - L -, b1e publ.: pe hn cmc,-[hlr I C.-.i-. J .jrJ inJ UrUu t..r t0 L mITPIr pa.. In IE h.O.-Lk- [al Ir t'L L..1rIC-11 - 1 ort [Kir 0 '.C I i1 i r.% d -J u1 .f pcn,ion r1.ormlin .1rc It.i, .4 -' .-rkcr a nd t0' ! ctLrc,:,. buf bLO-111.e C thi I..1'[ I Z-1 t;~- iC K-Crl US-Cd1_1~d li sciiin.3[c %J,.'kI LII rFicrc irc ro-lic rcctrc%, rhari ..r-k- inp,arr.nr n,,,!aL [hc Fr,I.[ 4 ru,.hrokl1 a, I1- r:. k .V.-,rid Bia1ril r ,ri. 4 -1 11 [". (A: IL.C b . L1 [0 hJlp ..u ri I[VlILILU" 111C CifC-C .,arid iB hIo.-Lip di; ,c nii i,o.l i ,iid ccr 1 .'rhcii LL,rru ra pcnl .it-n ~.'cr iIt, rclo'rni al[crn.1- .ie; (ilT a-rc a .f : d L-1ilPI s i h -i. '. a Ilz- I'c In fiiin:. L t, rhe-% n.. ih e, him.c becuML I r a rnC '11f -1fil I(h po1i1 r . 1 F. ..'rE ic pl.atorn1 f,r p hlic d hJL. '-11 pCrI -..n rot..rim aridi Jop'.*FI n.' ri . ' hell,pcd ri.n-, Lo.'irrl,, [LIdh chc V. L. t-'r Ib c.*. ernmvcnr r C'_1.V-n ILC [he pLIbL rh, 4 ic -[ -, I i nr d11Lr111rM.'n . 1I I ..C..-ir 54 AID CAN BE THE MIDWIFE OF GOOD POLICIES with statist policies, such as those of Julius Nyerere in Tanzania. Through experience, however, the world has learned that this approach does not lead to sustained development. At the same time such policy- makers as T.S. Chiang and T.C. Liu in Taiwan (China) were experi- menting with policies that were not in fashion with the international Where domestic political development community. factors have started a It was only in the early 1980s that development agencies started to reform movement, well- appreciate fully the value of sound macroeconomic management and trained economists and began to promote trade liberalization and encourage closed economies to civil servants are crucial learn from the success of more open ones. It is hard to assess the impact for its success. of this dissemination of ideas. Sachs and Warner (1995) identify 35 coun- tries that liberalized trade in the past 10 years. These countries almost cer- tainly were influenced by past successes in other countries. Did development agencies play an important part in disseminating knowl- edge about successful policies? Likely, yes-but it cannot be proved one way or another. Development assistance can also promote dissemination of knowl- edge about development by funding programs to send students abroad-to study economics, law, public administration, and so on. Returning students often play a key role in policy reform-either directly as government employees or officials or indirectly through work in domestic universities and the media. One famous example is the Berkeley group that designed Indonesia's reform in the late 1960s and early 1970s. Over the past 10 years many reforms in Latin America have been designed and implemented by ministers who studied eco- nomics abroad. In Botswana the U.S. Agency for International Development had a long-standing program that trained most govern- ment officials. One study of the political economy of reform concluded that few gen- eralizations could be made about successful developing country reform (Williamson 1994, p. 589). Of critical importance was a "coherent eco- nomic team." In most cases key members of economic teams in reform- ing economies have at least some overseas economics training, usually supported by donors or by nongovernmental organizations, such as the Ford Foundation. Overseas training is not, however, a panacea for reform. Instead, where other domestic political factors have started a reform movement, well- trained economists and civil servants are crucial for its success. There are also many cases where people have been trained but not put to effective 55 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY use. A recent study of aid in Africa by the Overseas Development Council includes both successful and unsuccessful examples: "Governments find it difficult to retain qualified staff because of Master's level staff in internal management problems and low wages in the public sec- government earn a fifth tor. At independence, civil service salaries were typically much of what they could earn higher than salaries in the private sector. ... In the Tanzania and working for one of Kenya of 1970, for instance, a government worker earned 14 and Nairobi's international 11-16 percent more, respectively, than a private sector employee management consulting with the same experience and education. By the late 1980s, how- firms. ever, reckless patronage practices, along with inflation and per- sistent fiscal crisis, had devastated individual salaries. In Tanzania, the number of civil servants had increased from 135,000 in 1970 to 302,000, but civil service salaries had lost 94 percent of their former purchasing power .... African govern- ments found it difficult to retain competent staff, particularly at higher grades and technical levels. Thus, one study describes a CIDA-funded project in Kenya that trained 13 economists to master's level between 1985 and 1991; within a year, ten had found jobs outside of government and the other three, freshly returned from Canada, were looking for better paid positions in the private sector. The problem is simply that master's level staff in government earn a fifth of what they could earn working for one of Nairobi's international management consulting firms or the resident mission of a donor agency. "The most able often leave: across Africa, civil servants have accepted better-paid jobs in the private sector or migrated out of the country. According to the United Nations, some 50,000 to 60,000 middle and high-level African managers left their country of origin between 1986 and 1990. Many went to work for the aid agencies themselves, lured by salaries often five to ten times higher than in the public service. In Kenya, for instance, a World Bank project hired eight Kenyans for a project it was financing in the Ministry of Agriculture, paying them between $3,000 and $6,000 a month compared with a total compensation package of approxi- mately $250 available to a senior economist in the civil service. "In a country like Botswana that avoided economic crisis, the presence of a virtuous cycle results in a very different dynamic. Regularly paid and given the means to perform their jobs, civil ser- 56 AID CAN BE THE MIDWIFE OF GOOD POLICIES vants stay in their positions. It is not unusual for a permanent sec- retary to remain in the same position for more than ten years. Although private sector wages have increased recently and some employees have left the civil service, few have left the country. The case study reports USAID's claim that all but three of the 1,300 or Stimulating debate so Botswanans who have received master's level training in the in civil society about United States since independence have returned to the country" policy is one way for (van de Walle and Johnston 1996, p. 89-92). development assistance to influence policy Clearly, well-trained officials can be an important input to good pol- reform. icy (witness Botswana). If other factors conspire to maintain poor poli- cies, however, trained officials are likely to leave the public sector (though they may still make a big contribution to development). But they often emigrate, in which case training will not have provided the intended ben- efits to the country. Stimulating debate in civil society about policy is an intangible way for development assistance to influence policy reform. This is not easy. Leaders in countries with poor policies have interests in maintaining those policies. Highly distorted trade regimes, exchange rates, and agri- cultural prices, for example, can lead to corruption and rent seeking among favored groups. In such cases donors should look for space to develop a dialogue with the middle civil service-usually more techno- cratic than political-and with elements of civil society. In Ukraine, for example, during an era of poor policies, the World Bank decided that lending would be counterproductive. It would post- pone reforms even further, and other interventions were needed-for example, public education for the government and civil society. The media, reformers within government, parliamentarians, nongovernmen- tal organizations, and the private sector were involved in major seminars, nationwide town meetings, and a weekly, high-profile roundtable with the media on key economic and institutional reform issues. One cham- pion of this program was the governor of Ukraine's central bank, who participated actively and remarked publicly that the most important things that the World Bank did in early transition to help promote reforms and development were to refrain from large-scale lending and implement the public education program. There are few hard research results to show that disseminating good ideas, sending students abroad, or stimulating policy debates in civil soci- ety leads to developing country policy reform and better performance. 57 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY But case studies suggest that these factors are often important. Moreover, such activities are inexpensive. It is not easy for If Commitment, Money-if Not, Ideas outsiders to generate reform. T HERE IS NO VALUE IN PROVIDING LARGE AMOUNTS OF MONEY to a country with poor policies, even if it technically commits to the conditions of a reform program. Providing adjustment loans to governments not serious about reform has been a major recent problem of foreign aid. For many low-income countries, initiating economic reform is crucial for future progress. But it is not easy for outsiders to generate reform. In countries with poor policies, donors should concentrate on activities that might support reform in the long run-overseas scholarships, dissemi- nation of ideas about policy reform and development, and stimulation of debate in civil society. In most cases these are relatively low-cost measures, so there is no contradiction with the strong message that most finance should go to poor countries that have made substantial progress with pol- icy reform. Our review also found that adjustment lending is most effective if a government is strongly committed to reform. This raises a question, how- ever: if reform programs must have strong domestic support to succeed, is there any point in having conditional lending? One study argues that there is still a useful role for policy-based lending, but that the way that it is man- aged must be reformed (Collier 1997). In the past donor agencies have tried to "buy reform" by offering assistance to governments that were not otherwise inclined to reform. This approach failed. Moreover, this use of conditionality undermines its potential value. Governments truly com- mitted to reform may agree to a conditional loan that binds them to a pol- icy and protects them from internal special interests. Such a conditional loan is a type of restraint that helps government resist the temptation to deviate from good policy in the pursuit of short-run interests. Conditional loans can also be useful as signaling devices. Uncertainties about policy retard investment. Conditional loans can indicate to the pri- vate sector that government is serious about reform and the new policy regime is likely to persist. In good-policy countries, aid is associated with higher private investment; the combination of reform and foreign assis- 58 AID CAN BE THE MIDWIFE OF GOOD POLICIES tance can boost investor confidence. Collier points out that the way in which policy-based lending has been managed tends to undermine its use- fulness as a restraint and signaling mechanism. These loans have not been useful as restraints (since funds have been disbursed regardless of whether reform has been carried out) and have not been good signals of the seri- To be effective, ousness of reform (before 1990 one-third of adjustment loans failed). adjustment loans need to The role of agencies such as the World Bank is not to arm-twist gov- focus on fewer, but truly ernments to do what they are reluctant to do. Nor is it true that if the important measures. World Bank works harder or puts more resources into an adjustment loan, a failed reformer can somehow be turned into a successful one. Instead, the role of international institutions should be to disseminate information that might influence public dialogue about policy reform- and to learn to read the signals about whether governments are serious or not. Mistakes are inevitable, but the success of adjustment loans should be closer to 85 percent than 67 percent. This target could be achieved by being more selective about the environments in which donors become involved in financing adjustment. Collier argues further that for this approach to be effective, adjustment loans need to focus on fewer, but truly important measures. (In the Dollar-Svensson sample the mean number of conditions is 46, and the maximum 193.) For such loans to be effective as a restraint, there has to be clear agreement between government and donor agencies about what is important and the grounds for nondisbursement. To be effective as a signal, loans would have to have a high success rate (say, 85 percent). They would then be accurate signals to the private sector that the government receiving the loans is seriously committed to reform. 59 CHAPTER 3 Money Matters-In a Good Institutional Environment ONEY MATTERS-BUT NOT IN THE WAY WE thought. Projects financed by foreign aid are often highly visible and important successes- roads and highways, schools and health clinics, irrigation infrastructure, power plants. But suc- cess can be assessed at two levels-at the micro or project level, which typically shows high rates of success or at the macro level of economywide growth and poverty reduction, where, as chapter 1 has shown, there is less visible success. If aid financing is fungible, the benefits of an aid-financed project are only loosely connected with the actual benefits of aid financing. Aid's true effect depends on the crucial (but difficult to assess) question: what would have happened in the absence of donor financing? Most aid goes to projects to assist particular sectors (figure 3.1). Do these targeted resources reach these sectors? Or is foreign aid fungible? Specifically, if donors finance a $1 million project in education, does the recipient country spend more on education than if the donors had sim- ply provided $1 million of budget support? The issue is not corruption, nor is it a question of administrative arrangements for aid flows. Instead, a person-or country-given resources "in kind" will rationally reallo- cate their other expenditures. If aid is not fungible, evaluating the overall effect of aid is easy: it is simply the collective effect of individual projects. If primary education is important, more projects in this area should lead to overall improve- ments. If aid is typically fungible, however, managing and assessing the effect of aid is much more complicated. What you see is not what you get. Evaluating aid's effect requires not just examining what happened in 60 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT the lifetime of a project, but judging what would have happened had there been no aid. The first section of this chapter explores in detail what is meant by fungibility. We then look at the available evidence on several dimensions Most aid goes to of fungibility. If a donor finances an investment for education, for projects-and most instance, what confidence does it have that the resources: project aid goes to social, economic, and a Stick with the government rather than finance, say, tax reductions? administrative " Finance investment rather than expanding consumption spending? infrastructure. . Finance education rather than some other sector? Figure 3.1 Distribution of Aid, Donors cannot be certain that resources actually go to the area favored by Type and Sector by the project. So, we examine the implication of fungibility for evalu- Program ating projects. Food assistance aid 6% One possibility is that in practice the actual effect of financing tied 3% to projects is to raise government spending exactly as if general budgetary support were given. This implies that, even when donors are financing projects, an assessment of aid's effect as a financial transfer depends on ther a judgment of overall government spending-whether expenditures are 26 allocated to activities likely to promote development and how effective Projects the spending is. This involves assessing the rationale for government spending. Again, what you see is not necessarily what you get, as the key question is: what would have happened without government interventions? What does fungibility mean for analyzing the allocation and effi- Industry and ciency of government spending and for managing and evaluating devel- other production 7% opment assistance? In countries with sound policies, appropriate Agriculture allocations of expenditures, and effective services, donors can provide 12% large amounts of assistance as general budget support, knowing the Soal and resources will be well used. In cases where there is agreement about allo- administrative cation but efficiency is low, aid projects and financing should be evalu- Inrrture ated not just as money for a particular project but also for what they Economic contribute to improving the overall efficacy of government expenditures infrastructure (see chapter 4). Where donors and governments do not agree on the allo- cation of expenditures and spending is not likely to be effective, the best approach is to reduce funding and increase support for policy dialogue Source: OECD 1998. and institution building-until donors are convinced that their funds will contribute to development. 61 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Fungibility? IRST, A SHORT BUT INFORMATIVE DETOUR. MANY GOVERNMENT programs transfer income to individuals-income that is tied to Fungibility is an issue specific purchases or that is transferred "in kind" rather than in only if the objectives of cash. A prime example is a food stamp program that provides vouchers donors and recipients are to be redeemed for the purchase of a limited list of food items. Even if different. the program is free of corruption (so that every food stamp dollar is redeemed for food), does such a program have a greater effect on people's Figure 3.2 Full Fungibility food consumption than if they had been given cash? Schools Compare two views. The naive "what you see is what you get" view is that, if food stamps bought food, food consumption increased dollar for dollar. The "fungibility" view is that food stamp recipients are sophisti- cated and have their own objectives, in which case the recipient would s- _ reduce expenditures on food from their non-food stamp income dollar- x for-dollar with increases in food stamp expenditures. So a dollar's worth of food stamps, even when spent on food, will increase food consump- tion by exactly the same amount as would have a dollar given in cash. Which view is right? For food stamps there is solid evidence. Puerto Rico "cashed out" food stamps in 1982, and evidence suggests that this had little or no effect on food expenditures (Moffitt 1989). In the United Roads States two randomized experiments of "cash out" found no effect on food Source: Hypcthetical illustration. expenditures in Alabama and only a small effect in San Diego (Fraker and others 1995).With food stamps the empirical evidence is overwhelming -that the naive view is wrong and the fungibility view is (almost exactly) right. But how does this relate to foreign aid? A simple diagram helps illustrate three points about the fungibility of foreign aid (figure 3.2). Suppose that a government spends on only two goods, roads and schools, and in its budget it chooses to spend Ron roads and Son schools (point X on the budget line). Now suppose that a donor is willing to make a grant of amount A to finance roads. If donor money is not at all fungible, this entire amount would be spent on roads. Thus the government would spend R+A on roads, and the same amount as before on schools (S). But if the money is completely fungible, the gov- ernment's new budget is effectively the pre-grant budget plus A (as long as the government meets the condition of spending more than A on roads). In this example the government would then spend R'on roads (more than the pre-grant R and more than A but less than R+A) and S' on schools. This illustrates three points. 62 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT First, all of the aid (A), or its equivalent, is spent on roads in an administrative sense in all cases, but the effect of aid spent on roads on the totalspending on roads could range anywhere from zero to the total amount A. Second, full fungibility does not mean that none of the aid gets spent on roads, just that the effect on total road spending of providing amount A for roads is exactly the same as if the amount were provided as general budget support. The practical importance of fungibility depends on whether the donor and recipient have the same objectives. If the donor simply wishes to expand the budget available to the government but it is Figure 3.3 Partial Fungibility administratively convenient to provide funding for roads, then fungibil- Schools ity is unimportant. If, on the other hand, the donor has different prefer- ences and really wanted the amount A to be spent on roads, then fungibility is a problem. Third, fungibility requires that the amount of aid provided for roads - - is smaller than that which the government would have spent on roads s - from its own resources. Suppose that the government would have chosen to spend at point X on roads and schools and then receives road aid of amount A (figure 3.3). Then actual expenditures will increase to point Y, but had the transfer been untied, the government would have chosen point Z. Thus fungibility is less likely if the amount of aid is large rela- tive to the government's budget (as in many aid-dependent economies) RR R"(R+A) or if the item to which the financing is tied is specific and only a tiny frac- Roads tion of the government budget (for example, project aid made available for an entirely new type of spending). Most foreign aid goes to projects in which the donor finances either a single large investment (such as a highway) or, more often, a cluster of related investments (rural roads, urban housing, systems of irriga- tion canals, water supply for a number of villages, programs of school construction). Yet the fungibility of aid financing has been recognized as an issue from the beginning of large-scale aid.' The analytical pos- sibility is not new, but what is new is the evidence on fungibility. Suppose that a donor provides $1 million to a government for an investment in a specific project in, say, education. The three relevant questions are: m Does government spending rise by $1 million? m Does government development spending rise by $1 million? " Does education spending rise by $1 million? 63 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Does Aid Increase Government Spending? Governments have several options when they receive aid. They can increase spending dollar for dollar. They can reduce taxes dollar for dol- There is apparently a lar. Or they can leave current spending and taxation unchanged and use greater "flypaper" effect the aid to reduce the deficit (equivalent to cutting future taxes)-or any with concessional loans mix of the three. than with grants. There is a large body of literature on the effect of transfers between one level of government and another in the same country-say, between the central government and provincial (or state) governments. The evi- dence is that money from higher-level to local governments tends to pass more into government spending and less into tax relief than a pure fun- gibility view would suggest. This tendency for nontax resources to "stick" to higher levels of government spending and not be passed on as tax relief is called the "flypaper" effect. With foreign aid there is a wide range of estimates of the size of the flypaper effect. Two studies based on sizable samples of countries find that $1 in aid translates into much less than $1 in government spending. Feyzioglu, Swaroop, and Zhu (1997) looked at 38 countries and found that $1 in concessional loans (such as those from the International Development Association) leads to 63 cents in additional government spending (figure 3.4). For concessional loans and grants combined, the effect is only 33 cents. Thus there is apparently a greater flypaper effect with concessional loans than with grants. In a sample of 46 countries, Cashel-Cordo and Craig (1990) found that bilateral loans had no effect on government spending. Studies based on smaller samples and individual countries find stronger flypaper effects. Feyzioglu, Swaroop, and Zhu (1997) use a subsample of 14 countries for which they had detailed data over time about government spending by sector (appendix 3). There was a complete flypaper effect: $1 of aid translated into roughly $1 of government spending (95 cents for bilateral loans and $1.24 for concessional loans). In a case study of Indonesia, Pack and Pack (1990) found that $1 in aid leads to $1.50 in additional government spending (figure 3.4). Thus, instead of providing tax relief by substituting for domestic resource mobilization, aid may increase taxation by "crowding in" government spending. This "crowding in" is possible if aid helps government mobilize other resources through improved tax collection or better access to commercial funds-or if aid relieves constraints that were limiting government spending. 64 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT Figure 3.4 A Dollar's Worth of Aid and Government Expenditure In large samples, a dollar's worth of aid Estimates from Large Cross National Samples leads to significantly Cents less than a dollar's 70 worth of government 60 expenditure ... 50 40 30 20 10 -10 ODA disbursements Concessional loans Bilateral loans 38 countries 38 countries 46 countries 1971-90 1971-90 1969-80 Source: Feyzioglu, Swaroop, and Zhu 1998; Cashel-Cordo and Craig 1990. Estimates from Individual Countries and Small Samples . . . but in individual Cents countries strong 200 "flypaper" effects are sometimes observed. 150 100 0-8 -50 14 countries 14 countries Sri Lanka Dominican Indonesia ODA disbursements Concessional Republic loans Source: Feyzioglu, Swaroop, and Zhu 1998; Pack and Pack 1990, 1993, 1996. 