Improving Effectiveness and Outcomes for the Poor in ... - World Bank

Improving Effectiveness andOutcomes for the Poor in Health,Nutrition, and PopulationConferenceedition

THE WORLD BANK GROUPWORKING FOR A WORLD FREE OF POVERTYThe 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), theMultilateral Investment Guarantee Agency (MIGA), and the International Centre for the Settlement of InvestmentDisputes (ICSID). Its mission is to fight poverty for lasting results and to help people help themselves and theirenvironment by providing resources, sharing knowledge, building capacity, and forging partnerships in the publicand private sectors.THE INDEPENDENT EVALUATION GROUPIEG: IMPROVING DEVELOPMENT RESULTS THROUGH EXCELLENCE IN EVALUATIONThe 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 onassessment of IFC’s work toward private sector development, and IEG-MIGA evaluates the contributions of MIGAguarantee 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 theBank Group’s work, and to provide accountability in the achievement of its objectives. It also improves Bank Groupwork by identifying and disseminating the lessons learned from experience and by framing recommendations drawnfrom evaluation findings.

Improving Effectiveness andOutcomes for the Poor in Health,Nutrition, and PopulationAn Evaluation of World Bank GroupSupport Since 1997http://www.worldbank.org/ieghttp://www.ifc.org/ieghttp://www.miga.org/ieg2009The World BankWashington, D.C.

©2009 The International Bank for Reconstruction and Development / The World Bank1818 H Street NWWashington DC 20433Telephone: 202-473-1000Internet: www.worldbank.orgE-mail: feedback@worldbank.orgAll rights reserved1 2 3 4 11 10 09 08This volume is a product of the staff of the International Bank for Reconstruction and Development / The World BankGroup. The findings, interpretations, and conclusions expressed in this volume do not necessarily reflect the views of theExecutive 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 TheWorld Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries.Rights and PermissionsThe material in this publication is copyrighted. Copying and/or transmitting portions or all of this work without permissionmay be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bankencourages 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 theCopyright 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.Cover images:Top left: As part of their efforts to reduce typhoid and save lives, the Junior Red Cross Club at Nasavusavu Public Schoolin Savusavu, Fiji, has been running a handwashing campaign. Part of this campaign has involved making affordable soapdishes to hang at the school taps and painting a student-designed mural on the wall of their bath house. These projectswere celebrated on October 15, 2008, the first ever Global Handwashing Day. Photo courtesy of Becky Huber Trytten,Peace Corps volunteer.Top right: Doctor treating boy in Ghana. ©Mika/zefa/Corbis.Bottom left: Indian woman with her children, who help her run her business. Photo by Curt Carnemark, courtesy of theWorld Bank Photo Library.Bottom right: Mothers and infants at a World Bank-funded health facility on the island of Nias, off Sumatra, Indonesia.Photo courtesy of Nathalie M-H. Tavernier.ISBN-13: 978-0-8213-7542-6e-ISBN-13: 978-0-8213-7543-3DOI: 10.1596/978-0-8213-7542-6Library of Congress Cataloging-in-Publication Data have been applied for.World Bank InfoShopE-mail: pic@worldbank.orgTelephone: 202-458-5454Facsimile: 202-522-1500Printed on Recycled PaperIndependent Evaluation GroupKnowledge Programs and Evaluation CapacityDevelopment (IEGKE)E-mail: eline@worldbank.orgTelephone: 202-458-4497Facsimile: 202-522-3125

ContentsviiixxixiiixxiiixxxixxxvAbbreviationsAcknowledgmentsForewordExecutive SummaryManagement Response SummaryChairperson’s Summary: Committee on Development Effectiveness (CODE)Advisory Panel Statement1 1 Introduction4 Rationale for World Bank Group Investments in Health, Nutrition, and Population5 World Bank Group Strategies in Health, Nutrition, and Population8 Objectives and Scope of this Evaluation9 Evaluation Design and Methodology10 Organization of the Report13 2 Evolution and Performance of the World Bank’s Country Support for Health,Nutrition, and Population15 The Health, Nutrition, and Population Portfolio and Its Performance23 Monitoring, Evaluation, and the Results Agenda27 Is Health, Nutrition, and Population Support Reaching the Poor?33 3 Lessons from Three Approaches to Improve Outcomes35 Communicable Disease Control41 Reforming Health Systems48 Sectorwide Approaches55 4 The Contribution of Other Sectors to Health, Nutrition, and PopulationOutcomes57 Intersectoral Approaches in Country Assistance Strategies58 Multisectoral Health, Nutrition, and Population Lending63 Health in the Lending Portfolios of Other Sectorsiii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION75 5 IFC’s Health Strategy and Operations77 Evolution of IFC’s Approach to Private Investment in Health80 IFC’s Investment Portfolio in Health85 IFC’s Advisory Services in Health88 IFC’s Institutional Arrangements for the Health Sector90 Social Impacts of IFC Health Investments93 6 Conclusions and Recommendations97 Recommendations101 Appendixes103 A: World Bank Group HNP Timeline123 B: Definition of the Samples Used for Portfolio Reviews and World Bank HNPStaff Analysis127 C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007133 D: IFC Health Investments, Fiscal Years 1997–2007137 E: World Bank Support for Population141 F: World Bank Support for Nutrition145 G: World Bank Support for Analytic Work on HNP147 H: Additional Figures on World Bank HNP Lending, Analytic Work, and Staffing153 I: Evaluation of World Bank Participation in Two Global HNP Partnerships155 J: Management Response165 Endnotes179 ReferencesBoxes7 1.1 Six Phases of World Bank Engagement in HNP10 1.2 Evaluation Building Blocks20 2.1 Analytic Work Supported Better Outcomes in Four Countries22 2.2 Family Planning Can Be Successful in Difficult Environments26 2.3 Early Childhood Interventions Improved Cognitive Development andNutritional Status in the Philippines28 2.4 Is Public Health Spending Pro-Poor?30 2.5 Links Between Health Reform Projects and the Health of the Poor WereComplex and Uncertain37 3.1 Successful Malaria Control in Eritrea39 3.2 Recommendations of IEG’s 2005 Evaluation of World Bank Support for AIDSControl Remain Relevant43 3.3 Shared Themes: Public Sector and Health Systems Reform46 3.4 Consequences of Inadequate Stakeholder Analysis47 3.5 Programmatic Lending Maintained Momentum on Health Reform48 3.6 Genesis of the Sectorwide Approach in Health: An International Consensus62 4.1 Quality-at-Entry for Multisectoral Projects Is Weak63 4.2 Greater Selectivity in Sectoral Participation Can Improve MultisectoralPerformanceiv

CONTENTS64 4.3 Poverty Reduction Support Credits: Multisectoral Development PolicyLending in Support of HNP66 4.4 Health Has Been Featured in World Bank Water Supply and SanitationStrategies Since 199368 4.5 Better Hygiene Behavior through Rural Water Supply and Sanitation in Nepal69 4.6 Health and Transport in the World Bank’s Sector Strategies73 4.7 What Accounts for Fewer Road Fatalities in Romania?84 5.1 Early Hospital Investments Provided Important Lessons86 5.2 What Are Public-Private Partnerships in Health?87 5.3 IFC Against AIDS: A Preliminary Assessment88 5.4 Outpatient Dialysis Services in Romania—A Successful Advisory ServicesPublic-Private Partnership in Health96 6.1 Evolving Value Added of the World Bank Group in HNPFigures3 1.1 Infant Mortality Rates Have Declined in Every Region, but Disparities acrossRegions Are Large5 1.2 Communicable Diseases Remain a Significant Share of the Disease Burden inMost Developing Regions6 1.3 Trends in World Bank Group Commitments and Project Approvals16 2.1 The Number of HNP-Managed Projects Has Risen Slowly, WhileCommitments Have Declined17 2.2 IDA and IBRD Commitments Declined; the Number of IBRD Projects Also Fell,But IDA Projects Rose in Number18 2.3 The Number of Sector Specialists Rose over the Decade19 2.4 Performance of HNP Projects Has Stagnated, while Outcomes in OtherSectors Continue to Improve21 2.5 IDA Project Outcomes in Africa Are Much Lower than in Other Regions25 2.6 Only a Third of Projects Bank-Wide, and Only a Quarter of HNP Projects,Have Substantial M&E26 2.7 The Discrepancy Between Plans, Evaluation Design, and Implementation ofPilots and Impact Evaluations in HNP Projects Approved in Fiscal Years1997–200631 2.8 The Share of Poverty Assessments with a Focus on HNP Declined36 3.1 AIDS Accounted for More than Half of Communicable Disease ProjectsApproved in Fiscal Years 1997–200637 3.2 HIV/AIDS Projects Have Performed Less Well than Other CommunicableDisease Projects42 3.3 The Share of HNP Project Approvals with Health Reform Objectives HasDeclined in Middle-Income Countries and in Most Regions44 3.4 Excess Inpatient Bed Capacity Declined across the Kyrgyz Republic49 3.5 After an Initial Spurt, Growth in World Bank Support for Health SWApsResumed after 200252 3.6 Neither Mortality nor Fertility Declined during the 10 Years of Ghana’s HealthSWAp58 4.1 Multisectoral Approaches to HNP in Country Assistance Strategiesv

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION59 4.2 Two-Thirds of Multisectoral HNP Projects Involve HIV/AIDS59 4.3 Multisectoral HIV/AIDS Lending Accounts for All of the Increase inMultisectoral HNP Lending60 4.4 Multisectoral HNP Projects Are Unevenly Distributed across Regions62 4.5 Distribution of Multisectoral HNP Projects by the Number of Assigned andDemand-Driven Implementing Agencies63 4.6 Multisectoral Projects Had Lower Performance than Other HNP Projects64 4.7 Approval of HNP Components Managed by Other Sectors Has GrownSteadily Since 1988, though the Commitment per Project Is Small67 4.8 A Large Percentage of Water Supply and Sanitation Projects Invested inEnvironmental Improvements That Could Improve Health Outcomes67 4.9 The Share of Water Supply and Sanitation Projects with a HealthPerspective Has Declined70 4.10 The Majority of Health-Related Objectives in Transport Projects Are forRoad Safety71 4.11 The Share of Transport Projects with Health Components Has IncreasedSharply71 4.12 Projects with Explicit Health Objectives Are More Likely to Measure HealthOutcomes72 4.13 Reduction in Fatalities per 10,000 Vehicles in Closed Transport Projects78 5.1 Timeline of Health Sector-Related Events in IFC (1997–2007)83 5.2 IFC Development and Investment Outcomes in Two Periods84 5.3 IFC Evaluation Results Show Substantial Improvement in the Second PeriodTables8 1.1 Objectives and Strategic Directions of Healthy Development, the 2007 WorldBank HNP Strategy9 1.2 Business and Developmental Objectives of IFC’s 2002 Health Strategy16 2.1 Objectives of HNP Projects Approved in Fiscal 1997–200618 2.2 Key Developments in the HNP Portfolio24 2.3 More Project Appraisal Documents Have Baseline Data, but There Is StillSome Distance to Go, Fiscal Years 1997 and 200729 2.4 Few HNP Projects Have Objectives That Explicitly Mention the Poor31 2.5 The Poverty Focus of HNP Analytic Work Declined42 3.1 Outcomes and Bank Performance Are Lower for Health Reform Projects thanfor Other HNP Projects in Middle-Income Countries50 3.2 HNP Projects Supporting Health SWAps in IEG Country Cases or ProjectEvaluations61 4.1 Distribution of Multisectoral HNP Projects by Management andImplementation Arrangements80 5.1 IFC Health Investments by Type of Investment and Period81 5.2 Geographic Distribution of IFC Projects by Period82 5.3 Profitability of IFC Health and Pharmaceutical Investments85 5.4 IFC Advisory Services in Health (1997–2007)88 5.5 Percent of Health and IFC Projects Rated High, PCR Pilots 1 and 2vi

AbbreviationsAIDSCASCASCRDOTSESWHAMSETHIVHNPIBRDICRIDAIDIIEGIFCM&EMDGMISPADPCRPEPFARPERPPARPPPPRSCQAGSISSTDSWApTBTFRUNAIDSUNDPUSAIDWDRWHOXPSRAcquired immune deficiency syndromeCountry Assistance StrategyCAS Completion ReportDirectly observed therapy, short course (for TB)Economic and sector workHIV/AIDS, malaria, sexually transmitted diseases, and TBHuman immunodeficiency virusHealth, nutrition, and populationInternational Bank for Reconstruction and DevelopmentImplementation Completion ReportInternational Development AssociationInstitutional development impactIndependent Evaluation GroupInternational Finance CorporationMonitoring and evaluationMillennium Development GoalManagement information systemProject Appraisal DocumentProject Completion ReportPresident’s Emergency Plan for AIDS Relief (U.S.)Public expenditure reviewProject Performance Assessment ReportPublic-private partnershipPoverty reduction support creditQuality Assurance GroupIntegrated Health Insurance (Peru)Sexually transmitted diseaseSectorwide approachTuberculosisTotal fertility rateJoint United Nations Program on HIV/AIDSUnited Nations Development ProgramUnited States Agency for International DevelopmentWorld Development ReportWorld Health OrganizationExpanded Project Supervision Report (IFC)vii

A girl is weighed during a routine checkup at the Santa Rosa de Lima clinic in Nueva Esperanza,Honduras, which specializes in health care for children. Photo by Alfredo Srur,courtesy of the World Bank Photo Library.

AcknowledgmentsThis evaluation is a joint product of the World Bankand International Finance Corporation (IFC) departmentsof the Independent Evaluation Group(IEG). Martha Ainsworth (IEG-WB) was the taskleader and coordinator of the joint evaluation,drafted the chapters of the report evaluating theWorld Bank, and was responsible for integratingthe World Bank and IFC results. Hiroyuki Hatashima(IEG-IFC) led the evaluation of IFC. Theevaluation was conducted under the overall directionof Vinod Thomas, Cheryl Gray, MarvinTaylor-Dormond, and Monika Huppi. The evaluationreport is based on contributions from ateam of IEG staff and consultants who conducteddesk reviews, inventories, and interviews and datacollection in the field.The seven health, nutrition, and population (HNP)Project Performance Assessment Reports (PPARs)conducted for the evaluation of the World Bankwere led and prepared by staff members DeniseVaillancourt (Egypt, Ghana) and Gayle Martin(Eritrea), and consultants Judyth Twigg (KyrgyzRepublic, Russia), Mollie Fair (Peru), and PatriciaRamirez (Peru). The PPARs on water supply andsanitation (Nepal) and roads (Romania) withhealth objectives were conducted by staff membersKeith Pitman and Peter Freeman, respectively.The country case studies were preparedby R. Paul Shaw (Nepal), A. Edward Elmendorfand Flora Nankhuni (Malawi), and Alejandra Gonzalez(Egypt).In-depth desk reviews of various aspects of theHNP portfolio over a decade were conducted byMollie Fair, Judith Gaubatz, Manisha Modi, andManuela Villar Uribe. Lisa Overbey contributed adesk review of the water supply and sanitation portfolioand Peter Freeman and Kavita Mathur provideda review of the transport portfolio. Additionalspecific reviews were provided by Mollie Fair (HNPstrategy and timeline, population); Judith Gaubatz(Country Assistance Strategy review); Gayle Martin(communicable disease portfolio and inventoryof HNP analytic work, with contributions fromManisha Modi and Marie-Jeanne Ndiaye); ManishaModi (HNP portfolio analysis, staffing, and formalBank analytic work, with inputs from Helena Tangand Saubhik Deb); Flora Nankhuni (health ofthe poor and staffing analysis); Shampa Sinha(Country Assistance Strategy review); Joan Nelson(health reform); Manuela Villar Uribe (monitoringand evaluation, impact evaluation and pilots,HNP in public expenditure reviews and poverty assessments);and Denise Vaillancourt (sectorwideapproaches). IEG is particularly grateful to ChristineAquino and Samina Amin of the Human Resourcesdepartment for facilitating access to andhelping to interpret the data on staffing of theHNP sector.The evaluation of IFC’s operations in health wasconducted under the guidance of Stoyan Tenevand Amitava Banerjee. The IFC evaluation teamincluded Maria-Antonia Remenyi, Aygul Ozen,Chaoying Liu, Miguel Rebolledo Dellepiane,Nicholas Burke, and Vicky Viray-Mendoza.The report team extends its gratitude to the hundredsof individuals in government, civil society,the donor community, the World Bank and IFC,and the private sector who were interviewed forthe background papers, PPARs, country case studies,and IFC field visits. They are acknowledgedindividually in these respective reports. The teamalso extends its appreciation to World Bank HNPstaff who served as peer reviewers for the portfoliowork, and to those staff members who attendedseminars on the findings of intermediateproducts and provided valuable feedback.ix

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONRuth Levine and Susan Stout were the officialpeer reviewers; comments from the following individualson the approach paper and various intermediateoutputs were also greatly appreciated:Anabela Abreu, Olusoji Adeyi, Anders Agerskov,Julie Babinard, Cristian Baeza, Eni Bakallbashi,Arup Banerji, Alain Barbu, Nigel Bartlett, Chris Bennett,Alan Berg, Peter Berman, Anthony Bliss, EduardBos, Mukesh Chalwa, Laura Chioda, SafiaChowdhury, Mariam Claeson, Daniel Cotlear,Rooske De Joode, Jocelyne do Sacramento, SheilaDutta, Maria Luisa Escobar, Shahrokh Fardoust,Armin Fidler, Deon Filmer, Ariel Fiszbein, SundararajanGopalan, Pablo Gottret, DavidsonGwatkin, Keith Hansen, April Harding, EvaJarawan, Timothy Johnston, Marc Juhel, ChristianKammel, Arthur Karlin, Nicole Klingen, Pete Kolsky,Anjali Kumar, Christoph Kurowski, KeesKostermans, Jerry LaForgia, Rama Lakshminarayanan,Jack Langenbruner, Bruno Laporte,Patrick Leahy, Maureen Lewis, Benjamin Loevinsohn,Akiko Maeda, Tonia Marek, John May, JulieMcLaughlin, Tom Merrick, L. Richard Meyers,Michael Mills, Norbert Mugwagwa, Mary Mulusa,Homira Nassery, Son Nam Nguyen, Elaine Ooi, OkPannenborg, Ron Parker, Miyuki Parris, ChristophePrevost, Sandra Rosenhouse, Michal Rutkowski,Fadia Saadah, Pia Schneider, Julian Schweitzer,Meera Shekar, Agnes Soucat, Verdon Staines, KateTulenko, Vincent Turbat, John Underwood, AdamWagstaff, Howard White, and Abdo Yazbeck.The External Advisory Panel for this evaluationconsisted of Augusto Galan (Former Minister ofHealth, Colombia), Anne Mills (London School ofHygiene and Tropical Medicine, UK), GermanoMwabu (University of Nairobi, Kenya), and SuwitWibulpolprasert (Ministry of Health, Thailand).The evaluation team is enormously grateful fortheir sage advice and perspectives on the intermediateoutputs and final report. The responsibilityfor interpreting the results and using thisadvice rests with the evaluation team, however.William Hurlbut and Caroline McEuen editedthe report and Marie-Jeanne Ndiaye providedlogistical and production assistance throughoutthe preparation process. Heather Dittbrennerassisted with the photographs. Juicy ZareenQureishi-Huq provided administrative and productionsupport. Rose Gachina assisted in theprocessing of consultants. Nik Harvey set up theevaluation Web site (www.worldbank.org/ieg/hnp).Finally, we wish to acknowledge the generous financialsupport of the Norwegian Agency for DevelopmentCooperation (NORAD).Director-General, Evaluation: Vinod ThomasDirector, Independent Evaluation Group-World Bank: Cheryl GrayDirector, Independent Evaluation Group-IFC: Marvin Taylor-DormondManager, IEG Sector Evaluation: Monika HuppiTask Manager: Martha Ainsworthx

ForewordThe global aid architecture in health has changedover the past decade, with the adoption of the MillenniumDevelopment Goals (MDGs), expansionof the amount and sources of development finance,and new modes of international cooperationreflected in the 2005 Paris Declaration on AidEffectiveness. Over this period, key health andnutrition outcomes have improved in every developingregion, but many countries are off trackfor achieving the MDGs, and progress has beenwidely uneven among and within countries. Enormouschallenges remain in reducing morbidityand mortality from infectious disease, high and risingnoncommunicable disease, and pervasive malnutrition,as well as in lowering high fertility.IEG’s assessment of the efficacy of the World BankGroup’s support for health, nutrition, and population(HNP) to developing countries since 1997aims to help improve results in the context of thenew aid architecture. The evaluation focuses on theBank’s policy dialogue, analytic work, and lendingat the country level and the performance of IFC’shealth investments and Advisory Services. Theevaluation also draws lessons from Bank Group experiencein supporting several major approachesadopted by the international community—sectorwideapproaches (SWAps), communicable diseasecontrol, health reform, and multisectoral action forhealth outcomes.From 1997 to 2008 the World Bank provided $17billion in country-level project financing, in additionto policy advice and analytic work for health,nutrition, and population. Bank support has helpedbuild government capacity in the sector, and analyticwork has contributed to better performance.About two-thirds of the Bank’s HNP lending has hadsatisfactory outcomes, often in difficult environments.The record also tells us that the Bank needsto intensify efforts to improve substantially theperformance of the HNP portfolio. Factors that inhibitstronger performance include the increasingcomplexity of HNP operations, particularly in Africa,but also in health reform in middle-income countries;inadequate risk assessment and mitigation;and weak monitoring and evaluation. Over thesame period, IFC has made important strategicshifts, committing $873 million in private health andpharmaceutical investments. IFC performance inhealth investments, mainly in hospitals, improvedmarkedly from a weak start. The Corporation needsto invest more in activities that both make businesssense and are likely to yield broader benefits for thepoor. The accountability of Bank and IFC-financedHNP projects to ensure that results actually reachthe poor has been weak. Non-health sectors, suchas water supply and sanitation and transport, haveweak incentives to obtain and demonstrate healthbenefits from their investments.Though the Bank Group now funds a smallershare of the global financial support for HNP, it hasstepped up its worldwide involvement, includingits catalytic, lead, and support roles in many respects.The Group has a unique and substantialability to help improve HNP results and outcomesfor the poor. In delivering on this potential, theWorld Bank Group needs to strengthen its countrysupport, renew the commitment to resultsfor the poor, and give greater attention to reducinghigh fertility and malnutrition. Better resultsalso hinge on supporting capacity and bettergovernance for efficient health systems, enhancingthe contribution of other sectors to HNP outcomes,and boosting monitoring and evaluation.Vinod ThomasDirector-General, Evaluationxi

Through the hygiene education and good sanitation practices supported in the Bank’s Rural WaterSupply and Sanitation Project, this Nepalese woman believes her children are now healthier.Photo courtesy of George T. Keith Pitman.

