THE WORLD BANK 38425 Annual Review of Development Effectiveness 2006 Getting Results THE WORLD BANK GROUP WORKING FOR A WORLD FREE OF POVERTY The World Bank Group consists of five institutions—the International Bank for Reconstruction and Development (IBRD), the International Finance Corporation (IFC), the International Development Association (IDA), the Multilateral Investment Guarantee Agency (MIGA), and the International Centre for the Settlement of Investment Disputes (ICSID). Its mission is to fight poverty for lasting results and to help people help themselves and their envi- ronment by providing resources, sharing knowledge, building capacity, and forging partnerships in the public and private sectors. THE INDEPENDENT EVALUATION GROUP ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELLENCE AND INDEPENDENCE IN EVALUATION The Independent Evaluation Group (IEG) is an independent, three-part unit within the World Bank Group. IEG-World Bank is charged with evaluating the activities of the IBRD (The World Bank) and IDA, IEG-IFC focuses on assessment of IFC’s work toward private sector development, and IEG-MIGA evaluates the contributions of MIGA guarantee projects and services. IEG reports directly to the Bank’s Board of Directors through the Director-General, Evaluation. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank Group’s work, and to provide accountability in the achievement of its objectives. It also improves Bank Group work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. WORLD BANK INDEPENDENT EVALUATION GROUP Annual Review of Development Effectiveness 2006 Getting Results 2006 The World Bank http://www.worldbank.org/ieg Washington, D.C. ©2006 The International Bank for Reconstruction and Development / The World Bank 1818 H Street NW Washington DC 20433 Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved 1 2 3 4 5 09 08 07 06 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Rights and Permissions The material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750-8400; fax: 978-750- 4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. Photo credit: Women working in the fields. Photograph by Hugh Sitton/zefa/Corbis. ISSN: 1520-9733 ISBN-10: 0-8213-6906-7 ISBN-13: 978-0-8213-6906-7 e-ISBN:0-8213-6907-5 DOI: 10.1596/978-0-8213-6906-7 World Bank InfoShop Independent Evaluation Group E-mail: pic@worldbank.org Knowledge Programs and Evaluation Capacity Telephone: 202-458-5454 Development (IEGKE) Facsimile: 202-522-1500 E-mail: eline@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 Printed on Recycled Paper Contents v Acronyms and Abbreviations vii Acknowledgments ix Foreword xi Executive Summary xvii Management Comments: Summary xix Chairman’s Summary: Committee on Development Effectiveness (CODE) 1 1 Introduction 5 2 Achieving Poverty-Reducing Growth 5 Economic Growth—the Recent Record 7 How Quality Growth Delivers Poverty Reduction 8 Reduction of Income Poverty in Bank-Assisted Countries 11 How Effectively Has World Bank Assistance Contributed to Poverty-Reducing Growth? 21 3 Achieving Meaningful Results at the Sector Level 21 How Have the Bank’s Projects Performed? 23 How Much Impact Do Projects Have on the Sector? 23 Essential Elements for Achieving Results 33 4 Strengthening Public Sector Accountability 33 Strengthening Accountability with Administrative Reforms 40 Fostering Local Control to Improve Accountability 45 5 Conclusions 49 Appendixes 51 A: Project Performance Results 61 B: Sector and Project Outcome Ratings in Country Assistance Evaluations 63 C: Project Outcomes and Policy and Institutional Quality 65 D: Poverty Data 71 E: Governance Indicators for Bank Borrowers with Public Sector Programs 75 F: Management Response 83 Endnotes 89 Bibliography A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Boxes 9 2.1 China and India Account for a Large Share of the World’s Poverty Reduction 12 2.2 Burkina Faso—Focused Country Assistance Strategy with Measurable Milestones Yields Results 14 2.3 Analytical Work on the Constraints to Poverty-Reducing Growth Helped Reshape Bank Assistance to Armenia 18 2.4 Conditional Cash Transfers Help Latin American Families Increase Human Capital Formation 26 3.1 Unrealistic Project Objectives Are Often Grounded in Inadequate Institutional Analysis 26 3.2 Understanding the Results Chain Was Key to Designing Interventions to Improve Girls’ School Enrollment in Egypt 37 4.1 Building Coalitions Made for Successful Regulatory Reforms in Senegal 38 4.2 Asset Declaration Can Reduce Corruption If It Is Enforced and Disclosed 41 4.3 Using Community Feedback to Enhance Accountability 42 4.4 Local Control of Water Systems Gets Better Results Figures 6 2.1 Growth Performance Has Improved, but Sustained Income Growth Remains a Challenge for Many Bank Borrowers 6 2.2 Middle-Income Countries Were More Likely to Experience Strong Growth than Low-Income Countries (1995–2005) 7 2.3 Countries with Higher Growth Have Stronger Economic Management and Better Policies for Social Inclusion and Equity than Slow Growers 8 2.4 Policies and Investments that Enhance Equity Strengthen the Growth Effect on Poverty 8 2.5 Poverty Reduction Remains a Significant Challenge, Even in Countries with Positive Growth Rates 10 2.6 Of 25 Sampled Countries, 11 Witnessed Poverty Reduction Between the Mid-1990s and Early 2000s 10 2.7 Changes in Distribution Reinforced the Poverty-Reducing Effect of Growth in Some Countries, but Dampened It in Others 16 2.8 Poverty Remains More Widespread in Rural than in Urban Areas in Most of the Sampled Countries 22 3.1 Project Outcomes Improved in over Half of All Sectors 22 3.2 Project Outcomes Are Better in Countries with Strong Sector Policies 23 3.3 Successful Projects Alone Do Not Ensure Sector-Level Impact 25 3.4 Setting Unrealistic Targets Leads to Perceptions of Failure, Even When Substantial Results Are Achieved 34 4.1 Government Process Quality Indicators Improved in About Half of 35 Countries with Bank-Supported Public Sector Reform Programs 35 4.2 But Governance Perception Indicators in 35 Bank Borrowers with Public Sector Reform Programs Did Not Change Significantly Table 11 2.1 Country Assistance Outcomes and Poverty Changes iv ACRONYMS AND ABBREVIATIONS APL Adaptable Program Loan ARDE Annual Review of Development Effectiveness BEEPS Business Environment and Enterprise Performance Survey CAE Country Assistance Evaluation CAS Country Assistance Strategy CDD Community-driven development CFAA Country Financial Accountability Assessment CPIA Country Policy and Institutional Assessment DECRG Development Economics Research Group EBRD European Bank for Reconstruction and Development ED Education EMT Energy and mining ENV Environment EP Economic policy EU European Union FY Fiscal year GDP Gross domestic product GIC Global information and communications FSP Financial sector policy HIPC Heavily Indebted Poor Countries (Initiative) HNP Health, nutrition, and population IBRD International Bank for Reconstruction and Development ICR Implementation Completion Report ICRG International Country Risk Guide IDA International Development Association IEG Independent Evaluation Group IFPRI International Food Policy Research Institute KDP Kecamatan Development Project KKM Kaufmann, Kraay, Mastruzzi LICUS Low-Income Country Under Stress LIL Lending and Innovation Loan MDG Millennium Development Goal NGO Nongovernmental organization OECD Organisation for Economic Co-operation and Development OED Operations Evaluation Department (renamed IEG, December 2005) PETS Public Expenditure Tracking Survey PIU Project Implementation Unit PMU Project Monitoring Unit PPAR Project Performance Assessment Report PPP Purchasing power parity PREM Poverty Reduction and Economic Management Network (World Bank) v A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S PRSC Poverty Reduction Support Credit PRSP Poverty Reduction Strategy Paper PSAL Programmatic Structural Adjustment Loan PSD Private sector development PSG Public sector and governance QAG Quality Assurance Group (World Bank) RDV Rural development SP Social protection SWAp Sectorwide approach TI Transparency International TR Transportation UD Urban development USAID United States Agency for International Development WDI World Development Indicators WDR World Development Report WSS Water supply and sanitation WUA Water users association OED changed its name to Independent Evaluation Group in December 2005. vi Acknowledgments This review was prepared by a team led by Stephen Knack, Gregory Kisunko, Stefan Monika Huppi. The core team comprised Koeberle, Frannie Leautier, Luc Lecuit, Danny Shonar Lala, Mirafe Marcos, and Rupa Leipziger, Richard Messick, Gisu Mohadjer, Julia Ranganathan. Shalini Ishwar Ahuja, Patricia Y. Nielson, Aloysius Uche Ordu, Praful Patel, Patti Chen, and Nara Meli provided research Petesch, Philip Schuler, Sándor Sipos, James assistance. William B. Hurlbut and Caroline Stevens, Eric Swanson, Richard Tobin, John McEuen edited the report. Yezena Yimer Underwood, Roberto Zagha, and Sally Zeijlon. assisted in the production of the report and Julia Akumu Ooro provided administrative support. Assistance by Shaohua Chen, Prem Sangraula, and Kathleen Beegle with the calculation of Contributions from numerous Independent poverty measures based on the Development Evaluation Group (IEG) staff are gratefully Research Group’s Povcalnet database is acknowledged. In addition, helpful comments gratefully acknowledged. were received from the following Bank staff: Surendra Agarwal, Ian Bannon, Deepak Peer review was provided by Carol Lancaster and Bhattasali, Francois Bourguignon, Louise Cord, Manuel Penalver-Quesada. Aline Coudouel, Sebastian Dessus, Shantayanan Devarajan, Brigitte Duces, Shahrokh Fardoust, The review was prepared under the direction of Marianne Fay, Ariel Fiszbein, Faezeh Foroutan, Victoria Elliott, Manager of IEG, Corporate Alan Gelb, Philippe Le Houerou, Homi Kharas, Evaluation and Methods. Director-General, Evaluation: Vinod Thomas Director, Independent Evaluation Group–World Bank: Ajay Chhibber Manager, Corporate Evaluation and Methods: Victoria Elliott Task Manager: Monika Huppi vii Foreword T he aim of the World Bank’s results agenda is to design country assis- tance programs and development projects that go beyond delivering assistance, to ensuring poverty reduction in socially and environmen- tally sustainable ways. Effective and sustained reductions in poverty result from a combination of sustained economic growth, policies and investments that improve income distribution, and the delivery of services to the poor. Achiev- ing this combination of outcomes, in turn, requires capable public institutions that are accountable to their stakeholders for the results they achieve. This Annual Review of Development Effective- The review identifies three important determi- ness (ARDE) assembles evidence from the nants of the Bank’s effectiveness in helping recent work of the Independent Evaluation countries reduce poverty. First, it is important to Group of the World Bank to explore the record understand the nature of growth and to identify of countries and the Bank in helping achieve the binding constraints to growth that creates results along the poverty reduction results jobs and reaches regions where many of the poor chain. It looks at three core questions. First, how are concentrated. It is also important to identify efficiently has growth translated into poverty factors that hinder intersectoral mobility of the reduction in Bank-assisted countries, and what poor. Second, the results chain is a valuable but factors have contributed to the effectiveness of underused tool that can help ensure that Bank assistance toward achieving this result? objectives are realistic and that the crucial cross- Second, what factors help interventions lead to sectoral constraints to their achievement are high-quality development results in sectors that considered. Finally, a realistic assessment of the deliver services to the poor? Third, what types of political economy of governance-related reforms Bank assistance have helped raise the accounta- increases the effectiveness of interventions bility of public institutions in charge of deliver- aimed at strengthening the accountability of ing and sustaining results? public sector institutions. Vinod Thomas Director-General, Evaluation ix Executive Summary A results-based approach to development tracks the contributions to progress in reducing poverty, going beyond questions concerning the delivery of development assistance. Effective and sustained reductions in poverty, the evidence shows, result from a combination of sustained economic growth and policies and investments that affect income distribution and the delivery of services to the poor. Achieving this combination of outcomes re- quires capable public institutions that are accountable to their stakeholders for the results they achieve. This Annual Review of Development Effective- • Effective programs have a twofold focus: they ness (ARDE) brings together evaluative evidence emphasize both the ingredients of growth and from the recent work of the Independent the measures that help the poor share in the Evaluation Group of the World Bank to address growth process. three questions surrounding this results chain • They build on a realistic and well-informed as- in countries, with a particular focus on the sessment of the political commitment and ca- Bank’s role in the chain: pacity of the recipient to deliver results, and they emphasize coalition and capacity building • How effectively has economic growth trans- to help attain results. lated into poverty reduction in Bank-assisted • They combine sustained engagement with countries, and what factors have affected these clear intermediate milestones. results? • Finally, they emphasize improved transparency • What factors have led to high-quality results in and local control of public institutions, factors areas that deliver services to the poor? that spur these institutions to deliver results. • What measures help raise the accountability of public institutions responsible for delivering Some Growth Patterns Reduce Poverty and sustaining results? More Effectively than Others Economic growth over the past decade has led The report identifies features that characterize to substantial poverty reduction in many East the country experiences and assistance pro- and South Asian countries, and more recently in grams that have delivered results: the transition economies of Eastern Europe and xi A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Central Asia. Impressive advances in the world’s it was less successful in bringing about job- most populous countries, China and India, have creating growth. In Madagascar, for example, the been at the forefront of the reduction in global Bank’s assistance strategy included putting the poverty. This progress notwithstanding, poverty overextended public sector on firmer ground reduction remains a formidable challenge for and establishing the preconditions for private many Bank borrowers. sector growth. It focused on sectors with high growth potential that would allow for relatively The growth performance of World Bank borrow- quick payoffs, but their impact on poverty was ers has strengthened over the past five years, but limited. In Georgia, the oil transport sector was achieving the sustained income growth essential a major driver of growth, but it created little to poverty reduction remains a challenge for employment. The Bank Group’s assistance many. Only 2 in 5 borrowing countries recorded helped reestablish macroeconomic stability and continuous per capita income growth during the contributed to growth in the oil transport sector, 5 years between 2000 and 2005, and just 1 in 5 but was less successful in helping to remove did so for the full 10 years from 1995 to 2005. obstacles to more broad-based growth. Countries that have posted strong growth have The Bank has found it challenging to help exhibited better policies and institutions than countries formulate and implement strategies slow growers. The strongest growers have better that effectively reduce rural poverty. Half of the records of both economic management and Country Assistance Strategy reviews completed policies for social inclusion than do moderate or by IEG over the past four fiscal years concluded slow growers. This indicates that high growth that the Bank’s assistance in rural areas had can be achieved alongside policies for social either not led to satisfactory outcomes or that inclusion. rural poverty reduction required increased attention. High and sometimes worsening income inequal- ity has dampened the poverty-reducing effect of To support growth strategies that more consis- growth in a number of countries. This effect was tently translate into poverty reduction, the Bank particularly noteworthy where growth was and its partners will need to further strengthen concentrated in sectors that generated little their understanding of what keeps the poor employment and where the poor lacked the from participating in growth in each country, basic skills or mobility to take advantage of the what prevents growth from reaching regions opportunities presented by growth. Growth and sectors where the poor are concentrated, delivers poverty reduction more effectively and how urban-rural linkages and intersectoral when it occurs in sectors and regions where mobility can be enhanced. most of the poor live and work and when it results in strong job creation. Achieving Results Requires Setting Realistic Objectives Strategies designed solely to boost overall The Bank’s assistance has been effective when it growth may miss opportunities to reduce has taken a realistic view of borrowers’ political poverty more effectively. In the countries and institutional capacity and has focused on reviewed by the Independent Evaluation Group well-specified objectives. But almost half of all (IEG) where growth did not result in poverty Bank Country Assistance Strategies reviewed by reduction, growth was concentrated in subsec- IEG in the past four fiscal years were found to be tors with low labor intensity and where few of overly ambitious in two distinct ways. They the poor could work. The Bank’s assistance in either lacked selectivity or they were founded these countries often effectively contributed to on unrealistic expectations for a reform program bringing the countries back to a growth path that was incommensurate with the country’s through improved economic management, but institutional capacity and political situation. xii E X E C UT I V E S U M M A RY Strategies that lacked selectivity caused the that contributed to the achievement of the Bank’s programs to spread their resources too country’s objectives for the sector. thinly across too many sectors, diminishing the impact of individual operations. Strategies based Balancing Long-Term and Short-Term on unrealistic expectations for reform led the Objectives Improves Results Bank to proceed with policy-based lending even Achieving high-quality development results when country conditions were not fully ready takes time, but pressure to show results quickly for the targeted reforms. Country Assistance can divert attention from the quality of results. Evaluations suggest that several factors can help For instance, the Millennium Development Goal determine ex-ante whether an assistance of ensuring universal completion of primary strategy is realistic or not, including the education by 2015 has spurred massive efforts country’s record with reform implementation to increase enrollments. These rapid increases and realization of the Bank’s assistance program, are welcome, but in many countries they have judicious analysis of the country’s political come at the expense of attention to learning economy and implementation capacity, and outcomes. In Uganda, for example, access to clear identification of country risks. education has improved greatly, but there are now 94 children per classroom and 3 students Unrealistic objectives can also be found in have to share a single textbook. Yet the experi- individual lending operations. For instance, ences of Ghana, India, and Uruguay have shown many financial sector loans in crisis countries that it is possible to combine increased access have had unduly ambitious objectives, driven by with gradual gains in learning outcomes. This an overestimate of the government’s commit- requires careful strategic planning and a strong ment to reform and a need to justify large loans. commitment to focus on learning outcomes Realistic and well-defined objectives, in contrast, from the outset. However, only about one-third can produce results when stakeholders focus on of operations in the primary education sector them. The Bank’s support for Bolivia’s health assessed by IEG explicitly aimed to improve sector, for example, focused for a decade on learning outcomes. infant and maternal health services and con- tributed to marked improvements in health In post-conflict countries, the pressure to show outcomes for poor mothers and children. quick results is especially intense, but haste may lead to the neglect of the institution building Achieving Sector-Level Impact Requires that is vital for recovery. In Timor-Leste, for More than Satisfactory Project Outcomes example, three community empowerment The performance of the Bank’s portfolio has projects supported by the Bank financed improved over the past five fiscal years, with impressive amounts of local infrastructure, but over three-quarters of completed operations too little attention was given to the development meeting their stated objectives. However, of durable local institutions.1 Country Assistance Evaluations show that satisfactory project outcomes alone do not A judicious combination of long-term objectives ensure country sector impact. Careful selection and interventions that yield quick and visible and phasing of interventions; long-term engage- results has proved effective. For example, Bank ment; and the complementarities of lending, assistance to the education sector in Ghana analytical work, and policy dialogue are factors combined support for policy reforms with that lead to impact on the sector as a whole. funding for school buildings, furnishings, and Bank-financed operations have yielded good teaching materials over 15 years. This sustained results when they have supported a country- approach has helped produce a stream of formulated, broadly owned sector strategy with physical improvements that have helped garner clear objectives, and when they have followed a support for reforms needed to expand access, distinct pathway designed to reach milestones while gradually improving learning outcomes. xiii A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S The long time required to achieve many of the multisector operations are not always an effective intended results underlines the importance of vehicle for achieving sectoral results. The sectoral continuity of donor engagement and of defining impact of multisector operations has tended to what is feasible for a single operation to achieve. be weaker than that of sector-specific operations, Frequent shifts in emphasis of Bank assistance partly because multisector operations allow for risk reducing its effectiveness. less-intensive engagement of Bank sector teams with country line agencies. In the financial sector, Strong Results Demand Attention for example, the outcome of loans overseen by to Cross-Sectoral Synergies the Bank financial sector departments was Achieving results in a given sector often requires substantially better than that of financial sector that constraints in other sectors be identified components of multisector loans. and removed as well. In Bangladesh, for example, Bank support for female secondary A combination of policy-based lending (which is schooling and rural electrification significantly often multisectoral) and sector-specific opera- contributed to reductions in child mortality, tions can deliver good results. In Armenia and alongside health sector interventions. Ghana, for example, the Bank effectively used development policy lending to support reforms The countries and the Bank need to pay more in the education sector, while parallel invest- attention to such complementary effects. The ment projects helped build the systems and impact of infrastructure investments financed by capacity to implement the reforms. community development projects, for example, has often been diminished by lack of attention Perceived Governance Quality Has Not to inputs such as teachers, doctors, and Yet Responded to Large-Scale Public medicines. Similarly, Bank-supported pension Sector Reforms reforms have sometimes failed to achieve the Achieving and maintaining results requires desired results because not enough attention public sector institutions that are accountable to was given to ensuring that the complementary stakeholders. Bank Country Assistance Strate- macroeconomic, financial, and institutional gies accordingly put substantial emphasis on conditions were in place. The workings of the strengthening performance and accountability Bank’s matrix management structure do not in the public sector. The bulk of the Bank’s provide staff with enough incentives to work support has taken the form of reform programs across sectoral boundaries and address cross- in public administration and public financial sectoral issues. management. More attention also needs to be given to the This assistance has led to improvements in the impact of reforms on different income groups, quality of public sector management processes because not all pro-growth policies are distribu- in some countries, but has not yet translated tionally neutral. In the area of trade reform, for into improvements in the perceived quality of example, the Bank often failed to conduct governance in most of these countries. Yet sufficient analysis to inform its policy advice and recent progress in perceived governance quality lending about the employment and poverty in some countries in Eastern and Central Europe effects of reforms. A full assessment of the distri- shows that it is possible to make progress in a butional impact of proposed reforms in a limited time when there is strong country country often requires analysis that reaches commitment. beyond the sector in which the reforms are carried out. Evaluation suggests that public sector reform initiatives have not always been aligned with Even though achieving a particular sector goal political circumstances. They have focused on may require a multisectoral approach, large new legislation and institutions, while overlook- xiv E X E C UT I V E S U M M A RY ing enforcement. They have also tended to capacity. Typically, Bank assistance programs overlook the interface between the public and have emphasized legal and regulatory frame- private sectors, even though regulatory reforms works for the financial sector, but they have have often been found effective against underestimated the time and human capacity corruption. required to enforce them. Civil Service Reforms Require Political Regulatory Reform Helps Beat Corruption Commitment The interface between the private and public Public sector reforms of a technocratic nature, sectors offers fertile ground both for corruption such as modernizing personnel practices, can and for combating it. Reforms to regulatory succeed when they build on political commit- regimes have made headway against corruption ment. Bulgaria’s achievement in professionaliz- even when they have not been part of compre- ing its civil service, for instance, has been the hensive anticorruption programs. In Turkey, for product of both donor-supported reforms in pay example, a Bank program for the energy sector and recruitment and broad political interest in supported the establishment of an independent meeting conditions for European Union (EU) regulatory agency that enabled sellers and buyers accession. But many reform programs have been of electricity to make contracts directly, without undermined by lack of political support. The involving government officials. It thereby sharply extent of political opposition is often underesti- limited the opportunities for officials to seek mated at the time of design. In Bolivia and kickbacks. Such sector-specific opportunities to Yemen, Bank-supported reforms in civil service combat corruption need to be more systemati- management achieved little, because there was cally exploited in Bank operations. no commitment to ending the traditional role of the public service as a vehicle for large- Transparency and Local Control scale patronage appointments. When political Encourage the Public Sector to Deliver conditions are not ready for wholesale reforms, Transparency is the foundation of good gover- it is advisable to proceed gradually, identifying nance, because access to information reduces the opportunities for less-contentious reforms in incidence of corruption, and transparent institu- order to build coalitions across affected interests tions earn the public’s trust. Bank operations have and to gradually gain momentum. helped bring more transparency to a variety of public management processes, including budget Anticorruption Measures Need formulation and execution, procurement, and Enforcement Mechanisms customs administration. In the Philippines and The Bank’s anticorruption efforts have helped Uganda, for example, the Bank has worked with support new laws and institutions in many governments to make the public procurement countries. But once established, they have often process more transparent. Civil society proved ineffective because they lack enforce- representatives have a mandate to observe the ment capacity. Anticorruption agencies, while tendering process in the Philippines, while important, have only a limited impact on the Uganda makes its final contract awards and prevalence of corruption when they are not fully related tendering information available on independent of those whose behavior they public Web sites. monitor. Local control and community participation can The need for enforcement capacity to properly make public sector institutions more account- implement legislation to improve transparency able. Bank operations support such local control and accountability reaches beyond anticorrup- in two main ways: by up-grading local govern- tion efforts. The implementation of prudential ment agencies and by channeling resources regulations and supervision in the banking directly to communities through community- sector has also suffered from low enforcement driven development projects. Such projects xv A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S have often established structures that parallel • A clearly articulated results chain: A well- those of local government, which has diluted articulated results chain allows Bank opera- efforts to foster decentralization.2 In Jamaica, for tions to ensure that objectives are realistic, instance, roads were built under community that cross-sectoral constraints to achieving development operations without adequate them are adequately considered, and that due involvement of the local councils that would attention is given to capacity building. Effective have to maintain them. There is now growing articulation and utilization of the results chain recognition in the Bank of the importance of also requires efforts to enhance country ca- strengthening the use of local systems while pacity to collect and use performance promoting community development.3 information. • A realistic assessment of the political economy Going Forward of governance-related reforms: The Bank can This ARDE finds three important areas where provide countries with the tools needed to the Bank can further strengthen its effectiveness strengthen government processes, and thereby in helping countries reduce poverty: to improve the governance environment, but effective use of those tools remains in the • A focus on the nature of growth: Poverty re- hands of country decision makers. Thus, re- duction will continue to require a strong focus forms to improve the accountability of public on growth. To ensure that growth translates ef- sector institutions require broad-based polit- ficiently into poverty reduction, the countries, ical support. When such support is absent, the Bank, and their partners will need to focus an incremental approach that allows mo- more on finding effective ways of helping the mentum for reforms to build can help deliver poor participate in this growth. This will require results. These reforms can be further en- country-level analysis of the binding constraints hanced with continued efforts to foster local to employment-creating growth and to growth demand for accountability through increased in regions where many of the poor live, as well transparency of government processes and as of the factors that hinder intersectoral mo- resource utilization. bility of the poor. xvi Management Comments: Summary W e very much value the 2006 Annual Review of Development Ef- fectiveness (ARDE) as part of the strong tradition of evaluation in the World Bank. Decision making within Bank management is made with the benefit of feedback from the Independent Evaluation Group (IEG), as well as from the Bank’s internal Quality Assurance Group (QAG). Man- agement prepared detailed comments on the 2006 ARDE as part of the back- ground for discussions with executive directors (first at the Board’s Committee on Development Effectiveness, and later at an informal meeting of executive directors). This note summarizes some of the main points of management’s comments, highlighting recent actions. Developing Country Growth. The ARDE amazing performance, given the challenges on paints an overly bleak picture of developing the continent, including HIV/AIDS, a high country growth and poverty reduction, failing to incidence of conflict-affected countries, and, fully reflect both strong global growth trends more recently, high oil prices. These impressive over the last five years and broadly favorable growth rates reflect the strong policy perform- prospects. Developing countries grew at 5–6 ance of developing countries, including in Africa percent between 2004 and 2006, even excluding (as noted in the 2003 ARDE and as has contin- fast-growing China and India. This year the ued since then). While management applauds number is expected to be 6.8 percent—a fifth their efforts, it knows that the World Bank must strong year of growth. Low-income developing do more still to help countries to reinforce and countries have done especially well, with high accelerate this trend. average annual income growth (more than 61⁄2 percent) over five years. Sub-Saharan Africa has Operational Quality. Management notes not been left out of this positive trend. Average with satisfaction the continued improvement in annual growth there in the last five years is 4.7 the performance of Bank-supported projects. percent. Over the past decade, two-thirds of The ARDE reports that for the cohort of Africans have lived in countries with a 5 percent operations completed in fiscal 2005 and or higher average growth rate. This is an evaluated so far, development outcomes were xvii A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S rated as satisfactory in 82 percent of operations World Development Report (WDR), Equity and (87 percent when weighted by disbursements), Development (World Bank 2005k), is under way sustaining a long-term upward trend in quality. and will reinforce attention to pro-poor growth. These strong results have been achieved with higher lending from the International Bank for The Challenge of Rural Poverty. Manage- Reconstruction and Development and increased ment shares IEG’s concerns with respect to rural credits and grants from the International poverty and has taken action in this area. Since Development Association (IDA). In fiscal 2006, the adoption in 2003 of the Bank’s new rural IDA provided a record $9.5 billion in support to strategy (World Bank 2003a), the Bank’s rural poor countries, with more than half going to staff report that CASs are stronger on rural Sub-Saharan Africa. IDA financing reflects a poverty, and, more important, country-owned performance-based allocation system that poverty reduction strategies now include rural rewards, and thus encourages, recipient poverty diagnoses. Lending for rural develop- countries’ effective use of funds for develop- ment is up; in fiscal 2006 it accounted for 14 ment and poverty reduction. percent of total lending commitments. IEG and QAG data show improvements in the quality of Successful Country Assistance Strategies Bank-supported rural projects. The 2007 WDR (CASs). The issues raised in the ARDE with currently under preparation will help hone respect to CASs, selectivity, and cross-sectoral further our knowledge on how best to support synergies are not new. Indeed, they reinforce rural and agricultural development. It will management’s determination to continue with provide more clarity on customizing agricultural its accelerated results agenda, and in particular strategies to specific country conditions and on to continue with its recently introduced results- dealing with the risks posed by heavy and often based CAS program (see World Bank 2006a). unpredictable government intervention in (Management would note that much of the agricultural export markets. analysis in the ARDE is based on outcomes of strategies, which raises a major question regard- Governance Issues. Management agrees with ing attribution: clearly, as the ARDE states, the much of the ARDE’s analysis on civil service Bank cannot take credit when poverty outcomes reform, transparency, and anticorruption. That are good; but the reverse is also true. Moreover, is why more recent operational support has the ARDE does not say much on donor focused on service-oriented approaches to harmonization and alignment—important improving governance and service provision. management objectives with the goal of improv- Management has also worked with development ing the quality of overall donor assistance to partners to devise and implement a public countries.) financial management performance measure- ment framework that is more objective than Distributional Impacts. Management takes previous perception measures (see World Bank the issue of the distributional impact of growth 2005g). Early experience with the tool is encour- seriously. That is the reason it has worked with aging. Management recently took the initiative other donors to pilot and mainstream Poverty to strengthen its strategy for Bank Group and Social Impact Assessments and has revised engagement on governance and anticorruption. its policy on development policy lending to take Consultations with partners are currently taking into account possible adverse poverty and social place, and management will be reviewing impacts and, as necessary, to include measures progress regularly with executive directors as to mitigate them. Operationalizing the recent the strategy is further refined. xviii Chairman’s Summary: Committee on Development Effectiveness (CODE) O n October 18, 2006, the Committee on Development Effectiveness dis- cussed the Annual Review of Development Effectiveness 2006: Get- ting Results (ARDE) and the Draft Management Response. 