65 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY This supports the view that aid is largely fungible. Fungibility means that a government can use increased resources as it chooses-to increase spending, fund tax cuts, or reduce the fiscal deficit (reducing future taxes). Different societies will make different choices, so it will be possible to find It is difficult to find the examples (such as Indonesia) in which government spending rises hand- link between aid and in-hand with aid. But this relationship is not seen uniformly in large sam- development spending. ples of countries, which suggests that although there is some flypaper effect, it is not generally true that $1 in aid increases government spend- ing by a full $1. Some goes to tax relief or deficit reduction. Does Aid Increase Development Spending? Donor financing and funding for projects is typically not structured to finance government spending in general. Instead, it is aimed at particular projects or specific expenditures, such as investment in infrastructure or social services. In developing country budgets these are usually captured in the distinction between "recurrent" and "development" expenditures. Many developing countries have weak data on how government spending is split between development expenditures and other spending (administration, subsidies, defense, and so on). Thus it is difficult to find the link between aid and development spending in a large sample ofcoun- tries. A general idea can be gained, however, by examining the relation- ship between aid and government consumption spending. In their sample of 14 countries, Feyzioglu, Swaroop, and Zhu find that $1 in foreign aid typically results in 29 cents of public investment (figure 3.5). These, remember, were countries with a strong flypaper effect, so aid went dollar-for-dollar to government spending. It seems, though, that aid financed the government in general, not the development expendi- tures that donors typically target. In this sample 29 cents was the exact amount of a typical dollar of government spending from all sources (aid and non-aid) that goes into investment. Thus an aid dollar had exactly the same effect as one from any other source of government revenue. The strong effect of aid (especially bilateral aid) on government con- sumption spending is seen in the same large sample of countries used in chapter 1 (figure 3.6). The strong association of more aid with higher government consumption spending (even after controlling for other determinants of government spending) suggests both a flypaper effect (that resources stuck) and fungibility (that allows aid provided for invest- ment to finance consumption spending). 66 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT Figure 3.5 A Dollar's Worth of Aid and Government Investment The typical aid dollar ("Development") Expenditures finances 29 cents of Cents public investment.. 100 60 40 20 I 0 14 countries 14 countries Indonesia Sri Lanka ODA disbursements Concessional loans Source: Feyzioglu, Swaroop, and Zhn 1998; Pack and Pack 1990, 1993, 1996. Figure 3.6 Bilateral Aid and Government Consumption ... most of the rest goes to government Government consumption as a percentage of GDP consu ptin. consumption. 40 35 30 + 25 *s * ** 4 ** ,4 * 20 * * * * 5 0 0 2 4 6 8 10 Bilateral aid as a percentage of GDP Source: Calculated frorn the Burnside-Dollar data set. 67 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Empirical studies of Indonesia and Sri Lanka also find that aid has a smaller effect on development spending than on total government spend- ing, so that not all aid passes into investment. Again, the differences across countries are large. In Sri Lanka an additional aid dollar raised govern- ment spending by 52 cents. Of that, 38 cents was devoted to develop- ment spending and 14 cents to consumption spending. In Indonesia a dollar's worth of aid raised spending by $1.58, of which development expenditures accounted for 89 cents. The evidence suggests that, even where aid goes mostly to government spending, the spending funded by aid is largely fungible between con- sumption and investment, regardless of whether the aid is administra- tively targeted to specific investment projects. Does Project Aid Finance Particular Sectors? Researchers have also tried to address sectoral fungibility-that is, does higher foreign assistance for a particular sector (for example, agriculture or education) raise spending for that sector? There are two ways to answer A dollar's worth of aid to Figure 3.7 A Dollar's Worth of Aid to Agriculture and Spending on agriculture leads to less Agriculture than a dollar's spending Cents on agriculture. 100 80 60 40 20 0- -20 Dominican Indonesia Sri Lanka 14 countries 14 countries Republic (public investment) (government expenditure) Source: Feyzioglu, Swaroop, and Zhu 1998; Pack and Pack 1990, 1993, 1996. 68 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT that question: by comparing spending over time within a country and comparing spending across countries. Whatever the method, the answer is the same: it varies. In some countries and sectors aid appears to be com- pletely fungible across sectors, while in other countries and sectors the money seems to "stick." Government commitment Pack and Pack (1990) show that for every dollar increase in donor to particular sectors is financing for Indonesian agriculture, expenditures in that sector increase more important than by 92 cents. In contrast, a cross-sectional study of 14 countries finds that targeting aid. agricultural expenditures decrease by five cents for every $1 in aid given to agriculture (figure 3.7). The net effect of an aid dollar on spending in a specific sector depends not only on the composition of the aid received, but also on how gov- ernment responds to aid flows. Take education and health, and the results for a cross-sectional sample of 14 countries. The direct effect on educa- tion spending of a dollar in aid depends on the allocation of aid among sectors: 8.7 cents of every aid dollar are allocated to education and health. With sectoral fungibility, however, that is not the whole picture, which depends on whether aid for education leads to a reduction in what the government otherwise would have spent on school programs, and whether aid for other sectors causes the government to spend more on education. In the estimates from the 14 countries, the net effect of a dol- lar's worth of aid is to reduce education and health expenditures by 6.5 cents, even though on average 8.7 cents was devoted to education and health (figure 3.8). The effect varies enormously by country. In Sri Lanka 5.9 cents of an aid dollar go to education and health, but the estimated net effect is to reduce spending by 1.9 cents. Indonesia receives a similar proportion devoted to education and health, 7.2 cents, but the effect is to increase spending by 18.9 cents; clearly money earmarked for other sectors was reallocated to education and health. The important point is that gov- ernment commitment to particular sectors is more important than tar- geting aid. These statistics do not tell the whole story of consequences. It may be that the observed changes in public expenditures are exactly what donors wanted anyway. The donor objective, for example, may not be to increase total education spending, but to increase spending on primary education while reducing spending on higher education. But a shift in the alloca- tion of spending requires agreement between donors and the government based on analysis and dialogue. 69 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY A dollar's worth of aid to Figure 3.8 A Dollar's Worth of Aid to Health and Education and Spending education may lead to on Health and Education little (or no) additional spending on education. Cents 20 1 Direct: Proportion of aid allocated to health and education *li Total: Net impact after accounting for sectoral fungibility 15 10 5r.I'II -5 -10 Indonesia Dominican Sri Lanka 14 country Republic sample Source: Feyzioglu, Swaroop, and Zhu 1998; Pack and Pack 1990, 1993, 1996. Budgetary Mechanics of Fungibility The magnitude of fungibility depends on a country's budgetary struc- tures, the degree to which governments are able to manage their finances, and the extent of donor involvement. In India's complicated federal fiscal system of transfers between the central government and individual states, for instance, the budgetary system was designed so that the donor financing coming through the central government to projects in specific states would not affect either the projects chosen or the total transfers received by the states. This made fungibility explicit. The Indian budgetary system was designed so that projects were chosen to be part of the public sector investment plan or not, and then financing was obtained for those projects. The evidence from an empirical study suggests that aid to the Indian cen- tral government was fully fungible, both between development and nondevelopment spending and across sectors (Jha and Swaroop 1997). Other countries, such as Mexico, have incorporated mechanisms into their budgetary system to create fungibility: the goal of their system is 70 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT to ensure that budget allocations are not to be affected by the avail- ability of finance. The Indian system also determined the amount of investment financ- ing from the central government to each state. In the past, if states received a larger amount of support from donors, their transfer for development Realized expenditures spending on capital account from the central government was corre- differ enormously from spondingly reduced (Jha and Swaroop 1997). budgets. One major weakness of aid in Africa is the lack of strong budgetary control by either government or society over public investment. A case study in Ghana, Malawi, and Uganda revealed that controls were too weak to accomplish any of the three main objectives of budgetary sys- tems: developing aggregate budget discipline, imposing strategic prior- ities, and encouraging technical efficiency (Campos and Pradhan 1996). The big problem is that budgets are often ad hoc-that is, realized expenditures differ enormously from budgets, and there is a disconnect between the recurrent budget and the development budget. Moreover, donor financing, particularly through projects, is a large component of investment and total expenditures. For instance, donor project financ- ing accounted for 71 percent of capital expenditures in Ghana and 87 percent in Uganda (table 3.1). Weak systems of budgetary planning, control, and management, com- bined with large donor contributions, often cause "the bulk of the Public Investment Program to be determined by donor-driven priorities, thereby fragmenting and undermining the priority setting process within the countries" (Campos and Pradhan 1996, p. 29). Lack of government con- trol over project selection and the budgeting process was one of the key problems in African aid identified in a recent study by the Overseas Development Council and implemented jointly with African researchers (van de Walle and Johnston 1996). Their case studies of seven countries found that donors often drove project selection and most projects had lit- tle recipient input. For example, in Senegal only 170 of 316 projects in Table 3.1 Donor Financing of Investment Budgets in Two African Countries (percent) -J P id, r. I - 71 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY the public investment plan had been assessed using the appropriate inter- nal procedures. Donor control likely meant that aid would be less fungi- ble, but it raised other problems-such as a lack of government ownership and commitments to recurrent finance for projects. Project Assessment with Fungible Money These differences in fungibility across countries mean that it is pretty certain that what you see is not what you get. Moreover, it is difficult to know either before or after donors provide finance the extent of fungi- bility, and so exactly what was financed. But if money is fungible, evalu- ating donor-financed projects is beside the point, as their effect is not the same as that of donor financing. To take a hypothetical example: suppose there were 11 discrete projects, each of which cost $10 million. Suppose, also, that the economic returns on them ran from 100 percent (for pro- ject A), to 90 percent (project B), and so on down to 0 percent (project K). Suppose, finally, the country will invest $50 million from its own resources, while a donor provides $10 million-in total, enough to finance six projects. There are several possibilities (table 3.2). Table 3.2 Evaluating the Effect of Projects with Fungibility slf.idinane psi 7. -Iofaid fanou P>oyet undeirko, Projecr Rerun Prolerr Rerunt [1--or hn inLr A. chc I,_hcr-rcrurn prI r..r, r-J rt w..-ir, ir. F i h. M.fc 111.11 P Ab DF F In p,urn.-..r F iln pdnrc.n.c red ,.:c 'r'.r,c LBEr r mll r- -jelr CNn-r hn i. crrin F,V Al Peru rkr F, 4k 1.Tr.I pcl r,' F''.~ r t"r.:d rKu'u.c Iman.:1- F'. i 1-.-rcrurn pr-+:,iI F F h '~.chl K 0pc~n ri,,rrr lir lC-1L.A. ch, hi_hc'r- rc[urrrprrur and Ir-id r.- otircc, hin ncc cc-n.Unipciojn ipfrn,. .r rtsd11-:C [A.L. ,'BCD LE 1 i pc rccur, r Nr-,nr Unrkno%% 1. could bu ". . Is .,J ..r, a:. .. Err ch bi. ,.rb_r!r l,.an.p' ... . b ..I h r.,. . 3 ..r e *t p I.. . p r.I.r ,.. ...r.. , * II u .I t r rh .rter.-,-Iec. r...r...c .* rorp. rh. . p .l JllE .i r,l 72 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT First, the government agrees that the donor will finance the highest- return project (A). Thus the apparent rate of return is 100 percent: that is, when the donor assesses the benefits it will find that its aid has been hugely successful. What really matters in evaluating the effect of aid, how- ever, is what the government does with the resources it otherwise would While the project have devoted to the project. It could do the next best project not yet approach has some financed, which would be F (with a return of 40 percent), since it is advantages for donors, it already committed to B,C,D, and E. Or the government could choose to does not ensure that aid do a white elephant project (K) with political payoffs but with a return is put to good use. to the economy of zero. It could also choose to reallocate the resources from investment to consumption. What is the return on additional con- sumption expenditures, high or low? Often consumption spending, such as that on recurrent inputs or operations and maintenance, is not financed by donors but has enormously high returns in keeping existing investments productive. So a shift from investment to consumption could have a huge return, say, 200 percent. Alternatively, the consump- tion could be the conspicuous consumption of the elite and have zero return. Or the aid could have financed tax relief, which as a "project" may have very high returns if it reduces distortionary taxes. Delivering aid through projects has several advantages for donors. There are administrative advantages to organizing tasks into discrete, manageable units. Investment projects fit well with the notion that increasing investment is the key to development. And by financing par- ticular items (family planning, education, and so on), projects appeal to single-interest coalitions in donor countries supporting foreign aid. But while the project approach has some advantages for donors, it does not ensure that aid is put to good use. Nor does it facilitate the eval- uation of aid effectiveness. If, say, aid finances a road project with a 20 percent rate of return, this information tells us next to nothing about aid effectiveness. Does fungibility imply that projects are irrelevant and should be abandoned as a way of organizing and delivering aid, in favor of broader financing mechanisms? Not at all. Project financing can be used to convey other value added. In many cases the project's value added was never intended to be the investment itself. Donors often are not try- ing to influence project selection or directly alter the sectoral composi- tion of spending but are using projects as a vehicle for technology transfer or to build institutional capacity or reform sectoral policies. Taking the road example, donor involvement may mean that those institutions responsible for roads are improved-or road policies are 73 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY changed for the better. Projects are a cauldron of money and ideas. Since the money does not necessarily stick with the sector, the design and evaluation of projects should focus on the contribution of ideas to alter- ing sector institutions and policies (see chapter 4). Donors are, more or less, financing whatever the government chooses Public Spending-Quality not Quantity to do. T HE SAFEST ASSUMPTION FOR DONORS IS THAT THEY ARE, MORE or less, financing whatever the government chooses to do. Since donors are financing government in general, they need to con- sider all public expenditures, in terms of both allocation and efficiency, when deciding the appropriate level and method of aid. So donors must each decide, what makes for an effective public spending program? Equity and Efficiency Governments ought to be doing things where there is a convincing rationale for public involvement. Twenty or so years ago, donors willingly financed almost anything in which the government chose to try its hand-textile plants, shoe factories, steel mills, and all sorts of manufac- turing. Not only were developing world parastatals financed through donor credits and loans; many government corporations were created because donor financing was available. There has been a clear shift in thinking about the value of government investments in shoe making or textile weaving. Today few people would dispute that these activities can just as easily (and with greater effect) be undertaken by the private sec- tor. In general, the rationale for government intervention is to improve equity and efficiency. But government action that achieves these improve- ments proves difficult in practice. Equity. Many government expenditures-even those traditionally jus- tified by their improvements in equity-are not well targeted toward poverty alleviation. For instance, an assessment in Brazil found that social insurance payments in 1980 reached only better-off formal sector work- ers, so the richest 20 percent of workers received 12 times the benefits that went to the poorest 20 percent. In education, too, equity implica- tions differ widely between primary and tertiary education, for instance. A study in Indonesia found that the richest fifth of households received 74 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT only 60 percent of the benefits per person from public spending on pri- mary education as the poorest fifth. Since children in richer households were more likely to remain in school, however, the higher the level of schooling, the greater the benefit to richer households. By senior sec- ondary school, the richest fifth received 17 times the benefits of the poor- The quality of est fifth (van de Walle 1995). government expenditures Efficiency. Justifying government actions to increase efficiency is also not is typically more always simple. The existence of market failures must be taken into account. important than the But so, too, must their magnitude, particularly when compared with the quantity. likely "government failure" in implementing a proposed remedy. The notion that since market failures were pervasive, they were complete and therefore demanded government action, has been replaced by a nuanced view-that both market failures and government failures are pervasive, but not complete. So, although working from common principles, a country- by-country assessment is needed to determine the magnitude ofmarket fail- ure and the depth of government capacity (World Bank 1997b). Nevertheless, the principle of efficiency often leads to clear recom- mendations. For instance, governments, sometimes supported by donor financing, have devoted substantial resources to manufacturing and trade that markets could have provided equally well and for which there was no underlying efficiency rationale. Moreover, governments have often used broad-based subsidies that lack rationale in either equity or effi- ciency-such as large-scale subsidies to agricultural products or fuels used mainly by richer consumers, which impose an explicit or implicit tax on poorer agricultural producers. Quality and Quantity The quality of government expenditures is typically more important than the quantity. While there is evidence about what kinds of public sec- tor outputs are crucial to growth and poverty reduction (basic education and infrastructure, for instance), there is often little or no evidence that increasing public spending would have an impact on these objectives. Just as the effect of aid differs across countries, so too does the effect of increased public spending on improving outcomes. Cross-national and sectoral evidence, as well as individual cases, suggests that inputs like money are often not the most important issue for public sector outputs. The empirical growth literature finds that government consumption has no positive effect on growth. Some studies find a negative effect, others 75 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY zero. The finding that aid largely finances government consumption, which has no positive effect on growth, again helps explain why aid is not fueling growth in many developing countries. One study used data on economic growth and the allocation of gov- The quantity of public ernment budgets for 43 developing countries to show that there was no spending on health bears relationship between economic growth and public capital spending little relationship to (Devarajan, Swaroop, and Zou 1996). The authors found that increas- lower mortality. ing by one percent the share of public spending devoted to capital (ver- sus current) spending would reduce economic growth by half a Figure 3.9 Public Spending on percentage point each year. Another study, using a specially created Health as a Percentage of GDP dataset on public investments, found that the share of GDP devoted to and Health Outcomes public investment had no association with economic growth, after con- Percent trolling for a standard set of determinants of growth (Easterly and Rebelo 1993). So while one intended purpose of aid might be to increase public investment, the quantity of investment alone bears little relationship to economic growth. Examining the effect of the magnitude of public investment across countries shows that differences in the efficiency with 2 which governments invest is as important in explaining growth as the amount they spend on investment (Pritchett 1996). The data are similarly agnostic about the relationship between sectoral expenditures and economic growth. Devarajan, Swaroop, and Zou find no association between growth and the share of public spending devoted to education or health. Moreover, they find that both education and health spending per capita are consistently negatively related to growth. Easterly and Rebelo also find that the share of GDP going to public 0- Ten Ten Jamaica Brazil investment in education and health is unrelated to economic growth. "obest" "worst" (-.79) {+.51) -cunr .oprsn thtmresed cuntries counties There is little evidence from cross-country comparisons that more spend- (-.46) (+.45) ing on public investment, or even reallocation across sectors, would auto- Nte: The number in parenthesis is the residual matically translate into better growth outcomes. from a regression ofthe natural log ofunder-5 This finding is consistent with estimates of public spending efficiency mortality on GDP per capita and other socio- economic determinants of mortality for three major sectors: health, education, and infrastructure. Comparisons Source: Filmer and Prichett i997.. across countries suggest that public spending has some (small) effect on health. Ranking countries by the difference between actual child mortality and the level expected for their economic and social determinants of health, the 10 countries with the best performance in health typically spend 1.3 percent of GDP on health-and child mortality is 40 percent lower than would be expected (figure 3.9). The worst 10 have child mortality roughly 40 percent higher than expected, but actually devote a larger share of GDP to public spending on health than the 10 best (Filmer and Pritchett 1997). 