Executive SummaryThe global aid architecture in health has changed over the past decade,with the adoption of the Millennium Development Goals (MDGs) anda major expansion of the levels and sources of development assistance,particularly for low-income countries. While key health outcomes such as infantsurvival and nutritional stunting have improved over the decade in everydeveloping region, nearly three-quarters of developing countries are eitheroff track or seriously off track for achieving the MDG for reducing under-fivemortality. Maternal mortality is declining at only 1 percent each year, a fifth ofthe rate needed to achieve the goal.There are important differences in progress acrosscountries, within regions, and within countries,with high levels of maternal mortality and malnutritionin Africa and South Asia. Enormous challengesremain in reducing morbidity and mortalityfrom infectious disease, the high and rising shareof noncommunicable disease, pervasive malnutrition,and high fertility, with its consequencesfor maternal mortality, maternal and child health,and poverty.The World Bank Group’s support for health, nutrition,and population (HNP) has been sustainedsince 1997—totaling $17 billion in country-levelsupport by the World Bank and $873 million in privatehealth and pharmaceutical investments by theInternational Finance Corporation (IFC). Beyondcountry-level support, the World Bank Groupparticipates in nearly three dozen internationalpartnerships in HNP, with indirect benefits tocountries. This report evaluates the efficacy ofthe Bank Group’s direct support for HNP to developingcountries since 1997 and draws lessonsto help improve the effectiveness of this supportin the context of the new aid architecture.Though the Bank Group now funds a smallershare of global HNP support than it did a decadeago, it still has substantial ability to add value if itcan do a better job of helping countries deliver results,especially for the poor. World Bank supportto countries in the form of lending, analyticwork, and policy dialogue has helped build governmentcapacity to manage the sector—criticallyimportant to improving aid effectiveness,given the increasing reliance of other donors ongovernment systems.About two-thirds of the World Bank’s HNP lendinghas had satisfactory outcomes, often in difficultenvironments, but a third has not performedwell. Contributing factors have been the increasingcomplexity of HNP operations, particularlyin Africa but also in health-reform support tomiddle-income countries; inadequate risk assessmentand mitigation; and weak monitoringand evaluation.The performance of IFC’s health investments hasimproved markedly from a weak start, but thereremain important gaps in investing in activities thatboth make business sense and are likely to yieldbroader benefits for the poor. Accountability forresults in these projects has been weak—the accountabilityof Bank and IFC-financed projectsto ensure that results actually reach the poor,xiii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONand the accountability of projects in the Bank’snon-health sectors, such as water supply and sanitationand transport, for demonstrating theirhealth benefits.For the Bank Group to achieve its objectives ofimproving health sector performance and HNPoutcomes among the poor, it needs to act in fiveareas: intensify efforts to improve the performanceof the portfolio; renew the commitment to deliveringresults for the poor, including greater attentionto reducing high fertility and malnutrition;build its own capacity to help countries to makehealth systems more efficient; enhance the contributionof other sectors to HNP outcomes; andboost evaluation to implement the results agendaand improve governance. By doing this, the BankGroup will contribute not only to meeting theMDGs but also to ensuring that the poor benefit,and that those benefits are sustained.Since the late 1990s, when the World Bank Groupwas the largest source of HNP finance to developingcountries, new aid donors and institutionshave emerged, and development assistance forHNP has more than doubled, from an annual averageof $6.7 billion in 1997/98 to about $16 billionin 2006. The international community hasadopted global development targets, the mostprominent of which are the MDGs, with a newemphasis on aid effectiveness, results orientation,donor harmonization, alignment, and countryleadership, reflected in the 2005 Paris Declarationon Aid Effectiveness and the 2008 AccraAgenda for Action.The World Bank Group, now one of many largeplayers in international HNP support, accounts foronly about 6 percent of the total—down from 18percent in the 1990s—and is reassessing its comparativeadvantage in the context of the new aidarchitecture. At the same time, a call for greaterengagement with the private health sector in developingcountries presents new opportunities forIFC to extend its support.The decline in its relative contribution aside, theWorld Bank Group commitment to HNP is stillsubstantial. Since 1997, the World Bank (InternationalBank for Reconstruction and Developmentand the International Development Association)has committed nearly $17 billion to 605HNP projects in more than 120 countries, sponsoredanalytic work, and offered policy advice. Thissupport aimed to improve health and nutrition statusand reduce high fertility; improve the access,quality, efficiency, and equity of the health system;reform health systems through changes in healthfinance, support for health insurance, decentralization,engaging the private sector, and otherstructural changes; and strengthen institutionalcapacity and sector management. In addition, theBank has greatly expanded its participation inglobal partnerships for health; as of 2007, it wasparticipating financially in 19 global partnershipsand engaged in others ways in 15 more.IFC has financed 68 private investment projectsin the health and pharmaceutical sectors of developingcountries—amounting to $873 millionin total commitments—and offered advisory serviceson health to the private sector, including supportfor public-private partnerships.The World Bank’s 2007 strategy, Healthy Development:The World Bank Strategy for Health,Nutrition, and Population Results, aims, amongother things, to improve HNP outcomes on averageand among the poor; prevent poverty dueto illness; improve health system performance;and enhance governance, accountability, andtransparency in the sector. It points to severalstrategic directions or actions for the Bank toachieve the objectives, including:• A renewed focus on HNP results• Efforts to help countries improve the performanceof health systems and to ensuresynergy with priority disease interventions,particularly in low-income countries• Strengthened Bank capacity to advise countrieson intersectoral approaches to improvingHNP results.The 2002 IFC health strategy defines the sector’sgoals as improving health outcomes, protectingthe population from the impoverishing effectsof ill health, and enhancing the performance ofxiv

EXECUTIVE SUMMARYhealth services. The strategy has both business anddevelopmental objectives, among them promotingefficiency and innovation in the health sector.It also calls broadly for increasing the social impactof IFC investments.The Scope of the EvaluationThis evaluation aims to inform the implementationof the most recent World Bank and IFC HNPstrategies to enhance the effectiveness of futuresupport. It covers the period since fiscal year 1997and is based on desk reviews of the portfolio,background studies, and field visits. The evaluationof the HNP support of the World Bank focuses onthe effectiveness of policy dialogue, analytic work,and lending at the country level, while that ofIFC focuses on the performance of health investmentsand advisory services before and after its2002 health strategy. The themes it covers aredrawn from the two strategies and the approachesadopted by international donors in the pastdecade. IEG has previously evaluated several aspectsof the Bank’s HNP support. IFC’s support forthe health sector has never been fully evaluated.Bank Support to the Public Sectorfor HNPOver the past decade, the World Bank directly supportedHNP outcomes in countries through lendingand nonlending services. The largest sourceof lending was from projects managed within theBank’s HNP sector units ($11.5 billion, 255 projects).Almost all HNP-managed projects were investmentlending. Beyond this, about $5 billionin lending for HNP outcomes was managed byother sectors. In nonlending services, since fiscal2000 the Bank has spent $43 million of its ownbudget and trust funds on HNP-related economicand sector work (ESW). The number of professionalHNP staff grew by a quarter, as did theshare of health specialists among HNP staff.The World Bank’s RoleAlthough the World Bank finances a smallershare of country-level development assistanceunder the new international aid architecture,it still has significant potentialto add value. But the value of that support iscontext-specific and depends on the Bank’sability to help countries deliver results. TheWorld Bank brings important institutional assetsto bear in helping countries make health systemswork better and ensuring that health benefitsreach the poor: long-term, sustained engagementin the sector; international experience; a historyof support for building country capacity toimplement programs; large-scale, sustained financing;strong links to finance ministries; and engagementwith many sectors other than healthwith potential to contribute to HNP outcomes.The Bank’s comparative advantage in a countryis context-specific, depending on health conditions,government priorities and resources, andthe activities of other development partners. Todeliver on its comparative advantages, the Bankneeds to improve the performance of its countrylevelsupport.The Evolution and Performance ofWorld Bank SupportWhile the overall level of HNP project approvalschanged little, the composition ofthe lending portfolio saw some major shifts.The number of HNP-managed projects approvedannually rose slowly, but new commitments declined.The share of communicable disease projectsdoubled over the decade, reaching about40 percent of approvals in the second half of theperiod, as did the share of multisectoral projects,reaching half of all approvals. The share of AfricaRegion projects in the HNP lending portfolio alsoincreased. These three trends were due primarilyto an increase in multisectoral projects addressingacquired immune deficiency syndrome(AIDS). Projects supporting sectorwide approaches(SWAps) in health rose to a cumulative total of 28operations in 22 countries, about 13 percent of theproject portfolio. In contrast, the share of lendingwith objectives to reform the health systemdropped by nearly half.Attention to population and malnutritionwas low, and support for population nearlydisappeared. About 1 in 10 projects had an objectiveto reduce malnutrition, which disproportionatelyaffects the poor, but the share of projectswith nutrition objectives dropped by half overthe decade. About two-thirds of nutrition projectsxv

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONwere in countries with high levels of child stunting,but Bank nutrition support reached only about aquarter of all developing countries with high stunting.Lending to reduce high fertility or improve accessto family planning accounted for only 4 percentof the lending portfolio, dropping by two-thirds betweenthe first and second half of the decade, at atime when the need for such support remainedhigh. Population support was directed to onlyabout a quarter of the 35 countries the Bank identifiedas having the highest fertility (with rates ofmore than five children per woman). Analytic workand staffing to support population and family planningobjectives nearly disappeared. Substantialanalysis of population and nutrition issues rarelyfigured in poverty assessments, even though bothissues are most acutely felt by the poor.Two-thirds of HNP projects have had satisfactoryoutcomes, and the portfolio’s performancehas stalled. The evaluation highlightsexamples of good performance based on fieldassessments. Support for reducing malaria inEritrea and schistosomiasis in the Arab Republicof Egypt, raising contraceptive use in pilot areasof Malawi, and reforming the health system inthe Kyrgyz Republic, for example, showed goodresults. However, roughly a third of the HNP lendingportfolio did not perform well, a share that hasremained steady, while performance in other sectorshas improved. The performance of HNP supportin Africa was particularly weak, with only onein four projects achieving satisfactory outcomes.Complex projects—multisectoral projects andSWAps—in low-capacity environments were leastlikely to achieve their HNP objectives. However,health reform projects in middle-income countriesalso performed less well and are complex andpolitically volatile.Poor-performing projects displayed common characteristics:inadequate risk analysis or technicaldesign, inadequate supervision, insufficient politicalor institutional analysis, lack of baseline dataon which to set realistic targets, overly complexdesigns in relation to local capacity, and negligiblemonitoring and evaluation. These problemsare similar to those cited in IEG’s 1999 evaluationof the HNP sector. The results of the recent DetailedImplementation Review of HNP projectsin India suggest that, even among projects thatachieve their objectives, field supervision needsto be intensified to ensure that civil works andequipment are delivered as specified, in workingorder, and functioning.Accountability of projects for deliveringhealth results to the poor has been weak. Improvinghealth outcomes among the poor isamong the foremost objectives of the 2007 HNPstrategy. Studies of the incidence of public expenditurehave shown that in most countries,public health spending favors the non-poor; mereexpansion of services cannot be assumed to improveaccess of the poor relative to the non-poor.While many projects targeted HNP support togeographic areas with a high incidence of poverty(including rural areas) or financed services or addressedproblems thought to disproportionatelyaffect the poor, only 6 percent of all HNP projectscommitted to deliver better health or nutritionamong the poor in their statement of objectives,for which they were ultimately accountable. Athird of projects with objectives to improve generalhealth status (such as maternal and childhealth) had no targeting mechanism for reachingthe poor. Among closed projects with objectivesto improve HNP outcomes among the poor, mostmeasured a change in average HNP status in projectareas. Very few actually measured whetherthe poor (individuals or poor project areas) havebenefited in relation to the non-poor or in relationto those in areas not reached by the project,and even fewer demonstrated that the poor haddisproportionately benefited. In some cases, improvementsin HNP status were only measured atthe national level.The Bank delivered several high-profile analyticproducts on HNP and poverty over the pastdecade—notably the Reaching the Poor withHealth, Nutrition, and Population Services projectand the World Development Report 2004:Making Services Work for Poor People. Nevertheless,the share of country poverty assessmentswith substantial discussion of health has declined,from 80 percent in fiscal years 2000–03 to only58 percent in 2004–06. Only 7 percent of povertyxvi

EXECUTIVE SUMMARYassessments had substantial discussion of population,and major discussion of nutrition declinedby more than half, from 28 to 12 percent. Abouta quarter to a third of Bank HNP analytic work—both ESW and research—was poverty-related,and this share has also declined in the decadesince 1997.Some aspects of monitoring have improved,but overall it remains weak, and evaluationis almost nonexistent, presenting a challengefor the HNP strategy’s results orientationand commitment to better governance. Since1997, an increasing number of projects have hadmonitoring indicators and baseline data whenthey were appraised. Yet, although nearly a thirdof projects supported pilot interventions or programs,or intended to evaluate the impact of a specificactivity or program, few proposed evaluationdesigns in appraisal documents, and even fewerevaluations were actually conducted. Pilot projectsor components without an evaluation design describedin the appraisal document were neverevaluated. Among the consequences of poor monitoringand evaluation and absence of baselinedata were irrelevant objectives and inappropriateproject designs, unrealistic targets—either toohigh or below the baseline value—inability to assessthe effectiveness of activities, and lower efficacyand efficiency because of limited opportunitiesfor learning. These findings are of great concerngiven the emphasis of the 2007 HNP strategy onresults and good governance.Approaches for Improving HNP OutcomesThe evaluation reviewed findings and lessons forthree prominent approaches to raising HNP outcomesover the past decade—communicabledisease control, health reform, and sectorwide approaches(SWAps). These approaches have beensupported by the Bank as well as the internationalcommunity and are not mutually exclusive. SWAps,for example, have included communicable diseasecontrol and health reform elements.Support for communicable disease controlcan improve the pro-poor focus of health systems,but excessive earmarking of foreignaid for communicable diseases can distortallocations and reduce capacity in the restof the health system. One of the strategic directionsof the 2007 HNP strategy is to ensure synergybetween priority disease interventions andstrengthening the health system. The rationale forinvesting in infectious diseases is that they disproportionatelyaffect the poor; their control haslarge, positive externalities; and control interventionshave been shown to be cost-effective inmany settings. Dedicated communicable diseaseprojects have dramatically increased as a share ofthe overall portfolio over the past decade, andBank support has directly built country capacityin national disease control programs. Supportfor communicable disease control, with the exceptionof AIDS projects, has shown better outcomesin relation to objectives than the rest of theHNP portfolio. Both equity and cost-effectivenessare particularly important to address in HIV/AIDSprograms, given the huge commitments to thatdisease and that, unlike tuberculosis and malaria,HIV does not always disproportionately strikethe poor. Care must be taken that, as the Bank enhancesits support to systemwide reforms andSWAps, progress on communicable disease controlremains a priority.Since the initial increase in Bank-supported communicabledisease control in the early 2000s,mainly for AIDS, the international communityhas also generously expanded funding through theGlobal Fund to Fight AIDS, TB, and Malaria andthe (U.S.) President’s Emergency Plan for AIDS Relief(PEPFAR), other bilateral contributions, andprivate foundations. In some low-income countrieswith high HIV prevalence, earmarked AIDSfunds from international partners account for30–40 percent or more of all public health funding.In an environment of scarce human resourcecapacity within the health system, care must betaken to balance the allocation of resources acrosshealth programs and budget lines, to ensure thatlarge earmarked funds for specific diseases donot result in lower efficiencies or reduced careelsewhere in the health system. There is little evidencein recently approved Bank support forHIV/AIDS or the other high-priority diseases thatthis issue has been considered in funding decisionsor in risk analysis.xvii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONHealth reforms promise to improve efficiencyand governance, but they are politically contentious,often complex, and relatively risky.Health system reform is central to the emphasisof the 2007 HNP strategy on strengthening healthsystems. About a third of HNP projects have supportedreform or restructuring of the health systemthrough changes in health finance, developmentof health insurance, decentralization ofhealth systems, and regulation or engagementof the private health sector. These objectives affectefficiency and governance, which are valid objectivesin their own right, even if they often donot directly affect health status in the short run.Reforms affecting health insurance help to preventthe impoverishing impacts of illness. Bank supportfor health system reforms has been mainly tomiddle-income countries, where health reformprojects represent about half of the portfolio.Many lessons have been learned over the pastdecade about the successes and pitfalls of supportfor health reform:• First, the failure to assess fully the politicaleconomy of reform and to prepare a proactiveplan to address it can considerably diminishprospects for success. Political risks, the interestsof key stakeholders, and the risk of complexity—issuesthe evaluation case studiesfound to be critical—are often neglected inrisk analysis in project appraisal documentsfor health reform projects.• Second, reforms based on careful prior analyticwork hold a greater chance of success, but analyticwork does not ensure success.• Third, the sequencing of reforms can improvepolitical feasibility, reduce complexity, ensurethat adequate capacity is in place, and facilitatelearning. When implementation is flagging, theBank can help preserve reform momentumwith complementary programmatic lendingthrough the Ministry of Finance, as it did in Peruand the Kyrgyz Republic.• Finally, monitoring and evaluation are criticalin health reform projects—to demonstrate theimpact of pilot reforms to garner political support,but also because many reforms cannotwork without a well-functioning managementinformation system.SWAps have contributed to greater governmentleadership, capacity, coordination,and harmonization within the health sector,but not necessarily to improved efficiencyor better health results. Sectorwide approaches(SWAps) represent a reform in the way that governmentand international donors work together(the approach) to support the achievement ofnational health objectives (the program). Theysupport the 2007 HNP strategy’s objective toimprove the organization, functioning, and sustainabilityof health systems. The approach promotesconsensus around a common nationalstrategy, country leadership, better harmonizationand alignment of partners based on theircomparative advantages, joint monitoring, thedevelopment and use of country systems, and, inmany cases, the pooling of donor and governmentfunds. The anticipated benefits include greatercountry sectoral leadership and capacity in managinghealth support, improved coordination andoversight of the inputs of all partners, reducedtransaction costs, more efficient use of developmentassistance, more reliable support for thehealth sector, and greater sustainability of healthprograms.The overwhelming focus of SWAps supported bythe Bank has been on setting up and implementingthe approach. Fieldwork found thatcountry capacity has been strengthened in theareas of sector planning, budgeting, and fiduciarysystems. However, weaknesses persist in the designand use of country monitoring and evaluationsystems; evidence that the approach hasimproved efficiency or lowered transaction costsis thin, because neither has been monitored. Experiencehas shown that adopting the approachdoes not necessarily lead to better implementationor efficacy of the government’s health programs:only a third of Bank projects that supportedhealth SWAps have performed satisfactorily onmeeting their health objectives. SWAps have oftensupported highly ambitious programs, involvingmany complex reforms and activities that exceedxviii

EXECUTIVE SUMMARYgovernment implementation capacity. An importantlesson is that programs need to be realisticand prioritized and that the process of settingup the SWAp should take care not to distract theplayers from ensuring the implementation and efficacyof the overall health program and a focuson results. SWAps have been most effective inpursuing health program objectives when thegovernment is in a leadership position with astrongly owned and prioritized strategy (as in theKyrgyz Republic). When this is not the case, thereis a risk that the health program implemented willbe less prioritized, reflecting the favored elementsof the diverse partners, weakening effectiveness(as in Ghana).The contribution of other sectors to HNPoutcomes has been largely undocumented;the benefits of intersectoral coordinationand multisectoral approaches need to bebalanced with their costs in terms of increasedcomplexity. Achieving the health MDGswill require complementary actions from sectorsother than health, an explicit activity proposedwithin the 2007 HNP strategy. The contribution ofother sectors to HNP outcomes has been capturedthrough multisectoral HNP projects (projectsthat engage multiple sectors in a singleoperation with an objective to improve HNP outcomes)and parallel lending in projects managedby other sectors, in some cases with explicit healthobjectives. Multisectoral HNP operations haverisen from a quarter of all HNP lending to half,greatly increasing the complexity of the portfolio.Most of the increase stems from multisectoralAIDS projects. The large number of sectors involved,the lack of specificity in design documentsabout the roles and responsibilities of eachparticipating sector, the relatively new institutionsput in charge, and other factors affectinglower performance in Africa all contribute tolower outcomes for multisectoral AIDS projects.Other multisectoral HNP projects with fewer implementingagencies have maintained stronger intersectoralcollaboration and better outcomes.Since 1997, the Bank has invested about $5 billionin smaller HNP components in 350 projects managedby other sectors, such as social protection,education, public sector management, water supply,and transport. Both the 2007 HNP strategy andits predecessor foresaw Country Assistance Strategiesas the instrument for coordinating intersectoralaction to improve HNP outcomes. However,this has not occured over the past decade. Lendingactivities in diverse sectors such as watersupply and sanitation and education have beenpursued—for the most part—independently ofeach other and of HNP operations, although thisdoes not mean that they have not contributed tohealth outcomes.Lending programs in other sectors may contributedirectly or indirectly to HNP outcomes, in somecases by including health objectives or healthcomponents in projects. For example, half of allwater supply and sanitation projects claim thathealth benefits will be generated, and 1 in 10 hasan objective to improve health outcomes. Butfewer water supply and sanitation projects includehealth objectives today than was the case 5–10years ago; in fiscal 2002–06, only 1 in 20 water supplyand sanitation projects had an objective toimprove health for which they were accountable.Interviews with water supply and sanitation staffsuggested that the sector has focused primarilyon what is perceived to be “their” MDG, namelyincreased access to safe water. Yet research hasshown that context matters; better access to safewater does not necessarily translate into betterhealth. In contrast, the health content of transportprojects has greatly increased, particularly in thefield of road safety and HIV/AIDS prevention.While trends in accident statistics are relativelywell documented for road safety components,there is very little in the way of documented outputsor results for HIV/AIDS components.Water supply and sanitation and transport projectswith health components or objectives rarelyinvolved collaboration with Ministries of Healthor the Bank’s HNP sector (for example, the RuralWater Supply and Sanitation Project in Nepal). Deliveryof health results in these and other sectorshas been generally weak, except when an explicithealth objective was identified at project appraisal.xix

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONThere were virtually no results reported for healthactivities that were retrofitted into active projects.IFC Support for Development of thePrivate Health SectorAbout three-quarters of health expenditures inlow-income countries and half in middle-incomecountries are private, and half of private healthspending among the poor is for pharmaceuticals.IFC has made support to private investment inhealth one of its strategic priorities. Health is a relativelysmall and recent sector of IFC operationsand involves the activities of two departments:Health and Education and General Manufacturing(for pharmaceuticals).The performance of IFC’s health investments,mostly hospitals, has substantially improved,following a learning process. Before1999, four-fifths of all health investments performedpoorly, and a majority of failed project businessescontributed to financial losses. The reasonsfor failure included the impact of financial crisesin certain regions, delays in obtaining regulatoryclearances from the authorities, and IFC’s weaknessesin screening and structuring health sectordeals owing to lack of sector-related experience.These experiences provided important lessonsabout hospital investments. More recent investmentshave realized good financial returns andperformed better in achieving intended developmentoutcomes. An evaluative framework forIFC’s Advisory Services was only recently launched,so very few health projects have been evaluated,and their results should not be used to infer theperformance of the whole portfolio. However,the few health Advisory Services projects thathave been evaluated have performed lower thanthe IFC portfolio overall.IFC has not been able to diversify its healthportfolio as quickly as anticipated. In 2002,the sector set objectives to diversify the portfoliobeyond hospitals and to improve the social impactof IFC health operations. IFC has continuedto finance private hospitals; the share of pharmaceuticalsand other life sciences investmentshas grown, though more slowly than envisionedin the strategy. IFC has also financed public-privatepartnerships in health and expanded health AdvisoryServices with a focus on Africa. Investmentnumbers and volume increased from 2005 onward.However, to date IFC has not succeeded infinancing any health insurance ventures and hasfinanced only one project in medical education.IFC’s health interventions have had limitedsocial impact, although efforts to broadenthose impacts are increasing. IFC’s investmentsin hospitals have targeted middle- and upperincomegroups. Linkages to public insuranceschemes will be necessary for IFC-supported hospitalsto meet the health needs of a wider population.Expanded support to public-private partnerships,jointly with the World Bank, such as a recentoutput-based aid project to improve maternal careamong some of Yemen’s poorest people, andmore strategic deployment of Advisory Services,such as recent efforts to assist social enterprisesin Kenya and India, could lead to broadening ofthe social impact of investments in the health sector.These investments are too recent to evaluate.Recent IFC health projects have had somepositive results for efficiency, governance,and affordability. State-of-the-art facilities insome IFC-supported projects have attracted professionalswith established, successful careers indeveloped countries. Many hospitals supportedby IFC have posted fees and introduced controlof doctors’ side practices outside of the institutions.The majority of IFC-supported pharmaceuticalprojects have resulted in significantdeclines in the prices of generic drugs, thus enhancingaffordability.The need to collaborate closely with theWorld Bank’s HNP sector is recognized asimportant in both the IFC and World Bankstrategies to promote greater efficiency in thehealth sector through finance of privatehealth care. The evaluation found some WorldBank–IFC interaction, particularly in middleincomecountries, but there is no real model ofhow that collaboration should occur in a situationwhere IFC health activities are few and very smallin relation to the entire World Bank Group HNPsector in a given country.xx

EXECUTIVE SUMMARYRecommendationsThe following recommendations for the WorldBank and IFC are offered to help improve the implementationof their respective HNP strategiesand further the mandate to reduce poverty andpromote economic growth in the context of thenew aid architecture.1. Intensify efforts to improve the performanceof the World Bank’s support forhealth, nutrition, and population.• Match project design to country capacityand reduce the complexity of support inlow-capacity settings, particularly in Africa.• Thoroughly and carefully assess the risksof proposed HNP support and strategiesfor mitigating those risks, particularly the politicalrisks and incentives of stakeholders.• Phase health system reforms to maximizethe probability of success.• Undertake thorough institutional analysisas an input into more realistic project design.• Support intensified supervision in the fieldby the Bank and the borrower to ensurethat civil works, equipment, and other outputshave been delivered as specified, arefunctioning, and are being maintained.2. Renew the commitment to health, nutrition,and population outcomes amongthe poor.The World Bank should:• Boost population, family planning andother support to reduce high fertility.• Incorporate the poverty dimension intoproject objectives.• Increase support to reduce malnutritionamong the poor, whether from the HNPsector or other sectors.• Monitor health, nutrition, and populationoutcomes among the poor.• Bring the health and nutrition of the poorand the links between high fertility, poorhealth, and poverty back into povertyassessments.IFC should:• Expand support for innovative approachesand viable business models that demonstrateprivate sector solutions to improve the healthof the poor, including expansion of investmentsin low-cost generic drugs and technologiesthat address problems of the poor.• Assess the external and internal constraints inachieving broad social impacts in the sector.3. Strengthen the World Bank Group’s abilityto help countries to improve the efficiencyof health systems.The World Bank should:• Better define the efficiency objectives of itssupport and how efficiency will be improvedand monitored.• Carefully assess decisions to finance additionalfreestanding communicable diseaseprograms in countries where other donorsare contributing large amounts of earmarkeddisease funding and additional earmarkedfunding may contribute to distortions inthe health system.• Support improved health information systemsand more frequent and vigorous evaluationof reforms.IFC should:• Support public-private partnerships throughAdvisory Services to government and industryand through its investments, and expandinvestments in health insurance.• Improve collaboration and joint sector workwith the World Bank, leveraging Bank sectordialogue on health regulatory frameworksto engage new private actors, andmore systematically coordinate with theBank’s policy interventions regarding privatesector participation in health.4. Enhance the contribution of supportfrom other sectors to health, nutrition,and population outcomes.xxi