2006 ARDE effectiveness: focus on the nature of growth to Building on the 2004 ARDE that examined the ensure that the poor can benefit from economic Bank’s contribution to poverty reduction, the growth; clearly articulate the results chain to 2006 ARDE considered the factors at the ensure that objectives are set realistically and country, sector, and institutional levels that fa- that key cross-sectoral constraints to achieving cilitate or hinder poverty-reducing growth them are adequately identified and addressed; and effective service delivery to the poor, and and build on a realistic assessment of the politi- how the Bank has taken these elements into cal economy of governance-related reforms. account in its operations. Drawing on recent IEG evaluations of World Bank operations, the 2006 Draft Management Response ARDE considered: (1) how efficiently economic Management found many of IEG’s findings and growth has translated into poverty reduction, recommendations useful as inputs into improv- and what factors contributed to the effective- ing the development impact of Bank support. At ness of Bank support towards achieving these the same time, it noted that the Annual Review results; (2) what factors have helped interven- portrays a pessimistic view on growth and tions lead to high-quality development results in poverty reduction. Management also suggested sectors that deliver services to the poor; and (3) that the ARDE could have been more balanced, what types of Bank assistance have helped to reflect the current Bank approach to country increase the accountability of public institutions assistance strategies, the difficulties around responsible for delivering and sustaining results. strategy development in uncertain environ- The Annual Review identified three key areas ments, the achievements in terms of continuing where the Bank can further strengthen its improvements in quality at entry and exit of xix A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S operations with a growing volume of lending, management’s view that the ARDE presented an and the focus on service-oriented approaches to overly pessimistic perspective of countries’ improving governance and service delivery, and performances in reducing poverty, and it could recent work to strengthen engagement with have presented a more balanced picture of the client countries on governance. Bank’s contributions. One member felt that the ARDE’s objectives were ambitious and it may Overall Conclusions and Next Steps have presented a simplified picture of the The Committee appreciated the candid report complex development issues. A number of and welcomed the findings on the sustained speakers indicated it could have further differ- improvements in project quality, although much entiated between the poor, in particular to take remained to be done. Indeed, many members into account the gender dimensions. IEG felt that management need not be defensive or pointed to a number of country sector embarrassed about the results and should examples in the ARDE that illustrate how maintain a positive approach to learning from gender issues were addressed. It also indicated experience. willingness to undertake another review of gender, which had last been evaluated in 2000. Members discussed the strategic issue of quality Management reminded the Committee of the of economic growth for poverty reduction, and Gender Action Plan, and the forthcoming the ARDE observations related to potentials of Global Monitoring Report will also cover growth resulting in job creation and productiv- gender issues. The tendency to group African ity for the poor, and the challenges of rural countries together and then generalize the poverty reduction. There was general support findings, which were too broad, was also noted. for ARDE’s call for setting more realistic Several speakers suggested the report might objectives in country assistance strategies, as have better defined the methodology, terms, well as in individual operations. In this regard, and data used, regarding which there were a members highlighted the importance of solid number of comments and questions. analytical work (including political economy) and the involvement of local institutions and Focus on Quality of Growth. The ARDE finding that experts in this area; the need to balance short- growth more effectively reduced poverty when and long-term objectives; and the effective use it occurred in regions where most of the poor of the results chain to set realistic objectives and live and in job-creating sectors resonated with a identify sectoral and cross-sectoral constraints number of speakers. However, some speakers to reach the stated objectives. Members cautioned against focusing too narrowly on job- supported addressing cross-sectoral constraints creating sectors or “cherry picking,” and noted for strong sector results, including the issue of the need to address broader development issues matrix management structure and organiza- that may be more challenging to address. Some tional incentives to support work across sectors. members supported more attention to job The chapter on public sector accountability creation for the poor, and observed that the drew interest in the context of the Bank’s new ARDE could have elaborated on the role of the governance and anti-corruption strategy. Some private sector in creating jobs, and on contribu- speakers noted that the ARDE could have tions of International Finance Corporation (IFC) incorporated gender dimensions and aspects of and World Bank Group synergies in this area. harmonization. IEG noted that ARDE covered the World Bank activities, while IEG-IFC prepares a separate The main issues raised during the meeting were review of IFC activities. It also indicated that it the following: was looking at ways to review the World Bank Group synergies in its evaluation work. Several ARDE’s Findings. The Committee appreciated the speakers commented on the need to strengthen candid report, although a few members echoed Bank support for rural development, taking into xx C H A I R M A N ’ S S U M M A R Y: C O M M I T T E E O N D E V E L O P M E N T E F E C T I V E N E S S ( C O D E ) account the rural-urban linkages. One member connection, speakers supported a better- sought information about the weaknesses and articulated and more realistic results chain. constraints of Bank assistance in rural develop- Management assured the Committee that it is ment, while another asked about the Bank plans using a results chain, which is a central ele- to improve outcomes. Management viewed ment of the Results-Based CAS. It also expected rural development as an important area for that the increased emphasis on impact analy- support, noted that the results of the new rural sis would also help strengthen the results chain. strategy cannot yet be measured, and also • Policy-Based Lending: A member consid- remarked on the need to address the overall ered realistic country assessment critical for trade environment. Deepening country-specific policy-based lending and sought to ensure knowledge of the constraints to the poor’s staff incentives and support in this area. An- participation in the benefits of growth and the other member urged closer monitoring of the distributional impact of growth-enhancing implementation and results of policy-based reforms was emphasized. lending. Management noted that the bulk of the Bank’s development policy support has Selectivity and Setting Realistic Objectives. gone to countries with higher Country Policy Generally, speakers supported ARDE’s call for and Institutional Assessment (CPIA) ratings, greater realism in country assistance strategies, indicating a better policy and institutional as well as in individual operations. The results- environment. based Country Assistance Strategies (CASs) were considered key to selectivity and setting realistic Enhancing Sectoral Impacts through Cross-Sectoral objectives. Management broadly agreed, but it Synergies. The Committee concurred with the also echoed the view of a few speakers that need to overcome cross-sectoral constraints for optimism and ambition are also needed in strong sector results, which IEG elaborated on, development. referring to a number of examples. Two members noted the transaction costs associated • Country analytical work: Speakers stressed with coordination and implementation, and the solid country analytical work (including polit- need to work across line agencies at the country ical economy) to underpin country strategies level. Of particular concern to most speakers and to support greater selectivity and proper were the constraints posed by the matrix sequencing of Bank assistance, especially in management structure and staff incentives to Low-Income Countries Under Stress/fragile work across sectors, and management action in states. Some speakers urged more involve- this area. Different views were expressed in ment of local institutions and experts to pro- response to a member’s question as to whether mote country ownership. Management cross-sectoral work was limited by the organiza- remarked on its increased efforts to consider tional structure or by lack of appropriate the political context and country ownership incentives. Management did not consider issues. In order to be effective, members em- structure as the main underlying issue limiting phasized the need to strengthen country ca- work across sectors. It commented on the need pacity for data collection and analyses, and for for behavioral changes, observing that limited monitoring and evaluation. A few remarked cross-sectoral work may also be attributed to on the importance of integrating technical as- the Bank hiring professionals with strong sector sistance for capacity building in the overall as- skills. sistance strategy. • Short- and Long- Term Objectives: The Strengthening Public Sector Accountability. Speak- need to balance short- and long- term objec- ers stressed the importance of country tives and the importance of sequencing, es- ownership, broad-based support, political pecially when countries face capacity issues commitment, and the long-term perspective and/or political resistance, was noted. In this needed in improving governance, which were xxi A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S highlighted in the ARDE. In this context, one of public financial management in the Africa member cautioned about Bank’s involvement in Region, management said that the actual coalition building when political conditions were results for public administration reform had not ready for broad-based public sector reform. been mixed, while that of public expenditure IEG clarified that better results have been associ- and financial management was stronger. A ated with an incremental approach to public member supported ARDE’s point that sector sector reform, taking into account the country’s initiatives having a positive impact on gover- political situation to allow governments to align nance should be further encouraged, even when support for its reforms. Other speakers stressed they are not labeled as “governance support.” the importance of working with local govern- ments in community-driven development initia- Harmonization and Coordination. Several speakers tives. Both management and IEG remarked that referred to the limited Bank resources, the need more recent community-driven development for the Bank to complement other donors’ operations are taking into account the need to efforts, and the importance of donor coordina- work closely with local governments. tion and harmonization. Some suggested ARDE could have considered whether harmonization, The unchanged perceptions of governance in line with the Paris Declaration on Aid quality at the country level, despite the Effectiveness, led to better overall results at the implementation of Bank-supported public country level. IEG recognized the importance of sector reforms, were of concern to some reviewing the extent to which harmonization is speakers. Some speakers questioned the useful- contributing to better results. It emphasized ness of the governance indicators, and one that such a review should be done jointly with member urged extra care in the underlying other partners, which is a challenge. IEG methodology of the CPIA and World Bank mentioned that a joint evaluation of donor Institute governance indicators. A few speakers harmonization efforts is being contemplated expressed interest in the factors that support under the auspices of the OECD/DAC. Manage- societal and attitudinal changes, for which one ment mentioned it was preparing a report on member proposed establishing process indica- harmonization and aid effectiveness, which tors. Responding to a question about the record will soon be considered by the Board. Pietro Veglio, Chairman xxii 1 Introduction T he desire to strengthen the effectiveness of development assistance and to demonstrate results has intensified over the past several years. High- level international meetings, such as the Monterrey Forum on Har- monization, the Marrakech Roundtable on Managing for Development Results, and the Paris Forum on Aid Effectiveness have highlighted the need to shift from a focus on development inputs to the achievement of verified outcomes. The World Bank’s way of measuring its corporate sector strategies. But availability of operational performance has also evolved, credible information needed to monitor reflecting this shift in the external environment. progress toward results, and ultimately to In the early 1990s, tracking of the number of measure the development impact of particular projects and lending amounts was supple- interventions, remains a formidable challenge mented by a focus on improving the quality of in many client countries. Efforts to help Bank operations. Attention has turned more countries build capacity to generate, analyze, recently toward the achievement of results as an and use data for decision making and policy indicator of the Bank’s effectiveness. In parallel, formulation have been stepped up to support the Bank’s lending has shifted from lending the results agenda. These efforts remain a work mainly for physical investments toward more in progress. emphasis on strengthening the quality of institu- tions as a core ingredient to achieving high- Despite the challenges of measurement, impact results. evidence is emerging from project, country, sector, and thematic evaluations about the Development assistance is only effective if it factors that increase the likelihood of achieving produces results that have a positive impact. meaningful results. Drawing on this evidence, Substantial effort has gone into setting targets the 2006 Annual Review of Development and developing indicators to monitor progress Effectiveness (ARDE) examines the effectiveness toward achieving results within the agenda of of Bank support in helping countries achieve the Millennium Development Goals (MDGs), high-impact results. Like previous ARDEs, this in the framework of IDA14 (International report draws primarily on Independent Evalua- Development Association, fourteenth replen- tion Group (IEG) evaluation findings for Bank ishment), and more recently for the Bank’s programs and projects. 1 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S The 2002 ARDE, Achieving Development translates into poverty reduction. Poverty Outcomes: The Millennium Challenge, found reduction also requires that the poor be able to that while the MDGs can help sharpen the avail themselves of services that equip them to development effort by focusing achievement on take advantage of the opportunities arising from quantified and time-bound development growth. Achieving poverty-reducing growth and targets, they also pose an inherent risk of effective service delivery requires capable public nonattainment. The 2004 ARDE, The World sector institutions that are accountable to Bank’s Contribution to Poverty Reduction, stakeholders for the results they achieve. Against found that linking the Bank’s country-level this backdrop, the ARDE considers three interventions to poverty reduction requires a questions: sharper results focus. This year’s ARDE builds on these findings by looking at what factors • How efficiently has growth in countries as- facilitate or hinder the achievement of high- sisted by the Bank translated into poverty re- impact results at the country, sector, and institu- duction, and what factors have contributed to tional levels and how the Bank has taken those the effectiveness of Bank assistance toward factors into consideration. achieving this result? • What factors help interventions lead to high- The World Bank’s overarching objective is to quality development results, particularly in help countries reduce poverty in socially and sectors that deliver services to the poor? environmentally sustainable ways. To achieve • What kinds of Bank assistance have helped this objective, countries must attain sustained raise the accountability of public institutions in growth and ensure that such growth effectively charge of delivering and sustaining results? 2 Chapter 2: Evaluation Highlights • Achieving poverty reduction remains a substantial challenge because growth has remained uneven and sporadic and has translated into poverty reduction with varying efficiency. • Strategies for poverty reduction need to consider where the poor live and how they earn income. • Strengthening urban-rural linkages and strategies to improve rural productivity requires more attention. • The distributional effects of growth-enhancing reforms need to re- ceive more attention. • Assistance strategies have been effective when they have taken po- litical and institutional capacity into account. When they have not done so, they have been overly ambitious and lacked focus. 2 Achieving Poverty- Reducing Growth W hen focusing on results at the country level, the World Bank’s over- arching objective is poverty reduction. Economic growth is essen- tial for poverty reduction, yet all growth is not equal in improving the welfare of the poor. Effective poverty reduction also requires that low- income groups be able to participate in economic growth. This chapter reviews how well countries assisted elude a considerable number of countries. The by the Bank have fared in achieving poverty- aggregate picture of growth and income in client reducing growth and examines the factors countries over the past decade has three main contributing to Bank effectiveness in getting this characteristics: result. The chapter draws on IEG Country Assistance Evaluations completed in fiscal • Middle-income countries were more 2003–06, as well as IEG analyses of Country likely to grow than low-income coun- Assistance Strategy Completion Reports from tries. Almost 90 percent of middle-income fiscal 2004–06, for a total of 48 countries. It also countries achieved a positive annual per capita uses poverty data available from the Bank’s income growth rate during 1995–2005, and Povcalnet database over the periods covered by about three-quarters of low-income countries the IEG evaluations. Such information is did so (figure 2.2). available for 25 of the 48 countries. The chapter • There are still many slow growers. Only will refer to the group of 48 countries as the full two of five countries were able to increase per sample and the group of 25 countries as the capita income at an average annual rate of 2.5 subsample. percent or better during 1995–2005, and one in seven countries had a negative average an- Economic Growth—the Recent Record nual per capita income growth rate during this Growth in the majority of Bank borrowers has time (figure 2.2). been stronger in the most recent five years than • Among countries that did grow, sus- in the first half of the decade from 1995 to 2005 taining growth was a challenge. Only 2 (figure 2.1). Nevertheless, achievement of of 5 countries recorded continuous per capita sustained increases in per capita income, income growth during 2000–05, and just 1 in essential for poverty reduction, continues to 5 did so for a full 10 years from 1995 to 2005. 5 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure 2.1: Growth Performance Has Improved, but Sustained Income Growth Remains a Challenge for Many Bank Borrowers Average annual GDP per capita >6 growth rate (%) 4 to 6 2.5 to 4 1 to 2.5 0 to 1 <0 0 10 20 30 40 Number of countries 1995–2000 2000–05 Source: World Bank 2006n. Figure 2.2: Middle-Income Countries Were More Likely to Experience Strong Growth than Low-Income Countries (1995–2005) 30 Number of countries 25 20 15 10 5 0 High growth Moderate growth Low growth Low income Lower-middle income Upper-middle income Source: World Bank 2006n. Note: High growth = average annual per capita GDP growth rate of > 2.5 percent; moderate growth = average annual per capita GDP growth of 0 –2.5 percent; low growth = average annual per capita GDP growth < 0 percent. High-growth countries Sustained and sustain- structural, social, and environmental policies and able growth is the result institutions. have stronger policies of multiple interlinked and institutions than factors, including a coun- The overall quality of policies and institutions (as countries with slower try’s stock of physical, measured by the Country Policy and Institutional growth. human, and environ- Assessment, or CPIA, score) has been stronger in mental capital and the moderate- and high-growth countries than in efficiency with which it is low-growth countries. The performance gap of formed and used. Efficient capital formation, in the slower-growth countries is most acute in turn, depends on the quality of macroeconomic, their quality of economic management. High- 6 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H Figure 2.3: Countries with Higher Growth Have Stronger Economic Management and Better Policies for Social Inclusion and Equity than Slow Growers 4.00 3.50 CPIA 3.00 2.50 2000 2005 2000 2005 2000 2005 2000 2005 2000 2005 Economic management Structural policies Policies for social Public sector Overall inclusion/equity management performance High growth Medium growth Low growth Source: World Bank CPIA database. Note: CPIA = Country Policy and Institutional Assessment. growth countries also have stronger policies for • An increase in inequality will lower the poverty- social inclusion than moderate- and low-growth reducing effect of growth. countries (figure 2.3), which demonstrates that • Poverty in richer and more unequal countries high growth can be achieved alongside policies responds more strongly to changes in in- for social inclusion. equality than to changes in income. • Poverty in poorer and more equal countries re- How Quality Growth Delivers Poverty sponds more strongly to changes in income Reduction than to changes in inequality. How effectively economic growth leads to the reduction of income poverty depends on how Cross-country regression The extent to which much the poor participate in growth. Increases suggests that for a given growth reduces poverty is in income and decreases in inequality can both level of income and lead to poverty reduction. The efficiency with inequality, growth trans- influenced by the initial which economic growth translates into poverty lates more efficiently into levels of income and reduction depends on three factors: the initial poverty reduction in coun- inequality and the level of income, initial inequality, and whether tries with high literacy growth is accompanied by changes in inequality rates, low regulatory bur- pattern of growth. or not (Bourguignon 2004a, 2004b; Lopez and dens, and broad access to Serven 2006; Ravallion 1997, 2004; World Bank credit for the private sector (Chhibber and Nayyar 2005f).1 Specifically: 2006). • A lower-income country will need to grow Higher literacy facilitates poverty reduction faster to achieve the same poverty reduction because it increases the share of the popula- as a higher-income country with the same in- tion that can take advantage of better employ- equality level. ment opportunities created by growth, while at • A country with high inequality will need to the same time providing entrepreneurs with a grow faster than a country with a more equal lar-ger pool of skilled labor. Broad access to income distribution to achieve the same credit and a lower regulatory burden can facili- poverty reduction. tate entrepreneurial investment, which in turn 7 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure 2.4: Policies and Investments that Enhance Policies and investments that take these factors Equity Strengthen the Growth Effect on Poverty into consideration can help enhance the efficiency with which growth reduces poverty (figure 2.4). The extent to which each of these factors constrains the ability of low-income groups to participate in growth depends on Pro-poor growth Initial inequality country-specific conditions, and thus calls for country-level analysis to help guide the formula- tion of growth strategies that translate efficiently into poverty reduction. Reduction of Income Poverty in Bank-Assisted Countries Poverty response to growth The share of people living on less than one dollar a day declined from 28 percent to 19 percent Source: Based on Chhibber and Nayyar 2006. between 1990 and 2002 (World Bank 2006n). The global reduction in the number of poor people has helps create employment. Access to credit can been driven by impressive advances in East and also help low-income groups hedge against South Asia, particularly China and India (box 2.1).2 risk. To see how well Bank-assisted countries have Other factors found to increase the efficiency of fared in achieving poverty-reducing growth growth in reducing poverty are investments that since the mid-1990s and how effectively Bank increase the poor’s access to infrastructure, assistance has contributed to this result, we use measures that ensure market access for rural the subsample of 25 Bank-assisted countries for producers, investments in productivity-increasing which IEG has recently assessed the outcome of agricultural technologies, and labor market the Bank’s country assistance and for which regulations that create attractive employment comparable poverty data are available from opportunities for poor workers (World Bank Povcalnet for at least two periods between the 2005f). mid-1990s and the early 2000s.3 Figure 2.5: Poverty Reduction Remains a Significant Challenge, Even in Countries with Positive Growth Rates 10 9 9 8 Number of countries 7 7 6 5 4 4 4 3 2 1 1 0 High growth Moderate growth Low growth Poverty stagnated or increased Poverty reduced Sources: Povcalnet, DECRG poverty database, World Bank 2006n. Note: High growth = average annual per capita GDP growth rate of > 2.5 percent; moderate growth = average annual per capita GDP growth of 0–2.5 percent; low growth = average annual per capita GDP growth < 0 percent between survey years. 8 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H Box 2.1: China and India Account for a Large Share of the World’s Poverty Reduction The reduction in the absolute number of the world’s poor has India largely been driven by poverty reduction in the world’s most In India, there is broad agreement that poverty declined in the populous countries, China and India. 1980s, although there is much dissent on the extent of poverty reduction in the 1990s. The official statistics indicate a 10- China percentage-point drop in poverty—from 36 percent in 1993/94 to No country has been more successful at reducing poverty in the 26 percent in 1999/2000. past quarter-century than China. Between 1990 and 2005 alone, India’s economy has performed impressively since the liber- the number of people living on less than $2 a day fell by over 400 alization of trade and industrial policies in the early 1990s. This million. While migration to urban areas has helped reduce growth has enabled India to reduce poverty but has led to in- poverty nationally, the bulk of the poverty reduction through the creasing income inequality that acts as a constraint to higher mid-1990s came from within the rural areas, fuelled by policy re- growth and stronger poverty reduction. The aggregate growth rate forms and technologies that increased productivity. during the 1990s could have reduced poverty in India even more Since the late 1980s, however, an increase in vulnerability and had growth been more balanced, sectorally and geographically. inequality has dampened the poverty-reducing effect of growth. Poverty is increasingly concentrated in lagging regions where The Bank made several contributions to poverty reduction through growth rates are substantially lower than in the rest of the its analytical and advisory services, poverty monitoring, and in- country. vestments in agriculture, health, and transport. Most important, the The Bank provided strong support for the reforms of the early Bank has helped establish successful models of targeted inter- 1990s. It expanded assistance to the social sectors and devoted ventions through integrated rural development projects. more attention to improving participation. After the mid-1990s the The Bank has been less successful in persuading the govern- Bank focused assistance on reforming states, with a notable ment of the implications of broader development policies for measure of success. In the late 1990s, it sharpened its focus on poverty and inequality. The mismatch between intergovernmen- poverty reduction and governance. Overall, however, the Bank had tal fiscal resources and responsibilities has exacerbated regional limited impact on fiscal and other structural reforms and failed to inequality, while migration restrictions have limited economic develop an effective assistance strategy for rural poverty reduc- integration. tion through much of the 1990s. Sources: Datt and Ravallion 2002; Deaton and Kozel 2005; Devarajan and Nabi 2006; IEG 2001, 2005f; Ravallion and Chen 2004; World Bank 2006e. Despite impressive advances in many East and average annual growth Poverty reduction South Asian countries over the past decade, rate in per capita gross remains a formidable poverty reduction remains a formidable domestic product (GDP) challenge in many of the countries assisted by of 2.5 percent or higher, challenge in many of the the Bank. Of the 25 countries in the subsample, except for Brazil. But in countries assisted by the only 11 reduced the incidence of poverty some countries growth Bank. between the mid-1990s and the early 2000s, did not efficiently trans- while poverty either stagnated or increased in late into poverty reduc- the remaining 14 countries. Lackluster and tion, because growth and changes in distribution volatile growth was a major reason for the did not work in the same direction (figure 2.7). limited poverty-reduction progress in the latter group (figures 2.5, 2.6). In the recovering transition economies of Armenia, Moldova, and Ukraine and in African Distributional changes are important for countries that saw substantial poverty poverty reduction reduction, the poverty-reducing effect of growth Growth has been an important driver of poverty was reinforced by a reduction in inequality. By reduction in all of the sampled countries where contrast, in China, Lithuania, Sri Lanka, and poverty dropped. These countries experienced an Romania and in several Latin American 9 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure 2.6: Of 25 Sampled Countries, 11 Witnessed Poverty Reduction Between the Mid-1990s and Early 2000s 20 poverty headcount (%) 15 Annual change in 10 5 0 Ϫ5 Ϫ10 Ϫ15 Ϫ20 ) ) ) ) 03) 01) 2) 3) ) ) ldo 6–01) ) ) 03) 01) ) ) 2) law –03) uan 7–04) 03) 3) 2) ) ) –03 –03 Bol 96–02 –02 –02 –03 –03 –03 –05 –03 –01 Lan 8–04 8–0 Geo (98–0 Nig (95–0 8–0 7–0 dag s (97– Pak r (93– – Arm l (94– 96– 6 4 (97 Hon u (96 (98 Cam ne (96 (98 2 (98 8 (96 (96 n (9 a (0 i (9 a (9 9 . (9 9 9 a (9 9 ia ( ( ( ( ( zil ( y rgia aso an key bia ivia on va na ka eria ega Rep a a gua Per ista ani eni ani asc Bra dur Dom Jord ai ero Chi Zam Tur aF Ukr Alb Sen Rom Ma Uru an Mo Lith kin Sri inic Ma Bur Sources: Povcalnet, DECRG poverty database. Note: Poverty changes based on per capita consumption/income and a poverty line of US$1.08/capita/ for low-income countries and US$2.15/capita/day (in 1993 purchasing power par- ity $) for middle-income countries and transition economies. Poverty figures may differ from country poverty estimates because of the use of different poverty lines and purchasing power parity exchange rates, as well as use of per capita rather than adult equivalent consumption. countries, the positive effect of growth was distribution translated into poverty reduction, dampened by worsening income distribution. despite the lack of income growth. While growth accounted for most of the poverty reduction In some of the countries where poverty across the sampled countries, even seemingly increased, such as Bolivia and Georgia, negative small changes in income distribution con- household consumption growth was accompa- tributed substantially to either reinforcing or nied by an increase in inequality. Brazil was an dampening the poverty effects of growth exception, because improvement in the income (appendix D). Figure 2.7: Changes in Distribution Reinforced the Poverty-Reducing Effect of Growth in Some Countries, but Dampened It in Others Change in poverty rate (%) 20 15 10 5 0 Ϫ5 Ϫ10 Ϫ15 Ϫ20 Cam va on kin al na- o an a u ka e a eria ia dan Dom Chin ia Rep l ic key Hon il as ivia bia y n dag a ar ra gua as ani Per ain eni z ista gi ubl g uan an asc Bra ldo Lan dur ero urb a-ru e Zam r Tur aF Bol Jor Nig Geo Sen Alb Ukr Arm Rom Uru Pak Mo Lith Sri Chi Ma an Bur inic Change in poverty due to growth Change in poverty due to change in distribution Sources: Povcalnet, DECRG poverty database. 10 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H How Effectively Has World Bank efforts to maintain—and Growth was an Assistance Contributed to Poverty- at times reestablish— important driver of Reducing Growth?4 macroeconomic stability To explore the correlation between the effec- and fiscal discipline, in- poverty reduction, but tiveness of Bank assistance and poverty reduction, cluding better targeting of even seemingly small the subsample of 25 countries was divided public expenditures. changes in income into groups (table 2.1), depending on how well Bank assistance programs achieved their stated What makes for distribution either objectives. Not surprisingly, the successful successful country dampened or reinforced outcome of a Country Assistance Strategy was assistance programs? the effects of growth on correlated with the extent of poverty reduction Three characteristics achieved in a country, though this does not mean mark most of the assis- poverty. that poverty-reduction results can be attributed to tance programs in the Bank assistance alone. This assistance often subsample countries where Bank assistance led represents only a small portion of the factors that to satisfactory results: influence overall country performance. Further- more, while poverty reduction is the World Bank’s • The country assistance programs were selec- overarching goal, individual Country Assistance tive and in support of a government-owned Strategies set varying intermediate objectives program that the Bank had often supported toward contributing to this goal, and IEG through prior analytical work and policy evaluates the outcome against these strategy- dialogue. specific objectives. This explains why the outcome • The pace of program implementation was of the Bank’s assistance strategy was assessed as aligned with government capacity. satisfactory by IEG in 40 percent of the subsam- • Assistance strategies were based on analytical pled countries that did not succeed in reducing work, often done in collaboration with local poverty, while it was unsatisfactory in two specialists, which helped tailor them to coun- countries where poverty was reduced (table 2.1). try conditions. Countries that reduced poverty and where the The experiences of Successful assistance outcome of the Bank’s assistance was satisfactory Brazil, Burkina Faso, and programs were selective, (countries in the upper-left quadrant of table 2.1) China illustrate how are distinguished from those with poor outcomes important it is for the aligned with government (lower-right quadrant) by substantially stronger Bank’s assistance pro- capacity, and tailored to economic management policies. Strong economic gram to be focused and country conditions. management helped spur growth, which translated tailored to country con- into poverty reduction with varying efficiency. In ditions. In Burkina Faso, the recently completed these countries, the Bank supported government Country Assistance Strategy was closely aligned Table 2.1: Country Assistance Outcomes and Poverty Changes (number of countries) Country assistance outcome Satisfactory/ Moderately Poverty change moderately satisfactory unsatisfactory/unsatisfactory Poverty reduced 9 2 Poverty stagnated or increased 6 6 Source: Based on Country Assistance Evaluations and IEG analysis of Country Assistance Completion Reports. 