76 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT In explaining any measure of mortality (infant, child) or life expectancy, regression analysis finds little or no correlation between public spending and health status (Musgrove 1996). Likewise, in education there is little evidence that higher spending means higher quality schooling. Using assessments of educational per- The actual effect of formance in math and science based on internationally comparable exam- spending in almost every inations, one study found that both total and recurrent spending on area of endeavor-from education as a fraction of GDP was negatively related to test score infrastructure to achievements (Hanushek and Kim 1995). While most agree that a rapid education-has been expansion of skills was key to the pre-1997 success of the fast-growing much less than the East Asian economies, these countries are not distinguished by the extent potential. of the resources devoted to education during its takeoff and after (public spending on education accounted for about 3 percent of GDP in the Republic of Korea) or by physical inputs such as the pupil-teacher ratio, which in 1985 was 38 in Korea, 31 in the Philippines, and 32 in Sri Lanka (Tan and Mingat 1992). In infrastructure there are similar differences in cost and performance across countries (World Bank 1994). While the flow of infrastructure ser- vices is important for economic growth, it is harder to show that infra- structure expenditures (or even stocks) are important (Hulten 1996). It will come as no surprise to anyone with experience in developing countries that comparisons show little connection between spending and outcomes. This does not mean, however, that spending cannot influence outcomes. The actual effect of spending in almost every area of endeavor- from infrastructure to education-has been much less than the potential. When the what and the how well come together and governments do the right things well, the results can be amazing. An obvious example is the educational accomplishments in East Asia, where countries such as Korea have gone from widespread illiteracy to educational quality that rivals (and surpasses) that of much richer countries. There are similar examples in health, where the mobilization of public health efforts to attack problems of communicable diseases, diseases carried by pests, and water-borne diseases led to a revolution in health status. Sri Lanka and the state of Kerala in India have life expectancies far exceeding that pre- dicted by their income and education. When the state of Ceara in Brazil was able to mobilize and energize government workers in the health sec- tor, it achieved enormous reductions in infant mortality in a few years. When the emphasis is on what government should do, and particularly how much it should spend, aid becomes a means to support spending, as 77 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY if spending were the end in itself. Thus aid flows independently of how effectively spending promotes poverty reduction. In many cases, even when donor resources increased investment in the right areas, it did not promote development. Thus it is not just what governments do, but how Donors must form an they do it and how well. opinion on the allocation and effectiveness of a country's public What Should Donors Do? spending. WHAT ARE THE IMPLICATIONS OF FUNGIBILITY AND OF DIF- ferences in the quality of public spending for the way donors do business? Many discussions of aid policy con- centrate on what donors ought to be financing: girls' education or rural feeder roads, agricultural extension or family planning clinics, power plants or vaccination campaigns. But if money is fungible among sec- tors (and it often is), then from the narrow view of expanding expendi- tures it simply does not matter what donors finance. Most development practitioners recognize fungibility, however, and have moved to a broader concept of the effect of donor financing: it is more than just an attempt to expand capital. This raises the question of how donors allo- cate effort and available resources to best assist each country. As with country-level policies, we return to two themes in choosing the magni- tude and mode of assistance: timing and selectivity, and the role of money and ideas. Form a View ... To pursue an effective development strategy, donors must form an opinion on the allocation and effectiveness of a country's public spend- ing. Donors can help only if they know what a country's needs are. In the past decade public expenditure analysis has emerged as a useful tool to assess the value of government spending in different areas. Reviews of public expenditures have focused on how well budgets meet the three criteria of budgeting: fiscal discipline, priorities across sectors, and tech- nical efficiency. Donors need to be able to form a judgment on how well government spending is working along each of the three criteria. These reviews led to the elimination of many "white elephants." Scrutinizing (and often 78 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT cutting) large planned investments that would later generate budget commitments was one fruitful result of public expenditure analysis. Public spending analysis has also brought to the fore attention on issues within sectors-across subsectors and across functional categories. When countries are faced with fiscal adjustment, two common tenden- One painful lesson of cies are across-the-board cuts in budget items and cuts in everything but experience is that payroll-approaches that are inefficient. Analysis has also been effective government and in bringing into focus subsectoral priorities-say, preserving primary community ownership of education, less so tertiary, from budget cuts. projects is crucial. The first generation of public expenditure analyses did not go far enough. The analyses generally did not use a framework for spending that established justification for public expenditures or broad guidance about the appropriate composition of spending across sectors (Swaroop 1997). A review of 113 World Bank public expenditure reviews found that few even looked at the role of the state and public-private boundaries (McCarthy 1995). Moreover, very few reviews went beyond expenditures and linked them to specific outputs. ... and Use the View Once donors have formed a view on the allocation and efficiency of public spending, what then? One common donor reaction to the fungi- bility of money is to try to avoid it by imposing additional conditions on aid financing beyond one project to ensure the additionality of resources at the sectoral level. Bad idea. One way to ensure additionality is for donors to finance only items or projects that governments are not financing. Common sense tells us, however, that this reasoning is flawed. It would mean that donors finance only what countries and governments do not want to do. And govern- ment would have little or no commitment to implementing or, much less, maintaining projects once donors depart. Some governments (purposely or by default) have given donors free rein to do whatever they feel like. Money for such projects would almost certainly be "additional," but with equal certainty the projects themselves would be pointless, except in those rare cases that the project itself creates commitment. Far better that both recipients and donors be strongly committed to the sustainability of pro- jects. But this is precisely when money will be fungible. One painful lesson of experience is that government and community ownership of projects is crucial. Trying to make financing additional, 79 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY however, implies that there is a disagreement between donors and gov- ernment about the allocation of expenditures but that donors have enough clout to ensure that governments will follow their recommenda- Donors should take it for tions on the overall allocation of spending. This is practically impossible. granted that their It is not just a question of saying that if a donor provides an additional financing is fungible $1 million that spending on the activity financed must increase by $1 because that is reality million because spending might have increased by that amount in the absence of the extra aid. To ensure additionality, donors would have to make governments promise that they are spending $1 million more than they would have had they not received the aid. But they would say that in any case, wouldn't they? In fact, donors should take it for granted that their financing is fungi- ble because that is reality. This, in turn, requires rethinking how aid is delivered. Efficient delivery in different countries depends on the nature of public spending and the efficiency of the public sector. One measure of public sector efficiency is the index of economic management. And an earlier recommendation was that of two equally poor countries, well- managed countries should receive more finance than poorly managed ones (see chapter 1). In a country with sound public expenditure management, a larger por- tion of aid can be in the form of general budget support. This recognizes the reality of fungibility and economizes on the administrative costs of aid. These are important-a typical World Bank project runs about $1 million in administrative costs. If a project is merely replicating what the country already does relatively well (building another power plant in a well-managed system), there is little value added from administrative costs. Providing general budget support would free these staff and donor resources for other, higher-value added activities. Problems with public expenditure management are usually one of two types: inefficiency in allocation of resources across activities or ineffi- ciency in the use of allocated resources. Composition ofSpending. Consider first a country that does things rea- sonably well, but continues to spend on many things that are inefficient because the private market could do these things just as well or the social returns are simply too low. Suppose that a country insists on putting large amounts of public finance into higher education. Donors traditionally want to fund primary education to expand its relative share of the bud- get. But with fungibility this approach is not rational. Possibly a better approach is "time slice" financing of the education budget. 80 MONEY MATTERS-IN A GOOD INSTITUTIONAL ENVIRONMENT Developed specifically for this purpose are such instruments as the Sector Investment Program. Donors and recipient governments reach a broad agreement on the goals for the sector-and develop projects and programs to achieve these goals. Donors then finance particular projects within the agreed program or provide general support to the sector bud- There are still wide get. Periodic reviews of the Sector Investment Program provide a forum divisions of opinion about where donors and recipient governments can frankly exchange views on the right role for the overall sector spending program. Donors can use this policy dialogue government between and to improve the allocation of expenditures. Another mechanism some within donor and have tried is a Medium-term Expenditure Framework that tries to pro- recipient countries. duce a strategic direction for government. Case studies of Ghana and Uganda, however, have found these less successful than hoped. In Ghana, some expenditures are designated as core. Pressures to include more items in the core led to the creation of a "supercore," which was just as pres- sured. Both concepts were abolished in 1990. One cannot expect mira- cles, however, and some early experiences with these innovations are showing natural implementation difficulties. While it is useful to have an open dialogue, it is unrealistic to expect complete consensus on the important activities for government to under- take. There are still wide divisions of opinion about the right role for gov- ernment between and within donor and recipient countries. Some people believe that governments should subsidize tertiary schooling. Others do not. Some believe that governments should play a strong role in allocat- ing credit. No, say others. The government should provide electricity, say some. Many disagree. These are not just disagreements about the extent and nature of market failure but about values, like the extent to which gov- ernments should take redistributive action. Technocratic advice will not be sufficient to resolve most disputes about allocations across sectors or within sectors across categories of spending. In 1990 Sweden spent three times Japan's 11.5 percent of GDP on social transfers (Lindert 1994). Given the enormous differences among donor countries in the allocation of spending and of private and public responsibilities, it is hard to believe that a sharp consensus will emerge from the democratic process even among donors, much less between donors and recipient governments. Effectiveness ofSpending. Suppose now that an aid recipient has spend- ing allocations that donors broadly support, but is inefficient in delivering services. If a country has carried out macroeconomic reforms, it should be getting a significant amount of aid. Macroeconomic reform is not techni- cally difficult, but developing an efficient public sector is more challenging. 81 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY In this case donors should largely stick with the project approach. The main objective of projects now is not to increase funding for a particular activity, but to help change the institutions and policies at the micro level that affect service delivery. An assessment of public sector efficiency is needed to make an informed decision about how to deliver aid. Once countries have car- ried out macroeconomic reforms, those with efficient public sectors can receive budget support, while those with inefficient public sectors would get relatively less money and more ideas. Where a country has both a severely distorted allocation of expendi- tures and an inefficient public sector, institutions and policies will be poor and finance will be of little value. The traditional donor response has, nevertheless, been to try to find something useful to finance. Fungibility helps explain why large amounts of aid have had no lasting effect in these highly distorted environments. There is no alternative to patience. But there are a variety of ways that donors can, without large flows, support the creation of an environment for productive public expenditures, the subject of chapter 4. Note 1. The World Bank's first Chief Economist, Paul Rodenstein-Rodan, empha- sized in the 1950s that even when financing was tied to projects, money often remained fungible. Assessments of aid as early as the 1960s raised the issue of fungibility (Little and Clifford 1965). 82 CHAPTER 4 Aid Can Be the Midwife of Good Institutions A S THE WORLD'S ECONOMIES MOVE INEXORABLY TO Ho0w can development embrace market-friendly policies, public sectors assistance be designed are reorienting themselves and downsizing. "Less is so that it helps better," has been the cry. But could this response governments carry out go too far? In many crucial areas there are essential better the activities public responsibilities-in developing human, essential to increasing institutional, and physical infrastructure-that government must growth and reducing continue to fulfill. In these areas, less is not necessarily better, but poverty? "better is better." One key to development is that government must do well those things that government must do. Since aid (both money and ideas) mostly sup- ports the public sector, one pressing question is: How can development assistance be designed so that it helps governments carry out better the activities essential to increasing growth and reducing poverty? Another is: How can aid combine financial support with help to create local knowl- edge so that governments can improve the quality and effectiveness of public services? Creating knowledge does not mean that donor agencies (or the experts they hire) have chunks of technical or engineering information that they simply transmit to aid recipients. In the public sector, devel- opment knowledge is needed to design and effectively run the institu- tions responsible for public services: primary schools in El Salvador, water supply in Guinea, road maintenance in Tanzania, or utility regu- lation in Argentina. This is not knowledge that exists somewhere and can be packed in a suitcase and carried to developing countries. To be effective, this is knowledge that must be created locally and internalized. Existing principles must always be adapted to new or local circumstances (or both), and developing country governments and citizens must take 83 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY the lead in creating this new knowledge. Development agencies can, however, bring unique value added to the table. A consistent focus on improving the quality of the public sector implies a radical shift in the way aid business is done-in the choice of Before choosing instruments, how those instruments are used and evaluated, and how instruments, aid donors donors relate to governments and civil society. first need to be clear Before choosing instruments, aid donors first need to be clear about about goals. goals. The role of aid varies depending on country circumstances; the mix of financial and nonfinancial development assistance must be tai- lored to specific needs. In the few poor countries where the public sec- tor is already effective, the task of foreign aid is straightforward: simply financing the expansion of public services is likely to be both success- ful and beneficial. In most developing countries, however, governments are not effective at providing public services, and financing more of the same is unlikely to be successful or beneficial. In these cases foreign assistance should focus on increasing the effectiveness of core public ser- vices. This requires the right mix of finance and ideas to produce the greatest value added. Money does not usually matter most. Aid agen- cies must look beyond providing finance to supporting the creation of knowledge. A combination of policy dialogue and financing is more likely to increase aid effectiveness than a narrow focus on successful implemen- tation of aid-financed projects. At times, donors have hindered the cre- ation of effective public sectors because they saw end runs around local institutions as the easiest way to achieve project success. Donors need to convince countries of the value of better policies and institutions rather than cocoon "their" projects from the worst consequences of those policies. Better Public Provision ROVIDING PUBLIC SERVICES EFFECTIVELY AND EFFICIENTLY requires that all stakeholders-governments, providers (whether public employees or private), and citizens-have reasonable incen- tives. These incentives can be manifold: from making governments responsive to citizens to ensuring the right structure of civil service pay, to designing concession contracts that fit country and sector circumstances. 84 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS Accountability No longer can governments be monopoly providers of services (educa- tion, health, roads, irrigation, water, and so on) with no accountability to citizens (directly or indirectly) for their performance. In recent years donors Donor finance can affect have moved to support accountability and governance as well as innova- accountability and the tions that support greater community "voice"and involvement. In a range quality of public services of sectors these reforms carry different names and take different forms but for good or ill. reflect the same impetus-rural water supply ("participation"), irrigation systems ("water user associations"), schools ("decentralization" and "school autonomy"), transport ("road management boards"), health ("community associations"), and environmental management ("community forestry"). Donor finance can affect accountability and the quality of public ser- vices for good or ill. In the irrigation sector examples abound of how an emphasis by governments and donors on quantitative (or investment) tar- gets and modern technologies led to the exclusion of the intended bene- ficiaries from planning, design, and implementation. Sometimes that exclusion can lead to almost absurd results. One major donor-financed irrigation system in Nepal was designed by technical staff on the assump- tion that the area was unirrigated (Ostrom 1996). A fortuitous delay in the project provided the time to discover that, in fact, there were 85 fully functioning farmer-managed irrigation systems. Beneficiary involvement would have saved red faces. Another Nepalese irrigation scheme actually lowered agricultural productivity by undermining preexisting arrange- ments among farmers (Hilton 1990, 1992). A recurring problem with irrigation projects has been inadequate rev- enues for operations and maintenance. Merely raising water fees does not solve the problem unless providers are accountable and the necessary rev- enues are devoted to maintenance. Even when revenues are available for maintenance, the organizational structure of the provider can be a prob- lem. One study compares irrigation systems in India and the Republic of Korea (Wade 1995). While the two systems are similar, the different design of their irrigation departments (responsible for delivering water) leads to enormous differences in performance. In India the department is a cen- tralized bureaucracy that relies on general treasury revenues for finance; in Korea irrigation officials are local and in constant contact with farmers. Such problems are not unique to irrigation, but plague service delivery in many other sectors-water supply, health care, education, road main- tenance-due partly to the way aid projects have been structured. The 85 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY emphasis on increasing the quantity of financial aid as the answer to increasing growth has led agencies to targets that centered on the size and speed of disbursements and a technocratic approach to the design of investments. Even when all parties know better, "moving the money'can easily and quickly become the paramount institutional objective (Tendler 1975). With that kind of thinking, agencies opt for large projects rather than small ones and prefer to deal directly with centralized agencies on efficiency grounds (box 4.1). Doing so makes accounting and adminis- tration easier. In many cases donor investment projects-with fixed sched- ules and budgets-have left no room for beneficiary involvement. In some cases this led to "approval cultures" in aid agencies, where success was judged by the volume of disbursements (World Bank 1992). Reviews of the Danish donor agency's 30-year experience in Tanzania's rural water sector found that the same problems (lack of maintenance, low community interest) occurred repeatedly, yet only modest changes were made in project design. The agency's procedures were dictated by a need to implement projects rather than to create supplies of water that met the needs of beneficiaries. A review of the World Bank's irrigation experience suggests that although the benefits of user participation are large, "irriga- tion professionals are wary of such participation, because they know it will lengthen the implementation period"(Jones 1995, p. 141). Some measure of the importance of beneficiary participation in pro- ject success can be gleaned from an evaluation of 121 rural water supply projects financed by donors and nongovernmental organizations in 49 countries (Narayan 1995; Isham, Narayan, and Pritchett 1995). As Bov 4.1 Aid and Centralization 0 t.. n wi unj I-u ' te . .u A . .. 'E a.-. ..i hblock Donoir . anee are c nrmall prepared to c p., Jl. in ch.. deli. cr.-otpubb.- s I. C.rm- lcnd duccdl, t lcd .ernmn i i ;. hout a gyaran- PIL hcalEh , oal acltrc agri ltr iI..r.l ,In. rnJ to.rr nmthe naninil-.er'nmnt Pt prs-iding'uch rUrd Intr.I'tr1ti urc pro% Iion. .1 . hich ha,C dll - guirinteec run ccoinier t.:. [h, principle of dooklinr hCLn de.,lk cd in ihe Philippin-i 1intui.c,. dilr- re Ip b . he kJI.'. em.cnment in the absencc Ai1. BLIE in pracEiCc dcnfa .tEln hay lc c. prob- loi 0ti1. t11A1n C*f1luiun, LA 1k lV[ins or l c ail Li, in [Li flrildari..*n of doinor pr11. n 'er Th. ga rnmn't. n ..i. er nmient ac-inCies tend tr, pi-diicni dI'mL no[ iuri e ;, niiich inI the :a,c of *,ranr tfllilj[ PWICEclita i'o.Ie dc%ok'.ed fumnctionw. Alfl'e .t in 1I-An-fLincLd 1 E h r the kILLI- un crninin,, thL d % IACon pro..es; IHabitu. in ni fI . * 1 condlitin" C:ll?J 1 GEun[lnn-. I()FC F.A\o'rld Baill I 11, p. 211. 86 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS shown in the overview among projects with a high level of participation, 68 percent were highly successful. But of those projects in which there was little beneficiary involvement, only 12 percent were highly effective. Another important finding was that not only was participation crucial to success, but also that it could be encouraged, or discouraged, through When agencies sought project design and government action. Implementing agencies which participation as a goal, actively sought to involve intended beneficiaries had 62 percent of their 62 percent of projects projects achieve success (figure 4.1). In contrast, government agencies were successful-when that did not actively seek beneficiary involvement ended up with less not, only 10 percent. effective projects and only 10 percent were highly successful. Another study found that investment projects have been more effec- Figure 4.1 Participation and tive in countries where citizens enjoy civil liberties (Isham, Kaufmann, Project Success and Pritchett 1997). This is not a measure of political democracy but of . . Successful projects people's freedom to express their views (free press, freedom of association (percentage of total) and assembly, freedom to petition governments). Civil liberties matter 70 even among projects assessed on the economic rate of return-projects that might be thought to be entirely technocratic matters for engineers 60. and economists and insulated from popular pressures (appendix 4). Indeed, the probability of a project failing (those with expost economic 5o returns below a cutoff of 10 percent) is 50 percent higher in countries with fewer civil liberties (figure 4.2). 