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONThe World Bank should:• When the benefits are potentially great in relationto the marginal costs, incorporatehealth objectives into relevant non-healthprojects for which they are accountable.• Improve the complementarity of investmentoperations in health and other sectors toachieve health, nutrition, and populationoutcomes, particularly between health andwater supply and sanitation.• Prioritize sectoral participation in multisectoralHNP projects to reduce complexity.• Identify new incentives for Bank staff towork across sectors to improve health, nutrition,and population outcomes.• Develop mechanisms to ensure that the implementationand results for small HNP componentsretrofitted into ongoing projectsare properly documented and evaluated.IFC should:• Improve incentives and institutional mechanismsfor an integrated approach to healthissues across units in IFC dealing with health,including the way that health in IFC isorganized.5. Implement the results agenda and improvegovernance by boosting investmentin and incentives for evaluation.The World Bank should:• Create new incentives for monitoring andevaluation for both the Bank and the borrowerlinked to the project approval processand the midterm review. This includes requirementsfor baseline data, evaluation designsfor pilot activities in project appraisaldocuments, and periodic evaluation of mainproject activities as a management tool.IFC should:• Enhance its results orientation by developingclearly specified baseline indicators andan evaluation framework that adequatelymeasures IFC’s health sector objectives andresults.xxii

Management ResponseSummaryWorld Bank Group management welcomes IEG’s evaluation of WorldBank Group work in the health, nutrition, and population sectorafter 10 years of implementation of the 1997 health, nutrition, andpopulation (HNP) strategy.As a global development institution dedicated tosupporting country and global efforts to achievethe better health outcomes that are central to reducingpoverty and to achieve the 2015 MillenniumDevelopment Goals (MDGs), managementappreciates the findings of this evaluation andother internal and external efforts to improvethe World Bank Group’s effectiveness in this keysector. (For the full-length Management Responseand the detailed World Bank Management ActionPlan, please see appendix J.)This IEG report is consistent with the findingsof the World Bank’s own self-assessment, whichprompted the Bank to adopt a new HNP strategyin mid-2007—Healthy Development: The WorldBank Strategy for Health, Nutrition, and PopulationResults. In fact, IEG staff shared the preliminaryresults of their work with the strategyteam that designed the subsequent 2007 strategy.The new Strategy sharpened the Bank’s focus onresults on the ground; concentrated Bank contributionson health systems strengthening, healthfinancing, and economics; supported governmentleadership and international community programsto achieve these results; and focused on enhancedengagement with global partners. Early implementationof the strategy has been promising.An example is the success of Rwanda’s resultsbasedapproach to improving service delivery,with malaria incidence declining by 62 percent andchild mortality decreasing by 30 percent.Main Findings and RecommendationsThe IEG evaluation covers a decade of Bank support,starting in 1997. The specific performanceto date of the Bank’s new strategy, adopted in mid-2007, is discussed in a separate Strategy ProgressReport (World Bank 2009). Bank managementhas taken many of the findings and recommendationsinto account in the Progress Report andthe Management Action Plan. While not detractingfrom the importance of the evaluation and itsusefulness for the Bank’s future work in health,nutrition, and population, management has observationson some of the findings.Quality of the Bank HNP Portfolio—Theprincipal finding of the IEG evaluation is that,while two-thirds of the Bank’s HNP projects in theperiod 1997–2007 achieved their developmentobjectives, one third, mostly in African countries,did not, clearly warranting close scrutiny. Currentdata on risky projects and programs in theHNP portfolio show that problems continue to bemost acute in HIV/AIDS projects and in programsin the Africa Region. The HNP sector’s performancein much of the rest of the world is nownear the performance of other sectors, and theAfrica Region has developed new strategic approachesto improving portfolio quality, includingxxiii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONa focus on results and emphasis on health systemsstrengthening.IEG notes the underperformance of HIV/AIDSprojects during the height of the epidemic insouthern Africa during the late 1990s, a periodcharacterized by regional conflict and instability,and the internal displacement of millions of refugeesacross borders in the Great Lakes Region ofthe continent. Current projects, focusing on highriskgroups, also constitute a disproportionatenumber of projects at risk in the current portfolio.In order to redress this situation, a comprehensiveapproach has been adopted to improvethe quality of the Bank’s HIV/AIDS operations, includingan umbrella restructuring of the Multi-Country HIV/AIDS Programs (MAPs).Investing in Health Systems—The 2007 strategyunderscores the need to focus on healthsystem strengthening to ensure better health outcomes,particularly for the poorest and the mostvulnerable. Over the past two years, projects witha primary focus on health systems have increasedtwofold. In line with the strategy, 67 percent of Bankprograms approved since fiscal year 2007 that focusedon priority disease areas also include strongcomponents on health system strengthening. AnAfrica-focused initiative started in 2008 will improvethe Bank’s capacity to provide rapid adviceand assistance on the ground, particularly in healthfinance, human resources, governance, supplychain management, and infrastructure planning.Doing More in Population and ReproductiveHealth—Bank management agrees withIEG that it should do more in this key area. TheProgress Report highlights plans for strengtheningsupport for population and reproductivehealth, using a health systems approach that is criticalto improving maternal and child survival rates.Investing in Nutrition Support—We also agreewith IEG’s findings that the Bank needs to focusmore on nutrition. The need for action is evenmore important today in the context of the ongoingcrises in fuel, food, and fertilizers, as wellas the escalating effects of the financial crisis. Weare therefore investing significant resources inthe next few years to ramp up the Bank’s analyticaland investment work and leverage resourcesfrom other donors. The agenda for scaling upnutrition is being catalyzed with additional budgetand external resources, starting in 2009 and continuingfor three years.Improving Monitoring and Evaluation(M&E)—As noted in the Strategy Progress Report,this is an important part of strategy implementation,(including the work on retrofitting projectsand improving the design of new projects).Routine health monitoring systems (includingsurveillance, facility reporting, vital registration,census data, resource tracking, and householdsurveys) may first need to be strengthened toprovide the data and indicators that are needed.The Bank is working with partners, such as theWorld Health Organization, to develop betterways to monitor the health MDGs, includingthe estimation of trends in child and maternalmortality, for which updates have recently beenissued.The International Financial Corporation’sHNP Footprint. IFC has considerably increasedits footprint in HNP over the past decade and isprepared to intensify collaboration within theWorld Bank Group. There is a growing acknowledgementof the role of the private sector inhealth care in developing countries. The periodunder review has seen a marked increase in IFC’sactivity in health. During this time, many lessonshave been learned, specialist knowledge has deepened,and performance has improved by any measureapplied. As in other sectors, IFC continuesto strive for greater development impact, and wetherefore welcome all input that could help us todo better.xxiv

MANAGEMENT RESPONSE SUMMARYManagement Action RecordRecommendationsManagement response1. Intensify efforts to improve the performance of the World Bank’s health, nutrition, and population support.(a) Match project design to country context and capacity andreduce the complexity of projects in low-capacity settingsthrough greater selectivity, prioritization, and sequencingof activities, particularly in Sub-Saharan Africa.Management agrees that complexity can be at least partially addressed by adopting IEG recommendations,such as thorough technical preparation, including solid analytical underpinning, politicalmapping, high quality at entry, including prioritizing interventions relative to the institutionalcontext, and establishing a good results framework, followed by in-depth supervision and parallelpolicy dialogue with client and partners.However, HNP operations are rarely institutionally or technically simple, since the desired outcomedepends on a complex and interacting set of social, cultural, and institutional factors. This isrecognized by donors and policy makers, whether in low-income, middle-income, or high-incomecountries. The inherent complexity of the sector may be attributable to the political economy ina multi-stakeholder environment, the need for extensive coordination and partnership with nationaland international agencies and civil society organizations, and the often difficult technicaland social nature of the subject. The recent international recognition of the need to invest in(complex) health systems in order to ensure the success of vertical disease control programs inlow-income countriess is testimony to the fact that there are few easy ways to avoid systemiccomplexity. Investing in simple programs would not necessarily provide for lasting impact.(b) Thoroughly and carefully assess the risks of proposed HNPsupport and strategies to mitigate them, particularly thepolitical risks and the interests of different stakeholders,and how they will be addressed.Management agrees in principle to carry out political mapping exercises prior to investments inthe sector where appropriate. As there are currently mandatory risk assessment and mitigationsteps built into the project cycle and approval process, we anticipate improved risk mitigationstrategies in newer HNP operations. These enable staff to identify major political and technicalrisks and devise with the client suitable risk-mitigation strategies. However, it is also evident thatdespite good assessments and risk-mitigation strategies, neither technical nor political risks canbe completely offset.(c)Phase reforms to maximize the probability of success.Management agrees with this recommendation and would note that more projects are now takingthis approach.(d) Undertake thorough institutional analysis, including anassessment of alternatives, as an input into more realisticproject design.(e) Support intensified supervision in the field by the Bankand the borrower to ensure that civil works, equipment,and other outputs have been delivered as specified, arefunctioning, and are being maintained.Management agrees with this recommendation.Management agrees in principle with this recommendation. Supervision requirements, both interms of staffing mix and budgeting, are being reassessed Bank-wide within the context of theongoing review of investment lending. Given the inherent dispersed nature of many HNP investments,care must be taken during project design to ensure that the client assumes responsibilityfor ensuring that civil works, equipment, and other outputs have been delivered asspecified, are functioning and being maintained, while the Bank audits/appraises/confirms thatsuch monitoring is taking place so that detailed supervision of projects can be properly conductedwithin likely budget norms.2. Renew the commitment to health, nutrition, and population outcomes among the poor.Management action: In response to the portfolio quality challenges, the HNP Sector Board hasintroduced a quarterly portfolio monitoring and benchmarking system, which is being used by theSector Board and regional management to improve portfolio performance.WORLD BANK(a) Boost population and family planning support in the formof analytic work, policy dialogue, and financing to highfertilitycountries and countries with pockets of highfertility.Management agrees with this recommendation for high-fertility countries and regions—in particularas those areas have received less attention from other development partners as well overthe past decade. Demand for stand-alone population (family planning) programs has declined overtime. The Bank should increase support for reproductive health programs, which are usually betterimplemented when they are fully embedded into public health/clinical services. We wouldxxv

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONManagement Action Record (continued)Recommendations(b) Incorporate the poverty dimension into project objectivesto increase accountability for health, nutrition, andpopulation outcomes among the poor.Management responsegenerally not support a return to stand-alone vertical family planning projects. Moreover, AIDScontrol projects (or components) and the work through UNAIDS have substantially contributedto reproductive health—and greatly expanded coverage of family planning for otherwise hardto-reachpopulation groups. In countries that are significantly advanced through the demographictransition, clients increasingly request advice and financing on financial protection, labor markets,and long-term care needs to address the ongoing demographic and epidemiological shiftsresulting in an aging society. Finally, Development Grant Facility–financed programs such as theSpecial Program of Research, Development and Research Training in Human Reproduction (HRP),into which the Bank has substantial technical, financial, and managerial input, are contributingto the global reproductive health agenda.Management generally agrees with the need to ensure that project design responds to the prioritiesand needs of the poor, and to measure the full impact of improved health services for thepoor. Management will therefore seek to ensure that adequate attention is given to poverty dimensionsin project design and supervision. However, direct assessments, where feasible, maybe technically complex and expensive. The Bank publication Attacking Inequality in the HealthSector—A Synthesis of Evidence and Tools (Yazbeck 2009) lays out a policy menu (pro-poor policyreforms along six dimensions) and a list of the analytical tools for understanding the constraintsto pro-poor targeting of public health investments in poor countries.As opposed to specific income groups, disease control programs must focus on the prevailing epidemiology(for example an AIDS program must focus on high-risk groups, irrespective of income).A malaria program focused solely on the poor would fail to eliminate malaria. Polio could only beeradicated from the Western Hemisphere by focusing on large, inclusive campaigns targeting allincome groups. Such investments in public health and control of communicable diseases areglobal public goods, generating positive externalities for society, irrespective of income status.Investments in health systems should result in increased access and better quality of services—also benefiting the poor. Investment in social security and social safety net systems prevents themiddle class from falling into poverty in case of a catastrophic health event.Management notes the substantive improvements over the past years in quantity and quality ofHNP involvement in Poverty Reduction and Economic Management Network (PREM)–led analyticalwork, and agrees that HNP must be fully included in all Poverty Assessments and fully examinedin the preparation of CASs.(c) Increase support to reduce malnutrition among the poor,whether originating in the HNP sector or other sectors.(d) Monitor health, nutrition, and population outcomesamong the poor, however defined.Management agrees with this recommendation. Particularly in the context of the global food crisis,the Bank needs to increase investments in nutrition, with a particular focus on maternal andinfant nutrition. Management is therefore investing significant resources in the next few yearsto ramp up the Bank’s analytical and investment work and leverage resources from other donors.The agenda for scaling-up nutrition is being catalyzed with additional budget resources, startingin 2009 and continuing for three years. The increased allocations are being utilized principallyin Africa and South Asia, two Regions where the malnutrition burden is highest. These funds willbe complemented by additional trust fund resources from Japan, and possibly from other donorsthat are currently engaged in discussions on this issue.Management agrees in principle with this recommendation, and will seek to ensure adequateprovision for data collection, where technically feasible, during quality-at-entry and supervisionreviews, in particular as far as poverty targeting is concerned. In order to accomplish this in asustainable manner, management believes that the first priority in many poor countries is to establishroutine health monitoring systems (surveillance, facility reporting, vital registration, censusdata, resource tracking, household surveys, and the like). These data systems need to bestrengthened in parallel to investing in project-specific management information systems, in orderto provide data and indicators that are needed for program targeting and monitoring for (but notlimited to) the poor.xxvi

MANAGEMENT RESPONSE SUMMARYManagement Action Record (continued)Recommendations(e) Bring the health and nutrition of the poor and the linksbetween high fertility, poor health, and poverty backinto poverty assessments in countries where this has beenneglected.IFC(a) Expand support for innovative approaches and viablebusiness models that demonstrate private sector solutionsto improve the health of the poor, including expansionof investments in low-cost generic drugs andtechnologies that address health problems of the poor.(b) Assess the external and internal constraints in achievingbroad social impacts in the sector.Management responseManagement agrees. Substantial progress regarding this recommendation has been made in someregions over the past years, both in terms of the analytical underpinnings, the need for capacitybuilding, and investment needs. Management will seek to improve cross-sectoral collaborationwith the PREM Network at country level as a precondition to further improvements.Management action: Recent major analytical work for staff and policy makers prepared by theHuman Development Network to improve effectiveness in reaching the poor includes: Reachingthe Poor with Health, Nutrition and Population Services—What Works, What Doesn’t and Why(Gwatkin, Wagstaff, and Yazbeck 2005) and Attacking Inequality in the Health Sector—A Synthesisof Evidence and Tools (Yazbeck 2009). It is expected that this work will help clients andstaff achieve better results in reaching the poor with health services.Concerning nutrition, in addition to disseminating the new Nutrition Strategy (Repositioning Nutritionas Central to Development—A Large Scale Action [World Bank 2006c]) the RegionalReprioritization Fund will allocate US$4 million over fiscal years 2009–11 to strengthen Bank capacityto scale up nutrition support and leverage resources from other donors.Management agrees with working on innovative approaches and helping private providers to movedown-market to serve lower-income groups and markets. IFC has several initiatives alreadyunder way to build on its work to date. Some examples include:• Health in Africa initiative.• Working with clients to invest expertise and capital from high-income to low- and lower-middleincomecountries, e.g., Saudi-German Hospitals, based in Saudi Arabia, opened hospitals inYemen, Egypt, and Ethiopia with IFC finance, creating a South-South investment.• Output-based aid projects in Yemen and Nigeria, in which poor people get subsidized care inIFC-financed private facilities that otherwise would not exist.• Creating finance facilities for health care small and medium-size enterprises in low-incomecountries by use of structured finance, combined with technical assistance, that IFC pioneeredwith banks in Africa and elsewhere to finance education facilities, and building on that knowledgeto apply it in health.• Working with clients to move down-market within their country, e.g., working with Apollo Hospitalsin India to create hospitals in secondary cities.A number of continually changing factors are enabling greater activity in the sector. Among others,the private sector partners with whom IFC must work are continually evolving and developingmore capacity and professionalism than was the case 10 years ago, due in part to IFC’s involvementwith them. Nevertheless, there is a lot still to be done and IFC needs more and larger partnerswith whom to work and is developing long-term partners and new approaches with a viewto greater scale and impact. IFC’s recent Health in Africa initiative also analyzed constraints andways to address them across multiple countries in a particularly difficult region. This is indicativeof the organization’s evolving approach as both its knowledge and resources for addressingthis recently entered sector expand.3. Strengthen the World Bank Group’s ability to help countries to improve the efficiency of their health systems.WORLD BANK(a) Better define the efficiency objectives of its supportand how efficiency improvements will be improved andmonitored.(b) Carefully assess decisions to finance additional earmarkedcommunicable disease activities in countries where otherManagement generally agrees with this recommendation. The efficiency argument is a key rationale,in particular for working with health systems in middle-income countries. Since nationalhealth expenditures rise with national income, improving sector efficiency makes an importantcontribution to fiscal sustainability. However, in the health sector there are important efficiencyequitytrade-offs. This points to the fact that efficiency gains should not be the sole objective ofBank-financed health programs.Management agrees that it is necessary to carefully assess the need for additional financewhere other donors are contributing substantial amounts. While fiscal space and potentialxxvii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONManagement Action Record (continued)Recommendationsdonors are contributing large amounts of earmarked diseasefunding and additional funds could result in distortionin allocations and inefficiencies in the rest of the healthsystem.(c) Support improved health information systems and morefrequent and vigorous evaluation of specific reforms orprogram innovations to provide timely information for improvingefficiency and efficacy.Management responsebudget substitution by ministries of finance should be closely monitored, the empirical evidenceof distortionary effects of large vertical disease programs is scanty. Proposals for Bank supportfor new disease-specific programs are closely coordinated with other donors and often fundcomplementary financing and institutional needs, for which financing was unavailable fromother donors.Management agrees partially with this recommendation: Technical support and financing for managementinformation systems as well as routine surveillance and vital statistics systems shouldbe ramped up. However, the outcomes of management information system investments may behard to evaluate fully within the timeframe of a project, and multiple determinants influence healthoutcomes. Management also notes that the Paris and Accra Declarations and the new OP 13.60emphasize the use of pooled funding and country-level M&E systems instead of ring-fenced fundingand stand-alone M&E systems. Hence, the standard should be that sufficient evidence on outputs,intermediate outcomes, and outcomes should be collected to establish a credible story lineto assess the link between Bank-financed investments and overall sector progress, including efficiencyand efficacy gains.Management action: The Human Development Network and the Regions have carried out majoranalytical work that will help policy makers and Bank staff to better understand challenges andtrade-offs in health financing, risk pooling and insurance, the issue of fiscal space and externalassistance (Health Financing Revisited—A Practitioner’s Guide [Gottret and Schieber 2006]).Furthermore, the Bank is a lead sponsor of the International Health Partnership (IHP+). This is acountry-led and country-driven partnership that calls for all signatories to accelerate action toscale up coverage and use of health services and deliver improved outcomes against the healthrelatedMDGs, while honoring commitments to improve universal access to health.IFC(a) Support public-private partnerships through AdvisoryServices to government and industry and through its investments,and expand investments in health insurance.IFC has supported pioneering health public-private partnerships (PPPs) in Romania and Lesothoand continues to work in this area. While health PPPs are a relatively recent development in emergingmarkets, there is increasing interest in health PPPs as a means to expand and improve servicesfor the public. The work in Lesotho is at the leading edge for emerging-country health PPPsin several aspects. From the investment aspect, partners in health PPPs to date have often beenconstruction companies rather than health providers and have not required capital from IFC. Insome of the few cases where it is truly health services, rather than construction and facilitiesmanagement, that have been provided by the private sector, IFC has financed providers of renaldialysis services and diagnostic services to public health systems.Many of IFC’s clients who provide health services have prepayment schemes for health care inoperation and this makes the most business sense. In some instances, by creating more low-costlocal capacity, IFC health-provider clients have made it possible for health insurers to offer newproducts with lower premiums. Experience to date has shown that the business case for directinvestment in stand-alone private health insurance does not exist to the extent envisaged whenthe 2002 IFC health strategy was devised. The few health insurance operations found to date thatactually needed capital have needed only very small amounts, too small to be viable transactions.This segment of the sector is intended to be addressed by the health-sector financing facilitiesnow being developed and in early implementation. For the balance, typically the health insureris one arm of a larger insurer that is well capitalized from its other operations such as life insuranceor is a subsidiary or joint venture of a well-capitalized foreign parent company.(b) Improve collaboration and joint sector work with theWorld Bank, leveraging Bank sector dialogue on regulatoryframeworks for health to engage new private actorswith value added to the sector, and more systematicallycoordinate with the Bank’s policy interventions regardingprivate sector participation in health.In some situations, such as the Health in Africa initiative and the Lesotho healthcare PPP, IFC andthe World Bank are collaborating very closely. In practice, there are times when this is practicaland possible and times when it is not. The imbalance in the size of human resources working onhealth in the two organizations, with the World Bank having many more people dedicated to health,requires IFC to be judicious in how it allocates its resources to work with the Bank.xxviii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONManagement Action Record (continued)RecommendationsManagement response5. Implement the results agenda and improve governance by boosting investment in and incentives for evaluation.WORLD BANK(a) Create new incentives for monitoring and evaluation forboth the Bank and the borrower linked to the project approvalprocess and the midterm review. This would includerequirements for baseline data, explicit evaluation designsfor pilot activities in project appraisal documents, and periodicevaluation of main project activities as a managementtool.IFC(a) Enhance its results orientation by developing clearlyspecified baseline indicators and an evaluation frameworkthat adequately measures IFC’s health sector objectivesand results.Management agrees. We face the challenge that countries may be unwilling to borrow for M&E,particularly for expensive impact evaluations. These large-scale evaluations may require external(grant) financing in most cases, which can be particularly difficult to obtain in a middle-incomecountry context. A second challenge is to ensure that data are readily available for public use.Some countries are not ready to fully share data and may be reluctant to include data in publicdocuments, such as PADs, and the like. This can delay the establishment of appropriate baselinedata and results frameworks prior to project approval.Management action: The HNP Sector Board and the Human Development Network have madeimportant progress to address M&E in the HNP portfolio: A number of Regions have carried outa complete portfolio review, including the retrofitting of all operations to assure an up-to-dateresults framework. Moreover, over US$2.8 million of Spanish Trust Fund (SIEF) resources are currentlyunder implementation, benefiting impact evaluations of 15 HNP projects in all Regions, focusingon Pay for Performance in Health (5 projects, US$1.2m), Malaria Control (5 projects,US$600,000), HIV/AIDS Prevention (3 projects, US$750,000), and Innovations and “quick wins”(2 projects, US$300,000).At the project level, IFC has implemented the Development Outcome Tracking System toward theend of the period under review. Over time, this is expected to improve such results orientationand specifying of baseline indicators.IFC also agrees that where there is a sufficient critical mass of projects in the health sector in aspecific country, it makes sense to try to asses IFC’s development impact in the sector beyondaggregating project-level results. While recognizing that attributing sector development to IFC’sintervention is an issue that needs to be carefully addressed, IFC is looking into ways of measuringresults beyond the project level. Among others, the work IFC is undertaking with the IDA-IFC Secretariat in reviewing the CAS results matrix could lead to the establishment of acountry-level sector development results framework that could be used in countries where IFChas a critical mass of projects in health. IFC is also exploring setting development impact andreach targets for investment departments, including health and education.xxx