11 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Box 2.2: Burkina Faso—Focused Country Assistance Strategy with Measurable Milestones Yields Results The Bank’s Country Assistance Strategy (CAS) in Burkina Faso interventions designed to improve the access of the poor to pub- was closely aligned with the government’s own strategy as out- lic services and to increase rural productivity. lined in the Poverty Reduction Strategy Paper (PRSP). To assess its effectiveness, the CAS identified 14 monitorable The Bank’s interventions in support of the PRSP were deliber- indicators and specific monitoring instruments for each indicator. ately selective and focused on well-specified objectives within each Progress was monitored against benchmarks, but IEG found that pillar of the government’s program. Assistance combined support further progress could be made by closely tracking intermediate for policy reforms to strengthen the competitiveness of the econ- indicators and ensuring measurement of the effects of IDA- omy, targeted public sector reforms, and specific poverty-targeted supported interventions on higher-order country objectives. Source: IEG analysis of World Bank 2005a. with the government’s own strategy and focused There are two main ways in which Country on well-specified objectives (box 2.2). Progress Assistance Strategies have been overly ambitious. was monitored against benchmarks and yielded They have either lacked selectivity or they have good results in the reduction of both income centered on Bank support for reform programs and non-income poverty. In China, the Bank that were not commensurate with institutional recognized that its lending accounts for only a capacity and the political situation in the country. small share of resource flow and relied mostly Forty-eight percent of the strategies IEG assessed on policy dialogue through analytical work and as overly ambitious fell in the former group and the demonstration effect of successful projects 36 percent fell in the latter, with the remainder to leverage policy outcomes (IEG 2005f). In suffering from a combination of the two. Brazil, the Bank’s strategy in the 1990s and early 2000s became more selective. The program’s Country Assistance Evaluations suggest that main component focused on the poor Northeast several factors can help determine ex-ante and on activities expected to address directly the whether an assistance strategy is realistic or not. roots of poverty, particularly lending in support These include past record with reform implemen- of human development and to improve access tation and realization of Bank assistance programs of the poor to basic infrastructure (IEG 2004d). (including dropped operations in previous assistance programs and implementation delays), Unsuccessful assistance strategies are judicious analysis of the political economy and frequently overly ambitious5 implementation capacity in a country, and clear Over-ambition is a widespread characteristic of identification of country risks. Bank assistance programs. About half of all the Country Assistance Strategies reviewed and Country Assistance Strategies that lacked evaluated by IEG over the past four fiscal years selectivity led the Bank to spread its resources were assessed as having been too ambitious. too thinly across too many sectors, which Ninety percent of assistance programs in the diluted its effectiveness. This was the case, for subsampled countries example, in Honduras and Malawi, where the Overly ambitious strategies where poverty was not Bank’s program covered more areas and institu- reduced and where the tions than could be handled effectively by either led to programs that either Bank’s assistance failed the Bank or the country (IEG 2006f, 2006i). were not sufficiently to deliver the expected selective or established results were overly The need to strategically focus Bank assistance ambitious and not suffi- on a selected number of important areas is unrealistic expectations for ciently tailored to coun- particularly strong in smaller countries and in reform programs. try conditions. fragile states. In Cambodia and São Tomé and 12 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H Principe, for example, IEG analysis found that resulted in better out- Assistance strategies that the Bank’s lending program stretched its comes. Collaborative work are not selective spread resources too thinly across too many sectors, also helps enhance local which prevented it from effectively engaging in capacity and build owner- the Bank’s resources too policy dialogue and supervision.6 ship of Bank-supported thinly and dilute its programs. effectiveness. Overly ambitious strategies have also led to unrealistic expectations for reform programs In Brazil, for example, the Bank’s strategy was supported by policy-based lending. In Malawi, underpinned by high-quality analytical work on for example, the Bank proceeded with policy- poverty and growth, done with substantial participa- based lending for reform programs that were tion by top Brazilian researchers (IEG 2004d). In not adequately scaled to country conditions and Armenia, the most successful periods of the Bank’s achieved few sustainable results (IEG 2006i). assistance were those underpinned by strong analyt- ical work on growth and poverty (box 2.3). A hiatus The Bank sometimes also proceeded with such in analytical work prevented recognition of lending under external pressure to help prevent important constraints to default. Its financing then perpetuated unsustain- growth and the strategy Assistance strategies built able fiscal situations, without addressing the failed to recognize the on analytical work done underlying causes. In Pakistan, two policy-based need to push the agenda loans had little sustainable impact on structural in new directions (IEG with local participation reforms, despite the large amounts of funding ad- 2004a). In Turkey, a broad have tended to be more vanced to avoid default. But a policy-based opera- program of analytical realistic and resulted in tion achieved satisfactory results in subsequent work, often done with years, when it was adequately tailored to local substantial participation of better outcomes. conditions and supported a government-owned Turkish experts, helped reform program (IEG 2006j). In Zambia, external build consensus for overdue reforms that proved pressure for balance of payment support essential to revitalizing Turkey’s economy (IEG weakened the design and supervision of succes- 2005n). sive policy-based operations and delayed privati- zation of mines and necessary structural reforms Achieving poverty-reducing growth— (IEG 2002c). In Bolivia, the prolonged high level challenges for Bank assistance of aid prevented adherence to hard budgetary Designing assistance strategies that are well constraints and, when a crisis arose, the Bank and grounded in country analysis, set out realistic other donors were compelled to provide further expectations, and can contribute to achieving assistance at a time when its likely effectiveness poverty-reducing growth poses a host of was most questionable (IEG 2005c). The Bank challenges. Recent IEG evaluations provide should be cautious about providing emergency evidence on some of them, including the need to: liquidity to avoid default in the absence of a strong reform program. • Deepen country analysis to understand what factors in a given country can lead to a pattern Analytical work with country participation of growth that efficiently translates into poverty helps tailor assistance strategies to local reduction. conditions • Find more effective ways to reduce rural A realistic Country Assistance Strategy requires poverty. an accurate assessment of the country’s political • Increase attention to the distributional ef- economy and of the main constraints to poverty- fects of growth-enhancing reforms. reducing growth. Assistance strategies built on • Find more effective ways to reduce poverty in analytical work done in collaboration with local low-income countries under stress and fragile specialists have tended to be more realistic and states (LICUS). 13 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Box 2.3: Analytical Work on the Constraints to Poverty-Reducing Growth Helped Reshape Bank Assistance to Armenia The study “Growth Challenges and Government Policies in Ar- labor. It showed that a poor business climate and weak private menia” demonstrated how poverty-focused analytical work can sector capabilities hindered the establishment or expansion of help shape Bank assistance in support of poverty-reducing private firms. It recommended policies and actions designed to growth. It analyzed why the steady rise in Armenia’s real GDP accelerate enterprise restructuring, attract investment, and en- during the second half of the 1990s had not led to significant im- courage the creation of new businesses in the medium term. provements in poverty and employment creation. It found that Largely as a result of this analysis, the Bank’s assistance was improvements in wage levels had disproportionately benefited redirected to support government efforts to improve the over- labor in a few sectors that employed a small proportion of total all business climate, rather than pursuing privatization in isolation. Source: IEG 2004a. They also illustrate the circumstances in which mobility of the poor, such as low skills or lack of the Bank can effectively help cushion the impact access to capital, infrastructure, or market of a crisis. outlets. Identifying such constraints and devising the proper response is an involved task Paying attention to the pattern of growth and will require country-level research and The ability of the poor to participate in growth analysis, which the Bank and other donors can depends on how much growth is driven by support. productivity increases in Strategies for poverty sectors where a large In the subsample of countries that experienced number of the poor positive growth without poverty reduction, reduction need to work, how much growth growth was concentrated in subsectors with low consider where the poor translates into job labor intensity and where few of the poor could live and how they earn creation, and how well earn their incomes. The Bank’s assistance often the poor are equipped contributed to bringing these countries back to their income. to take advantage of a growth path through improved economic such job growth. Growth management, but it was less successful in in more labor-intensive sectors and in sectors bringing about job-creating growth. In Georgia, where large numbers of the poor work, such as for example, the oil transport sector was a major agriculture and manufacturing, has been found driver of growth, but it created little employ- to lead to more poverty reduction than growth ment.7 The Bank Group’s assistance was in less labor-intensive sectors, such as mining or successful in helping to reestablish macroeco- utilities (Christiaensen, Demery, and Kühl 2006; nomic stability and contributed to growth in the Loayza and Raddatz 2005; Ravallion and Chen oil transport sector through related infrastruc- 2004). ture investments, but it was less successful in helping to remove obstacles to more broad- This does not mean that countries should not based growth. In Madagascar, the Bank’s invest in sectors with low labor intensity, but it assistance strategy sought to put the country does suggest that they should pursue a strategy back on a sustained growth path by putting the that balances growth in those sectors with overextended public sector on firmer ground growth in sectors that have the potential to and by establishing the conditions necessary for absorb a higher share of labor. Strategies for private sector growth. It focused on sectors with poverty reduction thus need to consider where high growth potential that would allow for the poor live, how they earn their income, and relatively quick payoffs, but their impact on what constrains growth in those areas. They also poverty was limited. The strategy succeeded in need to consider the constraints to intersectoral bringing the country back to a growth path, but 14 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H the growth was concentrated in a few export- Bank has been based on Many resource-rich oriented sectors that created only limited and the premise that the countries have performed cyclical urban employment. It thus made few engine of growth needs inroads into poverty, although preliminary data to be jump-started in below their potential. suggest that recent reforms in the rural sector urban areas. But insuffi- may have led to poverty reduction over the past cient attention has often been paid to strength- two years (IEG 2006h). ening urban-rural linkages and to implementing strategies that also help improve farm and off- Many resource-rich countries, among them farm productivity in rural areas. In Senegal, the Madagascar, Nigeria, and Zambia, have per- poverty focus of Bank lending came late, formed well below their potential over the past and with a strong bias decade. Underperformance in resource-rich in favor of urban areas Half of IEG’s country countries has been linked to unsound revenue (IEG 2006q). In Malawi, assistance reviews management and poor governance. While where poverty has re- Country Assistance Strategies suggest that the mained unchanged over concluded that assistance Bank is aware of the main reasons for underper- the past decade, the in rural areas either did formance in such countries, formulating a viable Bank attempted to ad- not yield satisfactory approach to address them remains a challenge dress rural development (IEG-World Bank, -IFC, and -MIGA 2005). The issues through adjust- results or that rural Bank is preparing guidelines for staff on the ment lending and largely poverty reduction challenges to consider when developing failed to help the gov- required more attention. Country Assistance Strategies for resource-rich ernment develop poli- countries. cies to address low agricultural productivity and food security (IEG 2006i). In Pakistan, the Bank’s Finding effective ways to reduce rural poverty rural assistance program lacked vision and remains a challenge insufficiently addressed rural poverty (IEG Poverty reduction was faster in urban than in 2006j). rural areas in the majority of the relevant subsample countries. The rural poor continue Even in transition economies and some middle- to outnumber the urban poor in all subsample income countries in Latin America, IEG Country countries except Brazil and Uruguay, which have Assistance Evaluations conclude that the Bank a high urbanization rate8 (appendix D). The in- should focus more on rural poverty. In Armenia, cidence of poverty remains higher in rural areas as in a number of other countries of the former than in urban areas in most of the subsample Soviet Union, increased focus on rural poverty countries (figure 2.8). reduction was highlighted as a priority for future Bank involvement because few inroads have Country Assistance Evaluations show that the been made into rural poverty, despite strong Bank has found it challenging to help countries growth (IEG 2004a). In Honduras, the Bank’s formulate and implement strategies that economic and sector work needs to inten- effectively reduce rural poverty. Of the full sify research on new economic opportunities sample of 48 Country Assistance Strategy re- for the poor, particularly in the agricultural sec- views completed by IEG over the past four fiscal tor, and future Bank support needs to include years, about half concluded that the Bank’s a focus on improving assistance in rural areas did not lead to satisfac- agricultural growth (IEG Strengthening urban- tory results and/or that rural poverty reduction 2006f). Even in Brazil, rural linkages and required increased attention. where the majority of increasing rural the poor live in urban Performance was particularly weak in Africa. In areas, the Bank and productivity require more some countries the strategy supported by the governments need to attention. 15 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure 2.8: Poverty Remains More Widespread in Rural than in Urban Areas in Most of the Sampled Countries 12.00 Ratio of rural to urban poverty 10.00 8.00 6.00 4.00 2.00 0.00 ar e rgia aso zil a n as key ivia i a u on na l eria a bia va dan law ega ic ani Per eni ain ani ista asc Bra dur ldo ubl ero Chi Zam Tur aF Bol Jor Nig Geo Alb Ma Rom Ukr Sen Arm Pak Mo Rep dag Hon Cam kin Ma Bur an inic Dom Source: World Bank, DECRG poverty database. find ways to increase agricultural productivity For example, some have found that financial outside the large commercial agricultural sector sector liberalization has a positive effect on to help curb rural poverty (IEG 2004d). income distribution by creating enhanced conditions for entrepreneurial investments that In countries where rural poverty remains generate employment (see, for example, pervasive, future Bank assistance will need to Chhibber and Nayyar 2006). Financial sector focus more on working with governments to liberalization can also positively affect income identify ways to establish stronger urban-rural distribution if the lower-income groups gain linkages and increase intersectoral mobility, as access to credit. Others have found that such well as strategies that improve productivity in the liberalization can result in increasing inequality rural sector. during the early stages of financial sector development if financial sector institutions are Some reforms entail undesirable distributional geographically concentrated or if gaining access effects9 to credit involves large setup costs for low- Not all pro-growth policies have a positive or income groups, so that they initially remain neutral impact on income distribution.10 The outside the formal credit markets. But they have empirical literature suggests that sound macroeco- also found that as the financial sector develops nomic policies, access to infrastructure, and further and a larger share of the population educational attainment tend to improve income gains access to credit over time, income distri- distribution. But the literature also finds that the bution eventually improves (see World Bank impact of financial sector liberalization, trade 2006h). reform, and the size of government on income distribution varies, depending on such factors as IEG’s evaluation of Bank assistance for trade existing distortions and policy reform found that the reforms had mixed Not all pro-growth the extent of complimen- effects on employment and poverty, depending policies have a positive or tary reforms undertaken on the extent of both complementary reforms (see World Bank 2006h and diversification in the economy. At times, neutral impact on income for a summary of the short-term adjustments led to worsening distribution. literature). income distribution, while the longer-term 16 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H effects on poverty were positive because they allowing children to stay in school or credit- enhanced growth (IEG 2006c). constrained poor people to take up productive opportunities. On this premise, several middle- The Bank has not always paid sufficient income countries have attention to the distributional effects of growth- initiated poverty-targeted The distributional effects enhancing reforms. As a result, the effects of the income transfers that are of growth-enhancing pro-growth reforms it supported were not conditional on beneficiar- always cushioned by safety net interventions. In ies investing in educa- reforms have not always trade reform, for example, the Bank often failed tion, and often on par- received sufficient to conduct sufficient analysis to inform its ticipating in maternal and attention. policy advice and lending about the employ- child health programs. ment and poverty effects of reforms. Con- sequently, trade-related projects did not The Bank has widely supported conditional cash adequately attend to these effects (IEG 2006c). transfer programs in Similarly, in many transition economies, price Latin America, and more Conditional cash and exchange rate liberalizations were not recently in such coun- transfers can help reduce accompanied by the necessary offsetting tries as Bangladesh, Pak- measures to protect food security and provide istan, and Turkey.12 Eval- poverty and invest in social safety nets (IEG 2004e). In its support for uations indicate that the human capital formation reforms of pension systems, the Bank’s transfers can be an of the poor. assistance focused primarily on improving the effective way to invest in fiscal sustainability of pension systems, but it human capital among poor households, while often failed to sufficiently address the pension also reducing poverty, and can even lead recipi- system’s primary goal of reducing poverty and ents to invest part of the transfer in income providing adequate old-age income within fiscal generation (box 2.4). constraints (IEG 2006k). Efforts to promote private sector participation in the power sector To attain the intended objectives, these opera- were found to require more focus on how the tions must be supported by improvements in poor can be assured of access to energy (IEG- the supply and quality of health and education World Bank, -IFC, and -MIGA 2003). services when the existing capacity is not sufficient to respond to the increased demand Given the demonstrated importance of distri- by program beneficiaries. Because these pro- butional changes to poverty reduction, an grams are costly, they need to be carefully increased focus on the distributional impact of targeted. It may also be necessary to make Bank-supported reforms at the country level is trade-offs between coverage and size of needed. The Bank has recently prepared guide- transfers. lines for staff on analyzing the distributional impact of reforms. The long-term impact of these programs remains to be assessed. The sustainability of behavioral Many Organisation for Economic Co-operation changes and the effectiveness of conditional cash and Development (OECD) countries use taxes transfers as a mechanism and transfers to smooth income inequality. to address chronic—as Many OECD countries use Transfers from the rich to the poor and from the distinct from transient— taxes and transfers to working to the nonworking have been found to poverty still need further smooth income lower the Gini inequality measure by up to 18 evaluation. Also unclear percentage points in some OECD countries.11 is the extent to which inequality. Income transfers can help the poor obtain such programs could be education and health services and protect effective in countries with weaker institutional against risks of losing income—for example, by capacity, including lower-income countries. 17 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Box 2.4: Conditional Cash Transfers Help Latin American Families Increase Human Capital Formation Conditional cash transfer programs in many Latin American Health— Mexico’s program increased child growth by 16 per- countries provide cash assistance to poor families in exchange cent among participating children ages 1–3 years. It also reduced for beneficiary compliance with human development actions, child illness by 12 percent, while participating adults were 17 such as school attendance or participation in health programs. percent less incapacitated by illness. Colombia’s program in- These programs have demonstrated positive impacts on edu- creased the height of participating children from 0–2 years by cation, health, and poverty reduction. .76 centimeters, weight by .3–.5 kilograms for children 2–4 years School Attendance—Mexico increased secondary enrollment old, and reduced the occurrence of diarrhea from 32.6 percent for girls by 14 percent and boys by 8 percent. Brazil achieved a net to 22 percent for rural children under 24 months of age. 3-percentage-point increase in school attendance within the tar- Nicaragua decreased the number of stunted children by 5.5 get group. In Nicaragua, school enrollment for targeted children percentage points and of child illness by 7 to 12 percentage increased by 18 percent and attendance by 30 percent. In Columbia, points, and increased immunization rates among beneficiaries school attendance for children between 12 and 17 increased by by 30 percentage points. 10 percentage points in rural areas. In Ecuador, primary school en- Poverty Reduction—In Mexico, there was a 10 percent re- rollment increased by 10 percentage points. duction in poverty and a 45 percent reduction in severe poverty Reduction in Child Labor —In Nicaragua, with increased en- among the target population. In Argentina, an extra 10 percent of rollment, the percentage of working children between ages 7 and the participants would have fallen below the food poverty line 13 declined by 5 percentage points. In Mexico, the probability of without the program in 2003. Colombia’s program helped reduce children of ages 8–17 working decreased by 10–14 percent. extreme poverty by 6 percent in one year. Sources: Cardoso and Souza 2004; Ferreira, Leite, and Litchfield 2006; Gertler 2004; Maluccio and Flores 2004; Attanasio and others 2005; Skoufias 2005; Skoufias, Ben- jamin, and De La Vega 2001. More effective ways need to be found to delivering physical infrastructure in the immedi- reduce poverty in fragile states ate post-conflict phase, but the Bank’s effective- Low-income countries under stress (LICUS), ness needs to be improved following this phase, also known as fragile states, are home to almost when structural change is needed. 500 million people, roughly half of whom live on less than a dollar a day. These countries face The Bank still has to sufficiently internalize poor governance, conflict or post-conflict transi- political understanding in its LICUS country tions, and a multiplicity of problems that make strategies. For example, the Interim Strategy for the achievement of development results partic- Papua New Guinea contained a good discussion ularly challenging. of the political system and recognized problems such as clan loyalties, political patronage, Awareness of the need to provide development corruption, and lack of capacity, yet it treated assistance differently in these countries has risen these problems as technical in nature and did in recent years. The Bank has improved its not directly use this knowledge to underpin the operational readiness to engage with fragile overall approach. states, and substantial progress has been made on donor coordination Strong donor coordination at the international The Bank has effectively at the international policy level still needs to be carried over to the contributed to policy level. Significant country level. In adopting state building as one macroeconomic stability challenges remain, how- of its two main objectives, the Bank has made and the delivery of ever. The Bank has an area of traditional weakness (capacity generally been effective development and governance) a part of its physical infrastructure in in contributing to macro- central focus in LICUS, but it still needs to post-conflict countries. economic stability and demonstrate how past weaknesses will be 18 A C H I E V I N G P O V E R T Y- R E D U C I N G G R O W T H avoided and better capacity development and analytical work. In But it needs to find better governance outcomes ensured (IEG 2006r).13 Turkey, development of ways to help build the knowledge base in Bank support can help cushion the impact of a the years before the crisis capacity and strengthen crisis allowed the Bank to governance in LICUS. The Bank’s assistance has yielded satisfactory respond quickly when outcomes, even in some countries where poverty the Turkish authorities recognized the need for stagnated or increased. This was, for example, decisive action on structural reforms following accomplished when Bank assistance helped the the financial crises of 1999 and 2001 (IEG 2005n). government weather a crisis and then supported Similarly, in India and Morocco, in-depth the implementation of structural reforms to bring knowledge of the economy acquired through the economy back onto a growth path. economic and sector analysis positioned the Bank to advise and support governments on the In Uruguay the Bank quickly changed its sequencing of macroeconomic, trade, and assistance strategy to support the government’s structural reforms following crises (IEG 2006c). adjustment program following the 2002 financial crisis, which helped to cushion the impact of the Where the Bank’s support to crisis-ridden crisis. The Bank’s program successfully countries had successful results, it supported a supported fiscal consolidation, financial sector reform program de- reforms, and measures to enhance the efficiency veloped and owned by When the Bank has of poverty reduction programs. At the same time, the government, often successfully assisted a it supported social and infrastructure invest- with elements based on crisis-ridden country, it ments that helped improve access of low-income prior analytical work has supported a reform groups to these services.14 In Turkey, a portfo- carried out with the lio cleanup and a concerted effort by the Bank Bank. In contrast, where program developed and to sharpen the strategic focus of its assistance the Bank’s assistance to owned by the program produced significant improvements crisis countries did not government, often in the Bank’s dialogue with the authorities yield satisfactory results, and allowed the Bank to provide timely policy the Bank, often under underpinned by prior and financial support in response to the 2001 external pressure to analytical work. financial crisis (IEG 2005n). provide liquidity, pro- ceeded with policy-based lending without To swiftly respond to a crisis, the Bank needs government commitment to fiscal adjustment strong knowledge of the economy based on prior and structural reforms. 19 Chapter 3: Evaluation Highlights • The share of projects with satisfactory outcomes has increased, but even satisfactory project outcomes do not guarantee sector-level results. • Achieving results in one sector often requires removing constraints in other sectors, which makes the use of a clear results chain essential. • Combining longer-term objectives with interventions that yield vis- ible short-term results helps achieve sector impact. • If pressures to show results quickly are not properly managed, the quality of results can be compromised. 3 Achieving Meaningful Results at the Sector Level T he World Bank’s contribution to poverty reduction is achieved through the impact of interventions in various sectors and the overall impact of these interventions on a country’s economic and sector-specific per- formance. This chapter will focus on the factors that enable Bank projects to contribute to sector-level results, particularly in sectors that deliver services to the poor. The chapter draws on the findings of IEG sector and thematic eval- uations, Project Performance Assessments, and Country Assistance Evaluations completed over the past four fiscal years. How Have the Bank’s Projects five fiscal years, up from less than 40 percent Performed? during the preceding five-year period (appendix The overall performance of the Bank’s portfolio A).2 has improved over the past five years. The outcomes of over three-quarters of Bank The policy environment affects project operations were rated moderately satisfactory outcomes or better by IEG in fiscal 2001–05, and project The quality of a country’s policies and institutions outcomes improved in 8 of 14 sectors (figure in a sector affects the outcomes of projects in that 3.1).1 The share of projects with outcomes rated sector. The share of projects with satisfactory moderately satisfactory or better ranged from outcomes is higher in countries with good sector lows of 62 percent and 66 percent in private policies, but the correlation of sector policy sector development and health, respectively, quality and project outcome varies widely across to above 80 percent in education, transport, sectors. It is highest for economic policy, public and global communications and information sector governance, and the financial sector, where technology. most Bank projects are designed to strengthen Project outcomes are Improvements are also observed in institutional the relevant policies. The correlated with the development. Over half of all projects have been correlation is weakest in evaluated as having achieved a substantial health, private sector quality of a country’s institutional development impact over the past development, and en- sector policies. 21 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure 3.1: Project Outcomes Improved in over Half of All Sectors 100 Percent satisfactory outcome 90 80 70 60 50 40 30 20 10 0 Bank GIC TR ED RDV FSP EP WSS EMT HNP SP PSG UD ENV PSD average Fiscal 1996–2000 Fiscal 2001–05 Source: World Bank database. Note: Satisfactory project outcomes are all projects rated as moderately satisfactory, satisfactory, or highly satisfactory by IEG. GIC = global information & communications technology; TR = transportation; ED = education; RDV = rural development; FSP = financial sector policy; EP = economic policy; WSS = water supply and sanitation; EMT = energy and mining; HNP = health, nutrition, and population; SP = social protection; PSG = public sector and governance; UD = urban development; ENV = environment; PSD = private sector development. vironment. In these sectors, projects perform outcome is strongly influenced by the quality of only slightly better in a good policy environment overall economic management in the country than they do in a weak one (figure 3.2). and the governance environment, particularly the rule of law. It shows that once income level, Regression analysis also shows that the like- economic management, and rule of law are lihood of a project achieving a satisfactory controlled for, there are no regional differences Figure 3.2: Project Outcomes Are Better in Countries with Strong Sector Policies 100 Percent satisfactory outcome 80 60 40 20 0 PSG SP ED FSP EP WSS ENV PSD HNP Good sector policy: sectoral CPIA >3.5 Weak sector policy: sectoral CPIA <=3.5 Source: World Bank database. Note: Sectors for which no sector-specific CPIA is available are excluded. This accounts for the smaller number of sectors in figure 3.2 compared with figure 3.1. PSG = public sector and governance; SP = social protection; ED = education; FSP = financial sector policy; EP= economic policy; WSS = water supply and sanitation; ENV = environment; PSD = private sector de- velopment; HNP = health, nutrition, and population. 22 A C H I E V I N G M E A N I N G F U L R E S U LT S AT T H E S E C T O R L E V E L in project success rates (appendix C). Assistance Evaluations Satisfactory project from fiscal 2001–03 (IEG outcomes do not ensure How Much Impact Do Projects Have 2005h). Rural and private on the Sector? sector operations country sector impact. The Bank intends to achieve more than to achieved a satisfactory simply finance individual projects that meet sector impact in only half of the countries; Bank their immediate objectives. It also aims to help operations in the public sector had a satisfactory countries strengthen their sectoral policies and sectoral impact in fewer than one-third of the institutions to permanently improve services countries assessed. The gap between the high and incomes for the poor. Not every project can share of satisfactory project outcomes and the be expected to have a sectorwide impact—this lower share of satisfactory sector outcomes is also often takes a long time and multiple initiatives. most marked in these three sectors, illustrating that satisfactory project outcomes alone do not IEG’s Country Assistance Evaluations, which guarantee a substantial country sector impact typically cover a period of about 10 years, permit (figure 3.3 and appendix B). a view of the cumulative effect of the Bank’s work in a sector. They reveal that it is careful Essential Elements for Achieving Results selection and phasing of interventions and the IEG Country Assistance Evaluations, Project complementarities of lending, analytical work, Performance Assessments, and sector evalua- and policy dialogue that lead to impact on the tions completed over the past four fiscal years sector as a whole. point to six elements that can help lead to stronger sector results: The sector-level impact of Bank assistance, as assessed in the 18 IEG Country Assistance Evalua- • The presence of a country-formulated sector tions completed during fiscal 2004–06, was lowest strategy in public sector, private sector, and rural develop- • Realistic objectives ment. This confirms earlier findings of 25 Country • A clear results chain Figure 3.3: Successful Projects Alone Do Not Ensure Sector-Level Impact 100 Percent satisfactory outcomes 80 60 40 20 0 Social Financial Education Infrastructure Health Enviroment Rural Private Public sector protection sector sector development Satisfactory sector outcome Satisfactory project outcome Sources: World Bank database and Country Assistance Evaluations. Note: Sector outcomes as assessed by IEG Country Assistance Evaluations (CAEs). The sample includes all CAEs completed in fiscal 2004–06, for a total of 18, but not all CAEs covered all sectors. Projects are those completed in countries where CAEs assessed sector outcomes. “Satisfactory” includes sector and project outcomes rated moderately satisfactory or higher. 