40 Donors are moving to encourage beneficiary involvement and local 30 ownership. To that end, they have embarked on numerous innovations, including social investment funds. These funds do not specify a particular 20 set of projects by sector (a road project, say, or new schools in particular vil- lages). Instead, they specify processes whereby communities can apply for ±o funds for any project they wish to undertake, subject to reasonable condi- tions on cost, need, sustainability, community contribution, and so on. 0 Participation Participation Such funds have been created in more than 20 countries, and a preliminary not a goal an important evaluation suggests that they have been reasonably successful. They force goal changes on donor agencies from project to process: the fund must specify Source: Isham, Narayan, and Pritchett 1995. a process for funding projects and assess the likelihood that this process will produce projects that generate sustained benefits. But the evaluation found that even social funds maintain features that are an end run around insti- tutional weaknesses and hence are only a temporary solution. Donors are also becoming more flexible about allowing midstream adjustments in projects and encouraging "structured learning" so that development objectives are met. An example is the World Bank's support to Brazil's PROSANEAR sanitation project. Sponsored in part by Caixa 87 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Economica Federal, this project uses an effective sewage collection strat- egy. "PROSANEAR boasts a high degree of community participation and shared responsibility. Communities are increasingly involved in monitoring household use and system performance, and in managing Civil liberties matter for their own repairs. The project's most striking feature is Caixa's commit- economic projects. ment to adjust design and works according to the lessons of experience" (World Bank 1995a, p. 6). Figure 4.2 Civil Liberties and the Probability of World Bank Project Public Sector Compensation and Performance Failure Probability of failure The approach to aid that emphasized the quantity of investment and so technocratic skills tended to underestimate the power of incentives in determining the actions of public sector providers. In the past donor and government technocrats assumed that once power plants were built, they 40 would be operated effectively; new roads would be maintained; health clinics, once constructed, would provide quality services. In reality ben- __ efits will flow only if there are adequate incentives for (mainly civil set- 30 vice) providers. Both bilateral and multilateral donors now recognize that incentives are crucial for success and have been providing support for 20 reforms in public sector management and the civil service. Many poor countries have found it hard to maintain a competent, effi- cient, and honest civil service because real public sector wages have tum- bled in the past 15-20 years. Underpaid (but not especially overworked) and with their morale at rock bottom, civil servants, especially high-level Fewer ones, have turned to moonlighting and corruption (World Bank 1995b). More civil ewer civil By 1986 the real average wage in Tanzania, for example, had fallen to a liberties liberties fifth of its 1969-77 value. The wage of top civil servants had fallen even Source: Ishain, Kaufmann, and Pritchett 1997. further: to just 6 percent of its previous level. The ratio of top level pay to the minimum wage was only five to one. Pay is not always the prob- lem. Even in countries where civil service pay is adequate, the structure of incentives is not. Guarantees of job tenure and lockstep pay rises based on seniority encourage none but the workaholic. Especially in poor and aid-dependent economies, donors have some- times done more harm than good. In their efforts to attract the best peo- ple for projects, donors can unwittingly deplete the civil service of its best and brightest, offering salaries and working conditions that governments cannot match. An independent study reported that for an agriculture pro- ject in Kenya the World Bank hired eight local people at salaries of $3,000-6,000 a month; of the eight, all but one was recruited from the 88 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS civil service, where the monthly salary for a senior economist was roughly $250. In Mozambique international organizations were paying five times the civil service wage for professional staff and 10 times the civil service wage for technicians (Fallon and Pereira da Silva 1995). The combina- tion of low civil service salaries and competition for skilled personnel When donor-financed among donors makes the strengthening of the civil service through train- projects fail, it is often ing a Herculean task. Workers leave as quickly as they are trained. One because of weak program in Kenya funded by the Canadian International Development institutions and public Agency trained 13 economists to the master's level; within a year, 10 had organizations. found jobs outside government, and the other three were searching (van de Walle and Johnston 1996). Civil service reform, too, often denudes the public sector of its most competent staff. Reforms usually include downsizing (jargon for elimi- nating redundancies). When this is "voluntary" (employees are offered a severance package, usually more generous than mandated by law), the most skilled are the first to leave. But downsizing does not have to be a problem. In 1991 Peru announced two staff reduction packages for its civil service. One was an across-the-board offer taken up by 250,000 workers, of which 163,000 had to be rehired later. In contrast, the down- sizing in tax administration targeted employees by skill level and was able to reduce staff by two-thirds while raising wages and doubling tax rev- enue (Haltiwanger and Singh forthcoming). Effective Public Institutions When donor-financed projects fail, it is often because of weak institutions and public organizations. There is plenty of evidence that institutional capability affects overall economic growth and the success of investments. One recent study found that over the 30 years to 1994, coun- tries with sound policies and capable and effective government institutions grew at 3 percent per capita each year-while those with sound policies but weak institutions grew at only 1.4 percent a year (World Bank 1997a). Using an index based on various dimensions of government perfor- mance, a recent staff review of World Bank-financed projects found that, whereas the average success rate on Bank projects was 68 percent, pro- jects in countries with sound policies and capable institutions had an 86 percent success rate. Countries with neither sound policies nor able insti- tutions had failures in 52 percent of projects, nearly four times the rate in countries with good policies and governance (see overview figure 5). 89 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY This study concludes that: "in settings where policies and institutions are seriously weak, the Bank would do best to focus on policy-based lending and non-lending services that support the strategic policy-making and implementation capacity of government"(World Bank 1997a, p. 13). The design of projects Would better-designed official development assistance mean more needs to adjust to the effective public institutions? Because aid has gone almost exclusively to reality that money or (or through) central governments, it has affected how public services are capital stock is less delivered. In general, government-provided services are those that are dif- important than good ficult to allocate via markets, either because they are pure public goods institutions or better that benefit all (national security, rule of law) or because they involve ideas. external effects that are hard to value (health services, environmental pro- tection). The nature of these services, however, implies problems in designing lasting institutions and incentives for efficient public sector delivery and performance. In the interests of effective implementation, donors often establish quasi-independent project implementation units outside the line min- istry (and sometimes outside the government). Donors do this to "cocoon" the projects they finance from the worst inadequacies of the public sector. When projects are "successfully"concluded and turned over for operation to the regular line ministry, they lack the commitment, competence, and resources needed to continue. While isolating projects can at times make sense for purposes of piloting, demonstration, or eval- uation, narrowly measured implementation "success" that comes at the expense of institutional capacity building generally is a Pyrrhic victory and inimical to the true success of aid. The design of projects needs to adjust to the reality that money or cap- ital stock is less important than good institutions or better ideas. The point of an education project is not to increase funding for the sector (this can be done without projects) but to help reformers change the ideas, institutions, and policies in the sector. A truly effective project is a bun- dle of activities that does not just build schools but, more important, helps to change how schools are run to provide high-quality education. Building the finest schools, hospitals, or roads is pointless if the institu- tional capacity does not exist to maintain and run them. This means that the most useful projects will often be innovative. If a country is competent at building schools and running them, donors should simply provide general budgetary support. The only rationale for a project is that things should be done differently than they are now. If existing schools are ineffective, a useful project might rehabilitate them and help 90 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS alter institutional arrangements (for example, increasing community input into decisionmaking). Such a project may draw on things that have worked elsewhere, but it is innovative in the setting where it is being implemented. An important corollary is that the success rate of financed projects is not particularly relevant-if success is narrowly defined. An effective If projects are deemed agency may finance a lot of innovative projects, and some of these may "successful" because "fail"-in the sense that they do not lead to better services. The impor- they merely replicate tant question is whether developing countries are systematically gaining past success, agencies knowledge from such experiments. Failed projects can often teach as much have the wrong (or more than) successful ones. If projects are deemed "successful" because incentives to provide they merely replicate past success, agencies have the wrong incentives to effective assistance. provide effective assistance. Managers will want to avoid risky, innovative projects in favor of things that are known to work. This leads to unseemly and counterproductive competition among donors to "skim the cream" and finance only those items in the public investment program that are likely to be successful-with or without donor involvement. In response, donor agencies have moved to broader measures of project success. They typically assess the effect of projects on the institutional capacity of sector agencies. Take a road project. Are better technologies for designing and building roads being introduced? Will pricing and other policies ensure maintenance? Are agency staff receiving necessary training? Assessing whether there is a "substantial institutional effect" is essentially asking a host of questions about whether the project is helping the recip- ient country to change the way that it manages its road sector. Note that a project could have a successful "outcome"-the road is built and has a high return-but no lasting effect. On the flip side, a pro- ject could have "failed" but have led to substantial institutional develop- ment. This is particularly true of innovative approaches to service delivery, to which a lot of aid now goes. The innovative approach may not work, but if it is systematically evaluated and the knowledge is fed into a broader reform program, the project is helping improve the man- agement of the sector. This kind of serious impact evaluation has often been missing from devel- opment assistance, yet it is potentially of the greatest value. Again, innova- tions may fail, but development assistance is strengthening the underlying institutions by assisting in the design and evaluation of new approaches and generation of knowledge. Increasingly, donors are emphasizing the crucial importance of this kind of knowledge creation. Witness the United Kingdom. A recent White Paper on International Development says: 91 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY "Research is an important weapon in the fight against poverty. Without research, many development interventions would fail or be much less successful; and research has significant multiplier effects-solutions to the causes of poverty in one part of the devel- Without research, many oping world may well be replicable in another. The principle of development shared knowledge is an important component of the partnerships interventions would fail which are essential to development. The Government sees the con- or be much less tinued investment in knowledge generation as a key element in successful. achieving its aims and objectives for international development" (U.K. White Paper 1997, p. 48). At the World Bank, each completed project is rated on the basis of whether it had a "substantial institutional development impact,"and this is viewed as the "most important evaluation criteria for long-term devel- opment effectiveness" (World Bank 1997a, p. 24). However, this compo- nent is also difficult. For a long time the share of World Bank-financed projects with a "substantial"institutional impact hovered around 30 per- cent, and only in recent years has it climbed to 39 percent. Thus it is pos- sible for development projects to help change significantly the way the public sector does business. But it still happens in too few cases. In our view, an effective development agency will be taking risks, sys- tematically evaluating the outcomes, and disseminating the knowledge gained. This generation and dissemination of knowledge is one of the biggest contributions that development assistance can make. In tandem with the old rationale for aid, in which donor financing addressed market failure in capital markets, donor activities need to address the market fail- ure in "knowledge" markets (Stiglitz 1988). But serious, rigorous evalua- tions that generate solid knowledge are expensive, and no one government has the incentive to undertake evaluations that will benefit other countries. Public Provision without Public Providers T O MAKE AID MORE EFFECTIVE, DEVELOPING COUNTRY GOVERN- ments are increasingly turning away from the traditional (and exclusive) use of public bureaucracies to provide services. Many public services can be provided effectively (often more effectively) by pri- vate organizations under contract or holding a concession. Donors and 92 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS governments have often turned to nongovernmental organizations. New sources of financing have also brought about important changes in providers' incentives, empowered intended beneficiaries, and so increased the effectiveness of aid. Many public services can Contracting and Concessions be provided effectively (often more effectively) Privatizations and divestments have reduced the drain on government by private organizations revenues and capacity in areas where public sector involvement is not crit- under contract or holding ical. But even when government retains responsibility for services, they a concession. need not be managed through the usual public sector agency. Many activ- ities can be contracted out to the private sector, and competition among potential providers can create pressures for cost-effectiveness. How does it work? Take AGETIPs (Agences d'Execution des Travaux d'Intert Public), not-for-profit associations that enter into contractual arrangements with governments to carry out infrastructure projects. The first was in Senegal, which (by contracting with individual suppliers through open bids) was able to reduce costs and delays. Other West African countries soon adopted this model (World Bank 1997c). A concession is a way to contract with private providers for the provi- sion of services without relinquishing public ownership of assets or pub- lic control of service delivery conditions. In Guinea a contractual arrangement under which a private management agency took over the operation of a water system was set up with the help of an International Development Association loan that initially paid the difference between costs and the revenues recovered from users. The subsidy was gradually reduced and the operation is now run commercially. This shows how aid can smooth the introduction of new arrangements. As important, donors can relay positive (and negative) experiences to other aid agencies and countries contemplating similar arrangements-be it power deregulation in Chile, water concessions in Argentina, or toll roads in Mexico. Disseminating experience and lessons from innovators is especially important in improving the effectiveness of aid. Take railways, a clear example of diffusion of information about innovations in management. In developing countries railways are often run down and costly, with out- dated rolling stock-all thanks to inefficient public management and competition from roads and airlines. Argentina was an innovator in rail- ways. It split its massive federal system into several freight and commuter rail networks and awarded the rights to run them to private firms, some- 93 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY times through "negative"concessions (where firms bid the lowest amount the government would have to offer them in subsidies). While some of these concessions have had problems, labor productivity quadrupled, prices fell, and the public sector saved $600 million (Thompson and Budin 1997). More important for aid, development agencies helped ensure that the pitfalls and pluses of Argentina's experience were incor- porated into programs in Brazil, Chile, Mexico, and countries in Africa, which have all moved to concession railways. NGOs as Service Providers The use of nongovernmental organizations (NGOs) as implementing agencies for donor-financed projects is another rising trend (box 4.2). Many NGOs also receive government contracts to provide services. NGOs have advantages in providing some services, over both govern- ment agencies and more profit-oriented suppliers. NGOs can often reach local and target groups more effectively than can traditional government agencies. But NGOs are no cure for the shortcomings of the public sec- tor. Moreover, external NGOs often have the same problem as donors in Box 4.2 A Useful Surge in Nongovernmental Organizations A 1b F T F'RI L)nJ- d ' I .EcoN.u NG0. [he Japan V id Bird Sociirv. to conduct an Coo011eraion Fund (OECF I noE: that the prcad enironmentad §urnc. on the hibiar of cranes a, of denocra,% and ihc broader pohual paracipa- pjrc If E hc OECF special fsistance [or proiect non rh.a hL c..m- with the end of chc (,old Var I'rnition for chc D,:,,elopinent Proiecc oF hai tmulatd NGO actuvi chroughmut the AgrIEukLVe on Sat Farm' in Hcilongitang ' orld. lhe pivnrem of both %c-loping ind Provin,c in (ina ianiian5 Plain). F-r ch, indu trial colnris haic a rclE l,,,-ood undcr- lamiund NlnLUpurpose Bridge Project in "[andmgof the l.rge role that NG( )s can pl in a Bangladesh. a local NGO is condlc]ung the ba'ic councr, dclopmctenr. The OE(CF has 'ollabo- sun e. on sit Lpropirianon. The r hae alio been rated -%iEh local NGO on preinplemcenranon nuirierous instanc, in which local and interna- studi and pr ject implemenration for sccral pro- nonal NG0, have conirbuted %aIlLable advice on lcts , InclUding~ the Arivallh Hilk Affore,,Eauo~n sector sur' c-Vs. prciniplci-inuarion studies, and PrTicc in Ind,a and clic Fori Nccrn Proiect in [le project appraisal .ind inipcmentaEion (DECF Philippins It ha ako commissioncd a lapanese 1 lo. p. O 94 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS creating sustainability and local ownership (World Bank 1998 a). So, while NGOs can be part of the service delivery system, they cannot replace gov- ernment and cannot be a permanent substitute for public sector capacity. User Fees, Effectiveness, and Demand-side Financing In recent years donors have financed innovative Government services (particularly those financed by aid) are often pro- ways to increase service vided free to users-health care, irrigation, water, extension services, effectiveness. schooling, roads, and so on. In many cases this approach is socially desir- able but raises three problems-two serious, one less so. Least important, free services tend to be overused, clogging the system-be it a road net- work or health clinic. The two more severe problems are institutional and closely related to aid. First, since the service does not generate revenues, its provider depends on the general treasury for funding. Especially in poor countries with chronic or recurrent fiscal crises, that is cold com- fort. There is little assurance that funds needed for, say, medicine or irri- gation maintenance will be forthcoming. Second, because funding comes from the treasury, it is difficult to empower beneficiaries or instill in them the sense of ownership so necessary for the success of aid projects. Users are frozen out, in setting priorities and planning and delivering services. In recent years donors have financed innovative ways to increase service effectiveness by creating charges directly linked to improvements in quality -from irrigation to education, health to agriculture. Health is a good example. Providing universal health care in poor countries is a laudable aim. But because of insufficient funding, morale is in the depths, and more important, clinics often have no needles, drugs, or medicine. That leads to a vicious cycle: patients, knowing the poor quality, seek care elsewhere, leading to underuse of clinics, which makes it harder to justify increased budgets to improve services (Filmer, Hammer, and Pritchett 1998). In 1988 the Bamako Initiative sought to reverse that cycle by intro- ducing user fees in Cameroon. A project financed by the U.S. Agency for International Development carried out a controlled evaluation of the combination of charging user fees and using them to improve quality by creating a reliable drug supply. The study found that not only did the use of health centers increase but that there were proportionately more poor users than rich (Litvack and Bodart 1993). Similar experiments have had similar results in other sectors, such as education (Birdsall and Orivel 1996). Examples like these reveal the valuable role donors can play in financing and evaluating innovations in service delivery. 95 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Making Aid Work for Better Public Services T HE SHIFT IN DEVELOPMENT THINKING FROM A GOVERNMENT-LED accumulationist strategy to a focus on fundamentals, effective- The shift in development ness, and efficiency in core public sector responsibilities requires thinking requires reorienting the instruments of aid. In particular, it means reconsidering reorienting the the methods of financial (project and program) and nonfinancial (tech- instruments of aid. nical assistance, policy analysis, training) assistance and how they can be adapted to support the new development strategy. Project Finance We have not yet discussed a topic that accounts for most of the analy- sis of aid effectiveness: the technical and administrative details of making aid projects more effective. There are two reasons for this. First, and most important, this area has been well covered in many past evaluations, reviews of experience, and research, and we have little of value to add. Many donor agencies review the performance of completed projects. The combination of individual project experience and comparisons across countries allows these evaluations to provide critical and constructive sug- gestions from the lessons learned to improve the organization's opera- tions. In addition, some donor agencies' technical departments and evaluation departments produce sector specific reviews of both the prob- lems faced and promising approaches in each sector.' This voluminous literature on project performance and sectors provides a wealth of infor- mation needed to implement successful projects. The second reason that we do not go further into the issue of effective aid projects is that we are more interested in the indirect than the direct effects of projects. Aid-financed projects can have spillover effects from individual projects to the general operation of the public sector. Almost from the beginning, projects financed by foreign aid were thought to have important indirect and broader benefits. Hirschmann (1967) argued that perhaps the greatest were institutional and personal learning from the conception, planning, and implementation of projects. He believed that projects were the "privileged particles" of development that created innovation by confronting problems as they arise. These ben- efits emphasize the value added of donor involvement, which allows a supportive "infrastructure" of institutional and technical support for implementation that would otherwise not be present. 96 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS Development projects can be a testing ground for ideas or concepts that are new to a country, demonstrating to the government and citizens what does (or does not) work--contracting out public services, using NGOs, involving user groups in management, and so on. The value added of donors is twofold. One is the ability to devote resources to careful evalua- Development projects tion of experience. Knowledge of this type is an international public good, can be a testing ground and no government has sufficient incentive to evaluate or disseminate it. for ideas or concepts Donor projects can also help break the mentality that locks the public sec- that are new to a tor into ineffective arrangements. Reform-minded governments find it country. hard to implement new ideas-especially those that challenge entrenched interests, particularly if there is no guarantee that they will work. That they will work can often only be proven by implementation. Donors can break this vicious circle by helping to implement such projects. In a joint meet- ing of officials of the Gambia and the World Bank to improve the working relationship, the Gambians expressed "the importance of the Bank's role in bringing experience and ideas from outside as the strongest plank in the partnership. Financing was vital, but more because it represented the pos- sibility to do-moving beyond talk" (Marshall 1997, p. 26). But the benefits depend on the project being implemented in a way that is different and on the evaluation of this difference. Unfortunately, both of these requirements are rarely met, as donor-financed projects often do not provide mechanisms for rigorous evaluation of outcomes- or did not until recently. Projects also present opportunities to engage in a sectorwide dialogue with government. Many investment project failures are linked to sector policies of government that often conflict and work at cross-purposes. For instance, in some countries agricultural extension, irrigation, and transport projects intended to encourage farmers to grow specific crops have been overwhelmed by macroeconomic and microeconomic poli- cies that effectively resulted in heavy taxation on agriculture. In other countries investments in roads, irrigation, and other infrastructure have been poorly maintained (or not maintained at all) because public agen- cies could not recover costs and were starved of funds. Changes in gen- eral policies that result from dialogue improve the sustainability of donor-financed projects and improve prospects for all projects in the sec- tor. This effect can be crucial value added from donor involvement. Just one example comes from the difficult area of resettlement caused by land acquisitions in public investments (see box 4.3). Because projects can have both a direct impact (a new road links communities, carries 97 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY people and freight) and spillover benefits (improving institutions and sec- tor policies), both need to be evaluated. Nonproject Finance Donor financing that is tied less to the implementation of particular projects-such as program aid, sector investment financing, sectoral adjustment programs, or other time-slice finance instruments-has advantages and disadvantages over project finance. In a country where spending allocations are sound and the effective- ness of expenditures is high, the potential value added of donor involve- ment in project design and implementation is minimal. Nonproject financing saves on project preparation and implementation costs, both for donors and recipient governments. One scarce resource of donors is staff with technical skills and experience from a range of countries. Since project preparation is staff intensive, however, using projects as a form of support in well-managed countries does not necessarily allocate staff where their skills are best used or needed. Scarce skills will probably be put to more productive use elsewhere. For recipients, nonproject assis- Box 4.3 Resettlement in Development A Q ii.11 .L. !