Chairperson’s Summary:Committee on DevelopmentEffectiveness (CODE)On January 28, 2009, the Committee considered the document ImprovingEffectiveness and Outcomes for the Poor in Health, Nutrition,and Population: An Evaluation of World Bank Group Supportsince 1997 prepared by the Independent Evaluation Group (IEG) and the DraftManagement Response.BackgroundFollowing a self-assessment of its support in thehealth sector, the World Bank (the Bank) renewedits focus on the health sector in 2007 with an updatedstrategy, Healthy Development: The WorldBank Strategy for Health, Nutrition, and PopulationResults. IFC outlined its health sector strategyin 2002.IEG EvaluationThe report evaluated the efficacy, specifically thehealth, nutrition, and population (HNP) outcomes,of the World Bank Group (namely IBRD/IDA and IFC) country-level support since 1997 anddrew lessons from that experience. It includes, forthe first time, an evaluation of IFC’s cumulativesupport for health. Based on the findings, theIEG made recommendations for the Bank and IFCgrouped under five broad areas, which include:intensify efforts to improve the performanceof the Bank’s support for HNP; renew the commitmentto HNP outcomes among the poor;strengthen the World Bank Group’s ability tohelp countries to improve the efficiency of healthsystems; enhance the contribution of supportfrom other sectors to HNP outcomes; and implementthe results agenda and improve governanceby boosting investment in and incentivesfor evaluation.The Draft Management ResponseManagement noted its agreement with many of thefindings and recommendations that confirmedthe Bank’s own self-assessment, undertaken beforeits health sector strategy was updated. At thesame time, it offered its views on several aspectsof the evaluation, including coverage of the evaluation,targeting issues, the importance of notonly focusing on health outcomes for the poor butalso on preventing poverty due to financial costsfrom poor health, emerging issues such as theaging population in developing countries, challengeswith respect to monitoring and evaluation,and actions being taken to address some of the issuesidentified by IEG.Overall ConclusionsThe Committee welcomed the discussion, notingthe importance of the IEG evaluation findings. Italso remarked on the centrality of the health sectorto the Bank’s mandate for poverty reductionand contribution to the Millennium DevelopmentGoals (MDGs), although the Bank’s role has shiftedto become a smaller financier in the sector. Con-xxxi

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONcerns were expressed about the main IEG findings,including the under-performance of the HNP portfolio,particularly in the Africa Region; the weak accountabilityfor ensuring that the results havereached the poor; and the continued weaknessesin monitoring and evaluation (M&E). Managementwas also asked to address the issue of excessivecomplexity of health programs whilerecognizing its multisectoral dimension.Speakers raised questions and comments on arange of issues, including the need for more Banksupport for nutrition and population, the importanceof addressing maternal health, HIV/AIDSmeasures in health system strengthening, andthe effectiveness of the sectorwide approaches(SWAps). They also remarked on the importanceof establishing realistic targets, project supervision,local capacity building, reliable data, and appropriatestaff skill mix and incentives to supportthe HNP strategy. Some speakers expressed interestin the development impact of IFC’s healthsector projects, particularly the inclusion of thepoor. The Committee asked management to reviseits response to address the various commentsmade at the meeting, including the requestfor an action plan to address the IEG recommendations.Several speakers also noted the needfor a sound communication strategy before thedisclosure of the IEG report.Next StepsThe revised Management Response will be circulatedfor information to the Committee in advanceof the Board’s informal meeting to considermanagement’s first report on the implementationof the new health sector strategy—Health, Nutrition,and Population Strategy Implementation—AnInterim Report. The Committee willrecommend to the Board that it consider the IEGevaluation report and the revised Management Responsealong with management’s interim report,which is scheduled for an informal Board discussionon April 9, 2009. Management proposed toprovide informal updates in response to speakers’interest in regular reports on the World BankGroup support for the health sector.Main issues raised at the meeting were thefollowing:World Bank Group’s Role. Members remarkedon the changing environment of the health sectorwith an increase in institutions providing HNPsupport. In this regard, they emphasized qualityover quantity of World Bank Group support. Theyalso noted that the World Bank Group has an importantrole in donor coordination, ensuring efficientallocation and effective use of resources inthe sector, and introducing innovations and generatingknowledge to strengthen health systemsand service delivery. Remarking on the inherentcomplexity and high risk of health sector support,a member emphasized the World Bank Group’scontinued role in this sector and also noted thatthe risks should be taken into account in assessingperformance, to avoid providing staff with disincentivesto invest in valuable but risky projects.Bank’s Portfolio Performance. The importanceof addressing the Bank’s comparativelylower portfolio performance in the health sector,including the poor results in Africa Region, washighlighted. Considering the Bank’s comparativeadvantage as providing policy advice, strengtheninghealth systems, and supporting institutionaland human capacity building, severalnon-members queried about the growing shareof programs addressing communicable diseases.A few members emphasized the importance of politicalanalysis, which could be done during thepreparation of country assistance strategies orthrough policy notes; project supervision; andlearning, including for local implementation capacitybuilding. Questions were raised about ensuringan appropriate staff skills mix to deliver HNPprograms, especially in the Africa Region, takinginto consideration the global competition forsuch skills. Bank management elaborated onits review of all at-risk HNP projects, and workwith the Quality Assurance Group to determinewhat actions are needed. It described the challengesand risks faced in the Africa Region andin the fragile states, and its efforts to improve performancein the Africa Region, including estabxxxii

CHAIRPERSON’S SUMMARY: COMMITTEE ON DEVELOPMENT EFFECTIVENESS (CODE)lishing two technical hubs and staffing themwith additional experts in epidemiology, healtheconomics, and health management to supportanalytical work and project implementation.It said that it is scaling up technical assistanceand capacity building at the country level.A few speakers expressed interest in understandingthe Bank’s HNP portfolio performanceacross time, as compared to other sectors, and inthe context of performance of other donor supportto HNP. One of them also sought informationon the Bank’s HNP support against the principlesof the 2005 Paris Declaration. IEG responded thatbetween fiscal years 1992 and 2001, the performanceof both the HNP portfolio and other sectorshad improved, but since fiscal 2002, theHNP portfolio performance has been flat, whilethat of other sectors has continued to improve.It referred to its evaluations of global programreviews in HNP, but noted that it did not have thesame level of data on the effectiveness of HNP supportby other donors.Reaching the Poor. Several speakers askedabout the Bank’s measures to strengthen thepoverty focus of HNP sector support. While agreeingon the importance of reaching the poor, amember observed the greater challenges of assessingoutcomes for the poor in the case of certaininitiatives such as for communicable diseases,where focus is on groups at the highest risk or investmentin vaccines. A few members expressedinterest in IFC’s efforts to improve the inclusionof the poor in its health projects, and in this regardalso requested IEG to elaborate on its recommendation.Bank management commentedon its renewed focus on results, including for thepoor, such as through the results-based financingmechanism. It also explained that investmentsin diseases of the poor, such as malaria,have a fully pro-poor targeted approach. IFC respondedthat it is supporting smaller health careinstitutions providing services to the poor throughwholesaling or other arrangements with financialpartners. It is also providing Advisory Servicesto promote public-private partnerships toprovide services to the poor, as well as workingwith large-scale service providers to achieve costefficiencies and to make services more affordablefor the poor. IEG clarified that the impact ofWorld Bank Group support on the poor is largelyunknown, because outcomes among the poorhave not been monitored. It suggested that IFCcould improve the social impact of health initiativesby supporting investments that havegreater benefits for the poor, such as supportingmanufacturing and distribution of lower-cost,higher-quality generic drugs and research anddevelopment to treat diseases that disproportionatelyaffect the poor.Project Design and Approach. A few membersnoted that the design of HNP projects could besimpler or phased, particularly in countries facinginstitutional and implementation capacity issues.At the same time, they also said the design shouldbe based on country context and achievableobjectives. Likewise, a member remarked that thelevel of multisectoral cooperation should besituation-specific, depending on capacity andavailability of resources. Bank managementelaborated on how a health project with a seeminglysimple objective and apparently easilymeasurable results (for example, providing mosquitonets to counter malaria) can require acomplex solution, particularly in countries withlimited infrastructure and capacity, and offerchallenges in monitoring results. IEG emphasizedthe importance of setting achievable objectivesand a clear results framework. It clarifiedthat there was no evidence in the evaluationthat simple health projects are less sustainable orthose with more complex designs are more sustainable.It also said that the World Bank Groupshould not avoid investing in worthwhile butrisky projects, but there is a substantial scopefor minimizing risks through better ex ante riskanalysis and mitigation measures. A few speakersexpressed interest in more analysis of themixed results of the SWAps and IEG’s recommendationsfor future use of this approach. IEGresponded that the context is important; SWApswork better in some contexts than in others. It alsoxxxiii

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONsaid that while SWAps have given much attentionto the process, there could be more focus on HNPoutcomes.Nutrition. A member welcomed the role of theBank in emphasizing the importance of nutrition.Another member noted the importance ofincorporating food security concerns in health sectorprojects.Population and Reproductive Health. Severalspeakers encouraged the Bank to strengthenits focus on population and reproductive health.A few members remarked on the developmentchallenge of high population growth from a globalperspective and in Sub-Saharan Africa. Othersurged the Bank to prioritize reproductive healthand identify maternal health as a key target in itshealth programs. In addition, some speakers notedthe importance of a gender-based approach to, andintegrating HIV/AIDS measures in, health systemsupport, particularly in the area of reproductivehealth. The need for adequate staffing and analyticalwork to support initiatives addressing populationand reproductive health was emphasized.Bank management said that it is currently consideringhow to strengthen the focus on populationissues in its health system support and itintended to elaborate on this in the forthcomingHNP interim report to the Board in April.Monitoring and Evaluation. Members and nonmembersunderlined the importance of improvingM&E to enable a better understanding of theBank’s performance in the sector. They commentedon the need for a clear results frameworkand M&E plan in all initiatives, for establishing andstrengthening country-based M&E systems, and forappropriate incentives within the World BankGroup as well as for promoting country ownership.The lack of reliable health data, challenges of datacollection, and reluctance of countries to sharedata were discussed. A member supported a pragmaticand realistic approach to M&E, taking intoconsideration the local capacity and financial resourceconstraints. Bank management elaboratedon the serious attention it is giving to M&E,noting that the Human Development Network hasthe biggest impact evaluation program in theBank. It described its work with other partnersto jointly strengthen data at the country leveland to strengthen the country’s M&E systems.IFC said it is setting project-level baselines andmonitoring relevant development impact indicatorsthrough the Development Outcome TrackingSystem.Response to the IEG Evaluation ReportMembers and non-members requested that managementrevise its response to the IEG evaluationreport to address members’ comments, includingthe requests for an action plan addressing IEG’srecommendations that may be monitored. IEGnoted that the evaluation report would be disclosedwith the revised Management Response.Several speakers stressed the importance of communicating,particularly to IDA donors, that management’sresponse will urgently and effectivelyaddress the evaluation findings and recommendations.In this context, the chairperson referredto another speaker’s observation about the issueof timing of the IEG evaluation with respect to theBank’s 2007 HNP strategy. It was widely felt thatthe IEG recommendations may be incorporatedin the interim report on health, nutrition, andpopulation strategy implementation, scheduled tobe discussed at the Board on April 9, 2009. In thisrespect, members and non-members requestedthat the Board consider the IEG evaluation reporttogether with the revised Management Response,alongside the management’s paper on Health,Nutrition, and Population Strategy Implementation—AnInterim Report. The general preferencewas to discuss the reports in one meeting.Giovanni Majnoni, Chairmanxxxiv

Advisory Panel StatementThe external advisory group welcomes this reporton World Bank Group support for health, nutrition,and population outcomes since 1997. In anera when health has been very high on the internationalagenda, it is vital that development agenciessuch as the World Bank Group rigorouslyscrutinize the effectiveness of what they do andlearn from such scrutiny to improve practices.From its involvement during the process of the evaluation,the advisory group was satisfied that theprocesses were transparent, independent, constructive,and evidence-based. The evaluation’sapproach and methods made the most of what evidencewas available, and the overall analyses, conclusions,and recommendations are sound. Belowwe highlight and comment on key findings.The decade has seen a remarkable increase ininternational assistance for health, and a markeddecline in the Bank’s share of total assistance,from 18 percent in the 1990s to 6 percent. Whilewe echo the view of the evaluation that the Bankstill has a very important role to play, we weretaken aback by the extent to which the Bank followedthe trend of increased support to communicabledisease control. At a time when diseasespecificprograms were getting greatly increasedsupport from elsewhere, we were surprised thatthe Bank did not provide a countervailing trend.Indeed, there was a fall of nearly half in the shareof projects with objectives to reform the healthsystem. We endorse the view in the report of theBank’s comparative advantage—that it can providelong-term, sustained engagement, a focus onbuilding country capacity in the sector, stronglinks to Ministries of Finance, and engagementacross many sectors—and fully agree that its focusshould be on making health systems work betterand ensuring that benefits reach the poor.In this context, it is a source of considerable concernthat the performance of the HNP portfoliooverall has been below average, and that withinthis the health sector reform type of projectshave tended to perform less well. The report providesmuch food for thought in exploring why thismight have been the case. We strongly endorsethe recommendations that project design shouldbe matched to country context and capacity, thatcomplex projects should be avoided in low capacitysettings, and that thorough institutionalanalysis and exploration of political economy issuesshould be part of project design and implementation.The strong preparatory analytical workthat the report calls for should help ensure thatprojects are relevant to country needs. Althoughthe advisory group agrees that it is important toseek to explore the determinants of project outcomeratings, the regression results summarizedin the text and presented in an appendix tableshould be interpreted with care.The evaluation did not undertake any extensiveanalysis of the past analytical work done in HNP.However, it is notable that 41 percent of the analyticalwork was on health system performance,and yet many projects in this area encountered difficulties.Further exploration of the analyticalwork would be valuable, to assess the extent towhich it was relevant to country programs and tosee whether lessons can be learned in terms of ensuringthat analytical work supports high-qualityproject design and implementation.Another area that would have benefited fromgreater attention is that of sustainability and buildingcountry capacity. Although sustainability wasan objective in both the 1997 and 2007 HNP strategies,this aspect has not been sufficiently exploredin the report. There are a number of aspects ofxxxv

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONcapacity that need building at the country level—human resources are absolutely key and yetwhether or not they were adequately addressedby projects was not explored. There are similarconsiderations with respect to health informationsystems and national monitoring and evaluationcapacity. No evidence is presented on the extentto which local capacity was used or built toensure projects are sustained into the future,or on whether projects were managed in themost efficient way. If there are inadequate attemptsto use or build internal capacity, it ishighly likely that the projects will not be effectivelysustained.The report comments that while half of all WaterSupply and Sanitation projects cited potential forhealth benefits, only 10 percent had an objectiveto improve health. This is presented as a weakness,but it need not be: the projects can improve otherthings that have large effects on health, so failureof a project to state that it is improving health orits failure to do things that directly improve healthmay not be bad for health. For example, SWAPsencourage projects to do things that indirectly improvehealth.The report highlights a clear problem of accountabilityfor results. Despite the Bank’s mandatefor poverty reduction, a very small share ofprojects had explicit objectives relating to improvingHNP outcomes among the poor, and ofclosed projects with these objectives, very fewwere able to demonstrate improvements. Similarly,many projects were termed pilots, implying theywere intended at least in part for learning, but fewprojects actually evaluated results. A widespreadweakness in monitoring and evaluation and inevaluation was identified.It is notable that these weaknesses have beenidentified in previous evaluations, and recommendationsmade to improve monitoring andevaluation. Improvements are noted—for example,in terms of availability of baseline data—butweaknesses remain. Failure to respond sufficientlyto previous recommendations suggests that thereare pervasive incentives in the Bank that workagainst investing sufficiently in monitoring andevaluation and evaluation. This is a vital area formanagement action to consider how to createstronger incentives. The issue of whether staffare drawing in the necessary expertise in evaluationresearch methods also needs consideration—issues of appropriate research design and dataanalysis methods are complex.The World Bank Group has many strong assets,including its analytical and synthesis capacity, itsstrong relationship to country financial policymakers, its extensive networks at the countrylevel in all regions, its massive financial and socialcapital, and its skill in managing developmentfunds. These assets are extremely important inhelping the group successfully formulate, implement,evaluate, and reformulate its HNP projects.If applied properly and efficiently, theseassets will allow the World Bank Group to buildup sustainable capacity within developing countriesand support health systems developmentthat is more pro-poor, more efficient, and moresustainable. However, given the now highly complexaid environment in health, it is vital that theWorld Bank position itself clearly with respect towhat others are doing. To address the problemof fragmentation within countries, it should alsoseek to support interagency coordination.Finally, there is a need to strengthen the IEG evaluationteam and resources. Insufficient human resources,financial support, and time inevitablylimited the work that could be done for the evaluationreport. The IEG staff have done their best,given limited resources, to come up with an excellentassessment with much rich detail. Butmore could be done with better support.Augusto Galán-Sarmiento, former Minister of Health, ColombiaAnne Mills, London School of Hygiene and Tropical Medicine, United KingdonGermano Mwabu, University of Nairobi, KenyaSuwit Wibulpolprasert, Ministry of Public Health, Thailandxxxvi

Chapter 1Evaluation Highlights• Because of fundamental changes inthe global aid architecture, the WorldBank is no longer the largest externalsource of health finance.• Health outcomes have improved inevery region, but averages conceal differencesacross and within countries.• The Bank has committed about $28.7billion and IFC about $951 million toHNP since 1970.• The 2007 HNP strategy, Healthy Development,aims to strengthen healthsystems, prevent impoverishment dueto poor health, and improve the healthof the poor.• This evaluation considers the effectivenessof both World Bank and IFCactivities in HNP over the past decadeand points to lessons of thatexperience.

Child being immunized in Liberia. Support for communicable disease control rose dramatically during the evaluation period.Photo courtesy of Melanie Zipperer.

IntroductionThe past decade has seen fundamental changes in the global aid architecturein health, with potential implications for the World BankGroup’s work and its comparative advantages relative to other sourcesof health, nutrition, and population (HNP) support.In the late 1990s, the World Bank was the largestsingle external source of finance for HNP in developingcountries, accounting for about 18 percentof global HNP aid. 1 Since then, new aiddonors and institutions have emerged, both publicand private (see Timeline in appendix A). Developmentassistance for HNP has risen from anannual average of $6.7 billion in 1997–99 to about$16 billion in 2006, most of it for low-incomecountries (Michaud 2003; World Bank 2008a).The international community has adopted globaldevelopment targets; the most prominent arethe Millennium Development Goals (MDGs), setfor 2015.There is also a new international emphasis onaid effectiveness, results orientation, donor harmonization,alignment, and country leadership,reflected in the 2005 Paris Declaration on Aid Effectiveness.Hence, the World Bank is now one ofmany large players in international HNP support,accounting for about 6 percent of the total in2006. 2 The Bank is in the process of reassessingits comparative advantage in the context of thenew aid architecture, while a call for greaterengagement with the private health sector in developingcountries presents new opportunities forthe International Finance Corporation (IFC) to extendits support.There have been improvements in somehealth outcomes over the past decade, butprogress is uneven and in many cases tooslow to meet the MDGs by the target dateof 2015. Key health outcomes, such as the infantmortality rate, have improved in every developingregion since 1990 (figure 1.1). The prevalenceof stunting among children under five has declineddramatically in Asia and Latin America since1980, though only modestly in Africa (Shekar,Heaver, and Lee 2006, p. 5). These improvementshave been attributed to rising average levels of incomeand education, coupled with improvementsin health technology and expanded public healthinterventions (see, for example, Jamison 2006;Deaths per 1,000 live birthsFigure 1.1: Infant Mortality Rates Have Declined inEvery Region, but Disparities across Regions are Large120100806040200Source: UNICEF 2006.59444343111 1118979554239351990 1995Sub-Saharan AfricaSouth AsiaMiddle East and North Africa105734937343010267443229262000 2004Europe and Central AsiaEast Asia and PacificLatin America and the Caribbean3

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONLevine and others 2004). Despite this progress,nearly three-quarters of developing countries areeither off track or seriously off track for achievingthe MDG of reducing under-five mortality by twothirds.Maternal mortality is declining by onlyabout 1 percent annually, a fifth of the rate neededto achieve the goal of reducing it by threequartersby 2015 (World Bank 2008a).Further, average outcomes conceal importantdifferences in progress across countries,within regions, and within countries.Under-five mortality rates in 30 countries have stagnatedor increased since 1990; 3 in some countries,high fertility rates have remained unchangedor even increased slightlyDespite improvement insome key health outcomessince the 1990s (Wagstaff and Claesonsince 1990, there are2004, p. 36). Despite some progressimportant differencesin Bangladesh and India, undernutritionremains very high in South Asia,across and withincountries.while in 26 countries, primarily in Africa,malnutrition is increasing (Shekar, Heaver, and Lee2006, p. 3). Maternal mortality remains extremelyhigh in Africa, where the average woman faces anearly 1 percent risk of dying from pregnancyand childbirth, and very high fertility repeatedlyexposes women to these high risks. 4The Bank committedabout $28.7 billion andIFC about $951 millionto HNP from 1970 tomid-2008.Communicable diseases account for about a third(36 percent) of the disease burden in developingcountries (Jamison and others 2006b); withincountries, the burden of morbidity and mortalityis greatest among the poor (Gwatkin and Guillot2000). In some countries, HNP outcomeshave improved disproportionately among thepoor, while in others they have improved primarilyamong the non-poor. The gaps betweenthe poor and non-poor, even when closing, oftenremain substantial, in part reflecting lower accessof the poor to public services (Gwatkin,Wagstaff, and Yazbeck 2005; Filmer 2003). Theburden of disease is distributed differently withindeveloping Regions, with communicable diseaseand maternal, perinatal, and nutritionalconditions as a group predominatingin Africa, while in the remaining fivedeveloping Regions, the burden ofnon-communicable disease is equal orgreater (figure 1.2).Rationale for World Bank GroupInvestments in Health, Nutrition,and PopulationThe mandate of the World Bank Group is toreduce poverty and promote economicgrowth. Poor health and malnutrition contributeto low productivity of the poor, so improvingHNP outcomes is seen as a major way of reducingpoverty. But poverty is also a prime cause ofpoor health, malnutrition, and high fertility. Thepoor have low access to preventive and curativecare (both physically and financially) and are morelikely to be malnourished, have unsafe water andsanitation, lack education, have large families andclosely spaced births, and engage in activities thatmay put them at heightened health risk.Within the World Bank Group, the WorldBank (International Bank for Reconstructionand Development [IBRD] and the InternationalDevelopment Association [IDA]) havecommitted about $28.7 billion to help governmentsimprove HNP outcomes in 132countries since 1970 (figure 1.3). 5 In addition,IFC has invested $951 million in the privatehealth and pharmaceutical sectors of developingcountries. 6 The World Bank supportsgovernment HNP policies and programs. TheWorld Development Report 1993: Investing inHealth (World Bank 1993c) highlighted the majorrationales for a government role in the healthsector; many of these also apply to nutrition andpopulation:• To provide public goods and invest in HNPservices with large positive externalities, whichthe private sector would have no incentive toprovide in adequate quantity• To enhance equity by ensuring provision ofcost-effective HNP services to the poor, who areotherwise unlikely to gain adequate access toessential clinical services or insurance 7• To address uncertainty and multiple marketfailures in health, including problems of adverseselection, moral hazard, and asymmetryin information between providers and patients. 8IFC supports investments and Advisory Servicesto the private sector in health and pharmaceuti-4

INTRODUCTIONFigure 1.2: Communicable Diseases Remain a Significant Share of the DiseaseBurden in Most Developing Regions10080Percent6040200AfricaSouth AsiaMiddle EastandNorth AfricaEast Asiaand PacificRegionLatin Americaand theCaribbeanEuropeandCentral AsiaHigh-incomecountriesInjuriesNoncommunicable diseasesCommunicable, maternal, perinatal, and nutritional conditionsSource: Jamison and others 2006b, table 4.1.Note: Based on 2001 data.cals to build institutional and systemic capacity andpromote efficiency and innovation in the healthsectors of developing countries.World Bank Group Strategies in Health,Nutrition, and PopulationThis evaluation aims to inform the implementationof the most recent HNP strategiesof the World Bank and IFC and to help enhancethe impact of future support. TheWorld Bank’s policies, strategies, and lending forHNP have evolved in phases (box 1.1). Its currentstrategy, Healthy Development: The WorldBank Strategy for Health, Nutrition, and PopulationResults, was launched in 2007, with objectivesto improve HNP outcomes on average andamong the poor and prevent the impoverishingimpact of illness by improving health system performance,including governance, and intersectoralapproaches (table 1.1). The strategy includesan increased emphasis on demonstrating outcomesby incorporating a detailed results frameworkfor the entire sector. The goals of IFC’scurrent health strategy, Investing in Private HealthCare: Strategic Directions for IFC, adopted in2002, are: improve health outcomes, protect thepopulation from the impoverishing effects of illhealth, and enhance the performance of health services(table 1.2) (IFC 2002, p. 3). These strategiesprovide a vision of the sector as a whole. Theextent to which their objectives are specificallyaddressed in a given country depends on theborrower’s interest and the country context.Since 1997, IEG has issued three evaluationsof the development effectiveness of theBank’s support for HNP; it has never evaluatedIFC’s cumulative support for thehealth sector. 9 Investing in Health: DevelopmentEffectiveness in the Health, Nutrition, andPopulation Sector (IEG 1999) foundthat the Bank had been more successfulin expanding health service deliverysystems (physical objectives) than inimproving service quality and efficiencyor achieving policy and institutionalchange. The lending portfolio hadgrown rapidly, and the most complexSince 1997, IEG hasevaluated aspects ofthe World Bank’s HNPsector three times; IFC’scumulative support forhealth has never beenevaluated.5