23 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S • A results-focused information system provinces. As a result, the ministry evolved from a • Choice of an appropriate lending instrument weak bystander to the owner and implementer of • Long-term engagement and striking a balance national health strategies in collaboration with between long-term and short-term objectives. international donors (IEG 2004i). Similarly, joint analytical work and support provided under a Build on a country-formulated sector strategy Bank-financed investment project helped Bank interventions have Armenia formulate a comprehensive education Bank interventions have yielded substantially sector reform strategy that the Bank helped better results when they implement through sector investment and policy yielded better results have supported a lending (IEG 2004a). when they have supported country-formulated and a country-formulated broadly owned sector Set realistic objectives and targets sector strategy. strategy with clear objec- An important step in obtaining meaningful tives. This is particularly sector results is to set realistic targets that are true for interventions grounded in country analysis, informed by past that support an overall sectoral reform agenda. performance and international experience. Unrealistic targets can generate a perception of The Bank and other donors can support the failure, regardless of what has actually been formulation of such a sector strategy through achieved. And perceived failure can lead to analytical work, policy dialogue, and institu- disillusion and negatively affect future efforts. tional capacity building, but to achieve sustain- able results, the strategy needs to be formulated In the Bank, as elsewhere, there has been a by the country itself. Even well-designed Bank tendency to adopt unrealistic goals. The Millen- projects can rarely turn around poorly designed nium Development Goals (MDGs), for example, sectoral programs, and are thus unlikely to are the most widely embraced sectoral targets achieve substantial results absent a strong sector and have been useful in galvanizing the support strategy. In Albania, for example, the Bank’s of the international community toward achiev- interventions achieved good results in the ing defined sector results. But many of these power sector, where the government had targets set unrealistically high expectations of adopted a clear sectoral reform strategy what a country can achieve by 2015. For supported by the Bank through its lending example, many countries that started with low operations. The results of the Bank’s assistance primary education school enrollments in 1990 in other sectors, where no clear government would need to perform as well or better than the strategy was in place and where the Bank itself historically best performers, such as Korea, to did not have the necessary analytical underpin- achieve the MDG of universal completion of ning, were more modest (IEG 2005a). primary education by 2015 (figure 3.4). Thus, it is not surprising that a considerable number of Donors can effectively support development and countries in Sub-Saharan Africa and South Asia adoption of a sector strategy, provided they that have made substantial progress toward recognize that it takes time and a series of MDG targets are still unlikely to reach those interventions. In Cambodia, for example, the goals, which puts at risk the momentum gained Bank’s first health sector project helped provide a so far. combination of capacity building and a revised incentive framework at Unrealistic targets have characterized many Bank- Unrealistic targets risk the central and local supported projects. Several loans with financial perceived failure, levels as a catalyst for sector components in crisis countries, for fundamental changes in example, had ambitious and unrealistic targets regardless of what might the relations between the driven by an overestimation of the government’s actually be achieved. Ministry of Health and the commitment to reform, combined with a need to 24 A C H I E V I N G M E A N I N G F U L R E S U LT S AT T H E S E C T O R L E V E L Figure 3.4: Setting Unrealistic Targets Leads to Perceptions of Failure, Even When Substantial Results Are Achieved Ethiopia Niger Net primary school enrollment (%) Net primary school enrollment (%) 100 100 80 80 60 60 40 40 20 20 0 0 1991 1995 2000 2005 2010 2015 1991 1995 2000 2004 2010 2015 Year Year Actual trend MDG-required trend Historical best: Korea (1951–75) Sources: Korea Educational Development Institute 2005, World Bank 2006n. Note: In the absence of historical data on primary school completion and net enrollments, gross enrollments are shown here for one of the historically best performers, Korea. Reaching a 100 percent primary school completion rate (MDG2) is substantially more challenging than achieving gross enrollment rates of 100 percent or higher. Thus, the primary education MDG would require even stronger performance than the historically best trend shown here. justify large loan amounts. In Indonesia, for the health sector focused Unrealistic targets have example, a series of loans following the 1997 crisis on areas where Bolivia characterized many addressed resolution of the banking crisis and lagged most, infant and corporate restructuring, including the resolution maternal mortality. A Bank-supported projects. and reprivatization of banks and disposal of assets. series of three credits But by 2003, although the pace of reforms had over a decade concentrated on improving the improved, the government still controlled over 60 quality and accessibility of essential maternal and percent of the banking system, and the banking child health services, particularly for low-income sector remained vulnerable to further shocks (IEG groups. It helped establish a system of preferred 2006g). access to health care for maternal and child health issues, supported the expansion of Similarly, education projects with designs to health facilities in rural areas, and established a improve management performance often set system that carefully monitored inputs and overly ambitious objectives. As a result, only outcomes. This resulted in dramatic improve- about a quarter of Bank projects reviewed by ments in the utilization of these services and in IEG’s recent assessment of assistance for maternal and child health outcomes (IEG primary education met their targets for improv- 2005c). ing planning, policy making, and budgeting activities (IEG 2006e). In many cases, project In Uruguay, two Bank-supported education sector objectives are not grounded in thorough institu- projects focused on increasing equity in the tional analysis, which can lead to setting unreal- education system through improved participation of istic objectives (box 3.1). the most disadvantaged students in preschool When objectives are set Clear and realistic objectives help achieve results education and in full-time realistically and When objectives are clear and realistic and schools. The projects’ stakeholders focus on them, they can be targeted interventions led stakeholders focus on achieved. In Bolivia, for example, Bank support in to substantial improvement them, they can be achieved. 25 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Box 3.1: Unrealistic Project Objectives Are Often Grounded in Inadequate Institutional Analysis Objectives for education sector management often have been Similarly, the government in Pakistan rushed into large Bank- overly ambitious and not sufficiently grounded in institutional supported social action projects, each allocating more than $100 analysis. million to basic education, without a clear understanding of their The Primary Education Quality Project in Peru aimed to improve complex financial management requirements. A good capacity as- educational management at the Ministry of Education and in the sessment would have revealed the low technical skills at the schools. But during preparation, the lack of consensus and polit- provincial and district levels, particularly in financial manage- ical will that eventually kept school autonomy and regional de- ment, as well as a culture of patronage that frequently led to the centralization from being implemented was not properly recognized. misallocation of funds. Source: IEG 2006e. in learning outcomes among students from low- (box 3.2). income families, which reduced the learning gaps • It assigns clear responsibilities. between social groups in the country (IEG 2006e). • It identifies the cross-sectoral interventions necessary to achieve the ultimate outcomes. Define a clear results chain The results chain helps identify actions to achieve Bank sector strategies can help focus on results. objectives Seven sector boards have results frameworks and The analysis needed to develop a clear results outcome indicators that provide a starting point chain is frequently overlooked. In Albania, for for Bank staff as they develop results chains that instance, the Bank did not pay enough attention link country development goals with specific to how ambitious goals for governance, health, Bank operations. Several other sector boards are and education would actually be achieved. developing such frameworks. A major challenge Conditions for the Poverty Reduction Support will be adapting these generic frameworks to Credit (PRSC) focused mostly on the passage country circumstances (IEG 2006a). A clear and enactment of laws, which were necessary results chain offers three benefits: but not sufficient to achieve the desired objectives in governance. Inadequate attention • It identifies constraints to achieving objectives was given to capacity constraints that hindered and specifies actions necessary to remove them effective implementation of these laws. In Box 3.2: Understanding the Results Chain Was Key to Designing Interventions to Improve Girls’ School Enrollment in Egypt In Egypt, gender disparities in primary school enrollment were with those problems by increasing the number of schools in thought to be cultural in nature, but a series of studies found sev- rural areas, improving teacher training, and including gender sen- eral other reasons why parents did not send their girls to school: sitivity in teacher training programs. This increased parental the distance of schools from home, the presence of male teach- demand for girls’ education and gave a second chance to girls ers, and the quality of education. In 1996, the government who may have dropped out. As a result of these interventions, launched the Education Enhancement Program, supported by the the gender gap in primary education has been substantially nar- World Bank and the European Union. The program dealt directly rowed over the past five years. Source: World Bank 2005h. 26 A C H I E V I N G M E A N I N G F U L R E S U LT S AT T H E S E C T O R L E V E L contrast, in the energy sector the government reductions in child A clear results chain and and the Bank agreed on a specific action plan for mortality, because a appropriate monitoring reform with monitorable indicators, and satisfac- child born to a mother tory results were achieved (IEG 2005a). The with secondary help assign Bangladesh Integrated Nutrition Project had education is around 80 responsibilities and disappointing outcomes for birth weight, even percent less likely to die measure the performance though participation was high. In this case, the than one born to a results chain missed two important links: it mother with no of different actors. neglected the role of men and mothers-in-law in education. Rural electri- nutritional choices, and it focused on pregnancy fication improved health outcomes as well, weight gain rather than pre-pregnancy nutri- because it helped raise incomes, improved the tional status (IEG 2005k). quality of health care, and expanded people’s access to health information through the media The results chain clarifies responsibilities (IEG 2005k). In areas of Morocco where the A clear results chain and appropriate monitoring Bank helped finance rural roads, agricultural help assign responsibilities and measure the activity picked up, enrollment in primary performance of the different actors. In Bolivia, education increased (faster than in control for example, the government, with Bank areas), and the quality of education improved. assistance, made significant progress on most of Improved accessibility was an important factor its objectives in heath and education, largely in retaining better- through well-designed interventions that qualified teachers and Achieving results in one established clear authorities and accountabili- lowering absenteeism ties for all involved, ranging from the sectoral among both teachers sector often requires line ministries, to regional departments, to and students in the removing constraints in municipalities and individual providers (IEG remote areas that other sectors. 2005c). The District Health Services Pilot and benefited from invest- Demonstration Project in Uganda, in contrast, ments in rural roads (Levy and Voyadzis 1996). had a weak results chain and unclear objectives. This caused confusion among national stake- Lack of attention to complementary inputs, holders and compromised national understand- however, has diminished the impact of ing and commitment. Consequently, the project infrastructure investments under some of the made little progress in improving health out- Bank’s community-driven development projects. comes (IEG 2006p). Although infrastructure construction in scattered communities increased access to The results chain helps identify cross-sectoral schools and health centers, it did not always constraints translate into effective service delivery because Achieving results in one sector often requires of the inadequacy of complementary inputs such identifying and removing constraints in other as teachers, doctors, and medicines (IEG 2005i). sectors. Analysis of the results chain can help Similarly, the effectiveness of the Bank’s support identify necessary cross-sectoral interventions for pension reform has at times been affected by that are often overlooked. Addressing cross- the failure to ensure that the necessary macro- sectoral constraints does not necessarily require economic, financial, and institutional conditions large cross-sectoral operations, but it does were in place for such require that the main constraints to achieving reforms to achieve the The Bank’s matrix core objectives be identified and addressed in a aspired objectives (IEG management structure synchronized fashion. 2006k). does not encourage staff In Bangladesh, for example, Bank support for The Bank’s matrix to work across sectoral female secondary schooling has contributed to management structure boundaries. 27 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S does not encourage staff to work across sectoral prove the competitiveness” of urban markets boundaries or to address cross-sectoral issues. for land, labor, credit, and infrastructure and Despite attempts by the Regions to create multi- housing inputs. But less than 10 percent of sectoral teams to coordinate community-driven urban projects address competitiveness be- development projects, cause city-level data are scarce for several pro- As the Bank and its for example, only 9 posed indicators. Moreover, practitioners are percent of Bank staff unsure of what competitiveness means in this partners incorporate surveyed reported being context and how to put it into practice (IEG sector-defined outcome satisfied or very satisfied 2004f). Many Bank-supported HIV/AIDS proj- indicators into with coordination across ects and national AIDS programs have used HIV operations, they face the sectors in these prevalence (the percent of the population that interventions (IEG is HIV-positive) as an indicator of the impact challenge of obtaining the 2005i). of prevention activities. HIV prevalence can information needed to rise or fall, depending on whether more peo- track progress and adjust Bank task managers re- ple become infected than die over a given pe- port that preparing and riod, so it is not an appropriate indicator to activities. implementing PRSCs has measure success in HIV prevention or treat- led to more cross-sectoral ment4 (IEG 2005g). Most indicators in interaction among Bank staff, but note that insti- community-driven development projects focus tutional constraints to working across sectors are on quantity of outputs rather than their qual- apparent in interactions around PRSCs (IEG ity, which makes it impossible to assess proj- 2004g). Similarly, the linkages between trade ect impact on strengthening local capacity reform and social and economic vulnerably have (IEG 2005i). not been addressed sufficiently in much of the Bank’s support for trade reform, partly because • Appropriate data are not collected. The stan- the complexity of the issue requires a multidisci- dard outcome indicators for the water supply plinary team that cuts across the Bank’s sector and sanitation sector are the numbers of peo- and network boundaries3 (IEG 2006c). Lack of ple with access to improved drinking water cross-sectoral coordination has also affected the and improved sanitation. About two-thirds of Bank’s assistance for pension reform. At times fiscal 2005 water supply and sanitation projects this has led to different Bank teams providing included these outcome indicators, but only inconsistent policy advice to the same country one-third of the projects had a baseline against (IEG 2006k). which to measure outcomes (World Bank data). Despite considerable emphasis on monitor- Use a results-focused information system ing and evaluation systems in Bank-supported As the Bank and its partners make progress in HIV/AIDS programs, projects are frequently incorporating sector-defined outcome indica- launched without baseline data that are criti- tors into operations, they now face the challenge cal to program design (IEG 2005g). of obtaining the information needed to track progress and adjust activities. Three kinds of • Data collection instruments are poorly obstacles inhibit the use of information to designed. Even when indicators are properly improve results. chosen and data are collected, the resulting in- formation may not allow assessment of the Enhancing country • Indicators are not impact of project or program interventions if capacity to collect and available or are inade- surveys are improperly designed. Repeated quately matched with the national surveys to assess the impact of donor- use performance objectives. The Bank’s supported HIV/AIDS programs in Cambodia, information helps achieve 2000 Urban Strategy, for Chad, India, and Uganda, for example, failed to better sector outcomes. example, aims to “im- ensure the comparability of questions across 28 A C H I E V I N G M E A N I N G F U L R E S U LT S AT T H E S E C T O R L E V E L surveys, which made it impossible to track dialogue and instrument to achieve results. In the changes in behavior over time (IEG 2005g). health sector, this led to weak ownership of the Where Bank operations have successfully health sector program helped establish country capacity to collect supported under the The shift from sector- and use performance information, that infor- PRSC by the Ministry of specific operations to mation has contributed to better sector out- Health.5 Ultimately, the comes. In Bolivia, for example, the Bank Bank decided to supple- sector components of promoted the idea that there was a need to col- ment the PRSC support multisector operations lect detailed statistics on inputs, outputs, and for health reforms with a has not always been outcomes at the regional level to judge the sector-specific operation. performance of regional administrators. Con- smooth. sistent application of this approach under Bank- The sectoral impact of supported health and education sector multisector operations has often been weaker than operations has contributed to substantial im- that of sector-specific operations, partly because provement in Bolivia’s health and education sec- multisector operations allow for less- intensive tor outcomes (IEG 2005c). engagement of Bank sector teams with country line agencies. In the financial sector, for example, Choose the appropriate lending instrument the outcome of loans overseen by Bank financial To enhance the sector impact of its operations, sector departments was substantially better than the Bank has gradually shifted emphasis from financial sector components of multisector loans specific investment projects toward supporting (IEG 2006g). Similarly, pension-specific loans and sectorwide initiatives through development policy pension reform and programmatic lending. Such sectorwide initia- components of develop- Sector-specific operations tives have taken three main forms: replacement of ment policy lending have tended to have sector-specific operations with components in operations led by the greater sector impact multisector development policy lending opera- social protection and than multisector tions; combination of multisector development financial sector depart- policy lending and sector-specific operations; and ments achieved better operations. dedicated sectorwide operations. These emerging outcomes than pension approaches have had varying results. components of multisector operations led by economic policy teams (IEG 2006k). In the edu- Multisector operations can dilute sectoral cation sector, multisector operations with educa- ownership and oversight tion components, such as PRSCs, have helped Where sector-specific operations were replaced boost educational enrollments, but they have by sector components of multisector operations, contributed little to improving educational quality the transition has often not been smooth and and learning outcomes (IEG 2006e). outcomes have been mixed. In the shift to budgetary support, the need for continued A combination of policy-based lending and sector- capacity building in the line agency often has not specific operations can deliver good results in been considered. The institutional reforms that countries where there is strong commitment to accompany policy adjustments usually take sector reforms but limited capacity to implement longer than the time allotted in quick-disbursing them. In Armenia and Ghana, for example, the operations. In Burkina Faso, for example, the Bank effectively used development policy lending transition from project to programmatic support to support reforms in the education sector, while resulted in only limited involvement of the sector parallel investment projects helped build the line agencies and Bank sector teams in the PRSC- systems and capacity to implement the reforms related sector dialogue, as both sides grappled (IEG 2004a, 2006e). The combination of policy- with the shift of attention from project-specific based and technical assistance lending has also issues to the sector budget as a vehicle for sector yielded better results in financial sectors in low- 29 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S income countries. The same has not been true for capacity development that traditionally has been financial sector support to middle-income coun- provided formally through the implementation tries, where the existence of a technical assistance of investment projects, and informally through loan in addition to policy-based lending for Bank supervision (IEG 2005e). The Bank and financial sector reforms may indicate lack of other donors need to strike a better balance government commitment to carry out the between providing support for strategy and reforms, rather than limited capacity (IEG 2006g). policy formulation and for building technical and implementation capacity. Sectorwide approaches To enhance the impact of sector-specific Engage for the long term support, the Bank is increasingly participating in Many of the intended results of even well- a sectorwide approach, or SWAp.6 While there is executed sector operations require more time not yet a comprehensive evaluation of this than a single project cycle. In education, for experience, an initial review of project outcomes example, international evidence shows that and lessons learned student performance often drops in the first five A better balance between from the first set of years after initiation of a sector reform, before support for strategy and completed SWAps the benefits of even a well-designed and well- suggests that they have implemented program take hold and learning policy formulation and generally succeeded in outcomes improve (IEG 2006e). Therefore, for building technical strengthening country expectations of what can be achieved in the time and implementation ownership for reforms span of a single Bank operation must be capacity is needed in and in enhancing managed, and feasible intermediate objectives capacity to set policies defined for each operation. sectorwide approaches. and strategic direction.7 The long time often required to achieve They have been less successful in building ca- intended results underlines the importance of pacity to manage the sector for results and to continuity of engagement in a sector. IEG’s improve service delivery. There are three main evaluation found, for example, that continued reasons for this: Bank support to the health sector in Bangladesh over three decades has contributed to remark- • First, the focus and interaction with govern- able improvements in the country’s maternal ment has largely been limited to the policy and child health outcomes, while such results level, with the Bank and other donors paying were not clearly discernable after the first less attention to technical issues and details of decade of assistance (IEG 2005k). An evaluation implementation. of the Bank’s urban portfolio found that projects • Second, because of their sectorwide focus, that followed on and integrated the lessons many of the operations have been too complex learned from previous projects achieved better in relation to local implementation capacity. results than other projects, which underlines • Third, the effort to harmonize approaches and the benefits of continuity of Bank engagement systems has required too much attention to (IEG 2004f). processes in relation to The long time often actual program Balancing short-term and long-term objectives implementation. improves results required to achieve Although achieving high-impact results takes intended results Some Bank clients in time, pressure to show results quickly is Africa have expressed inevitable. Such pressures, if not managed well, underlines the concern that a focus on can compromise the quality of results. In many importance of continuity strategic issues deprives low-income countries, the objective of achieving of engagement in a sector. local sectoral staff of universal completion of primary education by 30 A C H I E V I N G M E A N I N G F U L R E S U LT S AT T H E S E C T O R L E V E L 2015 has led to massive efforts to expand designed as a pilot to If pressure to show results coverage of the education system. While getting demonstrate the feasibil- quickly is not managed more children to attend school is an important ity of delivering an achievement, such rapid expansion has often essential health services well, the quality of results been at the expense of attention to learning package. But under can be compromised. outcomes. For example, in Uganda, enrollment pressure to quickly expansion was dramatic, but was not accompa- expand support during the post-conflict period, nied by sufficient expansion of physical facilities project activities were expanded countrywide and books, and learning outcomes plummeted. without the benefit of evaluating each phase and By 2005, there were an average of 94 students adjusting sector reforms based on experience. As per classroom and 3 students were sharing a a result, the project failed to achieve its overarch- single textbook (IEG 2006e). ing objectives of reducing disparities in access and improving maternal and child health Few Bank-supported education projects have outcomes (IEG 2006p). In Timor-Leste, the focused sufficiently on striking a balance between breakdown of government institutions, poor rapid expansion of coverage and building the governance, widespread suffering, and massive systems and institutional capacity to ensure that displacement of the population put pressure on learning outcomes are not negatively affected by the international community to respond rapid expansion. Only one-third of primary speedily. The Bank’s three Community education operations assessed by IEG explicitly Empowerment Projects (CEP-I, II, and III) aimed to improve learning outcomes. Yet the quickly transferred resources to communities experience of Ghana, India, and Uruguay shows and delivered massive amounts of infrastruc- that it is possible to expand access and improve ture. But speed came at the cost of the other student learning simultaneously, but only with project objectives, particularly community careful strategic planning, beginning with planning empowerment and the development of local to improve learning outcomes and ensuring politi- institutions (IEG 2006o). cal commitment to that goal. Where Bank support combined longer-term The approach taken to implement policies to objectives with interventions that yielded achieve these objectives differed across the visible short-term benefits, it has achieved good three countries, as did Bank support, highlight- results. In Cambodia, for example, the Bank ing the importance of tailoring support to successfully took a longer-term approach from country conditions. While Bank support focused the outset of its support to the country’s health primarily on providing learning inputs in Ghana, sector, investing heavily in government capacity in India it consisted of a mixture of learning and health infrastructure while supporting inputs and support of pedagogical renewal, programs to control communicable diseases while in Uruguay it targeted interventions, such that yielded visible as better access of disadvantaged children to results over a shorter Where the Bank combined preschool and extra instructional time to period (IEG 2004i). longer-term objectives improve learning outcomes (IEG 2006e). Similarly, in Ghana, the Bank’s support to the with interventions that The pressure to show results quickly can be education sector mixed yielded short-term particularly strong in post-conflict situations. In support for policy benefits, it achieved good Uganda, the Bank-supported District Health reforms with funding Services Pilot and Demonstration Project was for school infrastructure results. 31 Chapter 4: Evaluation Highlights • To succeed, technocratic reform programs need strong political support and a focus on enforcement capacity. • Process reforms need to be supported by efforts to raise the demand for accountability and increased transparency. • Bank operations have helped foster accountability by bringing transparency to a variety of public management processes. • Stronger community participation can raise the demand for accountability of public sector institutions. 4 Strengthening Public Sector Accountability A chieving and sustaining growth and poverty reduction require public institutions that perform well and are accountable to stakeholders for the results achieved. This chapter reviews how the Bank has worked with governments to strengthen the effectiveness and accountability of pub- lic sector institutions. It draws on emerging lessons from IEG Country Assis- tance Strategy reviews carried out over the past four fiscal years, as well as recent thematic evaluations on community development, extractive industries, and fiduciary assessments. The Bank’s Country Assistance Strategies put as civil service and administrative reforms, substantial emphasis on strengthening public improved public financial management, tax and sector institutions. For instance, among 48 customs reforms, and creating an enabling countries covered in IEG’s recent Country environment for the private sector. More Assistance reviews, 42 emphasized strengthen- recently, efforts have also been stepped up to ing public sector performance or reducing strengthen complementary aspects of gover- corruption, although an active Bank lending nance, such as public participation, information program in this area materialized in only 35 of disclosure, and reduction of corruption. these countries. Bank-wide, about one-quarter of all lending operations and 23 percent of Strengthening Accountability with commitments have targeted public sector Administrative Reforms governance and the rule of law over the past five The bulk of the Bank’s fiscal years. Nearly half of the prior actions for support has focused The Bank has focused on development policy operations are related to on strengthening pub- reforms in public governance, particularly public financial man- lic sector accountability agement (World Bank 2006m). through reform pro- administration and grams in public adminis- public financial The Bank’s work to improve the effectiveness tration and public fi- management to and accountability of public sector institutions nancial management. In has focused on initiatives that address institu- almost one-third of the strengthen public sector tional aspects of public sector governance, such 48 Country Assistance accountability. 33 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Strategies reviewed, the Bank-supported re- government processes necessary to improve the forms were aimed at making the public service governance environment, but effective utiliza- more professional, and almost three-quarters tion of such tools remains in the hands of contained efforts to improve public financial country decision makers. management, including procurement. Country Assistance Evaluations point to three These efforts have helped strengthen the quality factors that have attenuated the effectiveness of of government processes in some countries the Bank’s efforts to strengthen public sector (figure 4.1). But there is little evidence that the accountability through large-scale administra- process improvements have yet led to improve- tive reforms: ments in perceived quality of governance. In most countries assessed by IEG where the Bank • Reform initiatives have not been aligned with program included public sector reforms, political realities. governance perception indicators have changed • They have focused on the adoption of legisla- little since the mid-1990s (figure 4.2 and tion and the establishment of institutions, appendix E).1 while neglecting the need for enforcement capacity. This inertia points to the fact that improvements • They have not addressed the intersection be- in processes take time to manifest themselves tween the public sector and private sector, and affect perception. Yet, in at least two where there is most room for abuse. countries, Bulgaria and Lithuania, perceived governance indicators have improved. This Technocratic public sector reform programs demonstrates that it is possible to make rapid need political backing progress when there is strong country commit- Large-scale technocratic reform programs have ment to do so. The Bank can provide govern- yielded better results when they have built on ments with the tools to strengthen the strong country leadership and broad-based politi- Figure 4.1: Government Process Quality Indicators Improved in About Half of 35 Countries with Bank-Supported Public Sector Reform Programs 20 19 Number of countries 15 13 10 8 6 5 3 2 0 Quality of budgetary and financial Quality of public adminstration management Deterioration Improvement No change Source: World Bank CPIA database. Note: Quality of government processes indicators used are CPIA for budgetary and financial management and CPIA for quality of public administration from 1999 to 2005. Sample includes all countries where the Bank’s assistance program included active support for public sector reform and for which IEG completed a Coun- try Assistance Evaluation or an analysis of a CAS Completion Report in fiscal 2003–06. In 35 of these countries the Bank provided assistance for public finan- cial management, and in 16 it provided assistance for civil service and administrative reform. 