\ E iL . '. If 1 1iL \\VC- LL' ' ,:n bo_-rrwc. f do not genaind- concur %vib the Bank, c%pcrience r wiEh irnplementing 1i rccEilc- Barl: , r,LtlLment p ',.nit m ch, .'um . resct- lm c n -rh nor L rricd out well- 111cr' pui ilc i In rt pr.lcct' concid s that: rc'.zdLc 01 B n I: mo ;i:-n or iThc frcqucnrv% .-C f-latlll I lLn tih r CiB . 1 d I Iil F . ank thrc.at t Li U Jpnd d. Ibur,cmnin The Biink ha, bLcr tar n.,.rc effccrti.c c r- 'aop . .tn-airid inicdilac II'L'pc-n h a,c bcnct`ircd la pr*irtICd dUi hi IcL v irk Ehc B-rrtt' v.ILI-,h. ' . nFic-' n if I-n rec-ied .il,rcrment Ila.1h bor- p r f..'i,r, uldL 1n. I luncar- populm J,l *r-.lJ.,- in -C rncr rir - h e bic., n d meith O- nlr.1i. On in-iIan c t t L[h Brflflic n f r r d d teIi r lip'lL 1-',o no i-I iN that 'CIcral B rr Bank- .tr n e!-nd Ec ' s o m ,. than %1h.n it effl rts %kere in.nipwr c.%sled d.-mcr1c! lvdiclC irid C Ic Inc t .,I iercceiic n [,,r ind' idua- prw- c i!. I1 curri. ch 10I i .Es.l [ .il 0v1n in c T tn1c ent%A.1-irk; q h' n g'..crnr.iu-. tant V l tt hnd'hJ-I it ltar .i5tcCracnl a and the agreed u-, %..0,rk 1 l dil ai --a% L'tc niafi ll?T L%:ple1 thll ir C01 t11 ur-l)[S to p,sJ r,cctdcrocfr r Cc b i t iung pr*.iLtc p h *Ic h. _,..ntinmc - hormed the ha.ns ciLL ti rti[.. 1 MI 'it.-.a cap-. d cifif d at the - ')f discissins and iipr-,in: niore genebrar (do01nc Ec C0,J told pihe rimc% .o iWorld Bank Itij., p I J 98 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS tance helps with donor coordination. The administrative setup needed to monitor perhaps hundreds of donor-financed projects is burdensome to governments with limited public sector management capability. Often each donor and each project has its own procedures, plans, and report- ing and procurement requirements. Project financing brings Project financing brings its own biases into the incentives of aid agen- its own biases into the cies: favoring investment over recurrent spending, large projects over incentives of aid small ones, expansion of physical production over improvement in effi- agencies: favoring ciency, imported materials over domestic. Recipients often see projects as investment over being generated and implemented by aid agencies alone, with little local recurrent spending, large ownership. projects over small. The use of general budgetary support delinks projects and implemen- tation from overall targets for assistance-that is, the size of projects is often scaled to meet targets for total assistance disbursed rather than to maximize the value added of projects. At times the scale of the project works at cross-purposes with its intent. But all of these considerations apply mainly to those lucky few coun- tries and sectors in which public services are already effective and, per- haps, even efficient. If one of the main concerns of development is to improve policies and increase government's institutional capacity (and it should be), foreign assistance that simply expands the budget envelope of existing activities without changing the structure or incentives for ser- vice delivery is unlikely to see much of a payoff. The question is: What kind of support will best lead to public sector improvement? Sometimes projects are necessary. At other times a policy dialogue-connected to program assistance-is needed. And sometimes no financing at all is the most useful. An excellent study of aid in Africa sponsored by the Overseas Development Council, based on collaborative case studies in seven African countries, makes this recommendation: "[Program aid] is not likely to promote development in the absence of sound economic policies. In such situations, donors should maintain a policy dialogue with the government, but limit aid flows and direct them to project assistance, often focusing on non-state actors. When sound economic policies have been put in place, donors should expand program aid, perhaps in the context of sec- total investment strategies negotiated with the government" (van de Walle and Johnston 1996, p. 8). 99 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY This recognizes the complex factors in the choice of the method and level of development assistance. Effective aid depends not just on the type and method of aid chosen, but critically on government policies, institu- tional capability, and commitment to improve services. A society (or government agency) that has Nonfinancial Support generated its own reform program is receptive to Nonfinancial support can have a high payoff, but its record has been technical assistance and mixed. Lack of institutional capacity in developing countries has long institution building. been recognized as a big hurdle to development.2 One traditional solu- tion has been to use three types of technical assistance: employing foreign experts on a short- or long-term basis, training government officials (in country or abroad), and financing long-run educational programs. While donors have made much use of foreign experts to improve institutions (and there have been successes) the overall results have been disappoint- ing (World Bank 1996b). A United Nations Development Programme evaluation in 1993 suggested that "there is a growing sense that techni- cal cooperation does not work well, that as presently practiced it is inef- fective, that such benefits as it brings are extremely costly, and that in any case it has little lasting impact" (Berg 1993, p. 3-4). A society (or government agency) that has generated its own reform program is receptive to technical assistance and institution building. When not driven by domestic demand for particular expertise, however, foreign experts are often not integrated with ministries in a way that allows knowledge to be transferred. Free-standing technical assistance provided by international consultants and individual technical experts has served some purposes, but as a way to promote and improve public sector institutions they have usually been expensive failures. Training government officials in technical tasks has also had mixed results. In many countries training has been unsuccessful because public officials do not have the incentive to perform, are politically blocked from performing, or do not have the materials or resources to perform. In many countries there is no way to speed up the slow process that involves cre- ating not just the capacity for policy analysis and dialogue (both in gov- ernment and civil society) but also the demand from recipient governments for productive technical assistance. The impact of ideas is difficult to monitor, evaluate, and especially quantify because the ways in which policy analysis contributes to future performance are often intan- gible and indirect. 100 AID CAN BE THE MIDWIFE OF GOOD INSTITUTIONS One study undertook an empirical analysis of World Bank projects (mostly approved in the 1980s) to investigate the relative contribution of two staff inputs (Deininger, Squire, and Basu 1998). One is the input directly related to the project-preparation before the project begins and supervision while the project is being implemented and the finance is being The payoffs for analytical disbursed. The second is "economic and sector work,"which produces work can be enormous. reports on the economy and particular sectors for government but also cov- ers the dialogue with government and other development agencies. Figure 4.3 Cost and Benefit of an The study found that (even after controlling for country, sector, and Additional Week of Analytical Work economic conditions and staff preparation and supervision for a partic- Thousands of dollars ular project) prior analytical work improves projects. On average these $ $27,000 activities have a high payoff, as the benefit of one additional week of ana- lytical work by the World Bank is nine times the cost-including staff time, travel, and overhead (figure 4.3). Moreover, since analytical work affects many projects, the overall benefit is even larger. And these are just $3,000 the payoffs to projects financed by the Bank. If the changes made from 0 - policy analysis affect other donor-financed projects, or perhaps even all Average Benefit cost government projects, the returns to involvement in nonlending activities would be enormous. Source: Deininger, Squire, and Basu 1998. While analytical work has a high return, the same authors found that the World Bank underinvested in it, and overinvested in resources dedi- cated to ensure that a steady stream of projects went to the Executive Board (appendix 5). According to their estimates, a shift of resources from project preparation to analytical work would have led to lower commit- ments but a higher success rate of projects. In other words, fewer projects would have been approved, but they would have been better designed and better implemented. Ironically, a shift of resources from project preparation to analytical work would have led to higher-not lower- disbursements of funds. The reason is that problem projects are imple- mented slowly and hence disburse slowly. Methods of Assistance and Public Sector Efficiency Properly managed, aid can encourage better public sectors through both project and nonproject activities (and often through a combination of the two). Both can provide the knowledge countries need. Projects sup- port experimentation and reform, demonstration, piloting, evaluation, and innovation. There are also several ways in which donors can assist countries through nonproject activities, such as by creating and diffusing 101 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Properly managed, aid information drawn from their experience. Since aid agencies implement can encourage better many projects in different countries and with different institutional struc- public sectors through tures and have an established capacity to evaluate projects, they can draw both project and on cross-cutting evidence and experience that no single country could. nonproject activities. Policy analysis, bringing technical expertise to bear, can be useful. Training-that is, enabling domestic actors with the necessary capabili- ties to accomplish their objectives, whether by exposing policymakers to other experiences or providing new skill-also plays a key role. Notes 1. Even a partial list of the literature produced by just the World Bank gives some indication. From the Operations Evaluation Department have come reviews of, among other sectors, irrigation, agricultural extension, and adjustment lend- ing. Sectoral policy papers reviewing the experience with Bank involvement have been produced on, among others, energy, water, education, and adjustment lending. 2. Krueger, Michalopoulos, and Ruttan (1989) note that in the 1950s the U.S. Agency for International Development was home to a huge debate between advocates of capital assistance and advocates of technical assistance. 102 CHAPTER 5 Money, but More Ideas, Too OOD ECONOMIC MANAGEMENT MATTERS MORE TO The international developing countries than foreign financial aid does. community needs to get Policy and institutional gaps hold back economies beyond projects and spur that lag behind, not financing gaps. Aid money has a systemic change in big impact only after countries have made substantial whole sectors and progress in reforming their policies and institutions. Countries. These messages should not, however, be twisted to mean that countries with weak institutions and policies (call them difficult, weak, or troubled environments) cannot be helped. They can. Many things can be done usefully. But first we know from experience what is unlikely to work: a Large amounts ofmoney. Providing significant financing has not made much of a dent in poverty in countries with weak management. " Buying reform. Conditional lending has not generated reform in coun- tries with no domestic movement in this direction for reform. . Focusing on individual projects. In difficult environments donors have often fallen back on a core program of projects, mostly in the social sectors. But the success of individual projects (always difficult to achieve) does not have much impact unless it spurs systemic change. And because money typically is fungible, the whole (distorted) public sector is being financed, not just the favored sector. So, the international community needs to get beyond projects and spur systemic change in whole sectors and countries. Easier said than done. Almost nothing positive has happened in Myanmar (Burma) or Nigeria in the past three decades. There obviously is no panacea for the ills of such countries. Even so, there are examples of successful assistance 103 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY that has improved the lives of people living in countries with weak gov- ernance and poor policies. In some cases this assistance has contributed to a larger transformation that made a major difference. Here we present four case studies of effective aid under difficult con- In difficult environments, ditions. Running through them are four themes: effective assistance is more about ideas than . Find a champion. Countries, governments, and communities are het- money or projects. erogeneous. While it is fair to characterize Burma overall as "poorly managed," there are likely to be reform-minded elements in the com- munity and even in the government. If aid can find and support these reformers, it can have a big impact. m Have a long-term vision ofsystemic change. Successful reformers have a vision of how things could be different in, say, 10 years-different both in outcomes (more kids going to school, graduating, getting jobs) and in process (community involvement in schools, broad public sup- port for reform policies). n Support knowledge creation. While reformers typically have a long-term vision, they often need to develop the details of reform through inno- vation and evaluation. Moreover, for reform to take root requires a demonstration that it works. Financing and evaluating innovations is a key role for development assistance. m Engage civil society. In highly distorted environments the govern- ment is failing to provide supportive policies and effective services. That is why government-to-government financial transfers pro- duce poor results. Effective aid in such an environment often involves supporting civil society to pressure the government to change or to take service provision directly into its own hands (or to do both). The foregoing points deal with the characteristics of promising envi- ronments for change and the ways donors can enhance those character- istics. The lessons from successful assistance also provide guidance about donor behavior. Aid is more effective when: m Agencies focus on long-term reform. Successful aid in difficult environ- ments typically involves intensive staff input from donors and small disbursements of money. It also goes "beyond projects" to support sys- temic reforms. In difficult environments, effective assistance is more about ideas than money or projects. 104 MONEY, BUT MORE IDEAS, TOO m Donors work as partners rather than competitors. Studies of aid have long pointed to the proliferation of donors and the lack of coordination among them. Well-managed countries force coordination on donors, but in the weak environments donors often run amok. It is hard to explain this, except that different donors like to "plant their flags" on Well-managed countries something (anything) tangible. In cases of successful assistance, there force coordination on tends to be a strong partnership among donors with a focus on larger donors, but in the weak transformations, not on individual flags. environments donors often run amok. Vietnam: Adjustment without Lending IT WOULD BE HARD TO IMAGINE A TOUGHER ENVIRONMENT FOR development assistance than Vietnam in the mid-1980s. One of the world's poorest countries, Vietnam had to appeal for international food aid at several points in the decade. Despite Soviet aid (10 percent of GNP each year), there was virtually no progress in development (in poverty reduction or improved social indicators). Economic management was poor. The trade regime was closed. There was no scope in the econ- omy for the private sector. And a large fiscal imbalance was financed by printing money, leading to hyperinflation. Vietnam in 1985-86 is a good example of how a large amount of finan- cial aid can have no impact in an environment of distorted institutions and policies. There was widespread dissatisfaction within society, government, and the ruling communist party with these poor results. At a landmark meeting in 1986 the party decided to change direction and initiated doi moi (renovation). One of the leading champions of reform was Vo Van Kiet, who emerged as deputy prime minister responsible for the economy. The impetus for change came not just from the fact that Vietnam was doing poorly, but from the success of nearby economies that were more open to foreign trade and investment. Although the economic team had a general sense of the direction for economic policy-toward more private initiative and greater openness-it did not have a detailed reform program (Ljunggren 1993). The Soviet Union's aid dropped sharply in the late 1980s, leaving Vietnam with only limited assistance (less than 1 percent of GDP) pri- marily from the UNDP and Sweden (SIDA). While Vietnam was grap- pling with classic adjustment problems, it was receiving no financial 105 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY support from the IMF or the World Bank because of its political estrange- ment from the major shareholders of these institutions. The two were, however, contributing some staff time to advise the government. Furthermore, UNDP and SIDA invited the Bank and the Fund to man- Vietnam's reform opened age some of the technical assistance they financed. So, the IMF managed more space for civil UNDP-financed technical assistance for the central bank and con- society, and donor tributed some of its own staff time. In a similar arrangement the World activities encouraged Bank managed UNDP-financed assistance for the State Planning this as well. Committee (now the Ministry of Planning and Investment). In the early stages of reform this partnership of agencies provided important inputs. In 1991 the UNDP and the World Bank organized a meeting in Kuala Lumpur at which Vo Van Kiet and his team met with economic ministers from Indonesia, the Republic of Korea, and Malaysia. Each country laid out some key policies that had worked for them, and the Vietnamese team asked detailed questions about stabilization, trade liberalization, foreign investment, and other economic policies. At a more technical level, the same group of donor agencies organized training courses, policy workshops, and study tours for staff from Vietnam's eco- nomic ministries. All this was aimed at helping Vietnam learn about its neighbors' expe- riences and develop the knowledge it needed to manage its own economy successfully. The generation of knowledge was supported at a more micro level, too. UNDP, SIDA, and the World Bank teamed to help the govern- ment carry out the first representative household survey ever done in Vietnam-generating a wealth of information about household produc- tion, income and poverty, education, health, and child nutrition. This sur- vey, conducted in 1992-93, revealed that 55 percent of the population lived below a widely used international poverty line (Dollar and others 1998). Vietnam's reform opened more space for civil society, and donor activ- ities encouraged this as well. Two examples: The Asia Foundation and the World Bank organized the first public meeting between the domestic pri- vate sector and leading officials to discuss economic policies and key dis- tortions holding back private investment. And a group of donors organized rural workshops to discuss the poverty analysis that emerged from the household survey and how donors could support local initia- tives aimed at poverty reduction. It is impossible to measure the exact effect of these donor activities. But there undoubtedly has been dramatic improvement in economic policies in Vietnam and that has made people's lives better. Vietnam's 106 MONEY, BUT MORE IDEAS, TOO Figure 5.1 Vietnam: Aid and Policy Aid in Vietnam focused on policy advice and Policy Index Aid as a percentage of GDP technical assistance Good during adjustment and 3 reform. Large-scale finance arrived after a Aid good policy framework was in place. 2 Mediocre Poor 0 1988 1989 1990 1991 1992 1993 1994 1995 Source: Dollar and Easterly 1998. policy improved from truly terrible in the mid-1980s to merely poor by 1990 (figure 5.1). (The improvement captures the strengthening of property rights over agricultural land and the first stage of stabilization, going from 400 percent inflation to about 70 percent.) There was fur- ther steady policy improvement in the early 1990s as inflation was brought under control, trade liberalized, and the legal environment for the private sector clarified. Note that during 1989-93 Vietnam was receiving a small amount of aid (about 1 percent of GDP). Large-scale financial aid did not arrive until 1994-95, by which time the institutional and policy environment was quite good for a low-income country. Vietnam is a good example of systemic change. In the mid-1980s households were restricted in what they could do with land and labor. By the mid-1990s they had considerable freedom to work their land, start a household business, or look for wage labor. Innovations that had worked well locally were "scaled up" to the whole country. With the better pol- icy environment came the need for complementary investments in roads, power, education, and other sectors. In the wake of policy reform, growth accelerated. Of the 40 poorest countries in the world in 1986, Vietnam had by far the fastest growth 107 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY in the next decade. And benefits have been widely shared. A 1998 study of the households first surveyed in 1992-93 revealed that average household income had risen 39 percent in real terms and that the poverty rate had been cut nearly in half-from 55 percent of house- In recent years many holds to about 30 perceht. developing countries What are the lessons for foreign aid? Donors helped a reformer have experimented with develop policies for Vietnam and show that they could work. And the decentralization of clear objectives of Vietnam's leaders helped in coordinating donors. education. The focus during the critical 1989-92 period was clearly on ideas, not money. For reasons of international politics, major donors such as the World Bank could not provide money to Vietnam until 1993, so they had to do something different. Intensive staff time required little money and made a big contribution to the country's reform and development. El Salvador, Pakistan, and Brazil: Education's Decentralization and Reform GOVERNMENTS IN DEVELOPING COUNTRIES USUALLY PLAY A MAJOR role in the allocation and management of educational resources. Even when some authority is delegated to provinces or munici- palities, individual school administrators and parents have a limited voice. This centralized approach has supported many achievements in education, but it has not always reached groups that have traditionally had low levels of education (the poor and girls, for instance). And over- centralization can stifle initiative among those critical in affecting school outcomes-teachers, principals, and parents. In recent years many developing countries have experimented with decentralization of education in two ways: first, allowing more local autonomy and input into public schools, and second, creating more space for the nongovernmental sector. In some cases donors have helped finance and design these innovations. Most important, there has been support for government efforts to evaluate these innovations seriously, following pilot groups and comparing them with control groups not par- ticipating in the pilot. For reforms in El Salvador, Pakistan, and Brazil, there is now solid evidence that they work. El Salvador's community-managed schools program-more popularly known by the acronym, EDUCO, or Educacion con Participation de la 108 MONEY, BUT MORE IDEAS, TOO Comunidad-is an innovative effort for both preprimary and primary education to decentralize education by strengthening the direct involve- ment and participation of parents and community groups. A prototype of today's EDUCO schools emerged in the 1980s when public schools could not be extended to rural areas because of the country's civil war. El Salvador's community- Some communities took the initiative to organize their own schools, managed schools administered and financially supported by an association of households. emerged in the 1980s Although these early attempts were held back by the low rural income when public schools base, they demonstrated a strong inherent demand for education as well could not be extended to as a desire to participate in the governance of schools. In 1991 El rural areas because of Salvador's ministry of education, supported by such aid agencies as the the country's civil war. World Bank, decided to use the prototype as the principal method of expanding education in rural areas through the EDUCO program. Today's EDUCO schools are managed autonomously by an elected community education association (Association Comunalpara la Education, or ACE) drawn from the parents of the students. The ACEs take the cen- tral role of administration and management, and they are contracted by the ministry to deliver a given curriculum to an agreed number of stu- dents. The ACEs are then responsible for contracting and removing teach- ers' by closely monitoring their performance, and for equipping and maintaining the schools. The partnership between the ministry and the ACEs is expected to improve school administration and management by reflecting local needs. The EDUCO program is a way to expand educational access quickly to remote rural areas. But has this come at the expense of quality? No. Community-managed schools are not low in quality. Student achievement on standardized mathematics and language tests is the same for EDUCO schools and traditional ones (after controlling for household background). In fact, EDUCO schools may be superior. They have significantly fewer school days missed by students as a result of teacher absence (Jimenez and Sawada 1998). Based on this success, the government is now planning to introduce community management into all traditional schools. In Pakistan a long-standing problem has been the low participation of girls in schooling. In 1994-95 the provincial government of Balochistan began supporting nongovernmental schools, with an emphasis on increasing the enrollment of girls. Subsidies are paid to new nongovern- mental schools, based on the number of girls enrolled. After three years, the schools are expected to be financially self-sufficient through fees and private support. 