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 1.3: Trends in World Bank Group Commitments and Project ApprovalsA. World Bank2,50080Commitments (US$ million)B. IFCCommitments (US$ million)2,0001,5001,0005003503002502001501005000Board approveshealth lending1970197119721973197419751976197719781979198019811982198319841985198619701971197219731974197519761977197819791980198119821983198419851986Net commitments198719881989Fiscal year198719881989Fiscal year1997 HNPsector strategy19901991199219931994199519962002 IFChealth strategyProjects1990199119921993199419951996199719981999200020012002200320042005200620072008199719981999200020012002200320042005200620072008Source: World Bank Group data.Note: The commitment for the entire project is attributed to the fiscal year of approval. IFC data excludes investments made under the Africa Enterprise Fund/Small Enterprise Fund.70605040302010014121086420Projects (number)Projects (number)projects were approved in countries with theweakest institutional capacity. The evaluation recommendedthat the Bank be more selective in itsengagement and focus on improving the qualityof HNP operations, particularly through strongermonitoring and evaluation (M&E) and institutionalanalysis. In addition, the evaluation recommendedstrengthening health promotion and6

INTRODUCTIONBox 1.1: Six Phases of World Bank Engagement in HNPPopulation lending, 1970–79The Bank focused on improving access to family planning servicesbecause of concern about the adverse effects of rapidpopulation growth on economic growth and poverty reduction.A handful of nutrition projects was also approved following a 1973nutrition policy paper, and throughout the decade health componentswere included in agriculture, population, and educationprojects as important links between health, poverty, andeconomic progress were established.Primary health care, 1980–86The 1980 Health Sector Policy Paper (World Bank 1980a) formallycommitted the Bank to direct lending in the health sector with theobjective of improving the health of the poor by improving accessto low-cost primary health care. The rationale for this policychange was contained in the World Development Report 1980:Poverty and Human Development (World Bank 1980b), which emphasizedthat investment in human development complementsother poverty reduction programs and is economically justifiable.However, over this period systemic constraints were encounteredin providing access to efficient and equitable health services.Health reform, 1987–96Following the release of Financing Health Services in DevelopingCountries: An Agenda for Reform in 1987 (Akin, Birdsall, andDe Ferranti 1987), the Bank addressed two new objectives: tomake health finance more equitable and efficient and to reformhealth systems to overcome systemic constraints. The messagewas further refined by the World Development Report 1993: Investingin Health (World Bank 1993c), which highlighted the importanceof household decisions in improving health, advocateddirecting government health spending to a cost-effective packageof preventive and basic curative services, and encouragedgreater diversity in health finance and service delivery.Health outcomes and health systems, 1997–2000The 1997 Health, Nutrition, and Population Sector Strategy Paper(World Bank 1997b) focused on health outcomes of the poor andon protecting people from the impoverishing effects of illness, malnutrition,and high fertility. However, it continued to emphasizesupport for improved health system performance (in terms ofequity, affordability, efficiency, quality, and responsiveness toclients) and securing sustainable health financing.Global targets and partnerships, 2001–06The Bank’s objectives, rationale, and strategy remained unchanged,but major external events, the surging AIDS epidemic,and the Bank’s commitments to specific targets and to workingin partnerships led to an increase in finance for single-diseaseor single-intervention programs, often within weak health systems.System strengthening for results, 2007–presentIn the context of changes in the global health architecture,Healthy Development: The World Bank Strategy for Health, Nutrition,and Population Results (World Bank 2007a) emphasizes theneed for the Bank to reposition itself, with a greater focus on itscomparative advantages, to more effectively support countriesto improve health outcomes. It adheres closely to the 1997 strategy’sobjectives and means for achieving them, with increasedemphasis on governance and demonstrating results.Source: Fair 2008.intersectoral interventions, greater emphasison economic and sector analysis, a better understandingof stakeholder interests, and strategicalliances with regional and global developmentpartners.IEG’s 2004 evaluation of the Bank’s approach toglobal programs (IEG 2004a; Lele and others2004), including health programs, recommendedthat the Bank engage more selectively in globalprograms, favoring those that exploit the Bank’scomparative advantages and provide global publicgoods, and that the links between global programsand the Bank’s Regional and country operationsbe strengthened.IEG’s 2005 evaluation of the Bank’s support forthe fight against HIV/AIDS (IEG 2005a) foundthat the Bank had contributed to raising politicalcommitment and improving access to services.However, evidence of results for health behaviorsand outcomes is thin because of a failure to monitorand evaluate. IEG recommended that theBank be more strategic and selective, focusing on7

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable 1.1: Objectives and Strategic Directions of Healthy Development, the 2007World Bank HNP StrategyObjectives• Improve the level and distribution of key HNP outcomes(for example, MDGs), outputs, and system performanceat the country and global levels in order toimprove living conditions, particularly for the poor andvulnerable.• Prevent poverty due to illness (by improving financialprotection).• Improve financial sustainability in the HNP sector andits contribution to sound macroeconomic and fiscalpolicy and to country competitiveness.• Improve governance, accountability, and transparencyin the health sector.Strategic directions• Renew Bank focus on HNP results.• Increase Bank contribution to client country efforts tostrengthen and realize well-organized and sustainablehealth systems for HNP results.• Ensure synergy between health-system strengtheningand priority disease interventions, particularly in lowincomecountries.• Strengthen Bank capacity to advise client countries onan intersectoral approach to HNP results.• Increase selectivity, improve strategic engagement, andreach agreement with global partners on collaborativedivision of labor for the benefit of client countries.Source: World Bank 2007a, p. 7.8efforts likely to have the largest impact for theircost; strengthen locally adapted institutions tomanage the long-term response; and invest heavilyin M&E capacity and incentives as the basis forevidence-based decision making.Objectives and Scope of this EvaluationThis report evaluates the efficacy, specificallythe HNP outcomes, of World Bank Groupcountry-level support for HNP over the pastdecade and draws lessons from that experience.The objectives of the evaluation of theWorld Bank are to assess the effectiveness ofthe Bank’s support in improving HNP outcomesat the country level since 1997, particularly amongthe poor, and to identify lessons from that experiencethat can be employed to improve the efficacyof the Bank’s support in the next decade.Support to countries includes policy dialogue,analytic work, and lending. The evaluation focuseson assessing the effectiveness of supportunder the supervision of the HNP sector, as wellas support with likely health benefits supervisedby the water supply and sanitation and the transportsectors. 10The evaluation of the Bank addresses fourmain questions:• What have been the objectives, effectiveness,and main outcomes of the WorldBank’s country-level HNP support overthe past decade and what accounts forthis performance? How effective has theBank’s support been in helping countries improveHNP outcomes, on average and amongthe poor? To what extent has the Bank improvedits M&E performance and used evaluationto improve the evidence base for decisionmaking in HNP? These findings are intended tocontribute to more effective implementation ofthe 2007 HNP strategy’s objective of improvingHNP outcomes among the poor and of its resultsfocus.• What lessons have been learned about theefficacy, advantages, and disadvantages ofthree approaches to improving HNP outcomesin different settings: control of communicablediseases that disproportionatelyaffect the poor; programs to “strengthen” or “reform”the health system; and sectorwide approachesdesigned to improve ownership,reduce transaction costs, and improve the allocationof resources? These three approaches,which are not mutually exclusive, are closely relatedto the HNP strategy’s objectives of healthsystem strengthening and are relevant to the balancebetween investments in communicablediseases and health systems.• What has been the contribution of activitiesin other, complementary sectors toimproving HNP outcomes? In particular,to what extent have Country Assistance Strategiesbeen used as a vehicle to coordinate the

INTRODUCTIONTable 1.2: Business and Developmental Objectives of IFC’s 2002 Health StrategyBusiness objectives• Provide value-added financing.• Mobilize private resource flows.• Invest in financially viable projects.• Improve managerial and financial capacity.Developmental objectives• Contribute to institutional and systematic building of capacity in thehealth sector by expanding the capacity of the sponsoring institutionto serve patients, either through the creation of new facilities or theexpansion or improvement of existing operations. Capacity buildingis also served when IFC clients transfer technical expertise fromprivate providers to the public facilities.• Promote efficiency and innovation within the institution and thehealth sector.• Support country and World Bank health sector objectives.• Financial protection through:– Expanding access and quality of health services to those whowould otherwise receive inadequate or no care– Investing in health insurance schemes.• Reduce the brain drain by improving the quality and quantity of localfacilities, contributing to the supply of attractive local employmentopportunities.• Reduce pressure on the overburdened public sector.Source: IFC 2002, pp. 32–37.activities of other sectors to maximize the impacton HNP outcomes? Have projects in non-HNP sectors with plausible health benefitsdelivered on them? How effective has multisectorallending managed by the HNP sectorbeen at capitalizing on synergies between sectorsin producing health outcomes? This questionis intended to inform efforts of the recentstrategy to implement an effective multisectoralapproach and will also provide evidence on theextent to which the Bank has a comparative advantagein multisectoral action.• What have been the observed value added,comparative advantages, or contributionsof World Bank support for HNP in developingcountries over the past decade, andhow is that changing?Over the past decade, the World Bank increasinglyhas become involved in global partnerships. An appendixto the 2007 HNP strategy lists 19 globalhealth initiatives and partnerships to which theBank contributes financially and another 15 inwhich the Bank participates with no financial contribution.11 While it is beyond the scope of thisevaluation to assess the Bank’s participation andcontribution to these activities in the aggregate,IEG has recently reviewed 2 of the 19partnerships financed by the Bank—theMedicines for Malaria Venture and thePopulation and Reproductive HealthCapacity Building Program, summarizedin appendix I.The evaluation of IFC’s health activities hasthree objectives: assess the relevance, effectiveness,and efficiency of IFC’s health operationssince 1997; assess the design and implementationof IFC’s 2002 health sector strategy; and identifylessons from that experience to improve the efficacyof future IFC support to the health sector.Health operations in IFC include both lending andAdvisory Services for health and investments,both in health services, managed by the Health andEducation Department, and pharmaceuticals, managedby the Global Manufacturing Department.Evaluation Design and MethodologyThe evidence for the evaluation was distilled fromdesk reviews, background papers, country casestudies and field visits, IEG project assessments,and in-depth interviews, in addition to publishedand unpublished research and evaluation literature(box 1.2). Some of the evidence is compre-This report evaluates theefficacy of World BankGroup country support forHNP over the past decade.9

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 1.2: Evaluation Building BlocksWorld Bank• Review of strategy documents and construction of a timelineof World Bank HNP support and policies and internationalevents.• Desk review of Country Assistance Strategies (CASs) with respectto their prioritization of health, attention to health outcomesamong the poor, and planning of multisectoral operations.• Desk review of the objectives, strategies, and development effectivenessa of all 220 HNP projects approved during fiscalyears 1997–2006 under the responsibility of the HNP sector, aswell as analysis of projects with HNP objectives or componentsunder the responsibility of the transport and water supply andsanitation sectors.• Field evaluations (Project Performance Assessment Reports,PPARs) of completed HNP projects in Bangladesh, the ArabRepublic of Egypt, Eritrea, Ghana, Kyrgyz Republic, Peru, theRussian Federation, and Vietnam; a road safety project inRomania; and a rural water supply and sanitation project inNepal that had objectives to improve HNP outcomes. Fieldbasedcountry case studies to evaluate the entirety of WorldBank lending and nonlending support in Egypt, Malawi, andNepal.• Background papers reviewing the evidence from the portfolioreview and field studies on key evaluation themes, such as communicablediseases, sectorwide approaches, and M&E.IFC• Review of IFC health sector strategies.• Review of the objectives, characteristics, design features, andimplementation status of the portfolio of all 54 committed healthinvestment projects approved between fiscal 1997 and 2007 andan assessment of the performance of mature health projectsagainst established benchmarks and their stated objectives.• Desk review of all completed and ongoing Advisory Service projectsin health and follow-up interviews with World Bank andIFC staff, as appropriate.• Field visits to Argentina, China, Egypt, Philippines, and Turkeyto interview IFC clients and other stakeholders, and to reviewthe performance of 12 investment projects, IFC support forpublic-private partnerships, and IFC-World Bank collaboration.a. Development effectiveness is the extent to which a program has attained its major relevant objectives efficiently.hensive—representing 100 percent of the lendingportfolio—while other evidence is culled fromin-depth investigation of purposive samples. Thesamples of projects reviewed are described in appendixB, and the World Bank HNP sector projectsand IFC health projects included in theportfolio review are listed in appendixes C and D.The evaluation also draws on findings and lessonsfrom other IEG evaluations that are relevant to theWorld Bank Group’s HNP support, in particular, evaluationsof social funds (IEG 2002b), middle-incomecountries (IEG 2007b), public-sector reform (IEG2008f), economic and sector work (IEG 2008h), andan impact evaluation of maternal and child healthand nutrition in Bangladesh (IEG 2005b).Organization of the ReportThe evaluation results are presented in six chapters.Chapters 2–4 evaluate the World Bank’s supportsince 1997 with respect to the key evaluationthemes.• Chapter 2 examines the evolution and outcomesof the portfolio of HNP lending and analyticwork. Key issues include: trends in theobjectives, composition, and performance ofthe HNP portfolio; the poverty focus and outcomesfor the poor of the Bank’s investments;and whether M&E of HNP activities have improvedthe evidence base for decision making.• Chapter 3 distills the performance and lessonsfrom a decade of experience supporting threeapproaches that are closely related to the strategicactions to be taken in the 2007 HNP strategy:communicable disease control, health system reform,and sectorwide approaches (SWAps) inhealth, which aim to improve the efficiency andeffectiveness of donor assistance in support ofdeveloping country health objectives.• The 2007 HNP strategy maintains that multisectoralaction to improve HNP outcomes is acomparative advantage of the World Bank. 12Chapter 4 highlights the contribution of othersectors to HNP results: the extent to which10

INTRODUCTIONCountry Assistance Strategies (CASs) have beenused as a vehicle for generating synergies acrosssectors to improve health outcomes; the resultsfrom HNP lending operations that tried tocoordinate and bring to bear the actions of multiplesectors on improving outcomes; and thehealth impact of investments in the water supplyand sanitation and transport sectors.Chapter 5 assesses IFC’s health investments andAdvisory Services since 1997, the implementationof the 2002 health strategy, and the lessonslearned from IFC’s health investments and AdvisoryServices. Health is a relatively new area of investmentfor IFC, so the portfolio is small andrelatively young.Chapter 6 reflects on the value added of WorldBank and IFC support for HNP outcomes, as revealedby the experience of the past decade, andoffers recommendations for improving the effectivenessof both agencies of the World BankGroup.11

Chapter 2Evaluation Highlights• The Bank has committed nearly $17billion to HNP since 1997; $11.5 billionwas managed by the HNP sector.• Lending has shifted in favor of HIV/AIDS, multisectoral, and Africa Regionprojects.• Support and staffing for populationand nutrition have eroded.• About two-thirds of HNP projectsoverall—but only a quarter of HNPprojects in Africa—have had satisfactoryoutcomes.• Performance of HNP projects hasstalled, while it has improved in othersectors; complexity and low qualityat entry have contributed.• M&E remain weak, despite an increasein monitoring indicators andbaseline data.• The portfolio has a generally propoorfocus, but few projects couldshow improved outcomes for thepoor.

Family health clinics such as these in the Kyrgyz Republic, which cater to men, women, and children, increasingly replaced inefficient specializedcare as part of the Manas health reform program supported by the World Bank and other donors. Photo courtesy of Judyth Twigg.

Evolution and Performanceof the World Bank’sCountry Support for Health,Nutrition, and PopulationSince 1997, the World Bank has committed nearly $17 billion for HNP in605 projects, about 6.5 percent of all Bank commitments over that period.This includes projects managed by the HNP sector—about threequartersof all HNP commitments—and HNP components embedded inprojects managed by other sectors.Beyond this, the Bank has spent some $43 millionof its own budget and trust funds on economic andsector work (ESW) tasks in HNP that generated reports,policy notes, conferences, workshops, consultations,and country dialogue. 1 This chapterand the next focus on the commitments managedby the HNP sector, while chapter 4 assessesmultisectoral approaches and the HNP supportmanaged by other sectors.Two-thirds of HNP projects have had satisfactoryoutcomes, but the findings in this chapter pointto several challenges for the implementation of the2007 HNP strategy: the low performance of HNPsupport overall and in Sub-Saharan Africa, in partdue to the complexity of multisectoral operationsand sectorwide approaches; weak M&E that couldundermine the strategy’s results orientation andcommitment to improve governance; and a lackof evidence that the Bank’s HNP support is reallydelivering results to the poor.The Health, Nutrition, and PopulationPortfolio and Its PerformanceFrom 1997 through 2008, the portion ofWorld Bank HNP commitments managedby the HNP sector has amounted to about$11.5 billion in 255 projects. The number ofHNP projects approved annually has risen slowly,while new annual commitments have declined(figure 2.1). The projects managed by the HNP sectorare almost exclusively investment projects. 2IEG conducted an in-depth review of the lendingportfolio managed by the HNP sector and approvedduring the 10-year period from fiscal 1997through 2006, based on review of project appraisaldocuments (PADs), implementation completionreports (ICRs) for those that had closed,and tabulation of key characteristics. During thatperiod, 220 HNP-managed projects were approved,of which 110 had closed as of June 30,2008 (appendix C). This portfolio review formsmuch of the evidence base for the discussionhere, as well as in chapters 3 and 4.Improving health status was the most frequentlycited objective of HNP projects, followedby improving the access, quality,efficiency, or equity of the health care system,collectively cited by more than half ofthe projects approved from fiscal 1997through 2006 (table 2.1). A third of projects15

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 2.1: The Number of HNP-Managed Projects Has Risen Slowly, While Commitments HaveDeclined2,5002,0001997 HNPsector strategy504540US$ million1,5001,0005000Board approveshealth lending1970197119721973197419751976197719781979198019811982198319841985Commitments1986198719881989Fiscal year199019911992199319941995Projects1996199719981999200020012002200320042005200620072008Source: World Bank data.Note: The peaks in commitments in fiscal years 1996 and 1998 are due to a few projects in large countries—in 1996, one project each in Argentina, Brazil, India, Mexico, and Russia, and in1998, five projects in Bangladesh, Egypt, India, and Mexico. In both years the projects in these large countries accounted for 70 percent of commitments.35302520151050Number of projectsTable 2.1: Objectives of HNP Projects Approved in Fiscal 1997–2006Objective Number a PercentImprove health status 135 61Reduce the burden of communicable disease b 78 35Promote child growth/reduce malnutrition 21 10Reduce high fertility/promote family planning 8 4Improve access, quality, efficiency, or equity of the health system 126 57Improve access 70 32Improve quality 76 35Improve efficiency 61 28Improve equity 16 7Health system reform and financing 73 33“Health reform” 41 19Health financing 32 15Health insurance 16 7Decentralization 15 7Private sector 8 4Build/strengthen institutional capacity 68 31Improve management 39 18Improve participation c 26 12Learning 21 10Source: IEG portfolio review.a. Total projects = 220.b. Includes AIDS (29 percent); malaria (5 percent); TB (5 percent); and other communicable diseases such as leprosy, polio, and avian influenza (6 percent).c. Community participation and/or empowerment, multisectoral or intersectoral action.16

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTFigure 2.2: IDA and IBRD Commitments Declined; the Number of IBRD ProjectsAlso Fell, But IDA Projects Rose in NumberA. Commitments B. Number of ProjectsMillions of dollars3,5003,0002,5002,0001,5001,00050003,117Source: IEG portfolio review.2,376IDA2,5142,024Number of projects approved1003401997–2001 2002–2006 1997–2001 2002–2006Fiscal yearFiscal year80604020IBRD IDA IBRD Blend61358928aimed to reduce communicable disease, whileonly 1 in 10 had nutrition objectives and only 4percent had an objective to reduce high fertility.Health reform–related objectives collectively wereaddressed by a third of the projects.Trends in the Level and Composition of Health,Nutrition, and Population Support, Fiscal1997–2006Although IDA and IBRD commitments fell,the number of IDA projects rose, resulting ina larger number of small projects in the portfolioby the end of the decade. IDA’s share ofcommitments remained about the same, accountingfor 55 percent of total commitments inthe second period (figure 2.2A). The total numberof HNP projects approved increased from 99 to 121between fiscal 1997–2001 and 2002–06. All of thisincrease was created by IDA project approvals; thenumber of IBRD projects declined (figure 2.2B). 3There were other major developments inthe composition of the HNP portfolio overthe decade. Africa’s share of HNP project approvalsincreased from more than a quarter in fiscalyears 1997–2001 to more than a third in fiscal2002–06. The number of new HNP projectsin Africa rose by more than 60 declined over the period,While commitmentspercent (from 28 to 45), and in Latin the number of projectsAmerica and the Caribbean by 30 percent(from 20 to 26). The share of com-share in Africa.increased, as did themunicable disease and multisectoralHNP projects also increased dramatically. Theincrease in Africa Region, communicabledisease, and multisectoral projects reflecteda rise in HIV/AIDS project approvals; in thesecond half of the period, the HIV/AIDS projectsreached nearly 40 percent ofthe HNP portfolio. In line with internationalefforts to improve donor harmonizationand alignment, the shareof HNP projects financing SWAps inhealth also climbed. 4Although more HNP projects were approvedduring the second half of the decade, boththe share and the absolute number of projectsfinancing health reform–type objectivesdeclined (table 2.2, lower panel). 5 Thisincludes projects with objectives of reform, financing,insurance, decentralization, and governmentactions with respect to the private healthCommunicable diseaseprojects rose to 44 percentof approvals, while healthreform objectivesdeclined.17

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONNumber of HNP staffTable 2.2: Key Developments in the HNP PortfolioFigure 2.3: The Number of Sector Specialists Roseover the Decade100806040200Sector specialistsEconomistsOperations, program,and project officers1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Source: Nankhuni and Modi 2008.Fiscal yearPercent of the portfolioDevelopment 1997–2001 c 2002–06 cShare increasedIDA projects a 62 74Projects in Africa 28 37Multisectoral HNP projects 25 49Free-standing communicable disease projects 25 44HIV/AIDS projects b 17 39Projects supporting SWAps 9 15Share decreasedProjects with health reform/financing/insuranceobjectives 44 24Projects with objective to improve health carequality 49 22Projects with objective to improve efficiencyand cost-effectiveness of health care 37 20Projects with nutrition objectives 12 7Projects with population objectives 6 2Number of projects approved 99 121Source: IEG portfolio review.a. The share of IDA commitments was unchanged.b. Projects with an objective to prevent the spread of HIV/AIDS or mitigate its impact.c. Fiscal year of approval.sector. The share of projects with explicit objectivesto improve the quality and efficiency ofhealth care also dropped.Lending to reduce fertility, populationgrowth, or unwanted births remained lowand became scarcer, dropping from only 6percent of the portfolio to 2 percent (table2.2, lower panel). Over the decade, only 14 projectswere approved with objectives to lower fertilityor increase use of contraception, or with family planningcomponents. The Bank’s population supportwas directed to only about a quarter of the 35countries identified by the strategy (World Bank2007g) as having a total fertility rate of 5 or higher(appendix E).Lending with objectives to improve nutritionalstatus dropped by half, from 12 to 7 percentof the portfolio (table 2.2, lower panel).Bank support to reduce malnutrition was directedto only about a quarter of countries with childstunting of 30 percent or more; malnutrition is asignificant problem among the poor in manymore countries, though its causes are diverse(appendix F).Total staffing in the sector rose, particularlythe number of health specialists. Totalprofessional staff 6 affiliated with the HNP sectorgrew by a quarter, from 136 in fiscal 1997 to a peakof 185 in fiscal 2004, and then had declinedslightly, to 169, by fiscal 2007 (appendix H). Virtuallyall of this increase was among sector specialists,who increased by about 40 percent; thenumber of economists remained about the sameand the number of operations officers declined(figure 2.3).While the overall number of sector specialistsrose, the count of specialists in nutritionand population declined. The number of populationspecialists plummeted, from 24 in 1997 to7 in 2003, while the number of nutrition specialistsdwindled from 8 to 5 over the decade (appendixH). 7The increase in staff affiliated with the Regionswas not in proportion to the increase18