34 S T R E N G T H E N I N G P U B L I C S E C T O R A C C O U N TA B I L I T Y Figure 4.2: But Governance Perception Indicators in 35 Bank Borrowers with Public Sector Reform Programs Did Not Change Significantly 35 32 30 31 30 29 Number of countries 25 20 15 10 5 2 3 2 2 1 1 1 1 0 Government effectiveness Control of corruption Regulatory quality Rule of law Significant decline Significant improvement No significant change Source: Kaufmann, Kraay, and Mastruzzi 2005. Note: Governance indicators are Kaufmann, Kraay, Mastruzzi Indicators between 1996 and 2004. Classification for deterioration, improvement, and no change is based on a 90 percent confidence interval. Relaxing the confidence interval to 75 percent does not substantially alter the results (appendix E). Sample includes all countries where the Bank’s assistance program included public sector reform and governance activities and for which IEG completed a Coun- try Assistance Evaluation or an analysis of a CAS Completion Report in fiscal 2003–06 and for which governance indicators are available. Alternative governance indicators, such as those collected by the International Country Risk Group, show a similar picture (see appendix E). The European Bank for Reconstruction and Development–World Bank Business Environment and Enterprise Performance Survey (BEEPS) carried out in countries in Central Asia and Eastern Europe and the former Soviet Union, however, shows improvement in the governance environment in a number of countries in that region over the past three years (see appendix E). cal support. For instance, public sector reform in Sector Modernization Large-scale technocratic Bulgaria, with Bank support, has brought substan- and Adjustment Credit in reform programs have tial achievements. All new civil service recruitment Yemen demonstrates the is conducted through an external process; the pay limitations of techno- yielded better results gap between the public and private sectors has cratic changes. The credit when they have enjoyed been narrowed, especially at the higher echelons; aimed to improve the broad-based political and functional reviews of ministries are leading to efficiency, effectiveness, rationalized government structures.2 While a well- and poverty focus of support. designed reform program supported by a range of public expenditures. Al- donors has played its part, a substantial share of though all the reforms were at least partly the success of the Bulgarian reforms has been the implemented, they have yet to make a differ- product of broad-based political support elicited ence in civil service performance or public by the prospect of EU accession. Several other financial management. The civil service transition countries that were offered the possibil- ministry, for the first time, has a well- ity of entry to the EU have also improved their functioning personnel database—a significant governance environments, in contrast with transi- achievement—but what it reveals is continued tion countries that did not have such an opportu- rapid growth in numbers, and thus a failure of nity (Kaufmann 2006). establishment controls that casts doubt on the entire reform process. With the civil service When political support has been less forthcom- nearly twice the size it was when the policy- ing, the effectiveness of Bank-supported adminis- based operation started, the goal of a modern, trative reforms has been undermined. The Public well-performing civil service is now further 35 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S away. The management information system is least in the short term), establishment of new moving ahead, but it has had little impact on institutions with technical mandates, and public financial management, which continues improvements in basic budgeting, reporting, to be improper and lax. There has been an and accounting systems have made steady improvement in budgeting practices, but a progress (IEG 2005d). Medium-Term Expenditure Framework to discipline annual budgeting remains to be In heavily indebted poor counties (HIPCs), developed (IEG 2006s). substantial efforts have been made to track progress and build capacity for public expendi- In Honduras, two public sector reform credits ture management. More than half of the HIPCs and a PRSC in support of civil service adminis- have implemented at least 40 percent of the tration and public financial management actions in public expenditure management reforms yielded only limited results, because plans they adopted in 2002 (IEG 2006d). The capacity constraints and political opposition quality of public expenditure management in were regularly underestimated. The govern- over half of the HIPCs has improved, as ment’s willingness and ability to implement civil evidenced by a moderate upward trend in service and salary reforms were misjudged at a HIPC’s CPIA scores for budget management time when the govern- (see appendix E). The Bank has also relied ment team that negoti- on civil service and ated a credit in support The Bank has also relied on civil service and of these reforms simulta- public financial management reforms to reduce public financial neously undertook steps corruption. Often these have not taken into management reforms to to increase, rather than account the political context. For example, the reduce corruption. reduce, the number and use of the public service as an employment coverage of laws carving mechanism is a long-standing political tradition out privileged employment and compensation in Bolivia. To counter this, the Bolivian govern- regimes for a large share of public sector ment and the Bank joined together to envisage a employees. As a result, progress toward sustain- frontal attack on corruption called the National ably reducing the excessively large civil service Integrity Program. Its implementation relied on wage bill remained limited (IEG 2006f). civil service reforms and improved public sector financial management, with support from two The Bank has had mixed results from its support IDA credits and a half-dozen analytic and for public financial management. In about half advisory efforts. Aimed at developing a stable, of the countries for which IEG reviewed the professional civil service and establishing assistance strategy in the past four fiscal years, standards for procurement, these reforms also progress achieved through the Bank’s lending would have reduced opportunities for parties to in this area has been limited. Particularly in use public sector jobs and contracts as political Africa, where public financial management has rewards. The achievements have been modest in been a centerpiece of the Bank’s support for relation to objectives and have had a negligible institutional reforms, it has suffered from limited impact on incentives and behaviors in the public country ownership of the reform agenda and sector. The reforms have not been effective would benefit from a deeper diagnosis of because they were fundamentally incompatible underlying political and institutional constraints. with Bolivia’s patronage-based governance Efforts to build capacity for institutions have system (IEG 2005c). progressed slowly in areas such as procurement and auditing where country ownership has been Coalition building and incremental reforms can uncertain. Areas that have received political bring results support, such as tax administration (which When political conditions are not amenable to promises increased government revenue, at broad-based public administration reforms, 36 S T R E N G T H E N I N G P U B L I C S E C T O R A C C O U N TA B I L I T Y balancing technocratic reforms with coalition have been few indict- Balancing technocratic building across a broad spectrum of affected ments or convictions reforms with coalition interests can help bring results. This may require stemming from cases incremental changes, identifying reform that the Bureau has building helps bring opportunities that are politically acceptable, and investigated, because the results but requires creating momentum. In Senegal for example, cooperation of the incremental steps rather building political consensus and taking a gradual Director of Public Pros- approach made the difference between success- ecutions—a highly politi- than large-scale reforms. ful and unsuccessful regulatory reform in differ- cized office—is needed ent sectors (box 4.1). Similarly, experience with for a case to proceed. Recently, laws have been procurement reforms has shown that they need enacted to give the Anticorruption Bureau more than a single champion to succeed. They authority to handle its own prosecutions, but the require active and continuous engagement with impact of this has yet to be felt (IEG 2006i). key actors such as legislators, external auditors, civil society, and the media to form a critical Similarly, in Albania, a specialized small group mass of supporters for the reforms (IEG under the cabinet monitors a broad anticorrup- forthcoming). tion strategy but lacks enforcement power because it has to refer its findings to a New laws and institutions need enforcement prosecutor-general (IEG 2005a). In Madagas- capacity car, the government has—with IDA support— Bank-supported anticorruption efforts have put in place a package frequently focused on the adoption of legislation of reforms including Anticorruption efforts and the establishment of institutions and an anticorruption law, agencies, but frequently failed to ensure that agency, and Superior have often focused on there is adequate enforcement capacity to render Council; a procurement new laws and institutions these measures effective. In Malawi, where oversight agency; and but have given little entrenched and worsening corruption afflicts the mandatory asset declara- public service, the Bank has supported the tions for public servants. attention to enforcement government’s Anticorruption Bureau. But there These measures, while capacity. Box 4.1: Building Coalitions Made for Successful Regulatory Reforms in Senegal Consensus building has helped reforms in certain sec- In electricity, by contrast, privatization has failed. The tors in Senegal succeed. The telecommunications sec- power company is still fully state-owned, and service is tor is now privatized and competitive, despite the expensive and inefficient. The unsuccessful reform plans opposition of the former state phone monopoly. The re- were too far-reaching and too dependent on a hoped-for form was successful because it was gradual, involved the strategic investor. opposition political parties, and was based on a trans- In urban transport, the goal of fostering private bus parent bidding process. companies that would serve the poor has not been In the water sector too, supplies have been improved, achieved. Even though the Bank tried to achieve partic- at least in urban areas. Water reforms were successful ipatory solutions, the consultative process missed some because the Bank and government settled on a public- key leaders who could have helped secure ownership of private lease mechanism that was not the Bank’s first the reforms. choice but was politically acceptable within the Sene- galese context. Source: IEG 2006q. 37 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S important, have thus far had a limited impact on ically, Bank assistance programs have the prevalence of corruption, because the emphasized legal and regulatory frameworks for independence of the anticorruption agency and the financial sector, but they have underesti- Superior Council is in question. Corruption is mated the time and human capacity required to still perceived to be widespread and endemic enforce them. Fifteen of 24 countries that had (IEG 2006h). borrowed for legal or regulatory reforms in the financial sector or for strengthening banking A requirement that public servants declare their supervision had shortcomings in adherence to financial assets is a standard component of anticor- prudential regulations and lacked enforcement ruption programs. Asset declarations have been of the prudential framework by the supervisory implemented as part of Bank-supported programs authority. One of the constraints has been the in Albania, Argentina, and lack of institutional capacity, including the Regulatory and service Madagascar, for example. absence of special enforcement power and legal entities need to be To serve their purpose, immunity for supervisors (IEG 2006g, 2004e). such declaration require- independent of political ments must be enforced, Regulatory and service entities need to be influence. verified, and disclosed independent of political influence and adequately (box 4.2). In Albania, for funded to properly carry out their functions. one, their effectiveness has been limited by the lack Independence can be enhanced by techniques of enforcement mechanisms to track transactions such as clear reporting lines, transparency of and property ownership (IEG 2005a). operations and reports, merit-based personnel rules, and financial self-sufficiency. The need for enforcement capacity to properly implement legislation aimed at improving Lack of independence has proven an Achilles’ transparency and accountability reaches beyond heel in otherwise successful programs to anticorruption efforts. Implementation of upgrade public sector functions. For instance, prudential regulations state water boards in Nigeria are not financially The need for stronger and supervision in the autonomous. They depend on state govern- enforcement capacity banking sector have also ments for tariff approvals and many of them are reaches beyond suffered from low en- forced to have their staff salaries supplemented anticorruption efforts. forcement capacity. Typ- by state government budgets. This dependence Box 4.2: Asset Declaration Can Reduce Corruption If It Is Enforced and Disclosed Requiring public officials to declare their wealth and assets is Ability to prosecute: Perceived corruption is lower in countries an article in the United Nations, the African Union, and the Inter- whose declaration laws permit the government or anticorrup- American conventions against corruption. Research using Trans- tion body to prosecute officials accused of corruption. parency International’s Corruption Perception Indices identified Verification: Countries that verify officials’ statements have sig- the characteristics that asset declaration laws need to have to nificantly lower corruption than countries that do not. work effectively toward reducing corruption: Disclosure: The laws of some countries require that asset dis- Constitution: Placing the provision into the country’s closures be placed in the public domain, while others allow only constitution—intended to signal the high value placed on offi- designated government officials to view the declarations. Coun- cials’ integrity—does not necessarily translate into reduced tries that give public access to officials’ asset declaration have sig- corruption. nificantly lower corruption than those that restrict public access. Source: Mukherjee and Gokcekus 2006. 38 S T R E N G T H E N I N G P U B L I C S E C T O R A C C O U N TA B I L I T Y makes it impossible for water board managers to to improved service Regulatory reforms have improve their financial situation by demanding delivery and accountabil- made headway against payment from delinquent but politically ity of service providers. In connected customers. Indeed, less than half the Armenia, Bank lending corruption. water the boards supply is actually paid for. As a and analytical work was result, investments in the Kaduna and Katsina important in advancing energy reforms that Water Boards, financed by the Bank’s First Multi- included restructuring the sector, tariff increases, State Water Supply Project, are unlikely to be reduction of cross-subsidization, and privatization sustainable (IEG 2006n). of electricity distribution companies. Between 1993 and 2002, theft of electricity dropped from Similarly, in countries such as the Philippines 30 percent to 10 percent, collections reached 90 and Uganda, where procurement agencies have percent of billings, and supply interruptions a clear mandate, a fair amount of independence, virtually ceased (IEG 2004a). and are adequately resourced, they have successfully helped spur procurement reforms, The progress made in combating corruption while in countries where these conditions are through regulatory reforms suggests that sector- not fulfilled, they have proven less successful specific opportunities to strengthen public (IEG forthcoming). sector accountability should be exploited more consistently across the Bank’s sectoral The interface between the private and public operations. sectors is critical for reducing corruption Opportunities for irregularities occur when the The financial stakes and Disclosure of government public sector intersects with private firms, room for abuse are partic- revenues and contractual particularly in utilities, infrastructure, and ularly high in extractive extractive industries, where the financial stakes industries. Poor gover- arrangements helps are high. Reforms in these areas have made nance and unsound reve- strengthen governance in headway against corruption, even when they nue management have resource-rich countries. have not been part of high-profile anticorrup- been at the core of the tion programs. overall poor development performance of many resource-abundant countries. An important step Turkey illustrates how reforms aimed at greater in strengthening governance in such countries cost-effectiveness can also deliver improved is to require countrywide and industrywide integrity. During the 1990s, Turkey’s electricity- disclosure of government revenues from extrac- generation capacity was provided largely tive industries and related contractual arrange- through “take-or-pay” contracts between the ments, such as production-sharing agreements, government and electricity generators. These concessions, and privatization terms. These sole-source contracts not only resulted in issues require increased attention by the World inflated prices, but also offered opportunities Bank Group (IEG-World Bank, -IFC, and -MIGA for government officials to receive illicit 2005). payments. By supporting the establishment of an independent regulatory agency, Bank- Increasing transparency to foster financed adjustment operations paved the way accountability for direct contracting between sellers and Transparency is an essential foundation of good buyers of electric power. This, in turn, has governance and accountability. The benefits of helped to combat corruption at its source (IEG transparency are twofold. First, evidence shows 2005n). that the access to timely, relevant, and high- quality information reduces the incidence of Increased competition, when combined with the corruption (Svensson 2005; Recanatini, Prati, right regulatory reforms and supervision, can lead and Tabellini 2005). Second, transparent 39 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S institutions—in the private or public sector— surveys showing that the share of grants earn the trust and acceptance of the public. This reaching the schools had risen from 20 percent acquired legitimacy strengthens the institutions to 90 percent (IEG 2004j). These improvements while encouraging civil participation and emphasize the role of information in mobilizing promoting accountability. Transparency there- “client power” for better expenditure outcomes. fore acts as a regulator of the quality of governance. Lack of transparency in public Public procurement is a second important sector processes offers government officials target for increased transparency. In Turkey, a opportunities for irregular behavior and new public tendering system has introduced prevents the citizenry from holding them to more transparent bidding procedures, including account. World Bank operations have sought to the publication of winning bids. Such publica- bring transparency to a tion is intended to help eliminate the corrupt Bank operations have wide variety of public practice of the government accepting under- management processes, priced bids and then, in return for kickbacks, helped bring including public financial agreeing to cost escalations after contract award transparency to a wide management, public pro- (IEG 2005n). Transparency has also enhanced variety of public sector curement, tax and cus- the effectiveness of procurement agencies in the toms administration, the Philippines and in Uganda. In the Philippines, management processes. judicial system, and pub- civil society representatives have a mandate to lic service provision. observe the tendering process, and in Uganda procurement audits, final contract awards, and Measures to enhance budget transparency are other information are available on public Web typically supported by fiscal or macroeconomic sites (IEG forthcoming). policy–based lending operations. In Turkey, extra-budgetary funds that had undermined Bank operations have also helped improve fiscal discipline were brought into the budget transparency of border inspections and customs and subjected to parliamentary scrutiny as part administration. The Trade and Transport Facilita- of a major post-crisis reform program. Although tion Program in nine countries of southeastern not primarily aimed at corruption, the disman- Europe, which aims to streamline customs tling of the extra-budgetary funds was a signifi- procedures and improve efficiency of border cant step toward tackling corruption at its crossings, has introduced standard electronic source (IEG 2005n). In Albania, the budget customs forms showing duties payable. This preparation process has been debated in Parlia- simple measure discourages officials from asking ment, discussed with nongovernmental organi- for illicit payments. Some participating countries zations (NGOs), and analyzed by the media have added a telephone hotline for truckers to since 2002 (IEG 2005a). report problems. Recent surveys show that the percentage of businesses reporting customs Public expenditure tracking surveys are used in bribes has decreased in all participating countries a growing number of countries to identify but one over the past three years. In Bulgaria, for leakages of public funds and spark corrective example, the percentage of businesses reporting action. In Uganda, for example, an initial survey customs bribes to be a problem has decreased found that only about 20 percent of conditional from 26 percent to 16 percent, according to the grants allocated for schools actually reached Business Environment and Enterprise Perfor- them. This prompted a decision to publicize mance Survey (World Bank 2006b). financial allocations at the district and school levels and to improve accountability through Fostering Local Control to Improve public information campaigns. These measures Accountability resulted in a substantial improvement in the The most frequent interaction between state flow of funds to schools, with subsequent institutions and communities is at the point 40 S T R E N G T H E N I N G P U B L I C S E C T O R A C C O U N TA B I L I T Y where services are delivered. Strengthening Local control can also Stronger community voice community voice, participation, and oversight ensure that results are and oversight can raise can therefore also elevate the demand for sustained and maintained accountability of public institutions and reduce beyond the initial invest- demand for corruption. Mechanisms that promote channels ment period. Nigeria’s accountability and for citizen feedback can raise transparency and experience with promot- reduce corruption. accountability, as the experience of Malawi ing local participation in demonstrates. Local initiatives can even handle the financing and maintenance of water supply enforcement where conventional means of systems illustrates the advantages of local control redress are ineffective or corrupt (box 4.3). (box 4.4). Development interventions are more likely to Community-driven development can strengthen generate sustainable results when the local or undermine capacity of local institutions beneficiaries have authority and responsibility for Bank-financed operations support local and financing and operating them. In the education community control, both through direct support sector, for example, empowering communities to to local governments and through community- manage education funds has increased parental driven development (CDD) initiatives that give involvement in schools and brought improve- control over planning decisions and investment ments in facilities and teacher attendance, resources to community groups and local govern- although there is little evidence yet that it has ments. CDD projects span a broad range of sectors improved educational quality (IEG 2006e). and activities and pursue a variety of objectives. Box 4.3: Using Community Feedback to Enhance Accountability The ongoing third Malawi Social Action Fund Project uses com- ing more responsive to citizen complaints. The project is now munity scorecards to allow citizens to assess the performance working with local authorities and partners to adopt the process of local public services and agencies and thereby foster better as a regular part of local government. quality, efficiency, and accountability. Using participatory tech- The Kecamatan Development Project (KDP) in Indonesia in- niques, more than 500 communities throughout Malawi have cluded an innovative strategy to attack Indonesia’s deeply in- scored local agencies’ performance on their management grained corruption. One part of the strategy provides for systematic processes and the performance of project outputs such as procedures for local authorities to follow up instances of corruption water points and classroom blocks. At the same time, the agen- reported by villagers, including through a “complaints box.” In cies also do a self-assessment. The community and the agency many such cases, inferior work was rectified, and in a few cases then meet to agree on a joint action plan for reform or perform- funds were returned or (rarely) officials removed from office. ance improvement. Such sanctions are almost always the result of villager appeals The first round of scores was less than satisfactory. Commu- to higher-level officials. Complaints through the police and judi- nities complained about lack of transparency. They suggested cial channels have proved of little use; the police can be bribed that local authorities need to allow communities to participate and the judicial system is unable to act on complaints against of- more fully. In order to curb the problem of “ghost workers,” they ficials. The positive effects of improved community-based en- recommended that local project committees verify that the num- forcement may, however, not have succeeded in reducing the ber of people receiving wages equals the number who actually overall level of corruption. A recent impact evaluation found that worked. increasing grassroots participation in monitoring village-level While not yet independently assessed, the scorecard process KDP interventions altered the method of possible corruption but appears to be making communities aware of their role in the man- had relatively limited effects on the overall level of possible agement of subprojects and local authorities seem to be becom- corruption. Sources: Staff reports and IEG 2006m. 41 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Box 4.4: Local Control of Water Systems Gets Better Results The Bank’s support for water supply in Nigeria initially solvent. The 2000 Small Towns Project required the par- supported national-level institutions and then shifted to ticipating towns to contribute financially—through new state-level water boards. More recently, a new ap- Water Consumer Associations—to the project invest- proach was piloted: demand-driven investments in ments and to organize sustainable operation and main- small towns. tenance of the systems they now owned. Twelve of the The results have been dramatically different. The na- 16 towns now have functioning water systems, along tional- and state-level entities remain utterly ineffective with good monitoring and evaluation tools to track and at providing good quality water and are financially in- report on performance. Source: IEG 2006n. They generally combine delivering local infrastruc- project in Jamaica and CDD projects in Brazil’s ture or services with building local capacity for Northeast illustrate the risks of bypassing local decision making and governments and creating parallel structures. In CDD projects that resource management. Jamaica, the sustainability of investments in establish parallel How effectively these roads has been jeopardized by bypassing the structures and procedures mechanisms contribute local councils responsible for maintaining them. risk undermining efforts to better public services In Brazil, Bank-supported CDD projects in the depends on the extent to Northeast have contributed to a proliferation of to strengthen local which they reinforce uncoordinated and ad hoc municipal councils government capacity. legitimate institutions. that constitute a structure parallel to that of the planning process of municipal government (IEG To ensure efficient delivery of local infrastructure 2002b, 2005i).3 and services, CDD projects have frequently es- tablished structures, pro- cedures, and funding Local governments have often been bypassed mechanisms parallel to those of the government. on grounds that they lack capacity. Yet Where the projects have done so, it has tended Albania’s experience counters this argument. to hinder the long-run enhancement of local Even though local governments in that government capacity, and thus to dilute or country are of very recent vintage, they are key undermine efforts to foster decentralization and to realizing projects financed by the Albanian strengthen local control (IEG 2005i). Development Fund. Local governments make the final decisions on subprojects to be By sidelining local governments, CDD projects funded, and with the fund’s assistance they have often missed the opportunity to engage local contract out the works through public tenders officials in decision making and to enhance their (IEG 2002b). capacity to identify, appraise, and supervise subpro- jects. Furthermore, they There is growing recognition in the Bank of the The importance of have missed the opportu- importance of building partnerships between nity to strengthen local community groups and local government strengthening governments’ ability to organizations and of strengthening the use of partnerships between effectively engage with local systems. For example, a recent Bank- local governments and local communities and supported social fund project in Zambia is thereby increase their designed to integrate into the larger decentral- community groups is accountability to them. ization effort in the country. Zambian district increasingly recognized. The Bank’s social fund authorities are given increasing responsibility in 42 S T R E N G T H E N I N G P U B L I C S E C T O R A C C O U N TA B I L I T Y the project cycle for community-level subpro- An Urban Development Different projects aiming jects (IEG 2005i). Other social funds, such as the and Decentralization Proj- one in Honduras, have contributed to decentral- ect in Senegal helped to foster local control ization by helping to strengthen local govern- many urban municipali- within a country did not ment capacities in participatory budgeting, ties learn how to prepare always use consistent project planning, and implementation (IEG and implement invest- approaches. 2006l). ment programs and man- age assets. In doing so, it created parallel While support for enhancing local government institutional arrangements and financing mech- capacity under CDD projects has significantly anisms that bypassed the regular ones. In increased in recent years, the Bank has not Senegal’s rural communities, two different always followed a consistent approach across operations supported capacity building. One— different projects that aimed to foster local the Rural Infrastructure Project—worked within the control within a country. This inconsistency has existing administrative framework for transfer of arisen in countries where the Bank has funds, thereby reinforcing the decentralization supported projects under different institutional process. The other—the social fund—operated arrangements. in many of the same rural areas, but relied on nongovernmental agencies (IEG 2006q). Recog- In Senegal, for example, Bank operations were nizing the need to reconcile the different helping to build the capacity of representative approaches used under past projects, the re- local governments and communities to manage cently approved Participatory Local Develop- programs and raise resources. But the Bank’s ment Project now aims at consolidating the methods seemed inconsistent with its goal of experiences gained from the previous two developing strong local government. projects in rural areas (World Bank 2006k). 43 5 Conclusions T his ARDE has reviewed factors that affect the Bank’s ability to achieve results at the country and sector levels and its efforts to help strengthen the accountability of public sector institutions responsible for country- level results. It has found several factors that can further strengthen the Bank’s effectiveness in helping countries reduce poverty. Strategies need to identify growth patterns Strategies aimed only at boosting overall growth likely to reduce poverty. may miss opportunities to reduce poverty Economic growth over the past decade has led effectively. In countries that had growth without to substantial poverty reduction in many East poverty reduction, IEG found that growth was and South Asian countries, and more recently concentrated in subsectors that created little in the transition economies of Eastern Europe employment. The Bank’s assistance in these and Central Asia. But the kind of growth that countries often effectively contributed to delivers significant poverty reduction continues bringing the countries back to a growth path to elude a considerable number of developing through improved economic management, but countries. it was less successful in bringing about job- creating growth. Sustained growth has been shown to reduce poverty better than sporadic growth. Only two The poor tend to live in rural areas. The Bank has in five borrowing countries have recorded found it challenging to help countries formulate continuous per capita income growth over the and implement strategies that effectively reduce five years 2000–05, and just one in five did so for rural poverty. About half of the Country a full ten years, 1995–2005. High or worsening Assistance Strategy reviews carried out by IEG inequality has dampened the poverty-reducing over the past four fiscal years concluded that the effect of growth in some countries. This was Bank’s assistance in rural areas had either not led particularly the case in countries where growth to satisfactory outcomes or that rural poverty was concentrated in sectors that generated little reduction required increased attention. employment and where the poor lacked the basic skills or mobility to take advantage of To support growth strategies that more consis- opportunities arising from growth. tently translate into poverty reduction, the 45 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S countries, the Bank, and their partners will need achieved a moderately satisfactory or better to further strengthen their understanding of outcome rating during this period. what keeps the poor from participating in growth in each country, what prevents growth However, Country Assistance Evaluations show from reaching particular regions and sectors, that satisfactory project outcomes alone do not and how urban-rural linkages and intersectoral ensure country-sector impact. It is careful mobility can be enhanced. selection and phasing of interventions; long- term engagement; and the complementarities Achieving results requires realistic objectives. of lending, analytical work, and policy dialogue The Bank’s assistance has been effective when it that lead to impact on the sector as a whole. has been realistic about borrowers’ political and institutional capacity and has had well-specified Bank-financed operations have yielded good objectives. But IEG found that almost half of all results when they have supported a country- Bank Country Assistance Strategies it examined formulated, broadly owned sector strategy with over the past four fiscal years were overly clear objectives, and when they have followed a ambitious in two distinct ways. They either distinct pathway designed to reach milestones lacked selectivity or they were founded on that contributed to the achievement of the unrealistic expectations for a reform program country’s objectives for the sector. that was not commensurate with the institu- tional capacity and political situation in the Balancing long-term and short-term goals client country. improves results. Achieving high-quality development results Strategies that lacked selectivity caused the takes time, but pressure to show results quickly Bank’s programs to spread their resources too can divert attention from the quality of results. thinly across too many sectors, which For instance, the Millennium Development diminished the impact of individual operations. Goal of ensuring universal completion of Strategies that were based on unrealistic primary education by 2015 has spurred massive expectations for reform programs led the Bank efforts to increase school enrollment. This is to proceed with policy-based lending, even good, as long as rapid expansion in coverage when country conditions were not fully ready does not come at the expense of attention for the targeted reforms. to learning outcomes. In many countries, however, the rapid expansion did affect the Unrealistic goals are also found in individual learning environment, and only about one-third lending operations. For instance, many financial of primary education sector operations sector loans in crisis countries have had unduly assessed by IEG explicitly aimed to improve ambitious objectives, driven by overestimation learning outcomes. of the government’s commitment to reform and a need to justify large loans. At the same time, In post-conflict countries, the pressure to show Country Assistance Evaluations have also shown quick results is especially intense, but haste may that realistic and well-defined objectives can lead to the neglect of the institution building produce results when stakeholders are focused that is vital for recovery. The long time required on them. to reach many of the intended results underscores the importance of continuity of Achieving sector-level impact requires more Bank engagement and of defining what is than satisfactory project outcomes. feasible for a single operation to achieve. A The performance of the Bank’s portfolio has judicious combination of longer-term objectives improved over the past five fiscal years. Over with interventions that yield quick and visible three-quarters of Bank-financed operations results has been found effective. 