109 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY In urban areas a new private school was opened in every poor neigh- borhood randomly selected for the program. Girls' enrollment increased by 25 percentage points, compared with their enrollment in control group neighborhoods that did not participate in the program (figure 5.2). An innovative program to Moreover, even though the program provided a direct subsidy only to the support the start-up of enrollment of girls, it had a positive spillover effect on boys' enrollment. girls' schools in This success was not affected by income or other socio-economic char- Balochistan led to an acteristics of the neighborhood, suggesting that expanding it to other immediate leap in girls' poor areas of Pakistan would likely be successful. enrollment. In rural areas, girls' enrollment is particularly low. So, the program sup- ported the creation of girls' schools, often headed by a female teacher, a fac- Figure 5.2 Pakistan: tor believed to significantly increase parents interest in sending their School Enrollment of Five- to daughters to school. Again, results show that the program significantly Eight-year-old Girls increased girls' enrollment compared with villages without girls' schools. As Percent in the urban areas, there has been a spillover benefit in higher boys' enroll- ment rates. 100 In Brazil's State of Sdo Paulo the new secretary of education, Rose Neubauer, initiated decentralization and improvements in efficiency. The so traditional system involved using the same building for primary school 70 in the morning and secondary school in the afternoon, with teachers 60 commuting among different schools. There were also the usual problems .4s of low teacher salaries and morale. The result: low enrollment rates, many 40 repeaters and dropouts, and poor student achievement. The reforms were threefold. First, the system was rationalized. 20 Remapping allowed the concentration of primary students in primary schools and secondary students in secondary schools. Teachers were thus I iassigned to a single school, reducing commuting time and creating a more Pakistan Balochstan coherent faculty. This freed up resources to lengthen the school day and Average Program increase teacher salaries. The state education budget increased by 10 per- Source: im, Alderman, and Orazem 1998. cent, but teacher salaries nearly doubled-possible because of the effi- ciency gains. Second was decentralization. The state handed responsibility for pri- mary education to municipalities that had the resources, capacity, and desire to assume it. There was a further decentralization of authority to school heads. Previously, these were political appointees not always pos- sessing the requisite skills. In the new system these heads have to take a pedagogical exam and develop an action plan for the school. The third prong of the reform was a commitment to assessment and research-that is, to a "results focus." Serious evaluation enabled the gov- 110 MONEY, BUT MORE IDEAS, TOO ernment to determine quickly that the reforms were working. Within a few years primary school enrollments in municipalities doubled from 600,000 to 1.2 million. Repetition rates dropped from more than 10 per- cent to less than 4 percent. Dropout rates plunged from 9 percent to 5 percent in primary schools and from 20 percent to 12 percent in sec- In El Salvador, Pakistan, ondary schools. The ultimate test of reform will be its effect on student and Brazil decentralizing achievement, and the testing system will allow careful evaluation over the and involving civil society next few years. Early results are quite promising, and teachers are happier led to improvements in as well. Neubauer is the first secretary of education in more than 20 years public services. to go three years without a teachers strike! What can be learned from the El Salvador, Pakistan, and Brazil edu- cation reforms? In each case decentralizing and involving civil society led to improvements in public services-specifically, the broader availability of schooling to disadvantaged groups. A key issue in each case was to take a good idea and see if it worked, to create the potential for systemic change-that is, for the reform of the whole sector. In El Salvador and Brazil the government is moving in this direction. In Pakistan it remains to be seen if the innovation will be widely disseminated. But for donors, these provide clear examples of how assistance can help support reformers-in the initial piloting, in serious impact evaluation, and in mainstreaming if there is demand to "scale up." Moreover, it is possible that these efforts to stimulate decentralization and community participation in education will have larger spillover benefits leading to more community involvement in other areas (roads, health), an impor- tant area for future study. Cameroon: Health Financing and Delivery C AMEROON ENTERED THE 1980s WITH AN ANNUAL GROWTH RATE of 10 percent and exited them with an annual -10 percent growth. Partly the result of external shocks, the volatility was exac- erbated by poor economic policy. The government adjusted slowly to a changed international environment and allowed its fiscal deficit to climb to an unsustainable 12 percent of GDP in 1988. The fiscal imbalance led the government to adopt structural adjustment measures, including a 19 percent reduction in total spending. As in other African countries faced with similar problems, expenditure cuts fell disproportionately on public 111 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY investment and on nonsalary recurrent costs. The health budget declined by an even larger amount-46 percent in real terms between 1987 and 1990. These cuts were one factor leading to low-quality health care and poor In a period of fiscal use of public health centers. The centers were left with a staff of four or adjustment and five nurses but virtually no drugs or supplies. Not surprisingly, health per- expenditure cuts in sonnel were demoralized. Only a few patients visited the health centers Cameroon, health center each day for curative care; and opportunities to deliver preventive health use declined 16 percent. care decreased. Most people bypassed the health center and "self-treated" But in villages with an through traditional medicine or through purchasing drugs of dubious innovative program of quality from private vendors, or through traveling to a distant town to user fees and revolving buy expensive brand name drugs at a pharmacy. drug funds, use Then, in 1989 the ministry of health decided to decentralize the admin- increased 9 percent. istration and finance of rural health care delivery. The approach was simi- lar to that being tried, with UNICEF support, in several other African Figure 5.3 Cameroon: Change in countries. The ministry asked USAID to help design and make operational Health Center Use and Utilization, this reoriented approach to public health care in two provinces. Other 1990-91 donors and nongovernmental organizations (Germanys GTZ, UNICEF, Percent CARE, Save the Children) were focusing on different aspects of primary health care. But the USAID-funded project was the pilot in which the gov- ernment was experimenting with a comprehensive new approach to financ- 10 9 ing and delivering health services. Local health committees were formed to oversee the operation of health centers and tell villagers about the importance of preventive health -e16 measures. Health centers were supplied with curative drugs and supplies, which were replenished through consultation fees and drug sales. -s Preventive health measures (vaccinations, family planning, information on safe drinking water) were still provided free. By making good-quality, generic drugs available at health centers at a -15 reasonable price it was thought that use of the centers would rise. In fact, the use of health centers increased significantly for households in areas -20 Control group Pilot program with the new policy, but declined in the control group areas (figure 5.3). It was also thought that access of the poor might decline because of the Source: Lirvack and Bodart 1993. user fees. But this did not happen. The poor were not left out. Indeed, they seemed to be benefiting proportionately more than others. Why? Because they lacked any source of reliable quality drugs prior to this pol- icy, so it had a greater impact on them (Litvack and Bodart 1993). Although this very successful pilot came to a close in 1994, opportu- nities for scaling up have not yet been realized. Many of the innovative 112 MONEY, BUT MORE IDEAS, TOO policies implemented in the two provinces required a waiver of normal procedures, and to replicate the pilot throughout the country would require broader changes in policy and legislation. For example, it had been envisioned that every health committee would become a private, not-for-profit association with ministry representation to ensure its inde- pendence. The needed legislation was drafted during the course of the USAID-funded project, but it has yet to be presented to the National Africa's road sector Assembly. Other issues need to be resolved before scaling up. How to provides good examples address poorer areas; how to design transparent, predictable intergovern- of how money-in a mental transfers to create incentives for cost recovery; and how to improve weak institutional and the central procurement system for pharmaceuticals. policy environment-can What is the result of this partnership between reformers in the health have little impact. sector and donor agencies? At least two provinces in Cameroon are deliv- ering better health care to their people. Furthermore, there is a stock of knowledge that community participation with proper government sup- port can be a powerful way to improve rural service delivery, particularly for the poor. This could be an important input into a larger reform pro- gram in the sector and beyond. But as yet there has been no broader reform, and it is difficult for donors to do much until there is a larger domestic movement for change. Africa's Road Maintenance Initiative FRICAS ROAD SECTOR PROVIDES GOOD EXAMPLES OF HOW MONEY -in a weak institutional and policy environment-can have lit- tle impact. Almost a third ofAfrica's $150 billion in roads (much of it aid-financed) has been lost due to lack of maintenance. Entering the 1990s about half the region's paved roads and 70 percent of the unpaved roads were in fair to poor condition. The poor state of the transport net- work was a serious bottleneck that lessened the value of reforms or other public investments. Institutional and policy weaknesses prevented governments from pro- viding necessary road maintenance despite ample funding. In most coun- tries there was an almost exclusive reliance on public road maintenance companies, which tended to have a large capital stock (equipment) at low levels of use. The public sector companies were also overstaffed. Thus funds were directed at equipment and employment, but often were not available 113 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY for other critical inputs (parts, fuel). The underlying problem: operations were used to meet objectives (say, employment) other than maintaining roads. Moreover, general weaknesses in national budgeting often led to an uneven flow of operating funds, further hampering efficiency. In response to the mounting crisis, African transport ministers launched the Sub-Saharan Africa Transport Program in 1987, with the The core problem of road road maintenance initiative as a key component. The premise of the ini- maintenance is not tiative is that the core problem of road maintenance is not technical or technical or financial-it financial, but that the "real causes are weak and unsuitable institutional is weak and unsuitable arrangements for managing and financing roads" (Heggie 1994, p. 2). institutional The objective was to foster long-term systemic change, so that by 2000 arrangements. countries would be well on their way to developing workable systems of maintenance. What were the initiative's innovative features? First, it began by bring- ing policymakers in the transport sector and road users together for sem- inars. Their focus was on the consequences of neglecting maintenance and on the institutional and policy weaknesses that resulted in poor mainte- nance. Countries discovered that they had common problems, and that already there were important examples of successful reform. Second, the initiative focused on institutional weaknesses. The fact that road maintenance units were often part of the general civil service with inflexible employment policies, made it difficult to recruit, moti- vate, retain, or dismiss staff. Establishing autonomy for road agencies can provide more flexibility in employment. Financial autonomy can be important as well. In general, road maintenance is a public good, and road maintenance companies cannot directly charge the beneficiaries of their work. But taxes for vehicles and fuel can be structured to closely approximate "prices" for road construction and maintenance. In a country with strong budget management a government may not want to earmark funds from vehicle and fuel taxes for road maintenance. (Whatever the source of the revenue, it should ideally go to the use that has the highest social return.) But the best can be the enemy of the good. If the budget is not well managed and funds do not flow to road main- tenance-even when the return to the activity is extremely high-then it is wise to search for the institutional arrangement that best fits this case and provides a good outcome. Some countries in the road maintenance initiative have started "road funds" in which certain taxes and fees are earmarked for road mainte- nance. This arrangement has benefits: 114 MONEY, BUT MORE IDEAS, TOO " It can improve tax collection. Collection always relies to a certain extent on the willingness of people to pay. When road users are con- vinced that their taxes are not going to provide public services, avoid- ance is high. If users are convinced that taxes will finance services that benefit them, they are more willing to pay. The partnership between " It can provide a steady financing source for day-to-day operations. In the government and the the past expensive equipment and staff could sit idle while the gov- private sector has led ernment resolved the latest fiscal crisis. In the new institutional not just to better arrangement, there is a more stable flow of funds. collection of road taxes " It can improve spending. Road funds are overseen by road boards that but to more efficient use include both government staff from the sector as well as road users- of funds. farmers, businessmen, bus and transport companies. This partnership between the government and the private sector has led not just to bet- ter collection of road taxes but to more efficient use of funds. Road boards have also provided input into other institutional changes, moving away from exclusive reliance on the public sector to carry out road maintenance. There now are commercially run equipment pools, and both road maintenance tasks and equipment maintenance are being contracted out (Carapetis and others 1991). One key lesson is the importance of having a long-term game plan for reform. The initiative began in 1987, but it came to show major results only in 1992. It also demonstrates the importance of "knowledge creation," as we use the term in this report. In highly distorted environments people stop expecting services from the public sector and become skeptical about possibilities for reform. What is needed in this situation is not just good ideas, but demonstrations that they work. Now that there is tangible evi- dence of their value, good ideas from the African road maintenance initia- tive are spreading quickly from country to country. Also needed for the reform of public institutions is the participation of civil society. Conclusion OVERTY HAS ALWAYS BEEN WITH Us. DEVELOPMENT ASSISTANCE IS only 50 years old. It is encouraging that there have been huge suc- cesses in the aid business-and it is not surprising that there have been some spectacular failures. Many failures involved investments 115 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY based on strategic (not developmental) considerations influenced by the Cold War. But the failures of aid are not simply the result of pursuing nondevelopmental goals. Many of the world's poor live in countries with weak governance and poor policies, so that aid agencies work primarily in difficult environments. Early development thinking suggested that large dollops of capital into these countries would spur development. It is possible to assist This is not the case. The examples here show that it is possible to assist development in countries development in countries with weak institutions and policies, but the with weak institutions focus needs to be on supporting reformers rather than disbursing money. and policies, but the A number of common themes run through these examples of effective focus needs to be on assistance. There must be champions ofreform with long-term visions at the supporting reformers local or national level. Aid must help them create the knowledge that they rather than disbursing need for effective development. In the difficult environments, public ser- money. vices are poor or nonexistent. Since the government in these cases is func- tioning poorly, effective innovations typically involve engaging civil society-either to bring about governmental reform or to create substi- tute institutions. While it is possible to make a difference in countries with weak insti- tutions and weak policies, scaling up the assistance to have a large effect generally requires nationwide reform. In Vietnam national reform in the late 1980s set the stage for a dramatic reduction in poverty in the 1990s. But in Cameroon's health sector, a successful innovation in two provinces was not replicated nationally. Essentially, national economic management remained weak and did not take advantage of the successful innovation. National transformations often catch observers by surprise. No one in the late 1980s could foresee the political and economic transformation that would come over the Soviet Union. Reforms in Vietnam, India in the early 1990s, Ethiopia or El Salvador after their civil wars-all would have been difficult to predict a few years earlier. What this means for donors is that it is worthwhile to look for opportunities to support pock- ets of reform in difficult cases, but that they will not have much leverage over whether local successes are scaled up to nationwide transformations. Another common theme concerns donor behavior. In the successful examples of assistance donors were more cooperative than competitive, more focused on knowledge creation than on disbursing money. For donors to be more consistently effective requires changing the internal culture and incentives for line managers in donor agencies-which have often encouraged a focus on the volume of finance or on narrow mea- sures of project "success," not on the contribution of assistance to sys- 116 MONEY, BUT MORE IDEAS, TOO temic change. A recent study of aid dependency sponsored by SIDA ana- lyzes the problem: "Both donor and recipient have incentive systems which reward reaching a high volume of resource transfer, measured in relation to Donor agencies need to a predefined ceiling.... In many administrations, both bilateral and create internal multilateral, the emphasis is on disbursements and country alloca- mechanisms and tions. Non-disbursed amounts will be noted by executive boards or incentives that foster parliamentary committees and may result in reduced allocations for selectivity and that focus the next fiscal year. large-scale finance on "On the recipient side, officials of ministries of finance and plan- developing countries ning are rewarded for the commitments they can attract from with good policies. donors in bilateral negotiations or in donor meetings. Whether these commitments result in disbursements is often of less conse- quence. Results are measured against volume figures, with no regard for the quality of management. Quality is supposed to be measured by regular evaluations, but this dimension of aid transfers is very complex and can always be disputed. Besides, when the time has come to evaluate the actual outcome, most of those responsible for the project on both sides will have been transferred. "These practices have a direct relationship to program quality as well as to the feeling of program ownership. In the present aid- dependent countries, new programs have been pushed through for decades by donors as well as by planning ministries under a time pressure that appears completely unwarranted in retrospect, con- sidering the slow rate of subsequent implementation. In these often very large-scale programs quality has taken second place behind quantity, and the question of ownership has never arisen: it is the owner who pushes" (Edgren 1996, p. 11). Effective donor agencies need to create internal mechanisms and incentives that foster selectivity and that focus large-scale finance on developing countries with good policies. In countries with poor policies, donors need to be more patient and accept that the best assistance may involve activities that do not result in much disbursement. Donor agencies have learned from their successes and failures. In the 1990s all of the major agencies have instituted reforms aimed at strength- ening the focus on results on the ground (box 5.1). Most agencies have also formulated country assistance strategies so that individual activities now 117 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Box 5.1 Reforming Aid Agencies in the 1990s A . f '.F P F ..* IE .. ii1 i f f E 'll '.v.i. InsEiEucional taccor- trical lor Ai,ainablz deel- eClfr 13 % n l) atpE EICrIf iALiEEiifliln E3cuciIna. .pfnilCr IiTTpa,r had ecn ncglc,rcd (_ hangs arc now-: cLIIre;. and c.ipaciLS co \rk parrticipirorv udcr %;-it in cich i rK E nuIlilac,ral and kecro.dc de.c lopnent Ni. parcrrnit u tating. recrul T! nC. appr_.u,achc; irc rcplacin .ndoidUl projecEs. ThI And pcririncl prbiriniricc .."cE mci. te.. prkqecr \\rid Bank lia cinibdied it, nc':..b izcuiev and orea- 1111r. .ciTicEa .pprilchct. and ncv. criEcria tir iudging Q,.Paron.iJ ciiliurc in a partrier'hip strdrcg. that rcache, 'ULc-All ha- L CnL.g * pril?cs,i%i e. I--L tar ha. .uL i D,couIerpart, ''%ithin Ind :.MUrside *-.wnrmcn[. rhi; prct' *. .ini \.ihat rcimniin s - b donc Siniil.ir h.nce arc --idnr in hilaccral agencic;. 1.:.r niilrilaccral dck eliopnnr hank! clt-corn,ini- %hk.i h h3. c ]h: fr,n bcn diburcnti-drivfn Bur .ionLd rcpore led r.-- a fi.indaincal rcc.n;ider.in'n i- their main in'Eruin iE been grant; not loan,, iuh E11 1 ff L- Ln. Of tlh.f eirradiio.nal inurumcnts-pr.j- il.rl in prepar inio and chknitA i-miEafncc icE ltan Beqirning .-:irh chc \\ orld Bank [ he cmpha,i; has .hifrd froni capital projeth O VN -a1'nh±nf RLpOrr foll. ed b, nuIlar rcp.. ut rhr capacir buildini1 and ;upp.*rr for p-lis: reform. IniE-.An -lri.in D%celi-pntn[ bank. E 'ian in ll checc chang thc fc L,i --in dclopnmen[ Di c flipncnt Bank. and the Afri-an DCCl0Pn1Ln[ r,oilE--iin tlh hunin progrcm; and ihc cc,.nomiic, Bin!: h, co-nlu'ion ,as Lhac lian quancir, had bc1n pEicaI. nd i.il ad'.ancc gncrarcd b- the iCforti It i'. cc pro t 1;'. r c.ran Llualiik. '1ccurin tin u d.m . clpii . til and LiiT parnLr . .1pp!-,, '.i a irc , [ crful r hnio.r iiot f.r frat thin 0.11rl,iiv4 Eocnrsirc pri-eC E Lucc.s;Ir lar4cric clk piifi L . . .prira . .1 ri ncr. . r r .t must fit into a larger game plan for policy reform and institutional change. In the same vein, the focus of evaluation has risen above the level of the pro- ject to overall country program reviews. These essentially ask: Have agen- cies used their resources to stimulate institutional and policy changes that have led to improved services and a better quality of life? This is not an easy question to answer. But it is the right one to ask. The Overseas Development Council recently sponsored reviews of the overall impact of aid in eight African countries, conducted jointly by scholars from the devel- oping countries in question and from major donor countries-which seems to be a fruitful approach. In the same vein, the shareholders of the IMF recently turned to a group of outside scholars to assess the Fund's sup- port to low-income reformers. And the World Bank has initiated country assistance reviews with input from a wide range of stakeholders. Called for are independent reviews of development agencies with strong input from developing countries and focusing on two questions: Has the bulk of financing gone to sound institutional and policy environments? And have agencies contributed to policy reform and institutional change? Evaluating the right things should feed back into the management and 118 MONEY, BUT MORE IDEAS, TOO incentives within agencies. With better management and evaluation, devel- opment agencies should become: m More selective-putting more money into economies with sound management. m More knowledge-based---using resources to support new approaches to service delivery, expanding knowledge about what works, and dis- seminating this information as a core business. m Better coordinated-results-oriented agencies should worry less about planting their flags on particular projects and more about how com- munities, governments, and donors working together can improve services. m More self-critical-agencies continuously should be asking themselves: Why do we do what we do? And what is the impact? With a better understanding of development and aid effectiveness and with the end of Cold War strategic pressures, there is reason to be opti- mistic that the reform of aid agencies will succeed. 