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTin HNP projects. The number of HNP projectsin Africa increased by 60 percent from the first tothe second half of the decade, while the numberof HNP staff affiliated with the Region rose byonly about 12 percent (appendix H). The numberof HNP projects in the Latin America and CaribbeanRegion rose by 30 percent, and HNP staff roseby half. The number of HNP staff affiliated withEurope and Central Asia more than doubled, yetthe number of HNP projects in the two periodsremained about the same.Performance of Health, Nutrition,and Population SupportOver the past decade, about two-thirds ofcompleted HNP-managed projects had satisfactoryoutcomes. The performance ofHNP projects has stagnated, while the outcomesof projects in other sectors have continuedto improve (figure 2.4). 8 The growing gapbetween the performance of HNP and othersectors is not due to underlying shifts in the compositionof the portfolio. The share of developmentpolicy loans relative to investment projectsover time within the HNP sector and other sectorswas stable. 9 According to the 2008 Annual Reviewof Development Effectiveness (IEG 2008a), projectoutcomes in Africa trail those in other Regionsacross the board, not just in the HNP sector. 10The share of Africa Region projects among closingprojects also does not explain the divergencein outcomes between HNP and other sectors: theshare of Africa Region projects in other sectors declinedfrom 28 to 24 percent, but by an evenlarger percentage among exiting projects in theHNP sector, from 35 to 26 percent. 11 If one comparesthe outcomes of HNP lending to lending inall other sectors—excluding projects in Africafrom both—the performance of the HNP portfoliodeclines from 78 to 74 percent satisfactory betweenthe first and second half of the period,while the performance of other sectors collectivelyrises from 78 to 81 percent satisfactory. 12The quality of project preparation by the Bankand the borrower’s overall performance in preparingand executing the project strongly predict betteroutcomes. The Bank’s performance, and to alesser extent the borrower’s performance, alsoFigure 2.4: Performance of HNP Projects HasStagnated, while Outcomes in Other SectorsContinue to ImprovePercent of projects with outcomes ratedmoderately satisfactory or higher100806040200Source: World Bank data.6336(n = 14)shows a widening gap between HNP Support to reduceand other sectors (appendix H). 13 Multivariateanalysis of the outcomes of support to addressmalnutrition was low;HNP projects approved during fiscal high fertility nearly1997–2006 14 found that the performanceof the borrower (which includesdisappeared.government and the implementing agency),as well as the Bank’s role in quality at entry,are strongly correlated with better outcomes(appendix H). Eighty-three percent ofprojects with satisfactory borrower performanceultimately achieve satisfac-particularly healthStaffing in the sector rose,tory outcomes, compared with only 7 specialists.percent of projects with unsatisfactoryborrower performance. Seventy-six percent ofprojects with satisfactory quality at entry achievesatisfactory outcomes, compared with only 19percent of those with unsatisfactory quality atentry.Given the importance of quality at entryin determining outcomes, it is worrisometo note that both the costs andduration of project preparation droppedprecipitously over fiscal 1997–2001 and67641987–91 1992–96 1997–2001 2002–06Fiscal year project closedOther sectorsHNP sector7270While performance inother sectors hasimproved, HNP outcomeshave stagnated.796819

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONThe borrower’s have only partially recovered to previouslevels (appendix H). In contrast, av-performance and thequality of project erage annual supervision costs for HNPpreparation strongly investment projects have been risingpredict outcomes. since 2001, as have supervision costs forinvestment projects in other sectors,with HNP supervision costs exceeding those inother sectors in fiscal 2004–06 (appendix H). 15Projects with unsatisfactory outcomes are morelikely to cite certain Bank performance factorsthan are projects with satisfactory outcomes: inadequaterisk analysis (relative risk for unsatisfactoryprojects = 10:1); 16 inadequate technicaldesign (8:1); inadequate supervision (5:1); inadequatepolitical or institutional analysis (3:1); lackof baseline data that can be used to set realistictargets (3:1); overly complex designInadequacies in risk (2:1); and an inadequate M&E frameworkor poor data quality (2:1, appen-assessment, politicalanalysis, technical design, dix H). In addition, 12 percent of unsatisfactoryprojects cited inadequate priorand supervision areamong the factors analytic work, while none of the satisfactoryprojects cited this factor. IEG’slowering outcomes.recent evaluation of Bank-wide analyticand advisory activities found that prior analyticwork led to higher project quality at entry(IEG 2008h). Fieldwork for this evaluation confirmedthe value added and influence of analyticwork in the case study countries (box 2.1).HNP projects in Africa do not have lower performancesimply because of the concentrationof low-income countries in the Region. Only aquarter of the closed HNP projects in the AfricaRegion had satisfactory outcomes, compared withthree-quarters of the closed projects in other Regions.All closed projects in Africa were in lowincomecountries and were financed by IDA grantsor credits. However, HNP projects financed byIDA in other Regions managed to perform substantiallybetter, often as well as or better thanIBRD-financed HNP projects in middle-incomecountries (figure 2.5). The performance of AIDSprojects in Africa has been particularly weak—only18 percent have had satisfactory outcomes. Buteven if AIDS projects are excluded, only 27 percentof HNP projects in the Africa Region have hadsatisfactory outcomes. 17The complexity of HNP operations in Africaand among IDA recipients is contributing tolow outcomes. Multisectoral investment lending,in which multiple sectors are involved inmanaging and/or implementing activities, is associatedwith lower outcomes in IDA countries,but not in countries receiving IBRD support.SWAps 18 are also associated with lower outcomes,and most of them have occurred in Africa. Thus,the problem of highly complex operations inlow-capacity countries found in IEG’s 1999 HNPevaluation continues. The review of Bank performance,cited above, showed that while projectcomplexity raises the risk of low outcomes, somecomplex projects do have satisfactory results.Among the factors found not to affect HNP out-Box 2.1: Analytic Work Supported Better Outcomes in Four CountriesAnalytic work sponsored by the Bank—one of the majorforms of country support in addition to policy dialogueand lending—was highly influential in four of the countrieswhere IEG conducted in-depth fieldwork.In Nepal, a burden-of-disease analysis showing that70 percent of life years lost were taken by preventablecommunicable diseases that disproportionately affectthe poor helped the government to improve the propoorallocation of public health expenditures.In Malawi, the 2005 health public expenditure reviewled to a more realistic assessment of the resourceneeds for the first SWAp in health.In Peru, Improving Health Care for the Poor (Cotlear2000) provided the analytic framework and major recommendationsthat guided health reforms in that countryand was valued highly by government.Five major pieces of influential analytic work weresupported in preparation for the second health reformproject in the Kyrgyz Republic.Sources: Shaw forthcoming: Elmendorf and Nankhuni forthcoming; IEG 2009b, forthcoming.20

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTFigure 2.5: IDA Project Outcomes in Africa Are Much Lower than in Other Regions100100Percent of projects rated moderatelysatisfactory or higher806040202471786780 8075670African = 33 IDAn = 0 IBRDSouth Asian = 14 IDAn = 0 IBRDEurope andCentral Asian = 9 IDAn = 9 IBRDEast Asiaand Pacificn = 5 IDAn = 5 IBRDLatin Americaand theCaribbeann = 3 IDAn = 16 IBRDMiddle Eastand NorthAfrican = 0 IDAn = 3 IBRDSource: IEG portfolio review.Note: Two projects that had blended operations are excluded. Total: IDA = 64, IBRD = 33.IDAIBRDcomes were the source of finance (IDA versusIBRD, when Region is controlled for), the size ofthe support, the borrower’s population size, andthe year of approval (there was no time trendwhen other factors were controlled for).Although they comprised a small part ofthe overall portfolio, projects with populationand nutrition objectives also had lowerthan-averageoutcomes compared with therest of the portfolio. None of the projects withexplicit fertility or population objectives achievedthem: modern contraceptive use and fertilitywere barely affected in Guinea, India, Kenya, andMali. In Bangladesh, Gambia, and Senegal, fertilitydeclined somewhat, but it is doubtful that theresults stem from population programs or Banksupport. 19 Three projects with population andfamily planning components (but no populationobjective) partially achieved their objectives,showing that it is possible to raise contraceptiveuse in difficult environments when both demandand supply-side factors are addressed (box 2.2).Among the 15 nutrition projects that had closed,only 2—the Indonesia Iodine DeficiencyControl and the Senegal Nutri-projects often did notPopulation and nutritiontion Enhancement Projects—were able achieve their objectives.to substantially meet their objectivesand show a change in nutritional outcomes. Theperformance of the population and nutritionportfolios is discussed in greater depth in appendixesE and F, respectively.Complex designs, the lack of up-front riskand institutional analysis, and the absorptionof these services into a basic package alongwith other health care were all implicatedin the low results. Project complexity was particularlyacute in the nutrition projects, cited ascontributing to shortcomings in more than half ofthe weak-performing projects. Nutrition projectstend to involve several sectors in implementationand are often managed outside the Ministry ofHealth, in some cases by new or inexperienced institutions(as in Mauritania). In Uganda,the added complexity arose from expandedgeographical coverage of whatwas supposed to be a pilot activity.Project complexity in IDAcountries has contributedto low outcomes.21

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 2.2: Family Planning Can Be Successful in Difficult EnvironmentsSupport to Egypt and Malawi demonstrates that intensive effortsto raise contraceptive use and lower fertility in high-fertility ruralareas can show results.The Egypt Population Project (1996–2005) financed activitiesto increase the demand for smaller families and the supply ofcontraception in rural Upper Egypt. a Community development associationsdesigned and managed subprojects in 148 of the poorestvillages. The subprojects financed home visits, communityconferences on healthy behaviors, and microcredit and literacyclasses for women and promoted local maternal and childhealth/family planning services, in order to improve women’s familyplanning and reproductive health knowledge, stimulate demandfor smaller families, and raise modern contraceptive use.The project also improved the quality and supply of family planningservices offered in the health system in the project villages.The modern contraceptive prevalence rate (CPR) rose from 40 to45 percent between 2000 and 2005, compared with an increasefrom 56 to 59 percent nationwide, while the total fertility rate (TFR)declined by 0.8 children, from 4.7 to 3.9, compared with a nationaldecline from 3.5 to 3.1. Attribution of these results to the projectis difficult because M&E was limited and other donors (mainly theUnited States and the European Union) were supporting similarinterventions in the same part of Egypt. It was found that the projectcontributed to the collective impact of these efforts and thatits design influenced other donors in adopting demand-sideinterventions.The Malawi Population and Family Planning Project (1999–2004) aimed to demonstrate that a community-based distributionapproach to family planning could raise contraceptive use inthree high-fertility pilot areas in rural Malawi. Public sectorcommunity-based distribution agents (CBDA) provided familyplanning counseling and contraceptives and referred clients forlong-term or permanent methods. Information, education, andcommunication activities aimed to generate demand. The percentof contraceptive users with CBDAs as their source rose from 1to 24 percent and contraceptive use rose by twice as much inthe pilot districts as in the three matched control districts (seefigure, below). While the evaluation demonstrated that thiscommunity-based approach, emphasizing both demand and supply-sideactivities, could raise contraceptive use in high-fertilityrural areas, the model was not replicated more widely. Certainlessons were incorporated into the government program, but by2004, the government and donors were fully engaged in launchinga health SWAp.The Modern CPR Increased in Pilot Districts in Rural Malawi, Relative to ControlDistricts, 1999–200350Modern CPR (percent)40302018.322.726.038.526.231.820.535.729.935.422.632.624.729.723.535.6100Karonga (C)Chitipa (P) Mulanje (C) Chiradzulu (P) Dowa (C) Ntchisi (P) All controldistrictsAll pilotdistrictsMatched control and pilot districts1999 2003Sources: IEG 2008b; Elmendorf and Nankhuni forthcoming.a. The changes in the CPR and the TFR in rural Upper Egypt are both statistically significant.22

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTComplexity was also cited as a factor in poorperformingpopulation projects, in addition tolack of up-front risk analysis and mitigation and institutionalanalysis. Absorption of family planningprograms into basic packages of services supportedby SWAps and health reform projects alsomay have contributed; field visits in Egypt confirmedthe finding of a previous study that emphasison a basic package of services had oftenreduced the availability and quality of family planning(appendix E).Results from a Detailed Implementation Reviewof five health projects in India suggestthat even projects that meet their objectivesmay be performing at substantially lower levelsthan their outcomes would suggest. 20 Thereview found that the Bank often relies on the borrower’sreporting systems to confirm the deliveryof such key outputs as civil works and equipment,without independent verification by the Bank of thequality of project goods and services and their delivery.For example, more than half of the pieces ofequipment procured for the Food and Drug CapacityBuilding Project were not delivered or not installed,21 and “severe construction deficiencies”were found in the Orissa Health Systems DevelopmentProject in buildings that the constructionsupervisors had reported to be complete and performingaccording to specification. 22 The reviewfound that “supervision generally did not involvecomprehensive site visits or physical inspections.” 23It is not known how representative these Indiaprojects are of the overall HNP portfolio, but thereis an important lesson here about the need to conductsupervisory field visits to verify the implementationdata provided by the borrower. Bettersupervision has the potential to enhance the impactof Bank support on outcomes, if addressed.To summarize, while the overall levels of lendingin HNP have not changed much over the pastdecade, the composition of the portfolio hasshifted rather dramatically toward communicabledisease projects, particularly AIDS; projects inthe Africa Region; and, to a lesser extent, supportfor SWAps. Support for health reform (as measuredby the projects’ stated objectives) has declinedbut is still significant, particularlyin middle-income countries, while missing some keyBank supervision issupport for population and nutrition, implementationalready low, has declined further. Performanceof the lending portfolio hasdeficiencies.stalled, with only about two-thirds of projectsshowing satisfactory outcomes. While some ofthe activities are inherently risky, several factorsthat can be addressed were implicated—excessivecomplexity, particularly in low-capacity environments;lack of institutional analysis and relevantanalytic work linked to lending; inadequate baselinedata and M&E; and inadequate assessment ofthe risks and mitigation strategies.Monitoring, Evaluation,and the Results AgendaThe 2007 HNP strategy highlights deliveringresults and improving governance, both ofwhich depend critically on strong M&E forsuccess. Two previous IEG evaluations of HNPand HIV/AIDS support concluded that weak M&Ehad contributed to lower efficacy and less learning.The 2007 HNP strategy acknowledges theseshortcomings and maintains that inadequate M&Eprecludes a thorough analysis of the efficacy ofBank support over the past decade.An increasing share of projects An increasing share ofsince 1997 has had monitoring indicatorsand baseline data at the monitoring indicatorsprojects since 1997 havetime they are appraised, in part becauseof changes in the require-and baseline data.ments for PADs. Around the timethat the 1997 HNP strategy was approved, a required“logical framework” was added to thePAD. In fiscal 1997–98, all active projects wereretrofitted with performance indicators. Thus, itshould not be a surprise that HNP projects approvedin 2007 had more indicators and weremore likely to have baseline data than projects approvedin 1997 (table 2.3). 24Despite these improvements, too few projectshave baseline data at appraisal. In fiscal2007, the share of HNP projects that planned tocollect baseline data after the project was approvedwas about the same as a decade earlier (4023

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable 2.3: More Project Appraisal Documents Have Baseline Data, but There IsStill Some Distance to Go, Fiscal Years 1997 and 2007Percent of projects approved/fiscal yearCharacteristic 1997 2007Baseline values for all outcome indicators 14 47Baseline values for none of the outcome indicators 71 27Targets set for indicators 43 80Baseline data already collected at the time of the PAD 14 27Planned to collect baseline data after approval 43 40Number of projects 14 15Source: IEG portfolio review.Yet more than half of versus 43 percent). A quarter of projectsapproved in fiscal 2007 still had noindicators have nobaseline value and baseline values in the PAD, and only40 percent of projects about half had baseline values for allplan to collect baseline outcome indicators. A recent analysisdata after they are of 12 projects in South Asia found thatapproved. baseline data were collected for only 39percent of indicators and that therewas little evidence of improvement in M&E betweenprojects approved before and since 2001(Loevinsohn and Pande 2006).M&E of recently closed HNP projects—almostall of them approved since fiscal1997—has been weak. Since mid-2006, IEGhas been systematically rating the performance ofclosed projects on M&E, with a “quality of M&E”rating, based on M&E design, implementation, anduse of the data. It is rated on a four-point scale:negligible, modest, substantial, or high. The M&Eof 45 projects managed by the HNP sector hasbeen reviewed since mid-2006; M&EA significant number of was substantial or high for only 27 percentof the projects, slightly lower thanthe Bank’s ongoingimpact evaluation the rate Bank-wide (35 percent), whichinitiatives are addressing is also quite low (figure 2.6). 25 NearlyHNP themes. half of the ICRs for these HNP projectshad no baseline data when theprojects were approved, and baseline data werenever collected at all for five projects.Lack of M&E has had consequences for thedesign and efficacy of projects and for improvedgovernance, an objective of the 2007HNP strategy. According to a review of ICRs ofclosed projects with low M&E ratings, the lack ofbaseline data has reduced the relevance and feasibilityof projects’ objectives and design. Unrealistictargets were set—either too high, or belowlevels later found to prevail at the start of theproject. Weak M&E not only makes it difficult toassess the effectiveness of activities, but also contributesdirectly to lower efficacy and efficiency becauseit limits opportunities for learning andfine-tuning implementation for better results.The Bank has launched major initiatives inimpact evaluation, and a significant numberof these address HNP themes. These includethe Development Impact Evaluation Initiativefrom the Bank’s Development Economics Department;the Spanish Trust Fund for Impact Evaluation,which is dedicated to the evaluation ofinnovative programs to improve human development;and the Africa Impact Evaluation Initiative.In the databases for these initiatives, IEGidentified 101 ongoing or complete HNP impactevaluations—most with an experimental design. 26It is too early to assess the results and long-runimpact of the initiatives in institutionalizing orproviding greater incentives for evaluation. Moreover,many of the activities and reforms supportedby the Bank are not amenable to therandomized design or even quasi-experimental designsused by impact evaluations. The recentlylaunched Africa Results Monitoring System willprovide public access to information from bothcountry and Bank systems on data from countries24

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTFigure 2.6: Only a Third of Projects Bank-Wide, and Only a Quarter of HNP Projects,Have Substantial M&E100Percent of projects with M&E ratedsubstantial or high806040200Economic Policy (n = 21)Energy and Mining (n = 33)Environment (n = 29)Transport (n = 45)Social Protection (n = 27)Agriculture and Rural Development (n = 54)Financial and Private Sector Development (n = 28)Public Sector Governance (n = 34)HNP (n = 45)Sector Board27Social Development (n = 10)Education (n = 49)Water Supply and Sanitation (n = 19)Poverty Reduction (n = 13)35Urban Development (n = 15)All sectors (n = 425)Source: Villar Uribe forthcoming.and World Bank operations from national accounts,administrative sources, and surveys.Notwithstanding these initiatives, a largeshare of the Bank’s HNP support financespilot interventions or programs, or intendsto evaluate the impact of a specific activityor program, but few do. Sixty-five projects—about 30 percent of the total HNP portfolio—were labeled as pilot projects, had pilot interventionsor pilot regions, had an objective to testan approach or intervention, declared an intent toevaluate the impact of an intervention or program,or had an intent to conduct an impact evaluation,according to the PAD. 27 The objective wasgenerally to test the effectiveness or acceptabilityof an activity, and on this basis improve or expandit. However, only 17 projects mentioned a controlgroup, 12 described an evaluation design in thePAD, 8 had baseline data in the PAD, and 7 identifiedan explicit control group or planned an economicanalysis (figure 2.7). 28All of the pilot and impact evaluationsthat were eventually con-HNP support financesMuch of the Bank’sducted had an evaluation design in pilot interventions orthe PAD. About half of the 65 approved intends to evaluateprojects with planned pilot or impact impact, but few projectsevaluations had closed. Only 7 of them actually do so.had a detailed design in the PAD, andonly 4 actually conducted the proposed evaluationof a pilot (figure 2.7), including an impactevaluation of early childhood interventions inthe Philippines (box 2.3). 29 Among the 25 pilotand impact evaluation projects thatdid not have an evaluation design describedin the PAD, none conductedthe planned evaluation.There is great scope for improving the efficacyof the Bank’s HNP support by incorporatingrigorous evaluation into projectsthat already have some pilot feature. Projectscan also do more to ensure that public policy ismaking a difference by incorporating periodicThe evaluations thatdid occur had a designin the PAD.25

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 2.7: The Discrepancy Between Plans, Evaluation Design, and Implementation of Pilotsand Impact Evaluations in HNP Projects Approved in Fiscal Years 1997–2006706560Number of projects5040302010032Projects with pilotor impactevaluation178Mentioned aplan for acontrol group127Had a detailedevaluation designin the PAD85Had baselinedata in thePAD7Identified anexplicit controlgroup763 4Planned aneconomicanalysisConducted theevaluationSource: Villar Uribe forthcoming.Approved projectsClosed projectsBox 2.3: Early Childhood Interventions Improved Cognitive Development and Nutritional Statusin the PhilippinesThe objective of the Philippines Early Childhood Development(ECD) Project (fiscal 1998–2006) was to ensure the survival andpromote the physical and mental development of Filipino children,particularly the most vulnerable and disadvantaged. The projectimplemented expanded immunization, integrated management ofchildhood illness (IMCI), prevention and control of micronutrientmalnutrition, a service to inform parents on growth monitoring andECD, and a first grade early childhood education/development programin three regions. The project did not introduce new servicesbut sought to integrate and deliver existing ones through centerandhome-based interventions.The impact evaluation conducted longitudinal surveys annuallyfrom 2001 to 2003 on 6,693 children aged 0–4 in randomly selectedhouseholds of two project experimental regions, and onecontrol region without the interventions. The design includedbaseline and endline surveys: a general population-based samplesurvey of ECD status and services; nutrition surveys integratedinto the National Nutrition Surveys; surveys of ECD-relatedknowledge, attitudes, and practice; and provincial-level surveysof ethnic groups. The end-project survey in 2003 repeated the 1998baseline survey in full.The evaluation found a significant improvement in the cognitive,social, motor, and language development and the shorttermnutritional status of children residing in ECD program areascompared with those in nonprogram areas, particularly for childrenunder four. The share of children under four with worms anddiarrhea was significantly lower in program areas than in nonprogramareas, but higher for older children. These results, as wellas other studies sponsored by the project, helped guide improvementin ECD interventions, build capacity for evidencebasedpolicymaking, and assured the continuation of the ECDprogram.Source: Armecin and others 2006.26

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTevaluation of specific elements into the projectdesign and making continued finance of thoseelements contingent on their evaluation. For example,the efficacy of training programs in changingprovider behavior can be measured. It is alsoimportant to collect, in parallel with the resultsframework, information on other key determinantsof outcomes that are outside of the project.One of the main reasons for lack of M&E throughoutoperations is said to be a lack of incentives.The project appraisal and approval process is onesuch incentive that does seem to be operating—projects that do not have evaluation designs forpilot projects in the PAD, or those that do not havebaseline data collected at the PAD stage, often donot implement them. Making project approvalcontingent on evaluation designs for specific elementsof a program, and making baseline dataa requirement, should result in more frequentevaluation.To summarize, logframes and results frameworks,such as the one in the 2007 HNP strategy, have createdgreater incentives for monitoring inputs,outputs, outcomes, and impacts and offer guidanceto select the right indicators. However, theincentives are still not sufficient to ensure adequateincentives for baseline data and evaluation,which is key to understanding effectivenessand impact. Strengthening M&E is one of the keyelements of achieving the strategy’s objective ofbetter governance in the sector.Is Health, Nutrition, and PopulationSupport Reaching the Poor?In keeping with the Bank’s mandate forpoverty reduction, improving health outcomesamong the poor is one of the majorobjectives in the 2007 HNP strategy, as it wasin the 1997 strategy. Numerous studies haveshown that it would be dangerous to assume thatsimply expanding health services would extendbenefits primarily to the poor; instead, most havefound a strong bias in government health expenditurein favor of the non-poor (box 2.4).Over the past decade, most CASs and about halfof the health portfolio had a generallypro-poor focus, but evidence that thepoor actually benefited is weak. Thenumber of country strategies and projectsthat set explicit objectives regardingthe poor is small, and the evidencethat the poor have benefited absolutelyor relative to the non-poor is thin.Making project approvalcontingent on evaluationdesigns and baseline datashould result in morefrequent evaluation.Nearly three-quarters of CASs specifically discussedhealth issues among the poor (Sinhaand Gaubatz 2009). Attention to the poor increasedslightly from the first to secondhalf of the period—from 70 to 79 percentof CASs. However, only half artic-CASs discussed healthNearly three-quarters ofulated a definition of the poor, and issues among the poor.fewer than a quarter cited a method toreach them. Even among the subset of CASs thatemphasized health issues for the poor, only 15 percentset explicit targets for them. While this shareis low, it has increased over time, from 11 to 21percent.While two-thirds of HNP project support isfor the poorest countries, only about half ofthe lending portfolio could be said to havea poverty focus within countries. This wouldinclude projects that explicitly target the poor intheir objectives, those that have an objective toimprove access to health services more generally,those that seek to prevent and control communicablediseases known to disproportionatelyaffect the poor, 30 and projects Only about half of thewith objectives to improve health outcomesof the population but with a de-said to have a povertylending portfolio could besign that favors poor geographic areas focus within countries.or services for the poor. Using this definition,about half of the portfolio (107 projects)could be said to have a pro-poor focus. This definitionof poverty focus could be overly generous,as one could argue and it has been shown that increasingaccess to services often disproportionatelybenefits the non-poor (World Bank 2003b;Gwatkin, Wagstaff, and Yazbeck 2005). At the sametime, the PADs for almost all HNP lending operationsdiscuss the welfare of the poor or are linkedto poverty reduction, even if the project is not ex-27