46 CONCLUSIONS Strong results demand attention to cross- lending operations led by the social protection sectoral synergies. and financial sector departments have achieved Achieving results in a given sector often requires better outcomes than pension components of identifying and removing constraints in other multisector operations led by economic policy sectors as well. The countries and the Bank need teams. A combination of policy-based lending to pay more attention to these complementary (which is often multisectoral) and sector-specific effects. The impact of infrastructure investments operations can deliver good results. financed by community development projects, for example, has often been diminished by lack Large-scale public sector reforms require of attention to inputs such as teachers, doctors, political commitment. and medicines. Similarly, Bank-supported Achieving and maintaining results requires pension reforms have at times not achieved the public sector institutions that are accountable to desired results because insufficient attention stakeholders. For this reason, Bank Country was given to ensuring that the necessary Assistance Strategies emphasize strengthening macroeconomic, financial, and institutional the performance and accountability of the conditions were in place. The Bank’s matrix public sector. management structure does not provide staff with sufficient incentives to work across sectoral The bulk of the Bank’s support has taken the boundaries and address cross-sectoral issues. form of reform programs in public administra- tion and public financial management. This More attention is also needed to the distribu- assistance has led to improved processes in tional impact of reforms supported by the Bank some countries, but it has not yet translated into at the country level, because not all pro-growth improvements in the perceived quality of policies are distributionally neutral. In the area governance in most of them. Yet experience in a of trade reform, for example, the Bank often few countries shows that it is possible to make failed to conduct sufficient analysis to inform its progress rapidly when there is strong country policy advice and lending about the employ- commitment to do so. ment and poverty effects of reforms. A proper assessment of the distributional impact of Evaluations suggest that reform initiatives have proposed reforms in a particular country often not always been aligned with political circum- requires analysis that reaches beyond the sector stances. They have focused on new legislation in which the reforms are carried out. and institutions, while overlooking enforcement. They have also tended to overlook the interface Even though achieving a particular sector goal between the private and the public sectors, even may require a multisectoral approach, large though regulatory reforms have often been multisector operations are not always an shown to be effective against corruption. effective way to achieve sectoral results. The sectoral impact of multisector operations has Many reform programs have been undermined tended to be weaker than that of sector-specific by lack of political support. The extent of politi- operations, partly because multisector opera- cal opposition is often underestimated at the tions allow for less-intensive engagement of time of program design. Even reforms of a Bank sector teams with country line agencies. In “technocratic” nature, such as those of civil the financial sector, for example, the outcome of service personnel practices, can succeed only loans overseen by Bank financial sector depart- when they build on political commitment. When ments was substantially better than that of political conditions are not ready for reforms, financial sector components of multisector loans. IEG has found it advisable to proceed gradually, Similarly, pension-specific loans and pension identifying opportunities for less contentious reform components of development policy reforms in order to build momentum. 47 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Regulatory reform can help curb corruption. will continue to require a strong focus on The interface between the private and public growth. To ensure that growth translates effi- sectors is fertile ground both for corruption and ciently into poverty reduction, however, the for combating it. Reforms to regulatory regimes countries, the Bank, and their partners will have made headway against corruption even need to find effective ways to enhance the when they have not been part of comprehensive ability of the poor to participate in this growth. anticorruption programs. Such sector-specific This will require country-level analysis of the opportunities to combat corruption need to be binding constraints to employment-creating more systematically exploited in Bank operations. growth and to growth in regions where the poor may be concentrated, as well as analysis Transparency and local control encourage the of the factors that hinder the intersectoral mo- public sector to deliver. bility of the poor. Transparency is the foundation of good • A well-articulated results chain to achieve sectoral governance because access to information outcomes. A well-articulated results chain al- reduces corruption, and transparent institutions lows Bank operations to ensure that objec- earn the public’s trust. Bank operations have tives are set realistically, that cross-sectoral helped bring more transparency to a variety of constraints to achieving them are adequately public management processes, including considered, and that due attention is given to budget formulation and execution, procure- capacity building, particularly when the Bank’s ment, and customs administration. assistance focuses on policy-based lending. Ef- fective articulation and utilization of the re- Local control and community participation can sults chain also requires efforts to enhance make public sector institutions more account- country capacity to collect and use perform- able. Bank operations support such local control ance information. in two main ways: by upgrading local govern- • A realistic assessment of the political economy ment agencies and by channeling resources of governance-related reforms. The Bank can directly to communities. But community-driven provide countries with the tools needed to development projects can also dilute efforts to strengthen government processes, and foster decentralization when they establish thereby to improve the governance envi- structures parallel to those of local government. ronment, but effective use of those tools re- There is now growing recognition in the Bank of mains in the hands of country decision the importance of strengthening the use of local makers. Thus, reforms to improve the ac- systems in the course of promoting community countability of public sector institutions re- development. quire broad-based political support. When such support is absent, an incremental ap- Future directions proach that allows momentum for reforms This ARDE finds three important areas where to build can help deliver results. These re- the Bank can further strengthen its effectiveness forms can be further enhanced with contin- in helping countries reduce poverty. ued efforts to foster local demand for accountability through increased trans- • A focus on the nature of growth. Poverty reduction parency of government processes. 48 APPENDIXES APPENDIX A: PROJECT PERFORMANCE RESULTS This appendix presents long-term trends in ments and consist of 410 ICR Reviews and 92 project performance based on IEG project Project Performance Assessment Reports evaluations. Consistent with past ARDEs, the (PPARs).2 This newly evaluated cohort consists appendix uses the year 1990 as a starting point of 89 development policy lending operations in analyzing long-term trends. Following a brief and 412 investment operations, the vintage of description of the cohort of recent IEG project which is shown in figure A.1.3 The data for fiscal evaluations, analysis of the Bank’s lending 2005 exits represent a partial cohort of lending effectiveness is presented for each of IEG’s three exits (274 out of 317). key performance criteria: outcome, sustainabil- ity, and institutional development impact. Performance Trends Outcome Composition of the ARDE 2006 Exiting Projects in exit fiscal year 2004 registered an Cohort increase in performance outcome,4 which IEG has evaluated 502 closed projects since the jumped from 74 percent to 78 percent satisfac- last ARDE; 82 percent of these exited the Bank’s tory between fiscal 2003 and 2004, as shown in portfolio during fiscal 2004 and 2005.1 These figure A.2. Project performance in exit fiscal 2004 evaluations cover US$33.6 billion in disburse- exceeded the Strategic Compact target of 75 Figure A.1: ARDE 2006 Exiting Cohort by Approval Year 80 70 Number of evaluated projects 60 50 40 30 20 10 0 Pre- 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 1991 Approval fiscal year Development policy lending Investment Source: World Bank database. 51 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure A.2: Project Performance Continues to Meet Strategic Compact Target 100 Percent satisfactory outcome 80 60 40 20 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005* Exit fiscal year By project Weighted by disbursement Source: World Bank database. Note: 2005 data are partial (dashed line). percent satisfactory outcomes, which had not substantial or high institutional development been met by exit fiscal 2003 projects. impact.7 The institutional development impact rating represents a significant increase over the IEG had evaluated 82 percent of the 317 fiscal fiscal 2003 cohort rating of 48 percent. For the 2005 exits as of September 15, 2006. For this fiscal 2005 (partial) cohort, institutional partial fiscal 2005 cohort,5 outcome is rated development impact modestly decreased to 64 satisfactory for 82 percent of projects, and for 87 percent (disbursement weighted) substantial or percent weighted by disbursements. This better. Notwithstanding this modest decline, represents a significant recovery from the fiscal both sustainability and institutional develop- 2003 drop and a return to improved results that ment impact ratings represented, at a minimum, began in fiscal 2000. a 15-percentage-point increase for the fiscal 2005 (partial) cohort over the fiscal 2003 cohort. Sustainability and institutional development As with outcome ratings, these improvements impact may indicate a resumption of the upward trend Eighty-eight percent (weighted by disburse- in project performance (figure A.3). ments) of the exit fiscal 2004 cohort are rated “likely” or “highly likely” to be resilient to future Regional performance risks.6 The fiscal 2004 rating for sustainability Figure A.4 presents the percentage of satisfac- marks a significant increase from the fiscal 2003 tory project outcomes, weighted by disburse- cohort rating of 73 percent and represents a ment, for the fiscal 2001–05 (partial) cohort return to the upward trend that began in 1996. compared with the fiscal 1996–2000 cohort. The For the fiscal 2005 (partial) cohort, sustainability East Asia and Pacific, Middle East and North ratings weighted by disbursements continued to Africa, Europe and Central Asia, and South Asia improve, increasing to 90 percent. Regions are the top performers for the fiscal 2001–05 (partial) cohort, exceeding the Bank Sixty-eight percent (weighted by disbursements) average of 81 percent. The Latin America and of the fiscal 2004 cohort are rated to have a Caribbean Region, which was the best performer 52 A P P E N D I X A : P R O J E C T P E R F O R M A N C E R E S U LT S Figure A.3: Long-Term Trends in Sustainability and Institutional Development 100 100 Percent substantial or better Percent likely or better 80 80 60 60 40 40 20 20 1990 1993 1996 1999 2002 2005* 1990 1993 1996 1999 2002 2005* Exit fiscal year Exit fiscal year By project Weighted by disbursement Source: World Bank database. Note: 2005 data are partial (dashed line). for the fiscal 1996–2000 cohort, is the only one was below the Bank average in both periods and that declined in performance for the fiscal continues to lag behind all other Regions. The 2000–05 (partial) cohort. The Africa Region Europe and Central Asia and South Asia Regions improved in performance for fiscal 2000–05, but made the most progress in outcome perform- Figure A.4: Projects Improved in All but One Region East Asia Distribution of disbursements by Region and Pacific (fiscal 1996–2005*) Middle East and North Africa South Asia Africa 14% 13% Europe and Middle East Central Asia and North Africa 5% South Asia East Asia and Bank-wide Latin America Pacific and the 24% Latin America Caribbean and the 27% Caribbean Africa Europe and Central Asia 17% 40 50 60 70 80 90 100 Percent satisfactory outcome (weighted by disbursement) Fiscal 1996–2000 Fiscal 2001–05* Source: World Bank database. Note: 2005* data are partial. 53 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S ance, improving by 15 and 13 percentage points, social protection sector and in health, nutrition, respectively. and population. Outcomes for economic policy and water supply and sanitation were below the Sectoral performance Bank-wide average in both periods. Compared with the fiscal 1996–2000 cohort, the outcome performance weighted by disburse- Lending instrument performance ment for the fiscal 2001–05 (partial) cohort Outcomes of development policy lending improved in 10 of 14 sector boards.8 Figure A.5 operations rebounded in fiscal 2004, registering presents the sector boards’ outcome perform- a significant improvement to 94 percent from a ance in order of improvement. The biggest low of 71 percent in fiscal 2003. As figure A.6 improvements in outcome ratings were in shows, outcomes (disbursement-weighted) of energy and mining, environment, water supply development policy lending have fluctuated in and sanitation, and the financial sector. The recent years. However, exit fiscal 2005 (partial), largest declines in performance were in the at a high disbursement-weighted performance Figure A.5: Trends in Sectoral Performance Improvers Disbursements by Sector Board (Fiscal 1996–05*) GIC Water supply Global FSP Urban and sanitation information development 4% and tecnology PSG 1% 4% TR Transport Economic policy ED 11% 14% ENV Social Education EMT protection 8% 5% RDV Energy and Rural mining Bank-wide sector 13% WSS 12% Environment EP Public sector Financial 2% 40 50 60 70 80 90 100 governance Private Health, sector Percent satisfactory outcome (weighted by disbursement) 7% sector nutrition, 11% development population Fiscal 1996–2000 Fiscal 2001–05* 3% 5% Decliners HNP SP UD PSD Bank-wide 40 50 60 70 80 90 100 Percent satisfactory outcome (weighted by disbursement) Fiscal 1996–2000 Fiscal 2001–05* Source: World Bank database. Note: 2005* data are partial. The sector board classification applies to the whole project and enables outcomes to be matched to it. Key: EP = economic policy; ED = education; EMT = energy and mining; ENV = environment; FSP = financial sector; GIC = global information and technology; HNP = health, nutrition, and population; PSD = private sector development; PSG = public sector governance; RDV = rural sector; SP = social protection; TR = transport; UD = urban development; WSS = water supply and sanitation. 54 A P P E N D I X A : P R O J E C T P E R F O R M A N C E R E S U LT S Figure A.6: Long-Term Trends in Development Policy and Investment Lending Development policy lending Investment Percent satisfactory outcome Percent satisfactory outcome 100 100 80 80 60 60 40 40 20 20 1990 1993 1996 1999 2002 2005* 1990 1993 1996 1999 2002 2005* Exit fiscal year Exit fiscal year By project Weighted by disbursement Source: World Bank database. Note: 2005* data are partial (dashed line). of 96 percent, shows that outcome ratings for and 13 percent of all the disbursements that development policy lending operations may be exited during that period. on a steady improvement trend. For these instruments, in the aggregate, New lending instruments outcome is rated satisfactory for 82 percent of IEG has evaluated 172 operations employing the projects, and 95 percent weighted by disburse- Bank’s four new lending instruments— ments, exceeding the Bank-wide averages of 77 Adaptable Program Loans (APLs), Learning and percent and 81 percent, respectively (figure Innovation Loans (LILs), Programmatic Sector A.7). PRSCs and PSALs are the best performers Adjustment Loans (PSALs), and Poverty of the group, exceeding the Bank averages for Reduction Support Credits (PRSCs).9 All but two both projects and disbursement-weighted. APLs, of these operations exited the Bank’s portfolio while exceeding the Bank average for projects, during the fiscal 2001–05 (partial) period, are performing below the Bank average for amounting to $12.6 billion in disbursements, disbursement-weighted, while LILs are perform- and accounting for 12 percent of all the projects ing below the Bank average across the board. Figure A.7: Outcome Performance of New Lending Instruments Percent satisfactory outcome 100 80 60 40 20 LIL APL Bank-wide PSAL PRSC New operations (fiscal 2001–05*) By project Weighted by disbursement Source: World Bank database. Note: LIL = Learning and Innovation Loan; APL = Adaptable Program Loan; PSAL = Programmatic Sector Adjustment Loan; PRSC = Poverty Reduction Support Credit. 55 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Table A.1: Outcome, Sustainability, and Institutional Development (ID) Impact by Various Dimensions, by Project, Fiscal 1996–2005* Exit fiscal year 1996–2000 Sustainability ID impact Number of Share Outcome (% likely (% substantial projects (%) (% satisfactory ) or better) or better) Sector board Economic policy 96 7.3 65.6 63.4 32.3 Educaton 114 8.6 77.2 55.0 29.8 Energy and mining 154 11.7 63.2 52.3 38.8 Environment 55 4.2 72.7 66.7 41.8 Financial sector 79 6.0 62.3 56.0 42.1 Global information/ communications technology 21 1.6 81.0 76.2 57.1 Health, nutrition, and population 77 5.8 64.9 52.6 35.1 Poverty reduction 0 0.0 0.0 0.0 0.0 Private sector development 62 4.7 70.5 60.7 28.3 Public sector governance 86 6.5 78.6 75.0 47.6 Rural sector 253 19.2 65.7 45.3 39.6 Social development 0 0.0 0.0 0.0 0.0 Social protection 66 5.0 86.2 50.0 52.3 Transport 127 9.6 86.1 67.8 56.6 Urban development 67 5.1 77.3 51.5 33.3 Water supply and sanitation 63 4.8 59.7 41.9 27.4 Overall result 1,320 100.0 71.0 55.8 39.7 Lending instrument type Development policy lending 193 14.6 74.6 69.0 43.2 Investment 1,126 85.3 70.3 53.5 39.1 Not assigned 1 0.1 0.0 100.0 0.0 Overall result 1,320 100.0 71.0 55.8 39.7 Network Environmentally & socially sustainable development 308 23.3 67.0 49.2 40.0 Financial sector 79 6.0 62.3 56.0 42.1 Human development 257 19.5 75.8 53.0 37.1 Infrastructure 432 32.7 72.3 56.3 42.3 Poverty reduction & economic management 182 13.8 71.7 68.8 39.4 Private sector development 62 4.7 70.5 63.9 28.3 Overall result 1,320 100.0 71.0 55.8 39.7 Region Africa 385 29.2 57.3 38.9 32.4 East Asia and Pacific 195 14.8 78.6 62.6 44.0 Europe and Central Asia 214 16.2 80.5 73.2 50.5 Latin America and the Caribbean 273 20.7 79.3 63.1 45.7 Middle East and North Africa 102 7.7 69.3 55.0 30.7 South Asia 151 11.4 69.3 54.1 33.3 Overall result 1,320 100.0 71.0 55.8 39.7 56 A P P E N D I X A : P R O J E C T P E R F O R M A N C E R E S U LT S Exit fiscal year 2001–05* Exit fiscal year 1996–2005* Sustainability ID impact Sustainability ID impact Number Share Outcome (% likely (% substantial Number of Share Outcome (% likely (% substantial of projects (%) (% satisfactory) or better) or better) projects (%) (% satisfactory) or better) or better) 80.0 5.7 77.2 75.0 40.5 176 6.4 70.9 68.2 36.0 146 10.3 83.4 82.3 58.3 260 9.5 80.7 69.7 45.7 102 7.2 76.8 81.1 63.6 256 9.4 68.5 63.1 48.6 81 5.7 71.8 73.6 51.9 136 5.0 72.2 70.6 47.8 70 5.0 79.4 90.2 66.2 149 5.4 70.3 71.3 53.5 13 0.9 100.0 100.0 46.2 34 1.2 88.2 84.4 52.9 122 8.6 65.5 71.1 48.7 199 7.3 65.3 63.7 43.4 9 0.6 100.0 100.0 44.4 9 0.3 100.0 100.0 44.4 78 5.5 61.8 70.8 42.7 140 5.1 65.7 66.1 36.3 118 8.3 75.4 77.8 49.2 204 7.5 76.7 76.6 48.5 222 15.7 80.0 72.9 52.6 475 17.4 72.3 57.3 45.6 13 0.9 66.7 70.0 41.7 13 0.5 66.7 70.0 41.7 92 6.5 78.3 69.7 44.6 158 5.8 81.5 60.0 47.8 122 8.6 87.6 84.9 63.6 249 9.1 86.8 75.8 60.1 80 5.7 73.4 68.0 38.0 147 5.4 75.2 60.3 35.9 66 4.7 76.9 74.6 50.8 129 4.7 68.5 57.9 39.4 1,414 100.0 77.1 76.7 52.1 2,734 100.0 74.1 66.1 46.1 205 14.5 84.2 86.5 51.0 398 14.6 79.5 77.5 47.2 1,208 85.4 75.8 75.0 52.2 2,334 85.4 73.2 64.1 45.9 1 0.1 0.0 100.0 100.0 2 0.1 0.0 100.0 50.0 1,414 100 77.1 76.7 52.1 2,734 100 74.1 66.1 46.1 316 23.3 77.4 73.0 52.0 624 22.8 72.2 60.5 46.0 70 6.0 79.4 90.2 66.2 149 5.4 70.3 71.3 53.5 360 19.5 76.1 75.5 51.5 617 22.6 76.0 65.4 45.5 383 32.7 80.4 78.9 55.4 815 29.8 76.1 66.4 48.5 207 13.8 77.2 77.7 45.6 389 14.2 74.6 73.3 42.7 78 4.7 61.8 67.6 42.7 140 5.1 65.7 66.7 36.3 1,414 100.0 77.1 76.7 52.1 2,734 100.0 74.1 66.1 46.1 349 24.7 69.4 63.8 44.2 734 26.8 63.0 49.9 37.9 217 15.3 77.7 76.2 56.3 412 15.1 78.1 69.4 50.5 303 21.4 82.4 87.6 58.3 517 18.9 81.6 81.3 55.0 296 20.9 79.8 79.0 59.4 569 20.8 79.6 71.0 52.8 126 8.9 75.8 73.9 36.3 228 8.3 72.9 64.9 33.8 123 8.7 79.7 83.8 50.4 274 10.0 74.0 66.4 41.0 1,414 100 77.1 76.7 52.1 2,734 100 74.1 66.1 46.1 57 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Table A.2: Outcome, Sustainability, and Institutional Development (ID) Impact by Various Dimensions, by Disbursement, Fiscal 1996–2005* Exit fiscal year 1996–2000 Sustainability ID impact Number of Share Outcome (% likely (% substantial projects (%) (% satisfactory ) or better) or better) Sector board Economic policy 13,610 13.1 56.0 75.5 39.7 Educaton 7,108 6.8 82.9 64.4 36.1 Energy and mining 16,476 15.8 65.3 57.0 41.9 Environment 1,263 1.2 69.0 61.4 38.9 Financial sector 14,249 13.7 75.4 73.3 41.5 Global information/ communications technology 1,150 1.1 92.4 94.3 65.4 Health, nutrition, and population 4,134 4.0 80.4 68.4 42.7 Poverty reduction 0 0.0 0.0 0.0 0.0 Private sector development 3,442 3.3 84.9 75.8 42.2 Public sector governance 4,899 4.7 76.0 78.0 46.8 Rural sector 12,903 12.4 76.6 55.4 48.4 Social development 0 0.0 0.0 0.0 0.0 Social protection 4,833 4.6 97.2 72.6 54.7 Transport 11,012 10.6 89.2 70.1 57.9 Urban development 4,715 4.5 89.7 63.1 30.4 Water supply and sanitation 4,493 4.3 59.7 35.0 20.7 Overall result 104,286 100.0 75.0 66.0 43.4 Lending instrument type Development policy lending 35,612 34.1 74.0 79.7 44.9 Investment 68,674 65.9 75.5 59.2 42.6 Not assigned 0 0.0 0.0 0.0 0.0 Overall result 104,286 100.0 75.0 66.0 43.4 Network Environmentally & socially sustainable development 14,166 13.6 75.9 55.9 47.5 Financial sector 14,249 13.7 75.4 73.3 41.5 Human development 16,074 15.4 86.6 67.9 43.4 Infrastructure 37,846 36.3 75.4 60.0 43.3 Poverty reduction & economic management 18,509 17.7 61.3 77.6 41.6 Private sector development 3,442 3.3 84.9 75.8 42.2 Overall result 104,286 100.0 75.0 66.0 43.4 Region Africa 13,448 12.9 64.4 45.4 33.6 East Asia and Pacific 28,405 27.2 78.3 76.1 46.2 Europe and Central Asia 15,813 15.2 67.7 72.6 46.0 Latin America and the Caribbean 25,562 24.5 86.2 69.1 54.0 Middle East and North Africa 5,905 5.7 72.3 49.8 33.9 South Asia 15,154 14.5 68.0 60.8 30.5 Overall result 104,286 100.0 75.0 66.0 43.4 58 A P P E N D I X A : P R O J E C T P E R F O R M A N C E R E S U LT S Exit fiscal year 2001–05* Exit fiscal year 1996–2005* Sustainability ID impact Sustainability ID impact Number Share Outcome (% likely (% substantial Number of Share Outcome (% likely (% substantial of projects (%) (% satisfactory) or better) or better) projects (%) (% satisfactory) or better) or better) 12,479 12.7 66.0 60.8 26.5 26,090 12.9 60.8 68.7 33.4 8,957 9.1 84.9 92.7 65.4 16,065 7.9 84.0 79.7 52.4 10,156 10.4 83.4 83.3 64.4 26,632 13.2 72.2 66.9 50.4 2,781 2.8 64.1 79.9 41.4 4,044 2.0 65.9 74.1 40.6 8,661 8.8 91.5 96.4 74.9 22,910 11.3 81.5 80.8 56.0 855 0.9 100.0 100.0 57.5 2,005 1.0 95.7 96.7 62.0 6,308 6.4 67.0 77.6 55.3 10,442 5.2 72.3 73.9 50.3 662 0.7 100.0 100.0 33.5 662 0.3 100.0 100.0 33.5 3,243 3.3 81.0 89.1 73.7 6,685 3.3 83.0 82.2 57.4 8,967 9.2 90.6 91.2 52.8 13,866 6.9 85.4 87.3 50.7 11,144 11.4 82.4 79.5 59.2 24,047 11.9 79.3 66.6 53.5 405 0.4 29.8 92.1 22.2 405 0.2 29.8 92.1 22.2 4,788 4.9 77.3 81.9 52.4 9,621 4.8 87.3 76.8 53.6 10,874 11.1 90.2 91.6 59.5 21,885 10.8 89.7 80.1 58.7 4,131 4.2 79.8 73.3 40.6 8,846 4.4 85.1 67.8 35.2 3,581 3.7 76.4 70.6 50.9 8,074 4.0 67.1 50.6 34.1 97,994 100.0 81.1 82.6 54.9 202,281 100.0 77.9 73.7 49.1 36,490 37.2 82.3 83.2 51.6 72,102 35.6 78.2 81.4 48.4 61,504 62.8 80.3 82.2 56.9 130,179 64.4 77.8 69.7 49.4 0 0.0 0.0 0.0 0.0 0 0.0 0.0 0.0 0.0 97,994 100.0 81.1 82.6 54.9 202,281 100.0 77.9 73.7 49.1 14,330 14.6 77.8 79.9 54.7 28,497 14.1 76.9 68.0 51.2 8,661 8.8 91.5 96.4 74.9 22,910 11.3 81.5 80.8 56.0 20,053 20.5 77.4 85.3 59.1 36,128 17.9 81.5 77.2 52.1 29,597 30.2 85.0 83.6 57.4 67,442 33.3 79.6 70.0 49.5 22,109 22.6 77.0 74.5 37.4 40,618 20.1 69.8 76.0 39.3 3,243 3.3 81.0 89.1 73.7 6,685 3.3 83.0 82.2 57.4 97,994 100.0 81.1 82.6 54.9 202,281 100.0 77.9 73.7 49.1 13,200 13.5 75.2 70.2 41.9 26,648 13.2 69.8 57.0 37.7 19,678 20.1 88.7 88.3 69.6 48,083 23.8 82.5 81.1 56.4 17,782 18.1 82.8 91.9 63.0 33,594 16.6 75.7 82.4 55.0 28,545 29.1 76.9 75.1 52.3 54,107 26.7 81.4 72.2 53.1 4,955 5.1 82.8 85.4 36.3 10,860 5.4 77.1 65.7 35.0 13,833 14.1 81.3 88.6 48.3 28,987 14.3 74.3 73.2 39.0 97,994 100.0 81.1 82.6 54.9 202,281 100.0 77.9 73.7 49.1 59 APPENDIX B: SECTOR AND PROJECT OUTCOME RATINGS IN COUNTRY ASSISTANCE EVALUATIONS This appendix presents the sector outcomes and project outcome ratings are based on projects corresponding project outcomes of 18 countries that were included in the CAE and targeted the based on a review of each country’s Country relevant sectors. Columns I & IV represent Assistance Evaluation (CAE), completed by IEG countries that had both satisfactory sector and in fiscal 2004–06 (table B.1). project outcomes and both unsatisfactory sector and project outcomes, respectively. Not all CAEs covered all sectors, and the Column II represents countries that had sectoral outcome ratings are based on CAEs that satisfactory sector outcomes while the majority either had an explicit sectoral outcome rating of projects targeting the sector had unsatisfac- assigned by the CAE task team leader or had tory outcomes. Column III represents countries sufficient sectoral information for the ARDE that had unsatisfactory sector outcomes while team to make an assessment in consultation the majority of projects targeting the sector had with the original CAE task team leader. The satisfactory outcomes. Table B.1: Sectoral and Project Outcomes for Country Assistance Evaluations Completed in Fiscal 2004–06 I II III IV Countries with Countries with Countries with Countries with satisfactory satisfactory unsatisfactory unsatisfactory sector outcome & sector outcome & sector outcome & sector outcome & Total number >=50% satisfactory <50% satisfactory >=50% satisfactory <50% satisfactory of CAEs Sector project outcome (%) project outcome (%) project outcome (%) project outcome (%) rated Social protection 86 14 7 Rural 43 7 50 14 Private sector development 50 33 17 12 Environment 67 33 3 Financial sector 78 22 9 Public sector governance 30 60 10 10 Infrastructure 67 8 25 12 Health 70 8 23 13 Education 69 6 13 13 16 Note: “Satisfactory sector outcome” and “satisfactory project outcome” include sectors and projects rated as moderately satisfactory, satisfactory, or highly satisfactory. Likewise, “un- satisfactory sector outcome” and “unsatisfactory project outcome” include sectors and projects rated as moderately unsatisfactory, unsatisfactory, or highly unsatisfactory. 61 APPENDIX C: PROJECT OUTCOMES AND POLICY AND INSTITUTIONAL QUALITY This appendix explores whether the overall ment component of the CPIA and adds another quality of client countries’ policies and institu- indicator of institutional quality, the International tions affects the likelihood of project success. Country Risk Guide (ICRG) index of law and order, as well as a variable to control for external Using ordinary least squares (OLS) regression shocks, the coefficient of variation of the terms analysis, the hypothesis is tested that policy and of trade index. Economic management remains a institutional factors affect the likelihood of a Bank statistically significant explanatory variable, and project meeting its stated objective, as measured law and order is also positively related to the by an IEG project outcome rating of satisfactory likelihood of satisfactory project outcome. The (including projects rated moderately satisfactory, variation in terms of trade, however, does not satisfactory, or highly satisfactory). The depend- seem to affect the likelihood of project success. ent variable is defined as the percentage of projects with a satisfactory outcome rating in a Adding further explanatory variables related to given country over the period fiscal 2001–05. governance and institutional quality, such as an Only countries with at least three exiting projects indicator of perception of corruption (the ICRG between fiscal 2001 and 2005 are included. index of corruption), does not improve the explanatory power of the regression much As a first step, the World Bank’s Country Policy (column 4). There does not appear to be a statis- and Institutional Assessment (CPIA) ratings tically significant relation between the level of were used as explanatory variables, in addition perceived corruption and likelihood of satisfac- to controlling for income level (ln of GDP per tory project outcome. capita). The CPIA ratings are grouped in four clusters: economic management, structural Finally, adding Regional dummies for each of the policies, policies for social inclusion and equity, World Bank’s six Regions (with the Middle East and public sector management and institutions. and North Africa as the control Region) does not The regression used, for each of the four CPIA yield any additional insights. Once income per clusters, the average overall ratings for capita, the quality of economic management, 1999–2004. The results, shown in column 1 of and law and order are controlled for, countries table C.1, suggest that economic management in the Africa or Latin America and Caribbean and policies for social inclusion both positively Region, for example, are not any more or less affect the likelihood of satisfactory project likely to achieve satisfactory project outcomes outcome. However, because of high collinearity than countries in other Regions. To test further among the four CPIA clusters, alternative specifi- for Regional differences in project success rates, cations are used to further explore the relation- various iterations of regressions were also ship between the quality of institutions and estimated comparing each Region individually likelihood of project outcome. to all other borrowers. These iterations confirmed that there is no Regional difference in Column 3 shows the results of an alternative project outcomes once the other variables specification that retains the economic manage- discussed above are controlled for. 63 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Table C.1: Results of Ordinary Least Squares Regressions 1 2 3 4 5 GDP per capita 0.00 0.01 0.01 0.01 0.00 Ϫ0.23 Ϫ0.81 Ϫ0.57 Ϫ0.42 Ϫ0.18 Economic management 0.15 0.19 0.19 0.17 0.16 (3.91)** (7.62)** (6.79)** (5.83)** (5.17)** Structural policies Ϫ0.08 Ϫ1.33 Policies for Inclusion 0.16 (2.92)** Public sector management 0.01 Ϫ0.10 Law and order 0.04 0.04 0.04 (2.46)* (2.43)* (2.14)* Corruption 0.04 0.05 Ϫ1.43 Ϫ1.70 Terms of trade 0.17 0.11 0.11 Ϫ0.83 Ϫ0.56 Ϫ0.52 Africa Ϫ0.03 Ϫ0.42 East Asia and the Pacific 0.10 Ϫ1.33 Europe and Central Asia 0.02 Ϫ0.25 Latin America and the Caribbean 0.03 Ϫ0.46 South Asia 0.06 Ϫ0.59 Constant Ϫ0.14 Ϫ0.05 Ϫ0.17 Ϫ0.17 Ϫ0.09 Ϫ1.20 Ϫ0.43 Ϫ1.40 Ϫ1.46 Ϫ0.48 Observations 106 106 77 77 77 R-squared 0.50 0.44 0.55 0.56 0.59 Note: * Significant at 5%; ** significant at 1%. Absolute value of t statistics in parentheses. 64 APPENDIX D: POVERTY DATA This appendix provides statistical and method- deciles, the general quadratic and the beta Lorenz ological background to the poverty-related curves underlying the decile consumption/income information presented in chapter 2. distributions. Poverty and inequality measures are then derived from these parametrizations (for Data further details see www.worldbank.org/povcalnet). The analysis is based on a sample of 25 countries for which IEG completed a Country Assistance Table D.1 presents the survey years and poverty Evaluation (CAE) or analysis of a Country line used to calculate the poverty reduction Assistance Strategy Completion Report (CASCR) performance for each country, and table D.2 shows between fiscal 2003 and 2006 and for which the poverty and inequality measures thus derived. comparable household survey data for at least two periods are available from Povcalnet, Rural and urban poverty measures were also between the mid-1990s and the early 2000s. The calculated for each country from the same data data used to calculate changes in poverty sets. However, because Povcalnet does not make reduction, income, and inequality were obtained data available for urban and rural regions from Povcalnet, a database managed by the World separately, with the exception of China, the Bank’s Development Research Group. This poverty measures were calculated and provided database contains data on household income by the Development Economics Research and/or consumption derived from country Group (DECRG) from unit record data. Results household surveys for 82 countries, but data from are presented in table D.3. 2001 onward are only available for 57 countries. For 25 of these, IEG completed a CAE or analysis Changes in poverty reduction can be attributed of a CASCR during fiscal 2003–06 and poverty to changes in income and in income distribu- data is available for at least two years between the tion. The relative contribution of income growth mid-1990s and early 2000s from Povcalnet. and distributional changes to poverty reduction can be derived as follows for the headcount Data in Povcalnet is presented in deciles of per index (Datt and Ravallion 2002): capita consumption or income, depending on the household survey. Consumption and/or H1– Ho= (H1*–Ho) + (H1**–Ho)+ residual income means are converted into 1993 U.S. dollars using the Bank’s purchasing power parity where exchange rate. No adjustments were made for adult equivalencies. (H1*–Ho) provides the change in poverty because of a change in mean consumption/income, Povcalnet also makes available an interactive holding the first year Lorenz curve constant; computational tool that allows calculations of (H1**–Ho) provides the change in poverty because poverty and inequality measures from the data of a change in consumption/income distribution sets. It does this by calculating the parameters of (a shift in the Lorenz curve), holding mean two Lorenz curves from the consumption/ income consumption/ income of the first year constant. 65 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Table D.1: Country Listing, Survey Years, and Poverty Line Country Survey dates Poverty line Albania 2002, 2005 $2.15/day Armenia 1998/1999, 2003 $2.15/day Bolivia 1997, 2002 $2.15/day Brazil 1998, 2004 $2.15/day Burkina Faso 1998, 2003 $1.08/day Cameroon 1996, 2001 $1.08/day China 1996, 2001 $2.15/day Dominican Republic 1996, 2003 $2.15/day Georgia 1998, 2003 $2.15/day Honduras 1997, 2003 $2.15/day Jordan 1997, 2002/2003 $2.15/day Lithuania 1996, 2003 $2.15/day Madagascar 1993, 2001 $1.08/day Malawi 1997/1998, 2004 $1.08/day Moldova 1998, 2003 $2.15/day Nigeria 1996/1997, 2003 $1.08/day Pakistan 1998/1999, 2002 $1.08/day Peru 1996, 2002 $2.15/day Romania 1998, 2003 $2.15/day Senegal 1994/1995, 2001 $1.08/day Sri Lanka 1995/1996, 2002 $1.08/day Turkey 1994, 2003 $2.15/day Ukraine 1996, 2003 $2.15/day Uruguay 1998, 2003 $2.15/day Zambia 1996, 2002/2003 $1.08/day Note: Poverty lines: US$1.08/capita/day and US$2.15/capita/day in 1993 purchasing power parity dollars. 66 A P P E N D I X D : P O V E R T Y D ATA Table D.2: Poverty Headcount, Gini Inequality Index, and GDP per Capita Growth Rates Average Annual annual change change Negative in poverty in GDP growth Poverty Poverty headcount per capita episodes Starting End- headcount headcount Gini in Gini in between between between survey survey in starting in ending starting ending survey survey survey Country year year year year year year years (%) years (%) years Albania 2002 2005 10.8 7.7 28.2 28.7 Ϫ11.3 5.1 0 Armenia 1998/99 2003 38.3 31.3 36.0 33.8 Ϫ4.4 9.8 0 Bolivia 1997 2002 39.1 42.9 58.5 60.2 1.9 0.1 1 Brazil 1998 2004 22.7 19.8 59.8 57.0 Ϫ2.3 0.7 2 Burkina Faso 1998 2003 44.9 28.7 46.9 39.6 Ϫ9.0 1.6 1 Cameroon 1996 2001 32.5 17.1 46.8 44.6 Ϫ12.8 2.4 0 China, rural 1996 2001 72.5 71.0 33.6 36.3 Ϫ0.4 7.2 0 China, urban 1996 2001 9.7 6.5 29.1 33.3 Ϫ8.0 7.2 0 Dominican Republic 1996 2003 11.7 12.0 48.7 51.9 0.4 4.2 1 Georgia 1998 2003 12.9 25.8 37.4 40.4 13.8 6.0 0 Honduras 1997 2003 30.6 36.0 53.0 53.9 2.7 0.1 1 Jordan 1997 2002/03 7.4 7.5 36.4 38.9 0.2 2.0 0 Lithuania 1996 2003 7.7 7.4 32.4 36.0 Ϫ0.5 5.7 1 Madagascar 1993 2001 46.3 61.0 46.1 47.5 3.5 0.4 3 Malawi 1997/98 2004 21.9 20.8 39.0 39.0 Ϫ0.8 0.2 2 Moldova 1998 2003 51.2 29.1 39.1 35.1 Ϫ11.3 4.6 1 Nigeria 1996/97 2003 77.9 71.0 52.0 43.6 Ϫ1.4 0.7 3 Pakistan 1998/99 2002 13.6 17.8 33.0 30.6 7.7 0.7 1 Peru 1996 2002 28.4 32.1 46.2 54.7 2.1 0.3 3 Romania 1998 2003 12.8 12.7 29.4 31.1 Ϫ0.2 4.5 1 Senegal 1994/95 2001 24.0 17.0 41.4 41.3 Ϫ5.3 1.9 0 Sri Lanka 1995/96 2002 6.6 5.8 34.4 40.2 Ϫ2.1 3.3 0 Turkey 1994 2003 18.0 19.4 41.5 43.7 0.8 1.1 2 Ukraine 1996 2003 16.4 5.0 35.1 28.1 Ϫ16.9 4.6 2 Uruguay, urban 1998 2003 4.6 7.0 45.2 44.8 8.6 Ϫ4.6 4 Zambia 1996 2002/03 72.6 76.4 49.8 42.1 0.7 0.3 1 Note: Use of per capita income/consumption and US$1.08 or US$2.15 poverty line in 1993 purchasing power parity dollars may result in poverty indices that differ from country estimates or those calculated by Bank Poverty Assessments and used in other IEG evaluations. For Honduras, for example, recent Bank analysis using consumption rather than the above-reported income-based poverty measures suggests that changes in consumption-based poverty have not been statistically significant. Estimates for some countries may also differ from estimates reported in the World Development Indicators because of the recent updating of the Development Economics Research Group (DECRG) poverty database or different income aggregation. The estimates reported here are based on DECRG’s most recent poverty database and Povcal calculations. Data for Malawi 1997 were estimated by DECRG staff in conjunction with the Poverty Assessment rather than Povcal to ensure comparability with 2004 estimates. Data for Albania were provided by the Europe and Central Asia Region. 