119 Appendix 1 Estimating the Effect of Aid on Growth HE APPENDIXES PROVIDE TECHNICAL DETAILS OF SOME of the important original research results featured in this report. They are based on research papers that are available as journal articles or working papers that can be downloaded from the Assessing Aid website (http://www.worldbank.org/research/aid). The reader who wants more detail should go to these underlying sources. In most cases the datasets will also be available over the web. This first appendix deals with the effect of aid on growth. The theoretical foundation for recent empirical studies of growth relies on dynamic models of inter-temporal optimization. In these mod- els the accumulation of physical and human capital depends on initial conditions and the institutions and policies that affect the return to sav- ings and investment. Thus, growth is a function of initial conditions, institutions and policies, and external shocks such as changes in the terms of trade or in weather patterns. To introduce aid, this kind of analysis must take account of the fact that poor growth may induce donors to provide more aid. Instrumental variables techniques essentially split aid flows into per- manent and transitory components. Only permanent aid goes into the growth regression. The regressions in table Al.1 use the dataset developed in Burnside and Dollar (1997). The dependent variable is growth rate of per capita GNP, averaged over four-year periods, beginning with 1970-73 and ending with 1990-93. There are six four-year periods and 56 developing countries in the sample (table A1.2), though there are some missing observations where data were not available. 121 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Regression 1 explains growth as a function of initial conditions, the incentive regime, and an error term that captures external shocks. Following other literature, the measures of the incentive regime included are the inflation rate (Fischer 1993), the budget surplus (Easterly and Rebelo 1993), a measure of trade openness (Sachs and Warner 1995), and a measure of institutional quality (Knack and Keefer 1995). The level of government consumption is also considered, but this does not have a robust relationship with growth. The index of economic management used in this chapter is formed as a weighted sum of the inflation rate, the budget surplus, trade open- ness, and institutional quality, where the weights are the regression coefficients in regression 1. This index can be interpreted as the pre- dicted growth rate, given the quality of the incentive regime and assum- ing that the country had the average value of the initial condition variables. The index has a mean of 1.1 with a standard deviation of 1.6. (The phrase "good management" in the text refers to a level of 2.7 or above.) In regression 2 the index replaces the individual components; it can be seen to have a very strong association with growth. Regression 3 introduces aid relative to GDP and instruments for it with population and variables reflecting donor strategic interests. As in Boone (1994), there is no relationship between aid and growth. The picture changes, however, if aid is interacted with the management index. Regression 4 includes aid interacted with the index, as well as aid squared inter- acted with the index. (Potential endogeneity requires that instruments be used for aid as well as for both interactive terms.) The positive coef- ficient on aid times management and the negative coefficient on aid squared times management indicate that aid has a positive effect on growth in a good policy environment but that there are diminishing marginal returns to aid. It should be stressed, however, that the esti- mate of the diminishing returns is imprecise. To measure this requires cases in which there is good policy and large amounts of aid. There are only a few such cases, so the estimate depends on these few out- liers. If they are dropped from the dataset, there is no longer a signif- icant coefficient on the aid squared times management term. However, the positive coefficient on aid times management is quite robust (regression 5). The sample includes some middle-income countries, such as Chile and Mexico, that have received little aid and have access to international 122 APPENDIX 1 capital markets. Regressions 6-8 repeat regressions 3-5, but drop the middle-income countries. The results are stronger in two senses: first, the estimated impact of aid on growth in a good management environment is larger; and second, the statistical confidence has increased. For a given quality of policy and level of aid, regressions 4, 5, 7, and 8 each provide a different point estimate of the marginal impact of 1 per- cent of GDP in aid on growth. The average point estimates for different qualities of policy are: Marginal effect on growth of 1 percent of GDP in aid (percentage points) Poor policy (index = 0) Mediocre policy (1.1) Good policy (2.7) -0.3 0 0.5 While the estimated impact of aid in a poor policy environment is negative, the estimate is not statistically different from zero. The esti- mated impact in a good policy environment is significantly positive. These averages from the four regressions are the estimates shown in Figure 1.5. The positive coefficient on the interactive term also means that policy improvements are more potent if a country is receiving aid. Finally, note that government consumption has no significant relation- ship with growth-important because aid often finances government consumption (chapter 3). 123 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Table Al.l Estimating the Effect of Aid on Gmwth Dependent variable: growth rate of per capita GNP (four-year average) Sample: 56 developing countries, four-year periods (1970-73 to 1990-93) / 2 . 4 6 ~ 8 Fin in. i '1 .J prFi i 1 112 '~ . i ii II il < 1 iii -IS i i I 12', I I ' lI i½| i i I i2 i l i'- i ' ' I I ' "' i'.hr.. il 1111 ~Iit. -i -42 -i'-< -il - I .4 -Ii 34 --l ~2 i- li -0 ¯ I i>i 'I i 11I -I'. I i |v'il u i ilil i| 'V i| i: i .t.EonI. min iii menti - i 1 3 1 I -;I (i ¯ii'' I 21' 'liiil 2.. ' 1' 0'. >j30-ii L1.ni,.mient saidf)P1- - - - - 00 - - i-f' -- il iš '2 !1W .. I24 i mr ,r ri - 124 APPENDIX 1 ТаЫе А1.2 Cour�tries in the Dataset 11гг1�1/г Е,�: г .г+ггl �iгG-r�гh,�,,�1г.�ri..t L��lrг.�,,.г�:r.г '1-niтI1.-1(-�•ec.i l�,г,i.-1,..� �onrh.�i;г,i Е�.г.,•. 1П1 �г,�,�-�пг1.1 1IE.r11 1Пi{,•ii�•1i 1П�i.1 � 1П"�,гГ' п'П Е�.Г.�� i 1 EL'.'вГ ��. ,г�_ � T�11,.1•.fer� _'ис �1 1'11Г� t�ri:�1 111'•�"�_;'1 Гh�11рЕ,1г1�•. �г1 1.�1'1L.1 Ër1i1„I•.11 t Е'1L 1 иП1•11 11��.i11an•j 1 1•I�•"�1 1.1I._��-i1L•1.1 1� ri� �.I �11 .•1_1 ��.1�,1I�,1 � �., f.� k1_.1 1 1�Г1.�. 1��1J1i 1 L1"1111i-��1.1П �1.[1�. �l 1��_f111f•I�� 1��.1�.1� �1•.i. 1Г l �111�11�1 i' 1.11 �� • i t 1�•1•. ••1"г i.I�L 11i1дгсг�'1�1д j�1��.f 1�1.���i''1 N�,,,1{ Н,�г� 1г111'_.1I I�. �Г'1LL�i.1Г.1• �.11ff,1I'-'.'1��i �t1111i•..1 1 1'1�1 1l .�1. 1�_'.' 1 т; 1г,11 ��11� 1� �_11 � � 1- �� 1 Г �.'_ 11 �'� .'..11ff 1'�.-г�'. ._.1�r�F�1.i 1 r1r-11�j �._1 : 1т.1„1.•, � 11�1.j 1 „h„',. l_�г�1_i11, 1\ � Г1 С-. 1 1•. �� 125 Appendix 2 Explaining the Success or Failure of Structural Adjustment Programs ECENTLY, THE THEORY OF MACROECONOMIC POLICY and economic reforms has changed focus. Instead of viewing the making and implementation of economic policy as a control problem in which the issue is to find the optimal policy rule, the core of the analysis has 6.shifted to the actual policy process. The chosen eco- nomic policy is explained by appealing to binding incentive constraints facing optimizing policyrnakers. The theoretical literature in political economy has identified several factors affecting the likelihood of suc- cessful reforms (see Rodrik 1996 for a recent review of the literature). To test these theories, and to investigate if factors under the World Bank's control have any effect on success or failure of reform, a measure of the extent of policy reform is needed. Previous work on explaining reform progress has used different out- come measures (or changes in them) as proxies of reform, but that has obvious shortcomings. For example, outcome is partly driven by exoge- nous shocks which are difficult to disentangle from policy effects, there is lag between policy change and outcome, and reforms differ in objec- tives and may therefore not be captured by a single outcome measure. Dollar and Svensson (1998) avoid these problem by using a zero-one variable reflecting failure or success of reform programs supported by the World Bank (outcome). The binary evaluation variable (outcome) is determined expost by the Operations Evaluation Department (OED) of the World Bank. Using the zero-one reform measure as dependent variable, a probit regression could be specified relating the probability of successful reform 126 APPENDIX 2 to domestic political economy variables, World Bank effort variables and other controls. The regressions in table A2.1 use the dataset developed in Dollar and Svensson (1998), consisting of 182 completed adjustment loans during the period 1980-95, for which comparable data could be collected: 36 percent of these reform programs were judged not to have met their objectives. Regression 1 explains the probability of success of reform as a func- tion of only the core political economy variables. All variables enter sig- nificantly: success is associated with democratic government and with political stability. Ethnic fractionalization and length of time that the incumbent has been in power enter non-linearly: the basic message is that high degrees of fractionalization are bad for policy reform, and that long- term incumbents are not likely candidates for reform. Regression 1 pre- dicts correctly 75 percent of the observations. In regression 2, several Bank-related variables are added to the speci- fication, recognizing that there is an endogeneity issue that has not yet been addressed. Some of these variables are likely to be exogenous: whether the adjustment loan focuses on trade reform or sectoral reform depends on the nature of the policy problems in the country and the gov- ernment's desire to attack particular problems. What is clearly under the Bank's influence is the amount of preparation staffweeks and amount of supervision staffweeks. In regression 2, preparation is positively associ- ated with the probability of success and supervision, negatively associ- ated. Once controlling for these two variables, other Bank-related variables such as number of conditions, loan size and the allocation of conditions among tranches play no role. An implicit assumption underlying regression 2 is that the World Bank does not respond to exogenous shocks that reduce the probability of success-that is, the error term in regression 2 is assumed to be uncor- related with the amount of preparation and supervision. However, an exogenous shock that reduces the probability of success is likely to call forth more preparation and supervision resources. Thus, in order to esti- mate these relationships it is necessary to partition the World Bank effort into a predicted part that is independent of exogenous shocks-and an unpredicted part. This is possible to do by employing a two-stage probit technique developed by Amemiya (1978). Regression 3 reports the result of including the predicted part of the World Bank effort variables. Once these Bank-effort variables are treated as endogenous, there is no rela- tionship between any of them and the success or failure of adjustment 127 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY programs, while the relationship between the political-economy vari- ables and outcomes is stable. This finding is consistent with the view that there are institutional and political factors that affect the probability of success of a reform program. Given those factors, none of the variables under the World Bank's control significantly affects success or failure of adjustment programs. If endogeneity is ignored, there is a positive rela- tionship between preparation and outcomes, and a negative relationship between supervision and outcomes. That these relationships disappear in the two-stage regressions indicates that the associations reflect how the World Bank allocates resources. To further explore the last issue, regression equations for supervision and preparation are specified in columns 4-6. Regression 4 shows that preparation is strongly related to a number of variables (regional dum- mies, loan size, number of conditions, income, and population) that in turn have no relationship with the probability of success-but has very little relationship with the political economy variables. There is a broadly similar story for the allocation of supervision resources (regressions 5 and 6). Unlike the preparation equation, regional dummies are no longer important. The regional departments of the World Bank have different amounts to prepare loans, but once these loans are approved the regions devote similar resources to supervising a loan of given characteristics. In the supervision equation, it is also con- sidered that preparation may affect supervision. In the OLS regression (regression 5) there is a large, positive relationship between preparation and supervision. This reflects the fact that the error terms in the prepa- ration and supervision equations are certainly correlated. Anything unobserved that leads to higher (lower) than predicted preparation will almost certainly lead to higher (lower) than predicted supervision. The fact that the regional dummies seem to belong in the preparation equa- tion but not in the supervision equation means that it is possible to use them as instruments to correct for this simultaneity problem. In the two- stage least squares regression (regression 6), the relationship between preparation and supervision is no longer significant. 128 APPENDIX 2 Table A2.1 Estimating the Outcome of Adjustment Loans _r I r. I .31.ïr[I. ._ 1 } _ 14 - |i -. lan -bir rl,lj,... Il -- . Il - Il 1 -h-11 *. r1.r -1~~- -I '' -1 4* 4 Illr .|||1- - - -_ 1 ' , i I -i ' IIr 1 .., -% l'%[itI1 Ii- .il-| ·.. |.- 1 11 . .l 1 ' . L 1-- 1- ''IlIl I r I a I th 1 - '1l1 I, 1 1.11 -, :. | - 129 Appendix 3 Analysis of the Fungibility of Foreign Aid NE OF THE MAIN CHANNELS THROUGH WHICH foreign aid influences development outcomes is its impact on the recipient country's public expendi- tures. The link between foreign aid and public expenditures is, however, not straightforward because some aid may be "fungible." An aid recip- ient country could render ear-marked aid fungible by reducing its own resources in the sector that receives aid and transferring them to other sectors of the budget. Feyzioglu, Swaroop, and Zhu 1998 study this issue using annual data from 1971 to 1990 from 14 countries-Bangladesh, Costa Rica, Ecuador, Egypt, Honduras, Kenya, Mexico, Malawi, Malaysia, Peru, Sierra Leone, Thailand, Turkey, and Zaire. Two foreign aid variables were used: total aid to a country (the annual net disbursement of Official Development Assistance or ODA), and the sectoral composition of con- cessionary loans from all sources, over time and across countries. Other variables in the data base included total as well as sectoral classification of public spending, per capital real GDP, infant mortality, average years of schooling, school enrollment ratios, military expenditures of neigh- boring countries, and the share of agriculture in national income. The model in the study develops links between foreign aid and pub- lic spending assuming that the observed mix of public expenditures results from a combination of the government's utility maximizing choice using fungible resources-domestic and external-and the pur- chase of goods from the nonfungible portion of aid. In the empirical analysis the impact of aggregate foreign aid on total government spend- ing is first estimated to examine whether foreign aid affects the resource 130 APPENDIX 3 mobilization effort of the recipient country. The impact of earmarked sector-specific aid on components of government spending is estimated next and the fungibility hypothesis examined. While the problem of simultaneity exists in principle in this study, attempt is made to minimize it by (a) using aid disbursement numbers which in most part are predetermined; and (b) including a few economic, political and social indicators of the recipient country as explanatory vari- ables in the regression analysis. Tables A3.1 and A3.2 report the main regression results of the study. Regression 1 shows a positive and statistically significant relationship between the share of total government expenditure in GDP and the share of the net disbursement of ODA. The regression shows that a dollar increase in foreign aid leads to an increase of 0.95 cents in total govern- ment spending. There is no tax relief effect in this sample. Increases in the net disbursement of concessional loans, however, are far more stim- ulative of total government expenditures. Regression 2 shows that a dol- lar increase in concessionary loans leads to a $1.24 increase in government expenditures. The likely reason why concessionary loans have a relatively larger impact on government expenditures than overall ODA is that a portion of such loans have matching requirements-that is, for every dollar that a government spends on a specified activity it gets a matching amount in concessionary loans. Among the control variables, the share of agricultural output in GDP-a measure of level of develop- ment in a country-is the only variable that is statistically significant in both the equations. The negative coefficient suggests that countries that have a bigger share of their GDP from agriculture and are therefore rel- atively less developed, have relatively smaller government spending. Regression 3-which includes expenditure shares according to the eco- nomic classification-indicates that roughly three-quarters of ODA is spent on government's current expenditure. The coefficient of ODA in regression 5 shows that the remaining one-quarter of aid (after account- ing for current expenditure) goes for capital expenditure. Regressions reported in table A3.2 examine the link between the net disbursement of concessionary loans to a particular sector and public spending in that sector. In each of the six regressions-one each for edu- cation, health, energy, agriculture, transport and communication, and defense-the coefficient on the variable Government expenditure net of aid in GDP which is statistically significant in all regressions, indicates how the government distributes an additional dollar that it gets from all 131 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY resources net of concessionary loans. Regression 5 has a positive and statistically significant relationship between loans to the transport and communication sector and the public spending in that sector; the coef- ficient on the aid variable is 0.92, which is statistically not different from 1. Other estimates indicate that loans to agriculture and energy, for the sample countries, have been fungible. But for the education and health sectors no null hypotheses of interest can be rejected. The likely explanation is that based on the available data for these sectors, the power of the test is not enough to reject any reasonable hypothesis. In recent years, the donor community has been increasingly concerned that devel- opment assistance is being used to fund military expenditures. Data from the sample countries do not, however, support the hypothesis that for- eign aid is diverted for military purposes (regression 6). The results show that there is no consistent link between aid to a sector and increased spending in that sector-that is, aid tends to be fungible. 132 APPENDIX 3 ТаЫе А3.1 lmpact of Foreign Aid on Tota1, Current, and Capital РиЫiс Expenditures, 1971-90 Rг {�'7 г • •:u+r /,+r.rl;+rг.rlн+��++�;/+lifгllин /'иGh�+гнг�н�гti��гirгlгггнг+ l4rhlпl,г�+н•г1�1.�еыгllгиrг, D.�.Srde�rt г•vrьiW.' ! _' 3 -t � G �_,.п г �п[ 1 �и 111 �111 lг� .rгifii�гif .•.F�c1„11r�_и: �1 п, 11 ni 11 :, , i� ;, П�Г1�F 1i�I11-iIJI)�• iI`� ♦i1 ��i ��1 ..i �:..1 i �.1I1 �1i�11� � r�1�1.'�11�11.`�11' 11�1•� 1_1 _ 11 �'1 1'i �_ i � 1 I '.' 1) _ 1�' 1 ily�rc „f Сг,1�.= �,�п;г, 1_'i l_'_ 11 +- I1�iП� 11i 1.��1' �i 11�� 1� �1-� � � ��1� 1•'г:'11��,�-��f.1ri1�1�1 и11� �1П� -пш�' -11П1�� ии11' 11ии' •11�� ��I11i 111{ii ��11'�1 i11`.'�1 •11`111 N�1�F�f�.iг.n1111CлR 11� I1a� -11IГ1 -1I'i4 i111`. 1111д •�'ГlJ1f�11�_1П1iL1�'��i, -1� 1� 11�i �� '1.. 111 1.1 i111 � .1111��1 111 •11� .i�,crd�_г •��-„�•1��Пс iri I • 1 1' • а ' `1' - . ^} --� Ч'� I�k��irt�ir.�'1i�'-11� 1111fi �ип�i i11��� i'.11i1 ,�_-i 'п1•� �П�.11��1 li•.1f1I�f� Г7Г` II1_1-I1� 1111�� 11111� 11111� 1111I -1111^� -1111� 1� '�11 111•11 1� ���i I11 �1'•1 1� ��� ;11����1• 1�-,.��, ,�t •i!��и1ги1•_ -111,; _и �; .п 1` U 11'� 11 п- П �� ��Uff'•1Jг 1п 1,I11' �I �_i-1 �. 1' 1г.�. i• п•i in •�ri � in 1. � iu �, � • t i;i l-.r[!1 1гг_it �,i k1.��1CL:1I 11 i'� 11 i' -11 1 -11 +~ и 11� -П 11, in•� .1 1I �1Н•сГ[�1.. 111 i�i• i11 �11� 111 �11i 1I в' � i11 I� i11 j11� "�•:��� . •. ��- 1: У l1 ыЧ 11'.+- ll :1- +"� -•1 !1 ,� . Г �i .. .. , . _�� ! 'л л� �� , � _1/•� �i� . i 1 �.+ ! �.+ 1 � . 1 �.. i .•.� �. . _ l�1.ре-�]._ni • ��i �1.1: i.. j г� .•1 _ i F�: _ 1,� ,,г.. . 1�.Пi- ii- G,: „J��.1 •� 1��1 1. 1 1 dJ �п.1�•.,г. 1�_г6•.1 г1,. .,��п11�. .{�,п-�г��•.• 1�� 1.. �__ �.1, �.�,г,:_n� � 1� :.1 _11::1 �, � п•1,.��, :1г. . �_.1, 1 ' , i .. �•._ •. 1 �� rF� �г �,.t•�•. .г�� � г� ..,{.�1г. _ ��,,,,{, 1. , -t'ч ..1Г .1 , иигг, .1„� �,,, _ � _ �.г �•.�, гг,:.1 г-.г,1�.1�._ и, f. � .��г1,� . 133 ASSESSING AID: WHAT WORKS, WHAT DOESN'Т, AND WHY ТаЫе 113.2 lmpact of Sectoral Concessionary Loans on Sectoral Government Expenditure, 1971-90 �iг'У1Л51п1! Т7tгддЛqr! �7i7l� Егlггь'riltOrr Нг.7Ilh Е1гг'r� .��rn'гг![ги•, ,пи�нггигiiи�гип Drfгu;г� !).ргыдгиг гп,гиG(г ` 1 ? . -г � б l ПГ�:1�ПС �]� � 1`1 -П �ii -' 11 _' l1\ i�� �1�"1 1l'. 1114�� 1�'и� �i+�1 1и,.,.11 {,. .�1пп�сПf ,.-.F�c11•i1[Uf: п п� п 1i' 111.11 п п, и �п п 1 l [Jc[�'�fд1�1Г1{:[{1' i�`1�1 i� '1 ��.1.i1 _-, 1{.-� 11111� .�.. �. �., .. , , Г /1г(1i 1 t J 1�i l[ 1„Г1 � l l 1 1 1 1� 1 1 I 1,, 1 1 1 15 1 1 •_ 1 1 � 1�11:_1 •111Ii1 111,��1 i1111i 111i�1 �11�.11 Н�..1г.11 -_'1 -11 ?1 ; 11- ; а5 1 11.1 � j�1 1и,,,� �11 `i1 11 1,1 � � l'�11 ��1 '11 11,1 •�1 i Ёг1г(,_1 -П. 1 I�1 �' 11 н� Ii'� 11 j- 1�11' 11 _'1� �'� •,{1 1, _ i� ;�11 i-ii i111J-i ��11,.11�I11fC (111i 11 ��� 1111•� ._1111�, -1111� 11 '� ` i_' _' i 1_ д�.i 111 ��1 111 .'1 i111131 11.1 f,i1 j г i1�•i+��ri 111.Э -и i•1 п 1 а 11 1 �� п' � 11 1' 1j �Г, �,�Пi1Г�L1п1�дf1,�П ; и� � 1' а±� 1''1 i 1 � 1, •,�.,1 11.а�� l lthcг :,т�г„гs _1i 11� и 11' 11 �11 (111г. п 11i _п i1� , � 1�:..1 _�11i i1i -ц1 _ i 1 � IMi 111 �`i hc�1 Frr •_,р1г{ 1�;L1j� 11 пи11с -1i urn1� i� 11i1� п u1.�i1; _и 1nui� i1 rnu1' �1_1 �п1 i11 1`�1 1� {�1 113�1 111 �-1 i11 1`.1 '\.:1��f1�y�Г.П1iI1fiR -11�' 111111i 1111' -111111i -1111{ 11I1� г•Е�:�7.З1[игс 1п 1=L11' I�д., �_� 1� 11 _.1 �11 ]-i 1и а1 � �11 �'i �п Г.-i 1и 11�� �l� ci и с•�����,Iif"�� 1П -и �•1 _11,11ti -11 �� 11 д1_. -I ni -1I ''1 1,h�,1 h�.г��. �Ia,,'-11� 1и 1.�,� �1.1 ..11 1(1 г1� _, `.1 �{�,�1 1i�'�i f,�f;n[m,�i[i1�r•. i�[•: 1�>_:1-11� 11111 -11Г1и: ип11' 11п� -ii,i1, -17.111 11 -� �и�1�1 1115; �j�.u1 _=.5i �1.1_' 1h �гг ,-.f i�f�.:u1[11ic -11 П�. п Г111�. 11 11! -Г111и-{ _Г1 п� -1�1 (1 i ��иг�,�.1[ и� l,ll1' I1;. _11� 11 1 • 1п 1,;� ; 1 i'� 1Г1,��.1 � ��1�� �1i ;п1 l, i.r11 1Л�jс•. ,�1 F,�,L[�.i1 -и 1 -и 1h� -(1 и1, _1111' -V (1 -Ц 11; s{1�ь1'.1��1Е•�гГи• II �1•.1 1l'i�� �� �•• 111._ �11i � 1П.i� . {.'! ' • . . � f � - , 1 Н., r 1 ' 1 i 1 � 1 r 1 ь r г. � ��' ' ( / ' ' �_1,.i ' ' i ��. ._. .' '_.\ '_5 �'�И 1? .., iг., ,. „- Е' , . , .., . , ,{, . 1. . ., : „ , l,,, .;г,� _: 1 �h1�= 1 .1 { •k�.'��. �1 г1.h1,- i1,_ . F i. J 1 . •h,� . .1 � „ .f., ... � {�,. .���.,г i� �1 �1 h'.1��.iJ �,��1i.,i: 1�=i1��:r г11. �. •11пГг .1�� �-�п-пс �п г11, гг_rг iпп г._F,ri •,.п� , F1 .�J �_rr..-i; � i i г, , 1�л�� ��ht�..1 ri���•.1 .�.� f.� ia� i•Г� гF1г :j �.lf . Y�.�.J-_1}г_ � ��i�.1 .,�� f1i. 1 г.. 1i �. п� �- i: 1. �г ic� �гг�.,{ г..г��,.