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 2.4: Is Public Health Spending Pro-Poor?Assuring that the poor have access to health services is often anobjective of government spending, but studies of the incidenceof government health expenditure in developing countries haveconsistently found that the wealthiest quintiles generally benefitthe most from public health subsidies. aIn Ecuador, for example, the poorest quintile received 8 percentof health subsidies compared with a 38 percent share for therichest quintile. b In Uzbekistan the poor received 13 percent ofhealth spending, while the rich got 39 percent. bA comparison of 11 Asian countries found a pro-rich bias in 9countries and that the bias was particularly acute in the lowestincomecountries. c In most cases there was a strong pro-rich biasin the distribution of hospital care; there was less pro-rich bias inbasic health services and in higher-income countries, wherethere are private alternatives and the rich can self-select out of thepublic sector.Aggregation of findings in countries across Regions suggestthat the bias of public spending against the poor is most acutein some of the lowest-income Regions—Sub-Saharan Africaand South Asia—while the bias is also strong in Europe andCentral Asia (see figure below). Even in Latin America, where governmenthealth spending is pro-poor on average, there are countrieswhere this is not the case.Government Health Expenditure in Most Regions Is Biased toward the Non-Poor35Percent of government expenditure302520151050Latin Americaand the CaribbeanSouth AsiaEast Asiaand PacificEurope andCentral AsiaSub-SaharanAfricaPoorest quintileRichest quintileSource: World Bank 2008a, based on calculations from Filmer 2003.a. For example, Van de Walle 1995; Castro-Leal and others 1999; Mahal and others 2000; Sahn and Younger 2000.b. Filmer 2003, table 1, p. 1.c. O’Donnell and others 2007, table 2, p. 100.pected to yield direct benefits for the poor in theshort run.A remarkably small share of projects had objectivesto improve health outcomes amongthe poor. Only one in eight projects (13 percent)had an objective to target health status, access, use,quality, or demand, or to provide health insurancespecifically among the poor (table 2.4). 31Beyond this, an additional 7 percent of projects hadan objective to improve equity, most often expressedin terms of equity in the distribution ofresources in the health system, in access to healthservices, or in health status. 3228

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTTwo-thirds of projects with objectives to affectoutcomes for the poor used geographictargeting in design—that is, they aimed toimprove outcomes among the poor by implementingactivities in poor regions. About1 in 10 planned to reach the poor by providing servicesthat the poor would use disproportionately(basic health package, nutrition, primary healthcare, communicable disease control), and 14 percentplanned to target individual householdsusing administrative data. The remainder plannedto reach the poor by targeting specific populationgroups presumed to be poor or making healthcare more affordable. Surprisingly, 7 percent of theprojects with objectives targeting the poor advocatedaddressing the needs of the whole population—thatis, there were no plans to ensurethat the poor would be reached.Very few of the closed projects with pro-pooror equity objectives were able to demonstratean improvement in HNP outcomesamong the poor. Among the 108 closed projectsthat have ICRs, 12 (11 percent) had objectives targetingthe poor. 33 Of these, only 2 projects collectedoutcome data in both the (poor) projectand control areas, and in both cases improvementsin health were shown across the board. Inhalf of the cases, data were collected only in projectareas, with no information on comparisonareas (or, in two cases, the comparison was nationaldata). In two cases, no outcome data werecollected at all. In a project in Bangladesh, wherethe targeting mechanism was expansion of an essentialservices package that would most benefitthe poor, all of the outcome targets improved atthe national level, but no data were collected torepresent access of the poor. Whenever healthoutcomes were measured, the results showed animprovement. However, the attribution to projectoutputs in these cases was weak. Without informationon what happened elsewhere, it isdifficult to attribute these improvements to theprograms supported by the Bank.Fieldwork for the evaluation revealed bothsuccess and challenges in reaching the poor.In Egypt, the Schistosomiasis Control Project successfullyreached the poor because the diseaseTable 2.4: Few HNP Projects Have Objectives ThatExplicitly Mention the PoorObjectivedisproportionately affects the rural poor Of projects with objectivesin areas with limited water management to improve outcomes forinfrastructure. The Population Project the poor, two-thirds usednoted earlier targeted low-income rural geographic targeting tocommunities in Upper Egypt, where fertilityrates were higher than the nationalreach the poor.average. The Eritrea HIV/AIDS, Malaria, SexuallyTransmitted Diseases, and TB (HAMSET) ControlProject asserted that community-managed subprojectswould benefit the poor, but there was nospecific effort to monitor the incidence of benefitsamong the poor, and the TB control activities,thought to disproportionately benefit the poor,had the weakest performance. The Basic Healthand Nutrition and Health Reform Projects in Peruhad explicit objectives to improve the health ofpoor women and children over a decade. Healthinsurance was extended to the poor nationwide.Yet only average health outcomes could betracked, and these were not even forthe specific areas targeted by the projects.The links between health reformprojects more generally and benefits forthe poor were complex and uncertain(box 2.5).About two-thirds of poverty assessmentsdelivered over fiscal years 2000–07 featureda chapter or subchapter on health. An importantvalue added of Bank support for HNPanalytic work is in drawing the link between healthand poverty and policies to address both. However,about a third of poverty assessments haveno substantial discussion of health, as evidencedby a health chapter or subchapter, and the sharePercentAny specific HNP objective mentioning the poor 13Improved health status among the poor 6Increased access to health care among the poor 5Improved quality of health care for the poor 4Increase in demand or utilization of services by the poor 3Providing health insurance for the poor 2Source: IEG portfolio review.Note: Projects approved from fiscal 1997 to 2006.Few projects with propooror equity objectiveswere able to demonstrateimprovements amongthe poor.29

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 2.5: Links Between Health Reform Projects and the Health of the Poor Were Complexand UncertainThe first and second Health Reform Projects in the Kyrgyz Republic(1996–2006) improved the efficiency of the health system, butwere less successful in redistributing funds in favor of the pooror addressing their health needs. Primary care was strengthened,broadened, and made more available, with clear improvementsin access to care for the poorer populations. Copayments weremade more predictable, and the outpatient drug package madeprescription drugs more affordable. However, the guaranteedbenefits were not universally implemented because of a shortageof funds. The centralization of fragmented pooling arrangementsshould have enhanced opportunities for efficiency andcross-subsidization, but it is not clear that this actually benefitedthe poor. During the second project, anticipated redistribution ofresources from relatively rich and over-serviced Bishkek to thepoorer oblasts did not occur. Neither project tracked health outcomesamong the poor.The Egypt Health Reform Program (1998–present) intended toimprove the health of the poor, yet chose to concentrate initiallyon relatively affluent governorates to increase the chances of success.The poor within these areas would benefit by rationalizinghealth infrastructure investment with an emphasis on underservedneighborhoods. But fewer than 40 percent of facilities followedthe pro-poor rationalization guidelines; positive gains wereundermined by enrollment and service fees without proper mechanismsto exempt the poor. Public-sector health providers interviewedby IEG had a vision of a competitive market where potentialusers preferred other options (nongovernmental organizations orthe private sector) on quality grounds; they saw it as their roleto compete with these providers, as opposed to enrolling and servingthe poor. Concern for the failure to enroll the poor was notvoiced until 2004; tracking of the enrollment of the poor was notadded as an indicator for the project until late 2007.Sources: IEG 2008d, Gonzalez forthcoming.About two-thirds ofpoverty assessments overfiscal 2000–07 hadsignificant sections onhealth; few addressednutrition or population,which greatly affectthe poor.The poverty orientation ofHNP analytic work hasdeclined over the pastdecade.of poverty assessments with a healthfocus declined from 80 percent overthe period fiscal 2000–03 to 58 percentover 2004–07 (figure 2.8).Very few poverty assessments addressnutrition or population. Only19 percent of poverty assessments hada nutrition chapter or subchapter, andthis percentage declined by more than half overtime, from 28 percent in fiscal 2000–03 to 12 percentin fiscal 2004–07. The treatment of populationwas even lower: only 8 percent of povertyassessments had a chapter or subchapter on population,fertility, or family planning. Of the 20poverty assessments conducted for countrieswith a total fertility rate of five or more, only onehad a subchapter on population (Mozambique,total fertility rate of 5.2).The inventory of analytic work assembledby IEG also shows a declinein the poverty-orientation ofHNP analytic work over the pastdecade. Overall, poverty-related analytic workaccounted for a third of all analytic work duringfiscal 1997–2006 if multicountry studies are treatedas separate studies, and about a quarter if multicountrystudies are treated as a single study (table2.5). 34 Irrespective of the inclusion of multicountrystudies, the share of poverty-related analyticwork declined.To summarize, the Bank has supported HNP outcomesin developing countries over the pastdecade in an environment of shifting internationalpriorities and approaches. Composition ofthe lending portfolio has shifted in favor ofHIV/AIDS and multisectoral projects, and a greatershare of projects in Africa.The shift of the Bank’s lending for communicablediseases has added to a huge international influxof resources for HIV/AIDS, TB, and malaria.The share of projects with health system reformobjectives has shrunk both relatively and absolutely,though global resources for these objectivesdo not seem to have increased nearly as30

EVOLUTION AND PERFORMANCE OF THE WORLD BANK’S COUNTRY SUPPORTmuch. Attention to nutrition has remained relativelylow; attention to population dropped fromlow to negligible, but appears to be reversing.About half of the lending portfolio is focused explicitlyor implicitly on improving health for thepoorest people within countries, yet accountabilityfor actually delivering health results is lacking.We know very little about the success of Bank supportin delivering on its institutional mandate ofpoverty reduction within the HNP sector. Attentionto HNP in poverty assessments appears to beslipping.Only two of every three HNP projects have satisfactoryoutcomes, and there has been no improvement,in part because of the increasedcomplexity of projects in countries with low implementationcapacity, as well as shortcomings ininstitutional and risk analysis. M&E remain weakand evaluation is scarce, even in projects withevaluation objectives or pilot activities that wereplanned for evaluation.The Bank’s HNP support in many countries hashad significant positive impacts, and additional exampleswill be highlighted in the chapters that follow.Yet these findings for the portfolio at largePercent of poverty assessmentsFigure 2.8: The Share of Poverty Assessments with aFocus on HNP Declined1008060402008058Source: IEG review of analytic work.Health Nutrition PopulationFiscal years 2000–03 (n = 50) Fiscal years 2004–07 (n = 72)pose major challenges to the sector in turningaround performance and in pursuing the 2007HNP strategy, with its aim of delivering resultson health outcomes among the poor, health systemperformance, and better governance.281277Table 2.5: The Poverty Focus of HNP Analytic Work DeclinedFiscal years1997–2001 2002–06 1997–2006Multicountry treated separatelyNumber of publications 559 899 1,458Share poverty-related (%) 52 20 32Multicountry treated as oneNumber of publications 380 899 1,279Share poverty-related (%) 29 20 23Source: IEG inventory of analytic work.Note: Analytic work includes official economic and sector work, research, and Working Papers and publications by Bank staff.31

Chapter 3Evaluation Highlights• Results of support for TB, malaria,schistosomiasis, and leprosy controlhave often been substantial; thosefor HIV/AIDS control have been moremodest, limited by project complexity,especially in Africa.• Excessive earmarking of donor supportfor diseases can distort sectorresource allocations.• Health system reform projects—about a third of the portfolio—areconcentrated in middle-income countriesand tend to perform less wellthan other HNP projects.• Failure to assess and address thepolitical economy of reform has reducedthe performance of healthreform support.• Support for SWAps has strengthenedsector capacity, but M&E remainsweak, and the impact on transactioncosts is unclear.• There is no necessary relationshipbetween adopting the SWAp approachand better HNP outcomes.

Eritrean mother tends to her child, shielded by an insecticide-treated mosquito net. Eritrea made enormous gains in malaria control over theevaluation period. Photo courtesy of the Eritrea National Malaria Control Program.

Lessons fromThree Approaches toImprove OutcomesSince 1997, the Bank has supported a number of approaches to improvehealth status and health system performance that figure prominently inthe action plan of the 2007 HNP strategy. The rapid increase in supportfor communicable disease control was not only a strategy for improvinghealth outcomes among the poor, but also has led to efforts in the action planto ensure synergy between infectious disease programs and health systems.Experience with health reforms and sectorwide approaches (SWAps) are atthe heart of improving the performance of health systems, a second majortheme of the 2007 strategy.This chapter presents findings and lessons fromBank support for communicable disease control,the reform of health systems, and health SWAps.While discussed separately, it is important to notethat these efforts are not mutually exclusive, andBank support for one or more may be found inthe same country at the same time.Communicable Disease ControlEighty percent of the 15 million peoplewho die every year from communicablediseases live in developing countries. TheWorld Health Organization (WHO) estimates thatthere were 247 million cases of malaria worldwidein 2006 and 881,000 deaths, most of themamong African children (WHO 2008). Tuberculosis(TB) kills 1.6 million people annually. As of theend of 2007, 33 million people worldwide wereliving with HIV/AIDS, more than 90 percent ofthem in developing countries (UNAIDS 2008).Three-quarters of the 2 million people who diedof AIDS in 2007 were living in Sub-Saharan Africa.Communicable diseases account for about a third(36 percent) of the disease burden in the developingworld (Jamison and others 2006b), butthe burden of morbidity and mortality is greatestamong the poor (Gwatkin and Guillot 2000).The 1993 World Development Report: Investingin Health (World Bank 1993c) highlighted thisburden and advocated investing in apackage of cost-effective, basic publichealth measures that included communicabledisease control. The WorldBank has long supported communicabledisease control components aspart of health projects; freestandingdisease projects date back to 1988. 1World Bank commitments for communicabledisease control accounted for $3.5 billion,or about a third of all HNP-managedcommitments approved during fiscal 1997–2006. 2 The World Bank approved a total of 93communicable disease projects, including 63 free-Commitments forcommunicabledisease control were$3.5 billion—abouta third of all HNPmanagedcommitmentsover fiscal 1997–2006.35

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 3.1: AIDS Accounted for More than Half ofCommunicable Disease Projects Approved in FiscalYears 1997–2006Multipledisease12%Otherdisease2%TB2%Diseasecomponent20%Polio3%Malaria3%n = 93 communicable disease projects approvedAIDS57%Source: IEG portfolio review.Note: Includes projects with objectives to control communicable disease or with a communicable diseasecomponent. Other disease includes leprosy and avian influenza (one project each).Support forcommunicable diseasecontrol has produceddemonstrable resultsfor TB, malaria,schistosomiasis, andleprosy.standing single-disease projects, 11 projects addressingmultiple diseases, and 19 projects withcommunicable disease components. Communicabledisease control was thus funded in at least42 percent of the HNP projects approved over theperiod. Freestanding AIDS projects accountedfor nearly 60 percent of all communicable diseaseprojects approved over the decade (figure3.1) and a quarter of all approved HNP projects.Between the first and second half ofthat decade, the share of communicabledisease projects (including freestandingsingle- and multiple-diseaseprojects and components) among allapprovals rose from a third to half of allprojects and from 30 to 43 percent ofall HNP commitments.The most common rationales offered for supportingpublic communicable disease control involvedthe positive externalities and public-goodnature of disease control, the understanding thatmost communicable diseases disproportionatelyaffect the poor, and the cost-effectiveness of communicabledisease interventions. 3 AIDS projectsin Africa were largely launched in an emergencymode in light of exploding HIV transmission rates,the high mortality rate from AIDS, and the adverseeconomic and social impact of the disease in a verypoor Region. 4The surge in Bank support for communicabledisease control was in step with majornew international initiatives for communicabledisease control. In 2000, the G8Summit in Japan committed to implement anambitious plan to prevent infectious disease, includingAIDS, malaria, and TB, and the Bankagreed to triple IDA financing for those diseases. 5Communicable disease control was consideredan international public good, and the internationalcommunity rallied behind it. The UN MillenniumDevelopment Goals were adopted witha goal for AIDS and malaria. Initiatives werelaunched to eradicate polio and guinea worm. Anoutbreak of Severe Acute Respiratory Syndrome(SARS) and avian influenza reminded the worldof its vulnerability to new infectious agents. 6 TheGlobal Fund to Fight AIDS, TB, and Malaria wascreated in 2002; it has committed $15.0 billion anddisbursed $6.9 billion to date. 7 The U.S. President’sEmergency Plan for AIDS Relief (PEPFAR)committed $12.8 billion for AIDS from fiscal 2004to 2007. Support from bilateral aid agencies expanded,and major private foundations have becomeinvolved. Bank strategies on HIV/AIDS andon malaria have been developed, and the Bankcontinues to be involved in global internationalpartnerships such as Stop TB, Roll Back Malaria(RBM), and the Global Alliance for Vaccines andImmunization (GAVI), 8 in addition to its role asa cosponsor of the Joint United Nations Programon HIV/AIDS (UNAIDS). Communicable diseasesfeature prominently in the Bank’s global publicgood strategic theme, one of six identified in2007 by Bank President Zoellick.Efficacy of Support for CommunicableDisease ControlEvidence on the efficacy of Bank support for communicabledisease control can be gleaned from anin-depth portfolio review of the 93 approved projects(Martin 2009), IEG Project Performance As-36

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESsessments of freestanding communicable diseaseprojects or projects with communicable diseasecomponents, 9 and country case studies. In 2005,IEG evaluated the development effectiveness ofthe entire HIV/AIDS portfolio.Support for communicable disease control,both in freestanding projects and components,has produced demonstrable resultsfor TB, malaria, schistosomiasis, and leprosy.In Bangladesh, the Fourth Population andFamily Health Project supported the NationalLeprosy Elimination Program (IEG 2006b). Theprevalence of leprosy declined from 13.6 to lessthan 1 per 10,000 population. Bank supportthrough two projects contributed to dramaticresults in reducing malaria in Eritrea (box 3.1).Support for TB Control in India (1997–2006)expanded coverage of directly observed treatmentshort-course (DOTS) by training half a milliongovernment staff and enlisting the help of10,000 private practitioners, more than 1,600nongovernmental organizations, 100 privateenterprises, and a wide network of communityvolunteers. Information, education, and communicationinvestments spread the word thatTB is curable and sought to reduce stigma. Morethan twice as many people were treated asplanned, the cure rate rose from 35 to 86 percent,and death rates in areas with DOTS were reducedsevenfold relative to non-DOTS areas(World Bank 2006a). In Egypt, support for schistosomiasiscontrol helped to reduce the prevalenceof S. mansoni (the cause of intestinalschistosomiasis in Lower Egypt) from 14.8 to 1.2percent, and of S. haematorbium (the cause ofurinary schistosomiasis in Middle and UpperEgypt) from 6.6 to 1.2 percent over the period1993–2006 (IEG 2008c).In contrast, the performance of the HIV/AIDS portion of the communicable diseaseportfolio, which was responsible for theenormous growth in communicable diseaseprojects over the decade, has beenmodest. Only 29 percent of the freestandingHIV/AIDS projects approved and completedduring fiscal years 1997–2006 had satisfactoryoutcomes, compared with 89 percent for theBox 3.1: Successful Malaria Control in EritreaFollowing a severe malaria outbreak in Eritrea in 1998, two successiveBank projects a supported a three-pronged strategy to reduce malaria:distribution of insecticide-impregnated bednets; targeted indoor residualspraying; and reduction in vector breeding sites. Malaria morbiditydecreased by 74 percent, mortality by 85 percent, and the case fatalityrate by 78 percent. The number of cases per 100,000 population nationallydropped from 7,546 to 568 between 1998 and 2006. While the earlypart of the response also corresponded to a period of declining rainfall,which would have contributed to these trends, malaria cases continuedto fall even when the rains resumed in 2003.Source: IEG 2009a.a. Eritrea Health Project, Eritrea HAMSET Control Project. The Bank was the largest singlesource of funding to the National Malaria Control Program. Other important sourcesof funding were the U.S. Agency for International Development (USAID); the ItalianCooperation; and, since 2003, the Global Fund.Figure 3.2: HIV/AIDS Projects Have Performed LessWell than Other Communicable Disease ProjectsPercent of projects ratedmoderately satisfactory or higher10080604020052AllfreestandingCD ( n = 23)Source: IEG portfolio review.FreestandingAIDS( n = 14)Freestandingother CD( n = 9)CDcomponent( n = 12)freestanding projects for other diseases 10 (figure3.2). Thus, while the performance of the HIV/AIDSportfolio has been much lower than that of otherHNP projects, performance of freestanding proj-29Breakdown offreestanding communicable disease projects895061Other HNP( n = 64)37

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONThe performance of the ects for other communicable diseasesHIV/AIDS portion of the has been substantially higher. 11 AIDSportfolio has been projects in Africa—where the crisis ismodest. most acute—have had particularlymodest performance: Only 18 percentof the projects in the Africa Multicountry AIDS Program(MAP) have had satisfactory outcomes. 12The advantage of support for one or morecommunicable diseases is that the projectcan be less complex. Communicable diseasesupport can have better defined boundaries; theobjective is easier to understand; the results chainleading from outputs to outcomes is straightforward;and results can be seen in many cases in arelatively short period. Many communicable diseaseprograms have dedicated workers and facilities,although most still rely on the rest of thehealth system. All of the AIDS projects have facedthe challenge that the ultimate outcome for HIVprevention—HIV incidence—is difficult to observeand rarely measured (though the behaviorsthat spread it can be). Success in stopping thespread of HIV also depends on reaching marginalized,high-risk individuals, which can be technicallyand politically difficult.Africa Region AIDS While there are specific characteristicsof HIV prevention andprojects have morecomplex designs than control that present challenges,other projects. what sets the Africa Region AIDSprojects apart from the other communicabledisease projects is theirrelatively more complex design.As the findings of chapter 2 indicate, complexityof design in low-capacity settings is a strong predictorof weak performance. The large number ofimplementers in both government and civil societymake the Africa MAP projects highly complex. Ademand-driven design with so many actors makesit more difficult to assess coverage and quality ofkey activities. Beyond this, while other communicabledisease programs are often run from longexistingdirectorates within the Ministry of Health,Simplification of projectdesign is easily within thecontrol of the Bank andborrowers.the Africa Region AIDS projects havegenerally relied on newly created andinexperienced institutions, separatefrom the Ministry. Finally, many wereprepared in an emergency mode; theinvestment in M&E and “learning by doing” inthese projects (which was supposed to be 5–10percent of costs) has not been realized, and thelarge number of implementers would challengeany M&E system. Most of the completed AIDSprojects in Africa have been unable to establish aclear results chain linking outputs to outcomes orto address systematically the key drivers of the epidemic,although the latter is a weakness of mostAIDS projects in other Regions.Simplifying the design of HIV/AIDS projectsis under the control of the Bank andborrowers. While it is difficult to change capacityin the short run, and some of the inherent difficultiesconfronting HIV/AIDS cannot be fixed,HIV/AIDS programs do not require an overly complexdesign. The relatively more successful AIDSprojects in Argentina, Brazil, and India, in settingswith greater capacity, still do not have projectdesigns as complex as those in Africa. Findingsand lessons from Bank support for HIV/AIDS projectswere evaluated by IEG in 2005, with recommendationsthat remain relevant to support in allRegions (box 3.2).Notwithstanding the accomplishments ofmany of these communicable disease projects:• M&E has been weaker than for otherHNP projects, even though the resultschain is theoretically easier to constructthan for projects that address broadhealth issues. Only one in five closed communicabledisease projects had at least substantialM&E, compared with a third in the restof the HNP portfolio. 13 An example of this paradoxcan be seen in malaria control. While thenumber of bednets distributed or treated withinsecticides is generally available, none of theseven completed projects supporting malariacontrol could provide data on actual bednetuse, and only one had data on bednet ownership(the endpoint only). The low capacity,weak institutions, and lack of M&E have led togovernance problems in a number of theseprojects, both in Africa and India. 14• There is little evidence to confirm that thepoor—however defined—have dispro-38