67 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Table D.3: Urban/Rural Poverty Headcount Rural Urban Rural Urban poverty poverty poverty poverty Country Year headcount headcount Year headcount headcount Albania 2002 13.1 7.7 2005 10.3 4.4 Armenia 1998/99 31.8 44.7 2003 35.5 27.9 Bolivia 1997 64.7 24.0 2002 72.9 26.6 Brazil 1998 45.5 16.4 2004 37.8 16.7 Burkina Faso 1998 51.3 16.3 2003 31.8 14.2 Cameroon 1996 36.2 9.9 2001 23.9 2.6 China 1996 72.5 9.7 2001 71.0 6.5 Dominican Republic 1996 15.4 9.7 2003 16.1 10.0 Georgia 1998 10.7 13.6 2003 34.2 16.9 Honduras 1997 40.1 18.6 2003 49.4 20.1 Jordan 1997 2002/03 6.8 3.7 Malawi 1997/98 2004 23.4 6.1 Madagascar 1993 2001 69.8 35.3 Moldova 1998 2003 32.0 23.7 Nigeria 1996/97 84.0 71.2 2003 78.1 61.9 Pakistan 1998/99 16.0 7.4 2002 21.6 9.0 Peru 1996 2002 65.4 14.4 Romania 1998 19.5 7.7 2003 21.9 4.8 Senegal 1994/95 57.0 43.0 2001 51.8 48.2 Sri Lanka 1995/96 2002 5.6 0.0 Turkey 1994 2003 31.0 11.4 Ukraine 1996 9.2 21.8 2003 7.5 3.8 Uruguay 1998 4.6 2003 7.0 Zambia 1996 85.7 51.0 2002/03 82.9 64.6 68 A P P E N D I X D : P O V E R T Y D ATA Table D.4: Growth and Distribution Effects on Poverty Change in poverty Change in poverty due to change in due to change Country mean income/consumption in distribution Residual Albania Ϫ6.1 2.6 0.4 Armenia Ϫ3.8 Ϫ2.5 Ϫ0.6 Bolivia 3.0 0.9 Ϫ0.1 Brazil 0.9 Ϫ3.7 0.0 Burkina Faso Ϫ7.4 Ϫ9.2 0.4 Cameroon Ϫ14.5 Ϫ2.5 1.7 China, rural Ϫ2.2 0.4 0.2 China, urban Ϫ6.9 6.0 Ϫ2.3 Dominican Republic Ϫ1.2 1.7 Ϫ0.1 Georgia 9.2 2.3 1.3 Honduras 3.0 3.5 Ϫ1.0 Jordan Ϫ3.3 4.6 Ϫ1.2 Lithuania Ϫ3.4 4.7 Ϫ1.7 Madagascar 13.6 3.3 Ϫ2.1 Moldova Ϫ17.8 Ϫ1.8 Ϫ2.5 Nigeria Ϫ3.6 Ϫ2.3 Ϫ1.0 Pakistan 7.7 Ϫ2.9 Ϫ0.3 Peru Ϫ5.7 9.4 0.0 Romania Ϫ2.7 2.7 Ϫ0.1 Senegal Ϫ7.7 0.9 Ϫ0.2 Sri Lanka Ϫ4.1 6.1 Ϫ2.9 Turkey Ϫ1.1 2.5 0.0 Ukraine Ϫ3.9 Ϫ8.4 0.9 Uruguay, urban 4.1 Ϫ2.5 0.8 Zambia 4.0 Ϫ1.0 0.8 69 APPENDIX E: GOVERNANCE INDICATORS FOR BANK BORROWERS WITH PUBLIC SECTOR PROGRAMS Chapter 4 noted that the bulk of the Bank’s work The KKM indicators, like most other currently to improve the effectiveness and accountability of available governance indicators, are indicators public sector institutions has focused on reform of perception. While such indicators are useful, programs in public administration and public they are subject to a substantial margin of error financial management. Analysis of the World Bank and must be used with caution. Therefore, it is Institute’s Kaufmann, Kraay, Mastruzzi (KKM) useful to consider a range of indicators to look governance perception indicators illustrated that at perceived governance quality. To this effect, in the vast majority of countries assessed by IEG this appendix also presents alternate indicators where the Bank program included public sector from the International Country Risk Guide reforms, KKM governance perception indicators (ICRG) and from the European Bank for have remained unchanged since the mid-1990s, Reconstruction and Development (EBRD)– with a 90 percent level of confidence. The results World Bank Business Environment and are similar when the confidence interval is Enterprise Performance Survey (BEEPS). relaxed to 75 percent (figure E.1). Almost all However, it must be noted that different indica- countries that show an improvement under this tors are based on different survey methodolo- relaxed confidence criteria are in the Eastern gies and may therefore yield results that are not Europe and Central Asia Region. strictly comparable. Figure E.1: Governance Perception Indicators in 35 Bank Borrowers with Public Sector Reform Programs Did Not Change Significantly Between 1996 and 2004 30 28 27 25 25 22 Number of countries 20 15 10 8 4 5 5 5 3 3 3 2 0 Government Control of corruption Regulatory quality Rule of law effectiveness Significant decline Significant improvement No significant change Source: Kaufmann, Kraay, and Mastruzzi 2005. Note: Governance indicators are Kaufmann, Kraay, and Mastruzzi indicators between 1996 and 2004. Classification of significant changes versus no significant changes is based on a 75 percent confidence interval. Sample includes 35 countries where the Bank’s assistance program included public sector reform and for which IEG completed a Country Assistance Evalua- tion or an analysis of a Country Assistance Strategy Completion Report in fiscal 2003–06, and for which governance indicators are available. 71 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Figure E.2: Perceived Governance Indicators Have Not Improved Since 1996 in Selected Bank Borrowers with Public Sector Reform Programs 20 19 19 18 Number of countries 16 14 12 10 9 9 8 7 6 5 5 4 2 1 1 0 Corruption Law and order Bureaucracy quality Deterioration Improvement No change Source: International Country Risk Guide, 2006. Note: Sample includes 35 countries where the Bank’s assistance program included public sector reform and governance activities and for which IEG completed a Country Assistance Evaluation or an analysis of a CAS Completion Report in fiscal 2003–06. ICRG indicators are available for 25 of the 35 countries. Changes over time may not be statistically significant. Figure E.3: Indicators from the Business Environment Surveys Suggest Some Improvement in Governance in Selected Transition Economies Between 2002 and 2005 12 10 9 Number of countries 10 9 8 8 6 6 4 3 3 2 2 2 2 1 0 Corruption Customs Bribe Tax Business trade and frequency administration licensing regulation and permits Decrease in the percentage of firms indicating a problem in 2005 compared with 2002 Increase No change Source: World Bank 2006b. Note: Sample includes 35 countries where the Bank’s assistance program included public sector reform and governance activities and for which IEG completed a Country Assistance Evaluation or an analysis of a CAS Completion Report in fiscal 2003–06 and for which BEEPS indicators are available. BEEPS indicators are available for 11 of the 35 countries. Changes over time may not be statistically significant. 72 A P P E N D I X E : G O V E R N A N C E I N D I C AT O R S F O R B A N K B O R R O W E R S W I T H P U B L I C S E C T O R P R O G R A M S Figure E.4: The Average Quality of Budget and Financial Management Has Modestly Improved in IDA-Only Countries 4.5 4.0 CPIA score 3.5 3.0 2.5 2.0 1999 2000 2001 2002 2003 2004 Year 29 Decision point HIPCs 34 Other IDA-only Source: World Bank CPIA database. Note: Sample includes 29 HIPCs that had reached decision point under the initiative as of August 2006. The bureaucracy quality, corruption, and law and public sector reform through lending. These order components of the ICRG’s political risk indicators, which reflect reported firm behavior category for the same country group also suggest rather than more broad-based governance that governance perception has not improved, perception, suggest an improvement in selected and at times has even deteriorated, since the mid- governance areas between 2002 and 2005 in a 1990s (figure E.2). ICRG indicators are only number of countries (figure E.3). However, these available for 25 of the 35 countries. In the absence changes, too, may not be statistically significant. of information to construct confidence intervals, the ICRG analysis was carried out using a count of Chapter 4 also finds that, on average, the quality sign changes from 1996 to 2005, irrespective of of public expenditure management has statistical significance. Thus, changes presented improved modestly in 29 countries that have below may not be statistically significant. qualified for debt relief under HIPC, as measured by an increase in the average CPIA The BEEPS indicators are based on surveys score for the quality of their budget and financial carried out in countries in Central Asia and management (figure E.4). The quality of public Eastern Europe and the former Soviet Union. expenditure management has also increased in There are BEEPS data for 11 of the 35 countries 34 other IDA-only countries. Did HIPC countries assessed by IEG where the Bank supported improve their CPIA scores in this area signifi- Table E.1: Sign Test for Change in CPIA Scores from 1999 to 2004 for Quality of Budget and Financial Management in 29 HIPC Countries Number of countries Probability of an Maximum probability of observing Number of with increased ratings increase used in I or more increases if probability countries (I) null hypothesis (p) of an increase is <=p 29 16 0.44 (15/34) 0.16 Source: World Bank CPIA database, author’s calculations. 73 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S cantly more than other IDA-only countries? A between 1999 and 2004 are distributed as sign test indicates that the tendency for HIPC binomial, with the probability of an increase for countries’ CPIA scores to increase was HIPC countries equal to 0.44 (that is the somewhat higher than it was for other IDA probability of an increase in the 34 comparator countries at a 84 percent confidence level, but IDA-only countries). The probability of getting not at a 90 percent confidence level (table E.1). 16 or more positive differences if the probability of a positive difference were 0.44 or less is 0.156. Sixteen of 29 HIPC countries posted an increase Therefore, it can be concluded that HIPC in their CPIA score for budget and financial countries were more likely to post an improve- management, while 15 of 34 non-HIPC IDA ment in the CPIA for budget and financial countries did so. The sign test is calculated with management than non-HIPC IDA countries at an the null hypothesis that the differences in score 84 percent confidence level. 74 APPENDIX F: MANAGEMENT RESPONSE Introduction management’s views on IEG’s analysis of This year’s Annual Review of Development measures that could increase the effectiveness of Effectiveness (ARDE) looks at issues around Bank-supported programs. In general, manage- growth strategies and sustainable results. In the ment finds many of IEG’s suggestions to be review, the Bank’s Independent Evaluation useful as the Bank refines its efforts to support Group (IEG) discusses the Bank’s recent record countries in improving their development on growth and poverty reduction, using evalua- effectiveness, but believes that the report could tion findings to comment on the effectiveness of have presented a more balanced picture of Bank World Bank assistance in contributing to poverty- contributions. reducing growth. The review groups lessons around poverty-reducing strategies, achieving Record on Growth and Poverty Reduction results at the sector level, and strengthening The ARDE paints a relatively bleak picture on public sector governance through government growth and poverty reduction. Management commitment, accountability, and transparency. would like to note that for several years develop- This note provides a brief response to IEG’s ing countries have enjoyed very positive growth findings and suggestions. It is organized around rates. Figure F.1 illustrates the point, noting the two main issues: (a) developing countries’ recent recent narrowing of the gap between devel- record on growth and poverty reduction; and (b) oping and industrial countries. Developing Figure F.1: Convergence: Narrowing Gap Between OECD and Developing Countries Average annual per capita income growth 6 5 4 OECD average 1980–2006 3 2 1 0 Ϫ1 Ϫ2 Middle-income Low-income Sub-Saharan Africa 1960s 1970s 1980s 1990s 2001/6 Source: World Bank. Note: OECD = Organisation for Economic Co-operation and Development. 75 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S countries grew at 5–6 percent in 2004–06, even context, the President has appointed a Commis- excluding fast-growing India and China sion on Growth and Development, supported by (although, given the population of these two the Bank and other donors,1 and comprising countries, their inclusion is important). This leading practitioners from government, business, year developing country growth is expected to and the policy-making arena. Over a two-year be 6.8 percent for the fifth consecutive year of period, the Commission is expected to shed light strong growth. The outlook for 2007 is one of on the long-run forces underlying growth experi- continued strong developing country growth, in ences and highlight the actions—at the national spite of higher oil prices. Sub-Saharan Africa and international levels—most likely to improve deserves particular mention: it has had a full developing countries’ growth prospects. In decade of growth, and annual growth in 17 Sub- addition, the Poverty Reduction and Economic Saharan Africa countries exceeded 4 percent Management Network (PREM) has been develop- over the entire decade. These higher growth ing and implementing a multipronged work rates reflect sustained improvement in the program on growth: flagship analytic work (World quality of developing country policies and Bank 2005c, 2005f); capacity building within the institutions (confirmed by the 2003 ARDE and Bank (training, dissemination of good practices, continuing since then). and support to country teams, including “growth diagnostics” pilots); and efforts to bring outsiders’ Poverty Reduction. On poverty, an important views and analytic work into the Bank (PREM conclusion is that the developing world as a conferences and lecture series). whole is predicted to meet the poverty Millen- nium Development Goal (MDG). The latest Increasing the Effectiveness of Bank- projections indicate that the share of the Supported Programs developing country population living on less The ARDE makes a series of observations on than $1 per day will fall from 27.9 percent in 1990 Bank programs, at the level of the Country to 10.2 percent in 2015. Even in Sub-Saharan Assistance Strategy (CAS) and at the level of Africa, where conflict and HIV/AIDs have individual operations. Management would like affected poverty-reduction efforts, poverty fell to comment on country strategies, sectoral from 46.4 percent in 2001 to 44.0 percent in support and cross-sectoral synergies, and 2002. It is likely that the substantially positive governance and transparency. per capita income growth discussed in the previous paragraph has reduced poverty further. Successful CASs The ARDE notes that Bank assistance has been Challenges of Building on This Progress. effective when it has taken a realistic view of Notwithstanding this overall positive record, borrowers’ political and institutional capacity management recognizes that many countries— and has focused on well-specified objectives. In particularly in Africa and South Asia—are far from recent years, these issues have been at the achieving many of the MDGs. However, it is much center of Bank support. An overview of less of a stretch now for Sub-Saharan African borrower capacity is now a key element in the countries, for example, to make the jump from design of CASs and lending operations. The CAS the recent rates of growth to the 7 percent rate sets out the Bank’s diagnosis of the country’s estimated as necessary for them to meet the development situation (political and institu- poverty MDG. To help countries achieve these tional) and outlines a selective program of higher rates of growth and poverty reduction, the planned Bank Group support that is tailored to development community needs a better the country’s needs, supporting the govern- understanding of what works well under what ment’s own development objectives and circumstances and how we can better customize strategy, and against the backdrop of the Bank’s support to clients to create jobs and opportuni- ongoing portfolio and the activities of other ties (the theme of this year’s ARDE). In this development partners. 76 A P P E N D I X F : M A N AG E M E N T R E S P O N S E The Ambition of CASs. According to the issues in infrastructure, the public sector, social ARDE, IEG finds that when CASs overlook sectors, macroeconomic reform, and the capacity, they tend to be too ambitious. Design- agricultural sector (World Bank 2006i).2 Another ing strategies and supporting reform programs major change is tied to the updated policy on is not an exact science. CASs support country- development policy lending (OP 8.60). Since it owned programs. Projecting the success of was adopted in September 2004, the vast strategies requires balancing elements of institu- majority of development policy operations have tional and political capacity and commitment. A included a review of possible negative poverty very circumscribed program can end up missing and social impacts and, where necessary, mitiga- opportunities for reform, as IEG’s own evalua- tion measures (World Bank 2006d). Beyond tions point out. Even in an uncertain political operations, the Bank has produced major climate, a country that has strong and broad- analytic and research work on income distribu- based commitment across stakeholders may be tion and growth, notably the 2006 World able to undertake major reforms successfully. Development Report (WDR), Equity and Development (World Bank 2003b, 2005k, Attribution. The review rightly points out that 2006h). On July 11, 2006, management reported successful poverty reduction in a country cannot to the Bank’s executive directors on the activi- be attributed to Bank assistance alone. The same ties with which the Bank is operationalizing the is true when poverty-reduction efforts are not messages of the 2006 WDR: support to Regions successful. In management’s view, the ARDE on operations and analysis, larger-scale pilot analysis of the Bank’s impact on poverty does programs in support of several countries, and a not fully capture the effectiveness of Bank three-year programmatic research agenda. assistance in helping countries achieve pro-poor growth. This is so because the ARDE analysis The Challenge of Rural Poverty in CASs. uses IEG’s ratings of outcomes of Bank- The ARDE notes that in most countries examined supported country assistance programs drawn by IEG evaluations, poverty reduction has been from its Country Assistance Evaluations or greater in urban than in rural areas, and it links outcome ratings in undisclosed CAS Completion that finding to Bank support being skewed in Report reviews. Outcomes at the country level favor of urban areas. Again, recent information depend on more than Bank performance. They shows progress, notably since the adoption of depend on the performance of other donors, the Bank’s new rural strategy in 2003 (World luck—good or bad—and, most important (as Bank 2003a). First, countries are taking rural the ARDE notes), the quality of the country’s poverty seriously. A review by Bank rural staff own policies, notably economic management. shows that of the 17 country-owned Poverty The subjective rating of outcomes is also a Reduction Strategies prepared in fiscal 2004–05, difficult task. all included detailed rural poverty diagnoses and 13 mentioned national rural development plans. Distributional Impacts. The ARDE argues Within the Bank, 88 percent of CASs prepared in that the Bank has not paid sufficient attention to fiscal 2005 had a satisfactory rating for rural the distributional effects of growth-enhancing poverty diagnosis. Lending is up; lending for reforms. Management would note that issues rural development was 14 percent of total around income distribution are not new to the lending commitments in fiscal 2006. Finally, Bank: there is a tradition of work on these operational quality is now strong. Recent IEG issues, and that work has accelerated over the data show that 84 percent of rural lending last five years and is perhaps not yet reflected in operations recently exiting the portfolio were IEG findings. A notable example is Bank support rated satisfactory in terms of development for Poverty and Social Impact Assessments in objectives. The latest Quality Assurance Group countries in all six Bank Regions, with particular (QAG) quality-at-entry study shows 87 percent of focus on Africa. These analyses have covered rural operations as satisfactory at entry—a good 77 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S predictor of operational success. The 2007 WDR reviewed received a moderately satisfactory currently under preparation will help hone rating or better in the latest quality-at-entry further our knowledge on how best to support assessment (World Bank 2006j). IEG’s project rural and agricultural development. It will ratings at completion show a continuation of the provide more clarity on customizing agricultural longer-term improvement in project results (for strategies to specific country conditions and example, of the fiscal 2005 cohort of operations dealing with the risks posed by heavy and often rated so far, outcomes are rated satisfactory for unpredictable government intervention in 82 percent of projects by number and 87 percent agricultural export markets. weighted by disbursements). The ARDE does not note the substantial improvements in LICUS Strategies in Support of Fragile States. The operations. In 2005, projects in LICUS actually ARDE notes that in state building, which is now achieved higher levels of performance than a central focus of the Bank in low-income projects in non-LICUS low-income countries. countries under stress (LICUS), the Bank needs While this record may not be sustainable in these to demonstrate how past weaknesses will be difficult environments, the situation is a vast avoided. Regarding state building, management improvement over the 50 percent satisfactory notes that we are learning from the experience ratings of LICUS operations in fiscal 2002, before of the past, when the international community the LICUS Initiative began. was too ready to ignore the task of making state institutions more effective and accountable to Bank Lending Support. Management notes their people, focusing instead on delivering that this strong operational quality performance quick fixes through parallel and unsustainable has been achieved with higher lending from the structures. Management also notes that in- International Bank for Reconstruction and country harmonization among donors in fragile Development (IBRD) and increased credits and states—although it remains a challenging task— grants from the International Development is making significant progress. There are several Association (IDA). In fiscal 2006, IDA provided good examples, such as the use of shared transi- $9.5 billion in support to poor countries, more tional results frameworks in the Central African than ever before, with more than half going to Republic, Haiti, and Liberia, and joint country Sub-Saharan Africa. Because of IDA’s per- strategies completed or under way in gradual formance-based allocation system, those funds reform environments such as Cambodia and are likely to be used effectively by the recipient Nigeria, as well as in countries with more severe countries for development and poverty issues, such as the Central African Republic, the reduction. One of IDA’s successes has been to Democratic Republic of Congo, Somalia, and channel funds to better-performing countries Togo. In general, the ARDE does not recognize and, perhaps even more important, to help align the significant progress the international overall donor support behind country-owned community has made in the last three years in its growth and poverty-reduction programs. IDA support to fragile states. remains central to support for the development process in low-income countries. At the same Sectoral Support and Cross-Sectoral Synergies time, the IBRD’s fiscal 2006 lending to middle- and Complementarities income countries—home to two-thirds of the The ARDE notes that the overall performance of world’s poor people—was at its highest level in the Bank portfolio has increased over the last seven years at $14.2 billion. The new strategy for few years but does not highlight how important a stronger Bank Group engagement in IBRD that is. While country focus (doing the right countries is designed to help the Bank engage things) is key, there is no substitute for quality even more productively with these countries. operations (doing things right). There has been a substantial improvement in the quality at entry Sectoral Synergies. IEG’s analysis suggests of operations: 92 percent of the projects that in sectoral work the Bank has not paid 78 A P P E N D I X F : M A N AG E M E N T R E S P O N S E sufficient attention to the inputs required from country strategies; and the Results-Based CAS is other sectors to achieve the outcomes of its the tool for meeting that challenge. interventions. There have been several advances in recent years with regard to cross- Support for Education. IEG notes that in sectoral synergies. The IEG review points out many low-income countries, the objective of examples of operations that successfully ensuring by 2015 that every child completes achieved these synergies: In Bangladesh, Bank primary school education has led to massive support for female secondary education has efforts to build more schools and provide more contributed to reductions in child mortality; educational materials, but that rapid expansion rural electrification operations in general have in coverage has often been at the expense of contributed to improved health outcomes; and attention to learning outcomes. While the rural roads in Morocco have led to increases in importance of expanding enrollment was critical agricultural activity and increases in enrollment and should not be underestimated, it is true that in primary education. Sector groups have been improvements in learning outcomes have working to increase the set of operations that lagged, and countries now need to put as much achieve these synergies. In fiscal 2005 the Water effort into improving learning outcomes as has Supply and Sanitation Sector Board developed gone into getting children into school. The Bank a toolkit jointly with PREM to improve the is scaling up its work in this area to help integration of water supply and sanitation countries improve learning outcomes, as is the policy reforms and implementation arrange- overall donor community. Improvement of ments into poverty-reduction support credits. education quality is now a central part of the Then, in recognition of the fact that water endorsement for Fast Track Initiative support, supply and sanitation services are key to achiev- and countries have taken this on board, as noted ing improved health outcomes (for example, in at the recent Special Ministerial Roundtable on reducing child mortality), in fiscal 2005 the Education at the Annual Meetings in Singapore health and infrastructure sectors also launched (World Bank 2006l). Additional activities include an important initiative to assess the burden of programs to increase incentives to focus on disease and develop tools for setting priorities. learning outcomes and strengthening capacity The approach paper Strategy for Health, to measure outcomes and to benchmark them Nutrition, and Population, discussed by CODE internationally. in June 2006 (World Bank 2006o), proposed a diagnostic tool—the Binding Constraints on Governance and Transparency Outcome Improvement (to be piloted) that IEG concludes that public sector reform initia- would highlight the possible gains from tives have not always been aligned with political enhanced coordination between sectors to realities and that the progress of Bank-supported achieve health goals. More generally, the Bank’s reform programs in public administration and Results Secretariat has worked with networks in public financial management has been slow the development of over a dozen results- because of lack of political support. IEG also oriented sector strategies. The Secretariat seeks finds that the Bank has focused excessively on to strengthen the results chains in these strate- passage of laws and has given insufficient gies in several ways: (a) ensuring that objectives attention to enforcement. Management notes are realistic; (b) identifying important synergies that much has changed in recent years. between the Bank’s interventions in the sector and in other sectors to achieve the objectives, Transparency and Related Reforms. In the and designing a strategy to exploit these past decade, the Bank has recognized that reform synergies; and (c) ensuring that the results of the civil service administration can take a long chain is based on solid evidence about what time and that it is not enough merely to ask for gets results. The challenge, as IEG correctly passage of laws. Thus, more recent operatio- points out, is to adapt the results chains to nal support has focused on service-oriented 79 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S approaches to improving governance and service Engagement in Supporting Good Gover- provision: (a) using public expenditure reviews to nance and Anticorruption. There is a general highlight sectoral spending priorities; (b) consensus that better governance and stronger engaging via sectorwide programs; (c) using anticorruption efforts are central to meeting the information to improve accountability for service MDGs. The Development Committee noted that provision on the ground; (d) decentralizing to the principal objective of the Bank’s governance shift responsibility for service provision to the work should be to develop capable and account- entities that provide services; and (e) adopting able states to deliver services to the poor, community-based approaches to local infrastruc- promote private sector–led growth, and tackle ture investments. Management has also worked corruption effectively (Development Committee with development partners to put together and 2006). They supported the Bank’s engagement implement a public financial management in governance and anticorruption work. The performance measurement framework3 that is strategy set out in the Bank’s paper Strengthen- more objective than previous perception ing Bank Group Engagement on Governance measures. Early experience with the tool is and Anticorruption (World Bank 2006m), encouraging. building on a decade of global experience and evidence, implies a change in how the Bank Doing Business. Regulatory red tape is associ- Group does business: providing incentives to ated with poor governance and corruption. A managers and staff to engage proactively on the thriving, open, and competitive private sector ground on governance issues; addressing can be a strong source of demand for better staffing, skills, and resource needs to operate governance. The Bank’s Doing Business Report effectively in challenging governance settings; (World Bank 2006g) benchmarks business and developing a stronger results framework. regulation in 175 economies. Follow-up work The Bank will now further refine and implement generated by country discussions has led to the strategy and report periodically on results. policy reform, for example, in Bangladesh. Conclusions Community-Driven Development. Even As noted in the introduction, management finds when opportunities for governance reform at many of the suggestions by IEG to be construc- the national level are limited, the Bank has taken tive. Its recommendations are intuitive: focus on advantage of entry points at the local level the nature of growth, better articulate results through community-driven development, chains to achieve sector outcomes, and provide a especially when such an approach also supports realistic assessment of the political economy of the development of local government capacity governance-related reforms. The Bank is ad- and accountability. The ARDE cites two Bank dressing all three as it continues the process of operations, one in Zambia and the other in improving Results-Based CASs. With regard to Honduras, as good examples. In a related area, the pattern of growth, the work designed to management would dispute what the ARDE says operationalize the WDR on poverty and inequal- about the role of municipal councils in Brazil ity is feeding into the CAS process, which often (page 42). There is substantial evidence that already benefited from Poverty and Social Impact these municipal councils have contributed to (a) Analysis. IEG suggests a blend of Bank-supported engaging local officials in decision making; (b) activities across sectors to support CAS enhancing local governments’ capacities to objectives, and management agrees, while at the identify, appraise, and supervise subprojects; same time noting, as IEG also suggests, being and, most important, (c) strengthening local selective in the activities the Bank supports, governments’ ability to effectively engage local given budget constraints, and for IDA, a communities and increase government account- constrained financing capacity. With regard to ability to them. governance, IEG suggests ensuring that partner 80 A P P E N D I X F : M A N AG E M E N T R E S P O N S E countries own the reform approach and pace governance and anticorruption strategy. The supported by the Bank, and management agrees. upcoming review of sectoral and country strate- CAS preparation and consultations are key in gies in fiscal 2007 will address progress across making these assessments, and this process will these three important dimensions for getting be reinforced with the implementation of the results from Bank assistance. 81 ENDNOTES Executive Summary Chapter 2 1. Despite the gravity of events in Timor during 1. Bourguignon (2004b) and Lopez and Serven 2006, management would note that, outside of the se- (2006) show that the poverty elasticity of distribu- curity sectors, Timor’s institutions have proven re- tional changes increases with the level of income and silient in the face of political crisis and that the new decreases with the level of inequality when income is transitional government is committed to involving distributed in a log normal function. They test for log community institutions in recovery. The international normality of income distribution using large cross- community’s response also demonstrates that it has country data sets and find that the log normal distri- learned that a rapid and coordinated response is a crit- bution applies to household income. ical factor in supporting countries that meet setbacks 2. The most recent global poverty figures date back in fragile post-conflict transition processes. to 2002, although more recent data are available for a 2. Management would repeat its finding, noted in number of individual countries that have undertaken the Management Response to the 2005 OED report efforts to monitor living standards on a regular basis. on The Effectiveness of World Bank Support for Com- 3. The poverty data presented in this chapter munity-Based and -Driven Development: An OED cover approximately seven years, from the mid-1990s Evaluation, cited as IEG 2005i in this report. Man- onward, with the second observation being between agement finds that many of its most innovative op- 2001 and 2005, depending on country survey data erations are designed to combine decentralization availability. The number of years between surveys initiatives with CDD operations to improve local gov- varies by country, from five in many of the transition ernance. The programs are based on the operational economies to nine in Turkey. Appendix D provides fur- lessons that empowered local communities are able ther details. to hold local government accountable for service de- 4. Management notes that it does not agree that livery, improve local government capacity to deliver the analysis below captures the effectiveness of World effective and demand-responsive services. Manage- Bank assistance in supporting poverty reduction. The ment notes that the issue of strengthening local gov- analysis relates outcomes as rated by IEG in countries ernment is standard now in the preparation of CDD supported by the World Bank to poverty reduction. operations. As the review itself notes, poverty reduction cannot 3. Management notes that the chairman’s sum- be attributed to Bank support alone. The reverse also mary of the World Bank Executive Directors’ Com- holds. Poverty increases cannot be attributed to Bank mittee on Development Effectiveness meeting on support alone. Outcomes in countries are the result August 31, 2005, on The Effectiveness of World Bank of support from the Bank; support from other donors Support for Community-Based and -Driven Devel- (often larger in dollar terms than Bank support); opment: An OED Evaluation, cited as IEG 2005i in sheer luck (good or bad); but, most important, the this report, recognized the “important contribution country’s own performance. That point is effectively of CDD operations to empowering and helping the made in chapter 2 of the review. poor” and supported “scaling-up bank assistance in 5. Management notes that it believes that the response to country demand.” measure of selectivity used by IEG is not well defined 83 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S and is overly subjective. It is often the issue most dis- 11. See, for example, World Bank 2006h, which es- puted in discussions between management and IEG timates that income transfers, including pension sys- on IEG country evaluations. Management would also tems, lower the Gini coefficient in the core OECD note that IEG’s judgment of over-ambition of a CAS countries by 15 percentage points. Similarly, Atkinson may be based on virtually the same factors as its judg- (2004) finds that income transfers lower the Gini co- ment that the outcome was unsatisfactory, calling efficient of inequality in the United Kingdom by 18 per- into question its separate usefulness as a tool for centage points. analysis. 