п� i Fi , i,rh: ._ 134 Appendix 4 The Impact of Civil Liberties and Democracy on Government Performance 0 MEASURE "CIVIL LIBERTIES" THE STUDY UNDERTAKEN as part of the research on aid effectiveness used a vari- ety of existing indicators constructed by political sci- entists Gsham, Kaufmann, and Pritchett 1997). The two most general indicators were the Freedom House (1997) civil liberties index which ranks countries annually on a seven-point scale based on assessments of 14 factors, such as a press free of censorship, freedom of assembly and demonstration, freedom of religion, and free trade unions. A different index compiled by Humana (1996) ranks countries on a scale from 0 to 100 on the degree to which the country complies with the human rights agreed to in the 1966 UN Covenant on Civil and Political Rights, To measure "democracy," there are similar rankings constructed by political scientists that focus more narrowly on the extent to which a countrys leaders (executive and legislative) were chosen by elections. This includes a political index of democracy created by Freedom House (1997) as well as others. To measure the performance of governments we used a sample of the economic rates of return (ERR) of World Bank projects, which requires some explanation. Since the World Bank applies the same project selec- tion and implementation procedures across all countries, the differences across countries in returns are an indication of how effective the govern- ment is in implementing public projects. To assess the impact of the governance variables we began from a base specification of the determinants of ERR frorn Isharn and Kauf- mann (forthcoming), which included a set of policy variables (black 135 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY market premia, fiscal surplus), economic variables (terms of trade changes, GDP growth, the capital to labor ratio), a dummy for "pro- ject complexity" and a set of dummy variables for the sector of the pro- ject. For the time-varying variables, we used the average value in the three years prior to the year in which the project was evaluated (usu- ally the same as the completion year). We experimented with and with- out a set of regional dummy variables. Table A4.1 shows the base case results, estimated using a Tobit procedure to allow for the downward truncation of the dependent variable (the ERR is truncated as -5). To this equation are added the variables for civil liberties. Both civil liberties variables had a strong positive and statistically significant impact on the performance of Bank-financed government projects (table A4.2). The two variables produce qualitatively similar results and the range of the estimates suggests that improving from the worst to the best civil lib- erties would improve the rate of return on government investment pro- jects by between 8 percentage points (Freedom House index) and 22 (Humana index) (the average return in the sample was 16). Two other interesting results from this regression suggest that this impact reflected the impact of citizen voice on the performance of gov- ernment. First, indicators of civil strife (riots, political strikes and demon- strations) are positively related to the ERR on Bank projects, but that partial correlation is explained by the fact that there is more of this expres- sion of discontent in countries with more civil liberties. Once one accounts for the greater civil liberties, this eliminates any independent impact of civil strife, indicating that when civil liberties allow it there is greater expression of all types of citizen voice and that ultimately this voice is a force for improving government performance. Second, once the regressions explaining project returns included civil liberties there was no additional impact of electoral democracy. So, while electoral democracy and civil liberties are obviously closely linked, both in practice and in the data, the main channel of influence appears to be the availability of civil liberties rather than the more purely political mechanisms of choosing leaders. 136 「――■「「―:!:&:「:■「「,「―「「「:.「「「:-:「:―「:「:&,―■「―-;- & F컵 겹 Z U 딛 낚 너 屯 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Table A4.2 Impact of Civil Liberties Indicators on the Economic Rate of Retum of Govemment Projects, Controlling for Economic and Project Variables rar.,able, It ith, qvoaal ranable, 1 3' J 11 hh ru- 1'? Hurl,ri-, rl.. fr-J..- H - ii-- lph-.rri, i r- i.. r f h 138 Appendix 5 Estimating the Impact of Analytical Work HILE THERE HAVE BEEN VARIOUS ANALYSES OF the impact of aid in the aggregate, far less attention has been devoted to investigating the impact of different types of aid and the composition of the aid portfolio. Of partic- ular importance in this context is the rela- tionship between financial transfers and analytical work. This distinction, and the effort to measure the impact and economic return to analytical services, is of particular relevance in a context where, with increasing access of developing countries to nonconcessional sources of finance, it is often argued that the comparative advantage of international institutions such as the World Bank will shift toward the provision of such "non-lending" services. In the case of the World Bank, such analytical work comprises two distinct areas. Countrywide studies include Economic Memoranda con- taining a comprehensive account of economic performance and prospects, as well as topical reports such as Poverty Assessments, Public Expenditure Reviews, and so on that underpin the World Bank's policy advice. These are complemented by sectoral studies such as reviews of the Transport Sector, Health and Education Sector Reports-which pro- vide the broad framework that generally serves as a basis for subsequent lending operations in these sectors. Undertaking this range and magni- tude of work does not come cheap-annual expenditures on economic analysis and advice are more than $100 million. 139 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Issues, Approach, and Data The three key questions addressed in Deininger, Squire, and Basu (1998) are (1) whether economic and sector work (ESW) enhances pro- ject quality, thereby making a positive contribution to development impact; (2) whether reallocation of staff time from activities related to supervision or preparation of specific projects to ESW or vice versa could have enhanced overall project quality; and (3) whether other goals, such as a tradeoff between lending volume and lending quality, might have guided the allocation of scarce staff resources. Estimation of reduced form equations for project quality (as well as demand for resources in preparation and supervision of projects, respectively) that can be derived from the manager's assumed objective function provides a basis for empirical testing of these hypotheses. Information on project perfor- mance is based on ratings of the Bank's Operations Evaluation Department (OED) of either the rate of return as estimated after project completion (for the set of projects where such an estimation is actually feasible) or a zero-one rating that classifies a project as either satisfactory or unsatisfactory. These measures are available for 1,367 and 3,957 pro- jects, respectively, although availability of data on ESW inputs somewhat reduces the sample. Results This approach finds that ESW has a significant positive impact on various measures of quality of World Bank projects. As table A5.1 indi- cates, a one-staffiveek increase in the amount of time devoted to ESW before project initiation is associated with an increase in the economic rate of return for an individual project of between 0.02 and 0.04 per- centage points, translating into an increase of between $12,000 and $25,000 in the project's net present value for an all-inclusive cost (with overheads, travel, and so on) of no more than $3,000-a dollar of ESW yields four to eight dollars in development impact. To the extent that a staffweek of ESW benefits more than one project, this would be an underestimate. Indeed, examining the impact of ESW on a country's entire lending program indicates that a dollar of ESW yields between 12 and 15 dollars of development impact. And even this figure fails to cap- 140 APPENDIX 5 ture nonproject related benefits of ESW-as in influencing broader pol- icy formulation and analysis in specific countries. A question that arises from the analysis is whether the allocation of resources between different types of support has been appropriate. Assuming that staff resources are fungible across time and between dif- ferent uses, one would expect the marginal contribution of ESW to the quality of lending to be lower than that of lending services (preparation and supervision), because-in contrast to lending services-ESW can have benefits beyond its immediate impact on lending. Indeed, ESW is often undertaken to provide the basis for policy advice to governments and is not necessarily tightly linked to a particular project or lending program. Deininger, Squire, and Basu find the reverse, however-ESW has a systematically positive effect on the quality of the lending program, whereas neither preparation nor supervision turn out to be significant. This suggests that reallocation of staff time from lending services to ESW would have increased the quality of the lending program. Consistent with this, we find that task managers at the project level are able to reduce the time allocated to lending services by about 2.5 staffiveeks for every staffweek expended on ESW. We infer that ESW helps staff to identify and support new investment options (it expands the set of feasible projects) and design better projects ex ante (it improves the quality of projects already in the investment program). Preparation Table A5.1 Impact of Economic and Sector Work on Project Outcome and Economic Rate of Return DApendent M.e Pa pcr Offnrjuji Rare o/'Rerurn Oer. .n -. r :- ndI.:.j ir a rhi i p : r :nt l1.el Wnt i-rI I I [1 1 1 I d I-..l r c lJ n L1 4 1.1141 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY and supervision, by contrast, can improve the quality of a project (whether good or bad) only expost. Even though the preceding results suggest that higher levels of ESW will improve the quality of the lending program, it is still possible that, within a given resource envelope, shifting staff time from lending ser- vices to ESW will reduce the overall lending volume. Addressing this question, we find that this was indeed the case-lending services were between 40 and 50 percent more effective in increasing total commit- ments than ESW. But if disbursements-resource transfer-is the vari- able of interest, we find that managers could have increased both lending quality and disbursements by switching resources from lending services to ESW. This, together with the conclusion that there has been under- investment in ESW from the standpoint of project quality, suggests that the volume of commitments has (at least to some degree) been an addi- tional objective guiding the disposition of staff resources. These results provide some insight into this tradeoff between quality and quantity- the analysis suggests that on average a manager was indifferent between a decrease of $2 million in the net present value of a lending program and an additional $4 million of lending volume. If this estimate is broadly accurate, it suggests that managers were prepared to allow a sub- stantial reduction in program quality in return for only a small increase (2 percent) in commitments relative to the average program size. 142 Selected Bibliography Alesina, Alberto, and David Dollar, 1998, "Who Gives Burnside, Craig, and David Dollar. 1997. "Aid, Policies, Aid to Whom and Why?" NBER Working Paper and Growth." Policy Research Working Paper 1777. 6612. National Bureau of Economic Research, World Bank, Development Research Group, Cambridge, Mass. Washington, D.C. Amemiya, T. 1978. "The Estimation of a Simultaneous - . 1998. "Aid, the Incentive Regime, and Poverty Equation Generalized Probit Model." Econometrica 46. Reduction." Policy Research Working Paper 1937. World Bank, Development Research Group, Bauer, P.T. 1971. Dissent on Development. London: Washington, D.C. Weidenfeld and Nicolson. Campos, Ed, and Sanjay Pradhan. 1996. "Budgetary Berg, Elliott. 1993. Rethinking Technical Cooperation: Institutions and Expenditure Outcomes: Binding Reforms for Capacity Building in Africa. New York: Government to Fiscal Performance." Policy Research United Nations Development Programme. Working Paper 1646. World Bank, Policy Research Department, Washington, D.C. Birdsall, Nancy, and Franqois Orivel. 1996. "Demand for Primary Schooling in Rural Mali: Should User Fees Be Carapetis, Steve, Hernin Levy, and Terje Wolden. 1991. Increased?" Education Economics 4(3). The Road Maintenance Initiative: Building Capacity for Policy Reform. EDI Seminar Series. Washington, D.C.: Boone, Peter. 1994. "The Impact of Foreign Aid on World Bank. Savings and Growth." London School of Economics. Cashel-Cordo, P., and S. Craig. 1990. "The Public Sector Branson, William, and Carl Jayarajah. 1995. Structural Impact of International Resource Transfers." Journal and Sectoral Adjustment: World Bank Experience, ofDevelopment Economics 32(3). 1980-92. A World Bank Operations Evaluation Study. Washington, D.C. Cassen, Robert. 1994. Does Aid Work? Oxford: Clarendon. Bruno, Michael, Martin Ravallion, and Lyn Squire. 1998. "Equity and Growth in Developing Countries: Old Chang, Charles, Eduardo Fernandez-Arias, and Luis andNew Perspectives on the Policy Issues." In V. Tanzi Serven. 1998. "Measuring Aid Flows: A New and K. Chu, eds., Income Distribution and High- Approach." World Bank, Development Research Quality Growth. Cambridge, Mass.: MIT Press. Group, Washington, D.C. 143 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Collier, Paul. 1997. "The Failure of Conditionality." In C. Edgren, Gus. 1996. "A Challenge to the Aid Relation- Gwin and J. Nelson, eds., Perspectives on Aid and ship." In SIDA (1996). Development. Washington, D.C.: Overseas Develop- ment Council. Ehrenpreis, Dag. 1997. "The Changing Global Frame- work." In SIDA (1997a). Collier, Paul, and David Dollar. 1998. "Aid Allocation and Poverty Reduction." World Bank, Development El Salvador, Ministry of Education. 1995. EDUCO Research Group, Washington, D.C. Learns and Teaches. San Salvador: Algier's Impresores, SA de CV. Deininger, Klaus, Lyn Squire, and Swati Basu. 1998. "Does Economic Analysis Improve the Quality of Fallon, Peter R., and Luiz Pereira da Silva. 1995. Foreign Assistance?" World Bank Economic Review "Recognizing Labor Market Constraints: Government 12(3). -Donor Competition for Manpower in Mozambique." In Lindauer and Nunberg (1995). Devarajan, Shantayanan, Vinaya Swaroop, and Heng-fu Zou. 1996. "The Composition of Public Expenditure Feyzioglu, Tarhan, Vinaya Swaroop, and Min Zhu. 1998. and Economic Growth." Journal of Monetary "A Panel Data Analysis of the Fungibility of Foreign Economics 37. Aid." World Bank Economic Review 12(1). Dollar, David, and William Easterly. 1998. "The Search Filmer, Deon, Jeffrey Hammer, and Lant Pritchett. for the Key: Aid, Investment, and Policies in Africa." 1998. "Health Policy in Poor Countries: Weak Links World Bank, Development Research Group, in the Chain." Policy Research Working Paper 1874. Washington, D.C. World Bank, Development Research Group, Washington, D.C. Dollar, David, Jennie Litvack, and Paul Glewwe, eds.1998. Household Welfare and Vietnam's Transition. Filmer, Deon, and Lant Pritchett. 1997. "Child Mortality Washington, D.C.: World Bank. and Public Spending on Health: How Much Does Money Matter?" Policy Research Working Paper 1864. Dollar, David, andJakob Svensson. 1998. "What Explains World Bank, Development Research Group, Wash- the Success or Failure of Structural Adjustment ington, D.C. Programs?" Policy Research Working Paper 1938. World Bank, Development Research Group, Fischer, Stanley. 1993. "The Role of Macroeconomic Washington, D.C. Factors in Growth." Journal of Monetary Economics 32(3). Easterly, William. 1997. "The Ghost of Financing Gap." Policy Research Working Paper 1807. World Bank, Fraker, Thomas, Alberto Martini, and James Ohls. 1995. Development Research Group, Washington, D.C. "The Effect of Food Stamp Cashout on Food Expenditures: An Assessment of the Findings from Four Easterly, William, Michael Kremer, Lant Pritchett, and Demonstrations." Journal ofHuman Resources 30(4). Lawrence Summers. 1993. "Good Policy or Good Luck? Country Growth Performance and Temporary Freedom House. 1997. Freedom in the World: Political Shocks." Journal ofMonetary Economics 32(3). Rights and Civil Liberties. New York. Easterly, William, and Sergio T. Rebelo. 1993. "Fiscal Haltiwanger, John, and Manisha Singh. Forthcoming. Policy and Economic Growth: An Empirical Investi- "Cross-Country Evidence on Public Sector Retrench- gation." Journal ofMonetary Economics 32(3). ment." World Bank Economic Review. 144 REFERENCES Hanushek, Eric, and Dongwook Kim. 1995. "Schooling, . 1978. Labor Absorption in Asian Agriculture. Labor Force Quality, and Economic Growth." NBER Bangkok: ILO-ARTEP. Working Paper 5399. National Bureau of Economic Research, Cambridge, Mass. Jha, Shikha, and Vinaya Swaroop. 1997. "Fiscal Effects of Foreign Aid: A Case Study of India." World Bank, Heggie, Ian. 1994. Management and Financing ofRoads: Development Research Group, Washington, D.C. An Agenda for Reform. World Bank Technical Paper 275. Washington, D.C. Jimenez, Emmanuel, and Yasuyuki Sawada. 1998. "Do Community-Managed Schools Work? An Evaluation Hilton, Rita. 1990. "Cost Recovery and Local Resource of El Salvador's EDUCO Program." Impact Mobilization: An Examination of Incentives in Evaluation of Education Reforms Working Paper 8. Irrigation Systems in Nepal." Burlington, Vt.: World Bank, Development Research Group, Associates in Rural Development. Washington, D.C. . 1992. "Institutional Incentives for Resource Jones, William. 1995. The World Bank and Irrigation. A Mobilization: An Analysis of Irrigation Schemes in World Bank Operations Evaluation Study. Nepal." Journal of Theoretical Politics 4(3). Washington, D.C. Hirschman, A. 1967. Development Projects Observed, Killick, Tony. 1991. "The Developmental Effectiveness of Twentieth Century Fund. New York. Aid to Africa." Policy Research Working Paper 646. World Bank, International Economics Department, Hulten, Charles R. 1996. "Infrastructure Capital and Washington, D.C. Economic Growth: How Well You Use It May Be More Important than How Much You Have." NBER Kim, Jooseop, Harold Alderman, and Peter Orazem. Working Paper 5847. National Bureau of Economic 1998. "Can Private Schools Subsidies Increase Research, Cambridge, Mass. Schooling for the Poor? The Quetta Urban Fellowship Program." Impact Evaluation of Education Reforms Humana, Charles. 1996. World Human Rights Guide. Working Paper 11. World Bank, Development London: Hodder and Stroughton. Research Group, Washington, D.C. Isham, Jonathan, and Daniel Kaufmann. Forthcoming. Knack, Stephen, and Philip Keefer. 1995. "Institutions "The Forgotten Rationale for Policy Reform: The and Economic Performance: Cross-Country Tests Impact on Projects." Quarterlyjournal ofEconomics. Using Alternative Institutional Measures." Economics and Politics 7(3). Isham, Jonathan, Daniel Kaufmann, and Lant Pritchett. 1997. "Civil Liberties, Democracy, and the Krueger, Anne 0., Constantine Michalopoulos, and Performance of Government Projects." World Bank Vernon Ruttan. 1989. AidandDevelopment. Baltimore Economic Review 11(2). and London: Johns Hopkins University Press. Isham, Jonathan, Deepa Narayan, and Lant Pritchett. Levy, Victor. 1987. "Does Concessionary Aid Lead to 1995. "Does Participation Improve Performance? Higher Investment Rates in Low-Income Countries?" Establishing Causality with Subjective Data." World Review ofEconomics and Statistics. 69(1) Bank Economic Review 9(2). Li, Hongyi, Lyn Squire, and Heng-fu Zou. 1998. Ishikawa, Shigeru. 1960. Economic Development in Asian "Explaining International and Intertemporal Variations Perspective. Tokyo: Kinokuniya. in Income Inequality." Economicjournal 108. 145 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Lindauer, David, and Barbara Nunberg, eds. 1995. Mosley, Paul. 1987a. "Conditionality as a Bargaining Rehabilitating Government: Pay and Employment Process: Structural Adjustment Lending, 1980-86." Reform in Africa. Washington, D.C.: World Bank. Princeton Essays in International Finance 168. Princeton University, Princeton, N.J. Lindert, Peter. 1994. "The Rise of Social Spending, 1880-1930." Explorations in Economic History 31(1). . 1987b. Overseas Aid: Its Defence and Reform. Brighton, England: Wheatsheaf. Little, I.M.D., and J.M. Clifford. 1965. InternationalAid. London: George Allen and Unwin. Mosley, Paul, Jane Harrigan, and John Toye. 1995. Aid and Power. Vol. 1. 2d ed. London: Routledge. Litvack, Jennie, and Claude Bodart. 1993. "User Fees Plus Quality Equals Improved Access to Health Care: Musgrove, Philip. 1996. Public and Private Roles in Results of a Field Experiment in Cameroon." Social Health: Theory and Financing Patterns. World Bank Science and Medicine 37(3). Discussion Paper 339. Washington, D.C. Ljunggren, Borje. 1993. "Market Economies under Narayan, Deepa. 1995. The Contribution of People's Communist Regimes: Reform in Vietnam, Laos and Participation: Evidence from 121 Rural Water Supply Cambodia." In B. Ljunggren, ed., The Challenge of Projects. Environmentally Sustainable Development Reform in Indochina. Cambridge, Mass.: Harvard Occasional Paper 1. Washington, D.C.: World Bank. University Press. North, Douglass. 1990. Institutions, Institutional Change Lopez, Cecilia. 1997. Impacto de la Ayuda Externa en and Economic Performance. Cambridge: Cambridge America Latina, 1972-1992 (Impact of External Aid University Press. on Latin America, 1972-1992). Bogoti: Tercer Mundo. OECD (Organisation for Economic Co-operation and Development). 1998. Development Cooperation: 1997 Lucas, Robert. 1988. "On the Mechanics of Economic Report. Paris. Development." Journal of Monetary Economics 22(1). - . 1996. Shaping the 21st Century: The Role of Development Cooperation. Paris. Marshall, Katherine. 1997. "Vision and Relationships in the International Aid World: The Gambia-World . 1992. Principles for Effective Aid. Paris. Bank Kairaba Partnership Forum." HIID Discussion Paper 602. Harvard Institute for International OECF (Overseas Economic Cooperation Fund) ofJapan. Development, Cambridge, Mass. 1996. The Role of OECF toward 2010. Tokyo. Mauro, Paul. 1995. "Corruption and Growth." Quarterly OECF of Japan and World Bank. 1998. A New Vision of Journal ofEconomics 110. Development Cooperation for the 21st Century. Proceedings of a symposium held in Tokyo, September McCarthy, Desmond. 1995. "Review of Public 1997. Tokyo Expenditure Work." World Bank, Policy Research Department, Washington, D.C. Ostrom, Elinor. 1996. "Incentives, Rules of the Game, and Development." In Michael Bruno and Boris Moffitt, Robert. 1989. "Estimating the Value of an In- Pleskovic, eds., Annual World Bank Conference on Kind Transfer: The Case of Food Stamps." Development Economics 1995. Washington, D.C.: Econometrica 57(2). World Bank. 146 REFERENCES Pack, Howard, and Janet Rothenberg Pack. 1990, "Is Sachs, Jeffrey, and Andrew Warner. 1995. "Economic Foreign Aid Fungible? The Case of Indonesia." Reform and the Process of Global Integration." EconomicJournal 100(March). Brookings Papers on Economic Activity, no. 1. 1993. "Foreign Aid and the Question of SIDA (Swedish International Development Authority). Fungibility." Review ofEconomics and Statistics 75(2). 1996. Aid Dependency. Stockholm. . 1996. "Foreign Aid and Fiscal Stress." University _ . 1997a. Development Cooperation in the 21st of Pennsylvania, Philadelphia. Century. Stockholm. Pritchett, Lant. 1996. "Mind Your P's and Q's: The Cost . 1997b. SIDA Looks Forward. Stockholm. of Public Investment Is Not the Value of Public Capital." Policy Research Working Paper 1660. Summers, Robert, and Alan Heston. 1991. "The Penn World Bank, Development Research Group, World Table, Version V." Quarterly Journal of Washington, D.C. Economics 106. 1998. "Patterns of Economic Growth: Hills, Swaroop, Vinaya. 1997. "Experience with Public Plateaus, Mountains, and Plains." Policy Research Expenditure Reviews." World Bank, Development Working Paper 1947. World Bank, Development Research Group, Washington, D.C. Research Group, Washington, D.C. Tan, Jee-Peng, and Alain Mingat. 1992. Education in Ranis, Gustav. 1995. "On Fast-Disbursing Policy-Based Asia: A Comparative Study of Cost and Financing. A Loans." Yale University, Economics Department, New World Bank Regional and Sectoral Study. Haven. Washington, D.C. Ravallion, Martin, and Shaohua Chen. 1997. "What Can Tendler, Judith. 1975. Inside Foreign Aid. Baltimore: New Survey Data Tell Us about Recent Changes in Johns Hopkins University Press. Distribution and Poverty?" World Bank Economic Review 11(2). Thomas, Vinod, Ajay Chhibber, Mansour Dailami, and Jaime de Melo, eds. 1991. Restructuring Economies in Riddell, Roger. 1996. Aid in the 21st Century. ODS Distress: Policy Reform and the World Bank. New York: Discussion Paper Series. United Nations Development Oxford University Press. Programme, Office of Development Studies. New York: United Nations. Thompson, Louis S., and Karim-Jacques Budin. 1997. "Global Trend to Railway Concessions Delivering Riddell, R.C., A. Bebbington, and L. Peck. 1995. Positive Results." Viewpoint 134. World Bank, Promoting Development by Proxy: An Evaluation ofthe Finance, Private Sector, and Infrastructure Network, Development Impact of Government Support to Swedish Washington, D.C. NGOs. SIDA Evaluation Report 1995/2. Stockholm: Swedish International Development Authority. United Kingdom, Secretary of State for International Development. 1997. Eliminating World Poverty: A Rodrik, Dani. 1996. "Understanding Economic Policy Challenge for the 21st Century. White Paper on Reform." Journal ofEconomic Literature 34(1). International Development. London. Sachs, Jeffrey. 1994. "Life in the Economic Emergency van de Walle, Dominique. 1995. "The Distribution Room." In Williamson (1994). of Subsidies through Public Health Services in 147 ASSESSING AID: WHAT WORKS, WHAT DOESN'T, AND WHY Indonesia, 1978-87." In Dominique van de Walle and - . 1995. World Development Report 1995: Workers Kimberly Nead, eds., Public Spending and the Poor. in an Integrating World. New York: Oxford University Baltimore: Johns Hopkins University Press. Press. van de Walle, Nicholas, and Timothy Johnston. 1996. - . 1996a. "Resettlement and Development: The Improving Aid to Africa. Washington, D.C.: Overseas Bankwide Review of Projects Involving Involuntary Development Council. Resettlement, 1986-1993." Environment Department Paper 032. Washington, D.C. Wade, Robert. 1995. "The Ecological Basis of Irrigation Institutions: East and South Asia." WorldDevelopment . 1996b. "Technical Assistance." Lessons and 23. Practices 7. World Bank, Operations Evaluation Department, Washington, D.C. White, H. 1992. "The Macroeconomic Analysis of Aid Impact." Journal ofDevelopment Studies 28. . 1997a. "Annual Review of Development Effectiveness." Report 17196. World Bank, Williamson, John, ed. 1994. The Political Economy of Operations Evalution Department, Washington, D.C. Policy Reform. Washington, D.C.: Institute for International Economics. - . 1997b. World Development Report 1997: The State in a Changing World New York: Oxford World Bank. 1992. "Effective Implementation: Key to University Press. Development Impact." Report of the World Bank's Management Task Force. Washington, D.C. . 1998a. Global Development Finance 1998. Washington, D.C. -. 1993. World Development Report 1993: Investing in Health. New York: Oxford University Press. . 1998b. World Development Indicators 1998. Washington, D.C. -. 1994. World Development Report 1994: Infastructure for Development. New York: Oxford University Press. 148