12. A key feature of Bank operations in support of 6. Based on IEG analysis of World Bank 2005b; IEG conditional cash transfer programs is that they have 2006r. included impact evaluations in program design. Doing 7. Based on IEG analysis of World Bank 2005d. so allows for an assessment of targeting effectiveness 8. While the incidence of poverty remains higher as well as of the impact on human capital formation. in rural than in urban Brazil, the absolute number of The evaluation results can then be used to improve poor living in urban areas is higher than that in rural program design, increase transparency and account- areas, due to high urbanization. ability, and ensure broad-based political support. The 9. Management notes that the discussion below World Bank has been instrumental in promoting the seems to imply that the Bank has been inactive around use of evaluations in these programs. issues of equity in development. There has been a long 13. Management would note that an important tradition of work on these issues within the Bank, in- improvement acknowledged by the IEG report cited cluding support over the last five years for Poverty and in this paragraph relates to the Bank’s portfolio per- Social Impact Assessments in literally dozens of coun- formance in LICUS. The percentage of closed projects tries across the world (see World Bank 2006i). Since rated satisfactory by IEG increased from 50 percent in the adoption in September 2004 of the updated pol- fiscal 2002, before the LICUS Initiative, to 58 percent icy on Development Policy Lending (OP 8.60), the in 2003, 65 percent in 2004, and 82 percent in 2005. vast majority of these operations have included a re- Regarding state building, management notes that this view of possible negative poverty and social impacts reflects a need to learn the lessons of the past, when and, where necessary, mitigation measures (see World the international community was too ready to ignore Bank 2006d). Most recently, management has pro- the task of making state institutions more effective and duced several analytical and research reports on in- accountable to their people, focusing instead on de- come distribution and growth, notably the 2006 World livering quick fixes through parallel and unsustainable Development Report (WDR), Equity and Develop- structures. Management also notes that while it re- ment. Management reported to Bank executive di- mains challenging, in-country harmonization among rectors on July 11, 2006, on the set of activities that donors in LICUS is making significant progress. There are in motion to operationalize the messages of the are several good examples, such as the use of shared 2006 WDR. These activities include support to Re- transitional results frameworks in the Central African gions on operations and analysis, larger-scale pilot Republic, Haiti, and Liberia, and joint country strate- programs in support of several countries, and, im- gies completed or under way in gradual reform envi- portantly, a three-year programmatic research agenda ronments such as Cambodia and Nigeria as well as because despite IEG’s message in this section, there more severe LICUS such as the Central African Re- is still much that the development community needs public, the Democratic Republic of Congo, Togo, and to learn to understand what works and what does Somalia. In general, the comments on page 19 do not work in achieving equitable growth. not recognize the significant progress made by the in- 10. Good macroeconomic management, a higher ternational community in the last three years in its sup- physical and human capital stock, well-developed fi- port to fragile states. nancial markets, better-developed infrastructure, 14. Based on IEG analysis of World Bank 2005j. stronger institutions, and more trade openness are as- sociated with more robust growth. The relative im- Chapter 3 portance of various pro-growth policies depends, 1. The analysis excludes the poverty reduction however, on country-specific conditions. and social development sector boards because IEG has 84 E N D N OT E S evaluated very few projects managed by these sector 5. Based on an IEG analysis of World Bank 2004b, boards. For fiscal 2001–05, IEG evaluated 9 projects 2005e. managed by the poverty reduction sector board and 6. A Sectorwide Approach (SWAp) is an approach 13 projects managed by the social development sec- to support a locally owned, coherent program for a tor board. IEG did not evaluate any projects man- sector in a comprehensive and coordinated manner, aged by the two sector boards during fiscal 1996–2000. with greater use of country systems. SWAps embody 2. Management notes that the improvement in a process that is guided by a set of principles, pro- portfolio performance was particularly strong in LICUS gressing over time from government leadership toward environments. The percentage of closed projects rated increased harmonization of implementation mecha- satisfactory by IEG increased from 50 percent in fiscal nisms and use of country systems. SWAps encompass 2002, before the LICUS Initiative, to 58 percent in 2003, a whole sector or a major sub-sector (such as pri- 65 percent in 2004, and 82 percent in 2005. In 2005, proj- mary or secondary education). SWAps are character- ects in LICUS actually achieved higher levels of per- ized by sustained, country-led partnership among formance than projects in non-LICUS low-income development partners and key stakeholders in support countries. There is no guarantee that this will be sus- of country-owned sector policies and strategies. SWAps tained (and indeed it is surprising given the surrounding promote increasing reliance on country systems and policy environment), but it is a testament to the efforts procedures and employ a common framework for of country teams working under difficult conditions. planning, implementation, expenditure, and moni- 3. A recent IEG evaluation of the Bank’s support toring and evaluation. The Bank has supported SWAps for trade found that while earlier on the lack of at- through a variety of lending instruments, depending tention to the poverty and employment effects of on the country and sector. Thus the SWAp is an ap- trade-related interventions supported by the Bank proach to support a government’s program in a par- was due to low prioritization, this no longer seems to ticular sector, rather than a lending instrument per se. be the case. Bank staff now show a considerable in- 7. The review involved analysis of IEG ICR Re- terest in the links between trade and poverty, but the views and Project Performance Assessment Reports, difficulty in incorporating poverty considerations into IEG case studies covering SWAps in Africa under the analytical and advisory activities and project design is IEG Evaluation of Capacity Building in Africa, and caused by a combination of not having the right ex- World Bank ICRs of SWAps completed over the last pertise at the operational level; the complexity of the three fiscal years in health, education, agriculture, issue, which requires multidisciplinary teams cutting and transport. across the Bank’s sector network boundaries; and, in some cases, lack of data to underpin the required Chapter 4 analysis (IEG 2006c). 1. It must be noted that most governance indica- 4. Many Bank-supported HIV/AIDS projects and na- tors currently available, some of which are presented tional AIDS programs have used stable or declining in this report, are perception indicators. While such HIV prevalence (the percent of the population that is indicators are useful, they are subject to wide margins HIV-positive) as an indicator of the impact of preven- of error and must be used with caution. Changes in tion activities. However, HIV prevalence can rise or fall, perception could also come from greater awareness depending on whether more people become infected of underlying governance problems, rather than a than die over a given period. For example, HIV preva- deterioration or lack of improvement. Furthermore, lence can fall if the number of new infections is rising these perception indicators reflect symptoms of bad rapidly, but the mortality rate is rising even faster. It can governance. They do not directly reveal the quality of rise, even if the number of new cases is declining, if treat- government processes, which Bank-supported re- ment programs are effectively keeping AIDS patients forms often aim to improve. The impact of such alive much longer. Thus, to assess success in HIV pre- process changes takes time to manifest itself and af- vention, countries need to track the number of new in- fect perception. fections (HIV incidence) or proxies for it; for success in 2 Based on IEG analysis of World Bank 2006c . treatment, they need to track indicators that directly 3. Management notes that the observations here measure health outcomes, such as the mortality rate. are based on a case study of the CDD experience in 85 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Rio Grande do Norte, to which management has ex- safeguard and business policies. It also provides the pressed numerous substantive disagreements on the data and analysis to substantiate these assessments, and methodology and on the interpretation of the data. it identifies the lessons learned from implementation. On the specific issue of municipal councils, the Bank The borrower prepares and provides the Bank has not contributed to a proliferation of ad hoc coun- with its own evaluation report on the project’s exe- cils that bypass local governments, as the review cution and initial operation, its cost and benefits, the states. Rather, the Bank has helped to establish par- Bank’s and borrower’s performance, and the extent ticipatory municipal councils throughout the North- to which the purposes of the loan were achieved. east that (a) are quite uniform in their structure and The borrower’s report is attached, unedited, to the operating procedures, (b) engage local governments ICR. formally as active participating members, and (c) in Once sent to the Board of Executive Directors, most of the states are increasingly serving as an es- each ICR is evaluated by IEG, which validates or ad- sential decision-making vehicle for the delivery of justs the ratings based on the information provided non-project resources and services. Precisely because in the ICR and other operational documents. IEG of their established effectiveness and transparency, mu- summarizes its findings in an ICR Review. This review nicipal councils are being used increasingly by other conveys the IEG ratings, comments on the lessons to state and federal government programs as the vehi- be drawn and on the quality of the ICR, and suggests cle for prioritizing public investments and channeling whether the project is a candidate for a Project Per- resources to rural areas. Other councils and coordi- formance Assessment Report (PPAR). Bank Regional nating bodies have actually been integrated with the staff have an opportunity to review this summary be- Bank-supported municipal councils, and in that way fore it is completed. are serving as an instrument of integration, rather The purpose of a PPAR is to validate the findings than a cause of fragmentation, as the IEG report and augment the information in the ICR and to ex- states. In sum, management has substantial evidence amine issues and lessons of broad applicability. Some that municipal councils have actually contributed to PPARs are intended to serve as building blocks for (i) the engagement of local officials in decision mak- broad sector studies or Country Assistance Evaluations. ing; (ii) enhancement of local governments’ capaci- They provide independent, field-based, post-com- ties to identify, appraise, and supervise sub-projects; pletion verification of a project’s implementation and and, most important, (iii) strengthen the ability of results. They incorporate the views of the borrower local governments to effectively engage local com- and main stakeholders, and analyze the operation in munities and increase their accountability to them. its sectoral and country context. The operational staff and borrower representatives have an opportunity Appendix A to comment on the draft report. The final report is sub- 1. Data for this appendix include project evalua- mitted to the Bank’s Board and is widely distributed tions conducted through September 15, 2006. Evalu- within the Bank and the borrowing country. ated projects that exited the Bank’s portfolio in fiscal 3. One project was not classified as either devel- 2006 have been excluded from the trend analysis opment policy lending or an investment operation. given the low coverage of this fiscal year. 4. IEG’s measure of outcome considers three fac- 2. The Bank prepares an Implementation Com- tors: relevance, efficacy, and efficiency. Relevance pletion Report (ICR) for each lending operation it fi- measures the expected development impact of a proj- nances. The ICR is prepared at the time of project ect design by weighing the continuing relevance a completion by the staff of the responsible Regional of- project’s objectives. Efficacy refers to the extent to fice (within six months of the final disbursement of the which each objective was achieved, or is expected to Bank loan). It assesses: (a) the degree to which the proj- be achieved. Efficiency measures the cost-effectiveness ect achieved its development objectives and outputs of a project, based mainly on sectorwide best practices as set out in the project documents; (b) other signif- and indicators, where available. Combining these icant outcomes and impacts; (c) prospects for the three factors, overall outcome is rated on a six-point project’s sustainability; and (d) Bank and borrower per- scale, ranging from highly satisfactory to highly un- formance, including compliance with relevant Bank satisfactory. 86 E N D N OT E S 5. This partial coverage is noted with dashed lines evaluated very few projects managed by these sector in all the figures in this appendix and with an aster- boards. For fiscal 2001–05, IEG evaluated 9 projects isk (fiscal 2005*). Lending includes IBRD/IDA, Global managed by the poverty reduction sector board and Environment Facility, Montreal Protocol, and Special 12 projects managed by the social development sec- Financing. tor board. For fiscal 1996–2000, IEG did not evaluate 6. IEG’s sustainability measure assesses the re- any projects managed by the poverty reduction and silience to risk of net benefit flows over time by an- social development sector boards. swering the following questions: At the time of 9. Some of the PRSC outcome ratings are based on evaluation, what is the resilience to risks affecting fu- simplified ICRs, and are thus provisional. They are ture net benefit flows? How sensitive is the interven- being revisited with full ICRs. tion to changes in the operating environment? Will the intervention continue to produce net benefits as long Appendix F as intended, or even longer? How well will the inter- 1. These donors include the Swedish, Nether- vention weather shocks and changing circumstances? lands, and United Kingdom governments and the 7. IEG’s institutional development impact meas- William and Flora Hewlett Foundation. ure evaluates the extent to which an intervention im- 2. See World Bank 2006i, a series of case studies proves the ability of a country or region to make more illustrating PSIA in agriculture, energy, utilities, so- efficient, equitable, and sustainable use of its human, cial sectors, taxation, and macroeconomic modeling. financial, and natural resources. Such improvements can 3. See World Bank 2005g. PEFA is a multi-agency derive from changes in values, customs, laws and reg- partnership program sponsored by the World Bank, ulations, and organizational mandates. Accountability, the International Monetary Fund, the European Com- good governance, the rule of law, and the participation mission, the U.K. Department for International De- of civil society and the private sector are prominent char- velopment, the French Ministry of Foreign Affairs, the acteristics of an effective institutional environment. Royal Norwegian Ministry of Foreign Affairs, the Swiss 8. The analysis excludes the poverty reduction State Secretary for Economic Affairs, and the Strate- and social development sector boards because IEG has gic Partnership with Africa. 87 BIBLIOGRAPHY Atkinson, A. 2004. “Increased Income Inequality Collier, Paul, and David Dollar. 2004. “Develop- in OECD Countries and the Redistributive ment Effectiveness: What Have We Learned?” Impact of the Government Budget.” In G. The Economic Journal 114, June. Cornia, ed., Inequality, Growth and Poverty Datt, Gaurav, and Martin Ravallion. 2002. Is in an Era of Liberalization and Globaliza- India’s Economic Growth Leaving the Poor tion. Oxford, U.K.: Oxford University Press. Behind? Washington, DC: World Bank. Attanasio, Orazio, Erich Battistin, Emla Fitzsi- Deaton, Angus, and Valerie Kozel. 2005. “Data and mons, Alice Mesnard, and Marcos Vera- Dogma: The Great Indian Poverty Debate.” The Hernandez. 2005. “How Effective Are World Bank Research Observer 20(2):177–99. Conditional Cash Transfers? Evidence from Devarajan, Shantayanan, and Ijaz Nabi. 2006. Colombia.” IFS Briefing Note 54. London: In- “Economic Growth in South Asia: Promising, stitute for Fiscal Studies. Un-equalizing,...Sustainable?” World Bank, Bellver, Anna, and Daniel Kaufmann. 2005. Trans- South Asia Region, Washington, DC. parenting Transparency: Initial Empirics Development Committee. 2006. “Statements and Policy Application. Washington, DC: Submitted to the Seventy-Fourth Meeting of World Bank. the Development Committee.” DC2006-0021. Bourguignon, Francois. 2004a. “The Growth Elas- Singapore. ticity of Poverty Reduction: Explaining Het- Dollar, David, and Aart Kraay. 2001. Growth Is erogeneity across Countries and Time Good for the Poor. World Bank Policy Re- Periods.” World Bank Working Paper 28104. search Working Paper 2587. Washington, DC. Washington, DC. Ferreira, Francisco, Phillippe Leite, and Julie ———. 2004b. “The Poverty-Growth-Inequality Litchfield. 2006. The Rise and Fall of Brazil- Triangle, Proceedings of the AFD-EUDN Con- ian Inequality: 1981–2004. Washington, DC: ference 2003.” Notes and Documents 10. World Bank. Cardoso, Eliana, and Souza Andre Porte. 2004. The Gertler, Paul. 2004. “Do Conditional Cash Trans- Impact of Conditional Cash Transfers on fers Improve Child Health? Evidence from Child Labor and School Attendance in Brazil. PROGRESA’s Control Randomized Experi- Working Paper No. 04-W07. Nashville, TN: De- ment.” Health, Healthcare and Economic partment of Economics, Vanderbilt University. Development 94(2). Chhibber, Ajay, and Gaurav Nayyar. 2006. “Ex- IEG-World Bank (Independent Evaluation Group- plaining the Cross-Country Variation in the World Bank). Forthcoming. The Effectiveness Growth Elasticity of Poverty.” Independent of World Bank Fiduciary Assessment. Wash- Evaluation Group, World Bank, Washington, ington, DC: World Bank. DC. Photocopy. ———. 2006a. 2006 Annual Report on Opera- Christiaensen, Luc, Lionel Demery, and Jesper tions Evaluation. IEG Study Series. Wash- Kühl. 2006. The Role of Agriculture in Poverty ington, DC: World Bank. Reduction – An Empirical Perspective. World ———. 2006b. 2005 Annual Report on Opera- Bank Policy Research Paper 4013. Washington, tions Evaluation. IEG Study Series. Wash- DC. ington, DC: World Bank. 89 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S ———.2006c. Assessing World Bank Support for ———. 2006q. Senegal Country Assistance Eval- Trade, 1987–2004: An IEG Evaluation. IEG uation. Washington, DC: World Bank. Study Series. Washington, DC: World Bank. ———. 2006r. World Bank Support to Low- ———.2006d. Debt Relief for the Poorest: An Income Countries Under Stress: An IEG Re- Evaluation Update of the HIPC Initiative. IEG view. IEG Study Series. Washington, DC: World Special Study. Washington, DC: World Bank. Bank. ———. 2006e. From Schooling Access to Learn- ———. 2006s. Yemen Country Assistance Eval- ing Outcomes, An Unfinished Agenda. IEG uation. Washington, DC: World Bank. Study Series. Washington, DC: World Bank. ———. 2005a. Albania Country Assistance Eval- ———. 2006f. Honduras Country Assistance uation. Washington, DC: World Bank. Evaluation. Washington, DC: World Bank. ———. 2005b. 2004 Annual Review of Devel- ———. 2006g. IEG Review of World Bank As- opment Effectiveness. IEG Study Series. Wash- sistance for Financial Sector Reform. IEG ington, DC: World Bank. Study Series. Washington, DC: World Bank. ———. 2005c. Bolivia Country Assistance Eval- ———. 2006h. Madagascar Country Assistance uation. Washington, DC: World Bank. Evaluation. Washington, DC: World Bank. ———. 2005d. Capacity Building in Africa: An ———. 2006i. Malawi Country Assistance Eval- OED Evaluation of World Bank Support. IEG uation. Washington, DC: World Bank. Study Series. Washington, DC: World Bank. ———. 2006j. Pakistan County Assistance Eval- ———. 2005e. Capacity Building in Africa: An uation. IEG Study Series. Washington, DC: OED Evaluation of World Bank Support, World Bank. Country Case Studies. IEG Study Series. Wash- ———. 2006k. Pension Reform and the Devel- ington, DC: World Bank. opment of Pension Systems: An Evaluation of ———. 2005f. China: An Evaluation of World World Bank Assistance. IEG Study Series. Bank Assistance. IEG Study Series. Washing- Washington, DC: World Bank. ton, DC: World Bank. ———. 2006l. “Project Performance Assessment ———. 2005g. Committing to Results: Improv- Report, Honduras, Third Social Investment ing the Effectiveness of HIV/AIDS Assistance— Fund Project and Fourth Social Investment An OED Evaluation of the World Bank’s Fund Project.” Washington, DC, World Bank. Assistance for HIV/AIDS Control. IEG Study Se- ———. 2006m. “Project Performance Assess- ries. Washington, DC: World Bank. ment Report, Indonesia Second Village Infra- ———. 2005h. Country Assistance Evaluation structure Project and Kecamatan Development Retrospective. IEG Study Series. Washington, Project.” Washington, DC, World Bank. DC: World Bank. ———. 2006n. “Project Performance Assess- ———. 2005i. The Effectiveness of World Bank ment Report, Nigeria, National Water Reha- Support for Community-Based and -Driven bilitation Loan, First Multi-State Water Supply Development. IEG Study Series. Washington, Project and Small Towns Water Supply and DC: World Bank. Sanitation Project.” Washington, DC, World ———. 2005j. Improving the World Bank’s De- Bank. velopment Effectiveness: What Does Evalua- ———. 2006o. “Project Performance Assess- tion Show? IEG Study Series. Washington, ment Report, Timor-Leste Community and DC: World Bank. Local Governance Project and Second Com- ———. 2005k. Maintaining Momentum to 2015? munity Empowerment Project.” Washington, An Impact Evaluation of Interventions to DC, World Bank. Improve Maternal and Child Health and ———. 2006p. “Project Performance Assess- Nutrition in Bangladesh. IEG Study Series. ment Report, Uganda, District Health Ser- Washington, DC: World Bank. vices Pilot and Demonstration Project.” ———. 2005l. Mauritania Country Assistance Washington, DC, World Bank. Evaluation. Washington, DC: World Bank. 90 BIBLIOGRAPHY ———. 2005m. Romania Country Assistance ———. 2002a. Peru Country Assistance Evalu- Evaluation. Washington, DC: World Bank. ation. Washington, DC: World Bank. ———. 2005n. Turkey Country Assistance Eval- ———. 2002b. Social Funds: Assessing Effec- uation. IEG Study Series. Washington, DC: tiveness. IEG Study Series. Washington, DC: World Bank. World Bank. ———. 2005o. “Uganda, Sexually Transmitted In- ———. 2002c. Zambia Country Assistance Eval- fections Project.” Project Performance As- uation. Washington, DC: World Bank. sessment Report. Washington, DC, World ———. 2001. India Country Assistance Evalu- Bank. ation. Washington, DC: World Bank. ———. 2004a. Armenia Country Assistance IEG-World Bank, -IFC, and -MIGA. 2005. Ex- Evaluation. Washington, DC: World Bank. tractive Industries and Sustainable Devel- ———. 2004b. Bhutan Country Assistance Eval- opment: An Evaluation of World Bank Group uation. Washington, DC: World Bank. Experience. IEG Study Series. Washington, ———. 2004c. Books, Buildings, and Learning DC: World Bank. Outcomes: An Impact Evaluation of World ———. 2003. Power for Development: A Review Bank Support to Basic Education in Ghana. of the World Bank Group’s Experience with IEG Study Series. Washington, DC: World Bank. Private Participation in the Electricity Sector. ———. 2004d. Brazil Country Assistance Eval- IEG Study Series. Washington, DC: World uation. Washington, DC: World Bank. Bank. ———. 2004e. Economies in Transition: An Jha, Raghbendra. 2004. “Reducing Poverty and In- OED Evaluation of World Bank Assistance. equality in India: Has Liberalization Helped?” IEG Study Series. Washington, DC: World In Inequality, Growth and Poverty in an Era Bank. of Liberalization, Giovanni Andrea Cornia, ed. ———. 2004f. Improving the Lives of the Poor Oxford, U.K., and New York: Oxford Univer- Through Investment in Cities: An Update on sity Press. (Also available at http:// the Performance of the World Bank’s Urban rspas.anu.edu.au/economics/publish/ Portfolio. IEG Study Series. Washington, DC: papers/wp2002/wp-econ-2002-04.pdf) World Bank. Kaufmann, Daniel. 2006. “Myths and Realities of ———. 2004g. The Poverty Reduction Strategy Governance and Corruption.” In World Eco- Initiative. IEG Study Series. Washington, DC: nomic Forum, Global Competitiveness Re- World Bank. port 2005–2006. Washington, DC: World Bank. ———. 2004h. “Project Performance Assessment ———. 2003. Rethinking Governance: Empir- Report. Albania Rehabilitation Credit, Recovery ical Lessons Challenge Orthodoxy. Washing- Program Technical Assistance Project, Public Ex- ton, DC: World Bank. penditure Support Credit; Structural Adjust- Kaufmann, Daniel, A. Kraay, and M. Mastruzzi. ment Credit.” Washington, DC, World Bank. 2005. Governance Matters IV: Governance In- ———. 2004i. “Project Performance Assessment dicators for 1996–2004. Washington, DC: Report. Cambodia, Disease Control and World Bank. Health Development Project.” Washington, Korea Educational Development Institute. 2005. DC, World Bank. Statistics on Korean Education. Seoul: KEDI. ———. 2004j. “Project Performance Assessment Kraay, Aart. 2004. “When Is Growth Pro-Poor? Report, Uganda, Primary Education and Cross-Country Evidence.” World Bank Policy Teacher Development Project and Education Research Working Paper 3225. Washington, Sector Adjustment Credit.” Washington, DC, DC. World Bank. Levy, H., and C. Voyadzis. 1996. Morocco Im- ———. 2003. Dominican Republic Country As- pact Evaluation Report: Socioeconomic In- sistance Evaluation. Washington, DC: World fluence of Rural Roads. Report No. Bank. 15808-MOR. Washington, DC: World Bank. 91 A N N U A L R E V I E W O F D E V E L O P M E N T E F F E C T I V E N E S S 2 0 0 6 : G E T T I N G R E S U LT S Loayza, Norman, and Claudio Raddatz. 2005. Ravallion, Martin, and Gaurav Datt. 1992. “The Composition of Growth Matters for “Growth and Redistribution Components of Poverty Alleviation.” World Bank, Washing- Changes in Poverty Measures: A Decomposi- ton, DC. Photocopy. tion with Applications to Brazil and India in Lopez, Humberto. 2006. “Did Growth Become the 1980s.” Journal of Development Eco- Less Pro-Poor in the 1990s?” World Bank Pol- nomics 38: 275–95. icy Research Working Paper 3931. Washington, Ravallion, Martin, and Emanuela Galasso. 2003. DC. “Social Protection in a Crisis—Argentina’s ———. 2005. “Growth and Inequality. Are They Plan Jefes y Jefas.” World Bank Policy Research Connected?” Background paper for World Working Paper 3165. Washington, DC. Bank, Pro-Poor Growth in the 1990s: Lessons Rawlings, Laura, and Gloria Rubio. 2004. “Eval- and Insights from 14 Countries. Washington, uating the Impact of Conditional Cash Trans- DC, World Bank. fer Programs.” World Bank Research Observer Lopez, J. Humberto, and Luis Serven. 2006. “A 20(1): 29–55. Normal Relationship? Poverty, Growth and Recanatini, Francesca. Alessandro Prati, and Guido Inequality.” World Bank Policy Research Work- Tabellini. 2005. “Why Are Some Public Agencies ing Paper 3814. Washington, DC. Less Corrupt than Others? Lessons from Insti- Mallucio, J.A., and R. Flores. 2004. “Impact Eval- tutional Reform Survey Data.” Paper presented uation of the Pilot Phase of the Nicaraguan Red at the Sixth Jacques Polak Annual Research de Protección Social.” Food Consumption Conference, November 2005, Washington, DC. and Nutrition Division Discussion Paper 184. Skoufias, Emmanuel. 2005. PROGRESA and Its Washington, DC: International Food Policy Impacts on Welfare of Rural Households in Research Institute. Mexico. Washington, DC: International Food Mukherjee, Ranjana, and Omer Gokcekus. 2006. Policy Research Institute. “Officials’ Asset Declaration Laws: Do They Skoufias, Emmanuel, Davis Benjamin, and Ser- Prevent Corruption?’’ In Global Corruption gia De La Vega. 2001. “Targeting the Poor in Report. Berlin and London: Transparency Mexico: An Evaluation of the Selection of International. Households into PROGRESA.” World Devel- OED (Operations Evaluation Department) Re- opment 29(10). named Independent Evaluation Group in De- Svensson, Jakob. 2005. “Eight Questions about cember 2005. Please see IEG for publications. Corruption.” Journal of Economic Perspec- PRS Group. Various years. International Coun- tives 19 (3): 19–42. try Risk Guide. Syracuse, NY. Thomas, Vinod, Mansoor Dailami, Ashok Ravallion, Martin. 2004. “Pro-Poor Growth: A Dhareshwar, Daniel Kaufmann, Nalin Kishor, Primer.” World Bank Policy Research Working Ramon Lopez, and Yan Wang. 2000. Quality Paper 3242. Washington, DC. of Growth. Washington, DC: World Bank. ———. 2001. “Growth, Inequality and Poverty: World Bank. 2006a. Accelerating the Results Looking Beyond Averages.” World Bank Pol- Agenda: Progress and Next Steps. Washington, icy Research Working Paper 2558. Washington, DC. DC. ———. 2006b. “BEEPS-at-a-Glance.” (Various ———.1997. “Can High-Inequality Developing countries.) Washington, DC. Countries Escape Absolute Poverty?” Eco- ———. 2006c. “Bulgaria CAS Completion Report. nomics Letters 56: 51–57. Bulgaria Partnership Strategy Annex C.” Wash- Ravallion, Martin, and Shaohua Chen. 2004. ington, DC. China’s (Uneven) Progress Against Poverty. ———. 2006d. Development Policy Lending Washington, DC: World Bank. Retrospective. Washington, DC. ———. 2001. “Measuring Pro-Poor Growth.” ———. 2006e. East Asia Update: Solid Growth, World Bank Policy Research Working Paper New Challenges. East Asia and Pacific Region, 2666. Washington, DC. Report No. 35748. Washington, DC. 92 BIBLIOGRAPHY ———. 2006f. Global Monitoring Report 2006. ———. 2005e. “Implementation Completion Washington, DC. Report: Burkina Faso Poverty Reduction Sup- ———. 2006g. How to Reform. Doing Business port Credit IV.” Washington, DC. Series. Washington, DC. ———. 2005f. Pro-Poor Growth in the 1990s: ———. 2006h. Poverty Reduction and Growth: Lessons and Insights from 14 Countries. Virtuous and Vicious Circles. World Bank Agence Française de Développement, Bun- Latin American and Caribbean Studies. Wash- desministerium für Wirtschaftliche Zusam- ington, DC. menarbeit und Entwicklung, U.K. Department ———. 2006i. Poverty and Social Impact Analy- for International Development, and the World sis of Reforms: Lessons and Examples from Bank. Washington, DC. Implementation. Washington, DC. ———. 2005g. “Public Financial Management— ———. 2006j. Quality at Entry (QEA7): A QAG Performance Measurement Framework.” PEFA Assessment. Washington, DC. Secretariat. Washington, DC. ———. 2006k. Senegal, Participatory Local De- ———. 2005h. Reducing Poverty on a Global velopment Project. Project Appraisal Docu- Scale, Learning and Innovating for Devel- ment. Washington, DC. opment, Findings from the Shanghai Global ———. 2006l. “Special Ministerial Roundtable on Learning Initiative, B. Moreno-Dodson, ed. Education,” Singapore, September 17. Case Studies. Washington, DC. http://web.worldbank.org/WBSITE/EXTER- ———. 2005i. “Results Focus in Country Assis- NAL/NEWS/0,,contentMDKL21056301~menu tance Strategies: A Stocktaking of Results- PK:34476~pagePK:34370~piPK:34424~the Based CASs.” OPCS. Washington, DC. SitePK:4607,00.html ———. 2005j. “Uruguay CAS Completion Re- ———. 2006m. “Strengthening Bank Group En- port. Uruguay Country Assistance Strategy gagement on Governance and Anticorrup- Annex 1.” Washington, DC. tion.” Washington, DC. ———. 2005k. World Development Report 2006: ———. 2006n. World Development Indicators. Equity and Development. Washington, DC; Washington, DC. New York: Oxford University Press. ———. 2006o. “World Bank Strategy for Health, ———. 2004a. World Development Report 2004: Nutrition, and Population Results.” Back- Making Services Work for Poor People. Wash- ground Note for briefing to the Committee on ington, DC; New York: Oxford University Development Effectiveness on the preparation Press. of the new Bank Health, Nutrition, and Pop- ———. 2004b. “Implementation Completion ulation Strategy. Washington, DC. Report: Burkina Faso Poverty Reduction Sup- ———. 2005a. “Burkina Faso CAS Completion port Credit III.” Washington, DC. Report. Burkina Faso Country Assistance Strat- ———. 2003a. Agriculture and Rural Develop- egy Annex 7.” Washington, DC. ment: Reaching the Rural Poor. Washington, ———. 2005b. “Cambodia CAS Completion Re- DC. port. Cambodia Country Assistance Strategy ———. 2003b. Inequality in Latin America and Attachment 1.” Washington, DC. the Caribbean—Breaking with History. Wash- ———. 2005c. Economic Growth in the 1990s: ington, DC. Learning from a Decade of Reform. Wash- ———. 2001. World Development Report ington, DC. 2000/2001: Attacking Poverty. Washington, ———. 2005d. “Georgia CAS Completion Re- DC; New York: Oxford University Press. port. Georgia Partnership Strategy Annex 2.” Washington, DC. 93 E C O - AU D IT Environmental Benefits Statement The World Bank is committed to preserving en- Saved: dangered forests and natural resources. The • 7 trees Office of the Publisher has chosen to print • 5 million BTUs of Annual Review of Development Effec- total energy tiveness 2006: Getting Results on recy- • 642 lbs. of net cled paper with 30 percent post-consumer greenhouse gases waste, in accordance with the recommended • 2,664 gallons of standards for paper usage set by the Green Press Initiative, a nonprofit program sup- waste water porting publishers in using fiber that is not • 342 lbs. of solid sourced from endangered forests. For more in- waste formation, visit www.greenpressinitiative.org. IEG PUBLICATIONS Study Series 2004 Annual Review of Development Effectiveness: The Bank’s Contributions to Poverty Reduction Addressing the Challenges of Globalization: An Independent Evaluation of the World Bank’s Approach to Global Programs Agricultural Extension: The Kenya Experience Assisting Russia’s Transition: An Unprecedented Challenge Bangladesh: Progress Through Partnership Brazil: Forging a Strategic Partnership for Results—An OED Evaluation of World Bank Assistance Bridging Troubled Waters: Assessing the World Bank Water Resources Strategy Capacity Building in Africa: An OED Evaluation of World Bank Support The CIGAR at 31: An Independent Meta-Evaluation of the Consultative Group on International Agricultural Research Country Assistance Evaluation Retrospective: OED Self-Evaluation Debt Relief for the Poorest: An OED Review of the HIPC Initiative Developing Towns and Cities: Lessons from Brazil and the Philippines The Drive to Partnership: Aid Coordination and the World Bank Economies in Transition: An OED Evaluation of World Bank Assistance The Effectiveness of World Bank Support for Community-Based and –Driven Development: An OED Evaluation Evaluating a Decade of World Bank Gender Policy: 1990–99 Evaluation of World Bank Assistance to Pacific Member Countries, 1992–2002 Financial Sector Reform: A Review of World Bank Assistance Financing the Global Benefits of Forests: The Bank’s GEF Portfolio and the 1991 Forest Strategy and Its Implementation Fiscal Management in Adjustment Lending IDA’s Partnership for Poverty Reduction Improving the Lives of the Poor Through Investment in Cities India: The Dairy Revolution Information Infrastructure: The World Bank Group’s Experience Investing in Health: Development Effectiveness in the Health, Nutrition, and Population Sector Jordan: Supporting Stable Development in a Challenging Region Lesotho: Development in a Challenging Environment Mainstreaming Gender in World Bank Lending: An Update Maintaining Momentum to 2015? An Impact Evaluation of Interventions to Improve Maternal and Child Health and Nutrition Outcomes in Bangladesh The Next Ascent: An Evaluation of the Aga Khan Rural Support Program, Pakistan Nongovernmental Organizations in World Bank–Supported Projects: A Review Poland Country Assistance Review: Partnership in a Transition Economy Poverty Reduction in the 1990s: An Evaluation of Strategy and Performance The Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank’s Support Through 2003 Power for Development: A Review of the World Bank Group’s Experience with Private Participation in the Electricity Sector Promoting Environmental Sustainability in Development Putting Social Development to Work for the Poor: An OED Review of World Bank Activities Reforming Agriculture: The World Bank Goes to Market Sharing Knowledge: Innovations and Remaining Challenges Social Funds: Assessing Effectiveness Tunisia: Understanding Successful Socioeconomic Development Uganda: Policy, Participation, People The World Bank’s Experience with Post-Conflict Reconstruction The World Bank’s Forest Strategy: Striking the Right Balance Zambia Country Assistance Review: Turning an Economy Around Evaluation Country Case Series Bosnia and Herzegovina: Post-Conflict Reconstruction Brazil: Forests in the Balance: Challenges of Conservation with Development Cameroon: Forest Sector Development in a Difficult Political Economy China: From Afforestation to Poverty Alleviation and Natural Forest Management Costa Rica: Forest Strategy and the Evolution of Land Use El Salvador: Post-Conflict Reconstruction India: Alleviating Poverty through Forest Development Indonesia: The Challenges of World Bank Involvement in Forests The Poverty Reduction Strategy Initiative: Findings from 10 Country Case Studies of World Bank and IMF Support Uganda: Post-Conflict Reconstruction Proceedings Global Public Policies and Programs: Implications for Financing and Evaluation Lessons of Fiscal Adjustment Lesson from Urban Transport Evaluating the Gender Impact of World Bank Assistance Evaluation and Development: The Institutional Dimension (Transaction Publishers) Evaluation and Poverty Reduction Monitoring & Evaluation Capacity Development in Africa Public Sector Performance—The Critical Role of Evaluation All IEG evaluations are available, in whole or in part, in languages other than English. For our multilingual selection, please visit http://www.worldbank.org/ieg THE WORLD BANK ISBN 0-8213-6906-7