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESportionately benefited from communicabledisease control within countries.Even when the poor are at greater risk, the literaturesuggests that the non-poor have greateraccess to information and services, which makesit more likely that they will benefit (Gwatkin,Wagstaff, and Yazbeck 2005). Without data onthe socioeconomic distribution of benefits, itcannot be assumed that the incidence of diseasecontrol outcomes is pro-poor, particularly forHIV/AIDS, which in many developing countriesdoes not disproportionately affect the poor.There are equity dimensions to some of the programs—suchas in the distribution of free bednetsor malaria drugs and access to subsidizedantiretroviral treatment—that are rarely studied.• Very few Bank projects have attemptedto calculate the actual costs or effectivenessof interventions as implementedin the field as a basis for improvingperformance. A case in point is the EgyptSchistosomiasis Control Project. One of theobjectives was to strengthen the capacity to periodicallyassess the cost-effectiveness of theprogram and to adjust the strategy accordingly(IEG 2008c). However, most studies were notcarried out. Recent economic and sector workon AIDS treatment in Thailand and India pointsto tradeoffs in cost-effectiveness and sustainabilityof AIDS treatment programs undervarying assumptions (Over and others 2004; Revengaand others 2006). Yet this key informationis not available for low-income Africa,which has the greatest HIV burden and thefewest resources. Only two public expenditurereviews (PERs) issued in Africa since 2000 havediscussed the costs and sustainability of AIDSprograms, including the long-run cost of providingantiretroviral therapy to patients. 15Synergy Between Communicable DiseasePrograms and the Health SystemThe 2007 HNP strategy calls for synergy betweenefforts to strengthen health systems and the focuson priority disease results in low-income countries,noting that a well-functioning health system isalso essential for the success of these programs. 16This is clearly borne out in the experience ofsupport for communicable disease control: healthBox 3.2: Recommendations of IEG’s 2005 Evaluationof World Bank Support for AIDS Control RemainRelevantCommitting to Results: Improving the Effectiveness of HIV/AIDS Assistance(IEG 2005a) evaluated the World Bank’s country-level HIV/AIDS supportthrough mid-2004. It pointed to three main avenues for improvingperformance:• Help governments to be more strategic and selective and to prioritizeactivities that will have the greatest impact on the epidemic.• Strengthen national institutions for managing and implementing the longrunresponse, particularly ministries of health.• Improve the local evidence base for decision making through improvedM&E.Source: IEG 2005a.system problems with the procurement and distributionof pharmaceuticals and the understaffingor low performance of fixed facilities for treatmentand immunizations also affect the efficacy of communicabledisease control programs. Yet it alsocould be said that a well-functioning health systemdepends on effective communicable diseasecontrol, because prevention of communicablediseases is often low cost and reduces demand formore expensive treatment, freeing resources forother conditions.In a number of countries, the Bank’s supporthas strengthened the health system bybuilding capacity in national disease controlprograms. In Cambodia, for example, the DiseaseControl and Health Development Project (1996–2002) strengthened health infrastructure anddecentralized health management,while financing complementary diseasecontrol programs for malaria, TB, Bank has strengthened theIn many countries, theand HIV/AIDS. Prior to the project, internationaldonors were operating capacity in nationalhealth system by buildingcommunicable disease interventions disease control programs.outside the framework of government,and capacity was weak. The project funded thegovernment communicable disease programs,enabling the Ministry of Health to fulfill its mandate,while integrating the disease control programsinto the health system at the provincialand district levels. Bank support for multiple dis-39

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONease and endemic disease projects has generallysought to improve communicable disease surveillanceand cross-cutting disease-control functions,such as health promotion and laboratoryservice, to improve efficiency.However, the monitoring systems for communicabledisease programs often exist inparallel to, and are not well linked to, governmentmanagement information systems(MIS). This can lead to duplication and inhibitsharing of data between agencies. However, weaknessesin the overall health MIS can shake theconfidence of communicable disease control managers,who need timely informationAs support to systemwide for a rapid response, and such weaknessescan lead them to maintain sep-reforms continues toexpand, there is a risk arate monitoring systems, as in Ghanathat the capacity that has (IEG 2007d). This creates, in turn, abeen built in disease heavy burden at the district and servicecontrol will be lost. levels, which are required to supplyinformation through multiple, overlappingsystems (IEG 2007d). The problem hasbeen overcome in Eritrea, where the NationalMalaria Control Program maintained a paralleldata collection system, and it was reported tothe MIS with little delay.As the Bank and donors enhance their supportto systemwide reforms and sectorwideapproaches, there is a risk that the capacitythat has been built in many disease controlprograms will be lost. The first healthThe large amount of SWAp in Bangladesh included plansdonor funds earmarked to integrate the National TB Programfor specific diseases in (NTP) into the approach, yet neglectedsmall countries may be to include the NTP or any TB indicatorscreating distortions in the in the logframe (IEG 2006b). Supportrest of the health sector. for Ghana’s health SWAp provided districtswith increased financing to carryout communicable disease activities. But the districtsoften did not prioritize communicable diseaseinterventions, assuming that central-levelunits would continue to support them (IEG 2007d).In Egypt, the success of investments in schistosomiasiscontrol has not been sustained by the HealthReform Project that supported these activities aspart of an integrated basic package (IEG 2008c).In Malawi, there is a concern that the much-neededemphasis on a functioning health system pursuedby the health SWAp maintain the technical strengthof disease programs as an important element (Elmendorfand Nankhuni forthcoming).While the World Bank was among the firstto step up to increase dramatically HIV/AIDSresources to Africa (fiscal 2001–02), there aresigns that the high level of earmarked diseasefunds may be creating distortions in therest of the health system. Only about one infive freestanding AIDS projects is in a countrywith more than 50 million people; a third are forcountries with fewer than 5 million people, andhalf of those are for countries with populationsof less than 1 million. In many cases, the Bank isnot the only donor—the Global Fund and PEPFAR,in addition to bilateral funds, are also often providingearmarked funding to a specific disease(Martin 2009). In these cases, there is a risk of ahuge imbalance between investment in communicablediseases (including AIDS and any other earmarkedprogram) and investment in the rest ofthe health system. A recent study calculated thatHIV/AIDS funding represented from 33 to 45 percentof total public and donor health expendituresin four African countries. 17 The 2007 PER forMalawi reported that HIV/AIDS accounts for asmuch as 60 percent of health expenditures (WorldBank 2007f, p. 89).The problem of distortions created by earmarkingof external assistance is unlikely to be extensiveenough in the large countries receiving Banksupport for AIDS, TB, malaria, or other endemicdiseases (such as Bangladesh, Brazil, China, orIndia) to create such difficulties. But two problemsarise with relatively large shares of money earmarkedfor disease in the context of a small country.First, the large funds in a small program relativeto the needs of the rest of the system can pullscarce resources—such as nurses or doctors, whoare in short supply—from elsewhere in the healthsystem and reduce services elsewhere in the system.18 Second, as the availability of IDA money andBank supervision resources are constrained, it isvery likely that there is competition between freestandingcommunicable disease projects and projectsthat support the entire health system in smallcountries. A case in point is Malawi: because ofconstraints in the availability of Bank budget for40

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESsupervision, IDA funds were available for thehealth SWAp or the AIDS SWAp, but not for both.The Bank opted to drop support for the healthSWAp and continue support for HIV/AIDS, despitethe large share of earmarked funds as a percentageof overall health expenditure (Elmendorfand Nankhuni forthcoming). Despite these risks,the Bank has continued to approve new HIV/AIDSprojects in small countries that receive a lot ofother earmarked foreign aid.In sum, the past decade has seen an enormousincrease in the share and absolute amount ofWorld Bank support for communicable diseasecontrol, most of which has been for HIV/AIDS.Communicable disease projects other than forAIDS are more likely to be in large countries, performbetter than the rest of the HNP portfolio, andare generally somewhat less complex. The HIV/AIDS projects are almost all multisectoral and involvemultiple public sector entities and diversesegments of civil society, in addition to dealing witha highly stigmatized disease. Furthermore, mostare in Sub-Saharan Africa, and many are in smallercountries with a large donor presence. The complexityof the support and low capacity haveresulted in lower outcomes. M&E has been inadequatefor most communicable disease support,there is little evidence that the poor havedisproportionately benefitted, and the actual costeffectivenessof interventions as implementedhas rarely been calculated. There are signs that thehigh level of earmarked disease funds may becreating distortions in the health systems of somesmall countries.Reforming Health SystemsThere is no internationally accepted modelof how a health system should function;the specific activities or policy content ofhealth reform are context-specific. Generally,however, reform programs are about fundamentalchanges in structure, incentives, and allocationof resources. Improved efficiency of health caredelivery, coupled with improvements in health statusand reduction in inequities, is often at the coreof health reform programs. Health reform is thusdistinct from efforts to improve outcomes by increasinginputs—money, training, salaries, facilities,and materials—although increasing inputscan be, and has been, used to leverage and supportreforms.About a third of all World Bank HNP projectsapproved from fiscal 1997 to 2006 hadan objective to reform or restructure thehealth system. These included objectives involvinghealth finance, health insurance,decentralization, or regulation About a third of Bankof or enhancing the role of the private HNP projects had healthsector in service delivery. 19 Implicitly if system reform ornot explicitly, most of these projects restructuring as anaimed to improve the efficiency of the objective.health system. 20 The share of approvedprojects with health reform objectives has declinedby nearly half over time—from 45 percentof approved projects during fiscal 1997–2001 toonly 26 percent in fiscal 2002–06 (figure 3.3). 21 Theshare has declined in all Regions except South Asia,and in Latin America and the Caribbean and Africathe decline is statistically significant. 22Health reform projects are concentrated inmiddle-income countries. Three-quarters of allhealth reform projects were in middle-incomecountries and half of all HNP projects in middleincomecountries over the decade had healthreform objectives, compared with only18 percent of all projects in low-incomecountries. Health reform is an objective Health reform projects arein about half of the projects in lowermiddle-incomecountries and in 63 income countries.concentrated in middle-percent of the projects in uppermiddle-incomecountries. About two-thirds of theHNP portfolio in Europe and Central Asia andthe Middle East and North Africa is comprised ofhealth reform projects, as is 43 percent of projectsin Latin America and the Caribbean.Health reform projects have somewhatlower outcomes than do projectswithout reform objectives inThey have somewhatlower outcomes than domiddle-income countries, althoughprojects without reformthis difference is not statisticallyobjectives.significant. Seventy-one percent ofclosed projects with health reform objectivesin middle-income countries had satisfactoryoutcomes, compared with 86 percent of HNPprojects with other objectives (table 3.1). Theborrower’s performance is also slightly lower forhealth reform projects.41

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 3.3: The Share of HNP Project Approvals with Health Reform Objectives Has Declined inMiddle-Income Countries and in Most Regions90808074Percent of projects7060504030204526664020163362603142312520201070Allcountries(n = 220)Middleincome(n = 220)Lowincome(n = 102)Middle Eastand NorthAfrica(n = 8)Europeand CentralAsia(n = 40)LatinAmericaand theCaribbean(n = 46)East AsiaandPacific(n = 25)Africa(n = 73)SouthAsia(n = 28)Country income and RegionFiscal years 1997–2001 Fiscal years 2002–06Source: IEG portfolio review.Note: The decline in share of health reform projects overall and in middle-income countries is statistically significant at p < .01, and the decline in Latin America and the Caribbean and in Africaat p < .05. Eight Regional projects are excluded from the tabulation by country income.Table 3.1: Outcomes and Bank Performance AreLower for Health Reform Projects than for OtherHNP Projects in Middle-Income CountriesIEG rating– Health reform Other HNPsatisfactory projects (n = 29) projects (n = 14)Outcome 71 86Bank performance 69 a 93 aBorrower performance 72 79Source: IEG portfolio review.Note: In low-income countries there are only 8 completed health reform projects that were approvedduring fiscal 1997–2006 and 37 other completed HNP projects. Both outcome and Bank performanceare substantially higher for health reform projects than for other projects in low-income countries,but these differences are not statistically significant because of the small sample. Borrower performanceis similar for the two groups.a. The difference between health reform and other HNP projects is weakly significant at p = .08.Bank performance for health reform projectsin middle-income countries is substantiallyand significantly lower than forHNP projects with other objectives. TheBank’s performance—including quality at entry(preparation) and supervision—was satisfactoryin only 69 percent of health reform projects, comparedwith 93 percent of other HNP projects. IEGreviews of Implementation Completion Reports(ICRs) found the following factors associated withgood Bank performance:• A design based on strong sector work 23• Thorough institutional and stakeholderanalysis 24• Use of local experts, consultation with stakeholders,good communication, and full ownershipof the reforms 25• Creative use of other instruments to backstopreforms 26• Strong and active dialogue with governmentthroughout implementation.Poor Bank performance was often linked to weaknessesin these same areas, as well as inadequateinstitutional analysis or risk mitigation plans, over-42

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESoptimism about capacity or political commitment,project complexity, and the failure to restructureprojects that are performing poorly. 27Health care reform usually takes place over an extendedperiod, beginning with the reforms withthe least political opposition and quickest gains,and often proceeding in fits and starts. Thismeans that the efficacy of the support for the laterstage,more difficult activities is not yet tested inmany instances. However, the varying experienceof the countries where IEG has conductedin-depth fieldwork highlights some of the achievements,shortcomings, and lessons to date. Healthreforms are often linked to more general civil serviceand administrative reforms across all sectors;the obstacles to public sector reforms arein many ways similar to those for health reforms(box 3.3).The Bank supported health system reformsin the Kyrgyz Republic, Peru, and Egypt thatshared some common elements, with differentoutcomes. In all three countries, the reformstrategy called, first, for improving access toand quality of primary health care, which Bank performance inwould disproportionately help the poor health reform projects isand reduce the demand for inpatient lower than for HNPcare, and for setting up output-based reimbursements,before attempting more objectives.projects with otherpolitically controversial and technicallychallenging reforms, such as rationalizinghospitals and public sector contracting withthe private sector or reimbursements with the privatesector to deliver services. Implementationand results on improved efficiency were uneven,although health status nevertheless continued toimprove in all three countries. Only in the KyrgyzRepublic, however, is there evidence of efficiencygains and any plausible link of health outcomes toany of the reforms. 28In the Kyrgyz Republic, the Bank and otherdonors supported implementation of a homegrownreform agenda:• Set up an institution to serve as a single payerfor health insurance.• Separate the financing of health care fromhealth care delivery.Box 3.3: Shared Themes: Public Sector and Health Systems ReformThe obstacles to many health system reforms are similar to thoseblocking improvement in public administration more generally, asare the measures to overcome them. The recent IEG evaluation,Public Sector Reform: What Works and Why? (IEG 2008f) foundthat activities aimed at strengthened financial management andtax administration were far more effective than those focused oncivil service and administrative improvements. Ministries of financetended to strongly support this type of reform, which wasusually underpinned by extensive technical analysis, and oftengenerated quickly observable results, expressed in terms of clearpublic expenditure and financial accountability indicators. TheWorld Bank’s expertise in these issues is widely recognized.In contrast, measures affecting personnel policies, includingdownsizing, pay decompression, and merit-based reforms wereparticularly unsuccessful. They often failed because of lack of politicalcommitment, discontinuities in leadership, politicians’ resistanceto measures diluting their control over patronage, andunion opposition. The World Bank’s analytical tools are less developedin these fields and, more fundamentally, there is a “lackof consensus around the ‘right’ civil service model for developingcountries, or indeed for developed countries.” (IEG 2008f, p. 54)Basic data are often lacking, and political leaders may not beable to identify tangible benefits. Almost identical language couldbe used to describe health sector reforms entailing changes in organizationand personnel policies.Public Sector Reform nevertheless identified six factors associatedwith comparatively successful administrative reforms:good analysis and diagnosis; pragmatic opportunism in selectingreforms; realistic expectations; appropriate lending packages(usually including technical assistance); tangible indicatorsof success; and effective donor coordination. Successful majorhealth reform programs, as in the Kyrgyz Republic, feature similarcharacteristics.Source: IEG 2008f.43

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 3.4: Excess Inpatient Bed Capacity Declined across the Kyrgyz Republic100Hospital beds per 10,000 population9080706050403020101996 1997 1998 1999 2000 2001 2002 2003 2004Kyrgyz RepublicIssyk-KulTalasBatkenNarynChuiJalad-AbadOshBishkekSource: IEG 2008d.• Use output-based financing that pays for servicesdelivered, rather than line budgets.• Increase autonomy for health facilities in budgetallocations.• Reduce excess inpatient capacity.• Adopt a guaranteed package of services.• Create a new type of family practice that reducesresort to specialists, provides more costeffectiveprimary health care on an outpatientbasis, and serves as a gatekeeper for referralsto higher and more specialized facilities.The share of the population covered by theMandatory Health Insurance Fund rose from 25to 85 percent and excess inpatient capacity wassignificantly reduced (figure 3.4). Hospitals usedtheir remaining capacity more efficiently, and thequantity and quality of outpatient services increased.By 2005, almost 99 percent of the populationhad access to primary health care, and theshare of government health spending on thatcare had risen from 11 to 25 percent. Hemorrhagicinsult among patients with high bloodpressure in one region dropped by three-quarters.Yet, despite Bank intervention and policy dialogueat key moments, political opponents stillmanaged to block the restructuring of tertiaryhospitals in the two major cities. Private paymentsremained a significant share of health spending.The finance ministry reallocated the savings fromimproved health sector efficiency to other activities,dampening the enthusiasm of health reformers.Since then, the government has met allcommitments to increase public financing ofhealth. However, low pay scales have led to outmigrationof family physicians, jeopardizing theachievements of the reforms.In Peru, the World Bank and Inter-American DevelopmentBank support to the government’s reformprogram saw a huge expansion in accessto a package of basic services among the poorthrough the Integrated Health Insurance (SIS)program, despite significant political changes overthe period. 29 Adopted nationwide in 2002, the SISreimburses Ministry of Health care providers for thevariable costs of the use of services by its beneficiaries,increasing the incentives for efficiency. In44

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESaddition, support was provided to expand and improvethe quality of community-managed healthfacilities, organized under a quasi-private managementscheme in which the community participates.However, the SIS still does not reimburseprivate providers or other public providers andthere are no major new revenues for the system.The extent to which hospital costs have been reducedis unclear, and ESSALUD, the entity providinginsurance to formal sector workers and their families,30 has resisted reforms that would separatefunding from service delivery and that would providebetter coordination with the Ministry of Health.In Egypt, the Bank, the U.S. Agency for InternationalDevelopment (USAID), and the EuropeanCommission have supported a health reform programover the past decade, with uneven ownershipby key government players. There has beenprogress in strengthening primary health careand family services and defining a basic packageof primary health care and public health servicesfor universal coverage. However, efforts to transformthe Health Insurance Organization into asingle-payer plan and to separate finance fromhealth delivery have met opposition from the Ministryof Health and Population and other entitiesthat want to retain control over finance. Rationalizationof hospitals has not occurred and the systemremains highly fragmented, with nearly twodozen different public or parastatal entities providingand financing health care, as well as overuseof and excess capacity in expensive tertiary care.The most pervasive lesson from the Bank’sexperience with health reform is that failureto fully assess the political economy of reformand to prepare a proactive plan to addressthis issue can considerably diminishprospects for success. One source of politicaleconomy risk is that reforms with high-level supportwill be abandoned with a change in government.Among the five countries that undertookhealth reform and were studied in-depth by IEG,four experienced changes in leadership with thepotential to affect the reform agenda (Bangladesh,Egypt, the Kyrgyz Republic, and Peru 31 ). Giventhe long period over which reforms to health systemstypically take place, the starting assumptionshould be that there will certainly be a change ingovernment or in leadership of the sector that canaffect the ownership of and commitment to reform.This underscores the importance of generatingevidence that reforms work, so they cantranscend a single administration, and of enlistingkey stakeholders in the system who are vested inthe reforms and likely to remain in place.The impact of reforms on the interests andincentives of key stakeholders also constitutesa significant political economy risk.Health reform projects create winners and losers;it is important that their interests be understoodfrom the outset. High-level commitment is noguarantee that key stakeholders in the healthsystem or the general public will go along with areform; stakeholders who have a role in implementingany reform can simply not cooperate.The general public may perceive that a reductionin excess hospital capacity is reducingtheir access to health care. Even within Failure to assess thean institution, the interests and incentivesmay vary according to whether reform and to plan topolitical economy ofthe person is a manager or delivers address it can diminishservices. The Kyrgyz health reform experiencewas an exception in this re-prospects for success.gard. It was based on a reform strategy that wastotally owned by a group of reformers in the Ministryof Health and a considerable prior analysisand a strategy for navigating the winners and losers.Even then, Bishkek and Osh, the two wealthiestregions, remain outside the reforms of tertiarycare. The experiences of Bangladesh, Egypt, andRussia are more typical (box 3.4).The political risks and the risk of complexity—twoof the issues found to be most criticalin the case studies—are often missingin the risk analysis for health reform projects.Review of the risk analysis of the appraisaldocuments for closed health reform projectsapproved since fiscal 1997 found that only 59 percentcited risks of a change in government commitmentto reforms (following turnover in ministrystaff or elections of top leadership), and only 43percent pointed to the risk that specific stakeholders—physicians,hospital managers, patients,managers of decentralized facilities—would be45

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 3.4: Consequences of Inadequate Stakeholder AnalysisThe Bangladesh Health and Population Program (1998–2005),which involved many donors in a sectorwide operation, soughtto merge the Health and Family Planning Directorates within theMinistry of Health and Family Welfare, starting from the subdistrictlevel and moving downward, a reform to improve efficiency.While some family planning workers were pleased to become regularpublic employees through this reform, mid-level workersfeared that their jobs would be abolished and took their case tothe Supreme Court. The government won the lawsuit, but the implementationof the reform had been delayed and unificationwas reversed by a new government. Other proposed reforms inthe project were tainted by association.In Egypt, there was high-level commitment and participatoryanalytic work from the Minister of Health and Population forhealth reforms in the mid-1990s, but the incentives of nearly twodozen other public and parastatal providers of health care werenot taken into account. The ministry would have lost control overfinancial resources with a single payer arrangement; it thereforewelcomed investment in primary health care, but resisted reformingthe Health Insurance Organization. The insurance reformswere subsequently dropped from the Health Reform Project,while expansion of family health services continued.The Russia Health Reform Pilot Project (1997–2004) aimed tomake changes in two regions. It was to introduce output-driven,cost-conscious provider payment mechanisms, accompanied byinformation-based quality assurance, and to reorient health careby strengthening primary and outpatient care, centered on a networkof family physicians, and reducing inpatient care services.The new models were not uniformly implemented; where introduced,they coexisted with the old and had little staying power.Family practitioners were trained, their offices equipped, and astandard Russian family medicine curriculum developed, but obstetrics,gynecology, and maternal and child health were excludedbecause of the influence of specialists in those fields.Sources: IEG 2006b, 2007f; Gonzalez-Rosetti forthcoming.Political risks andcomplexity are oftenmissing in the riskanalysis of health reformprojects.resistant to or undermine the reforms.The risk of excessive complexity wasmentioned in only 22 percent of theappraisal documents. Overall, only athird of projects with health reformobjectives were assessed at appraisal ashaving substantial or high risks. 32To have a chance at success, both the politicaleconomy risks and complexity needto be taken into account and planned for inhealth reform operations. The projects reviewedfor this evaluation have used a number ofstrategies to deal with these issues:• Institutional and stakeholder analysesare essential during the design of the reformsand to inform decisions during implementationabout when to intervene and when not to.Stakeholder consultations can help to flag possibleresistance and solutions.• Information and outreach to all stakeholdersand the public to explain the reforms,the benefits, and how they will be affected.• Sequencing of reforms, so that some constituencieswin up front, before the more difficultchoices—on hospital rationalization, forexample—must be decided. Four of the fivecase studies on health reform followed a deliberatelysequenced approach, whereby thefirst step was to improve access to and qualityof outpatient and primary health care, whilein most cases introducing changes in the paymentsystem. 33 This is also a strategy for avoidingcomplexity, a big problem in health reformprojects.• Piloting reforms in specific geographic areasto demonstrate their feasibility and efficacy beforethey are expanded. The Issyk-Kul regionof the Kyrgyz Republic piloted most reforms,sponsored by USAID, generating interest inother regions.• Creating new institutions, like the MandatoryHealth Insurance Fund in Kyrgyz and theSIS in Peru, can often result in less resistancethan attempting to reform existing ones—suchas ESSALUD and the health care units of thearmed forces and the national police in Peru or46

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESBox 3.5: Programmatic Lending Maintained Momentum on Health ReformKyrgyz RepublicGovernance Structural Adjustment Credit (2003). The objectivesof the project were to improve the transparency and responsivenessof the public sector and enhance the ability of externalstakeholders to hold it accountable, and to increase efficiency,effectiveness, and accountability within the public sector. The operationsupported ongoing reforms to improve service deliveryin health.PeruProgrammatic Social Reform Projects I-IV (2001–04). The objectivesof this series of projects were to improve the antipovertyfocus of public expenditure, increase the access of the poor toquality health and education services, and enhance the transparencyof social programs, while empowering beneficiaries intheir design and implementation. They supported health reformobjectives, including trying to promote the reform of ESSALUD.Sources: IEG 2008d, 2009b.the Health Insurance Organization in Egypt,which was eventually dropped from the project.• Complementary programmatic lendingwith the finance ministry has helped sometimesto maintain momentum when implementationof reforms through health sectorinvestments has flagged, as in the Kyrgyz Republicand Peru (box 3.5). The implementationof health reform may be piecemeal, but theBank can help ensure support that extends beyonda specific project. The Programmatic SocialReform Projects in Peru were effective in maintainingthe reform agenda, despite the politicalcycle.Eliciting the participation of stakeholderscan be important for ownership of reforms,but it is still necessary to prioritize. In Bangladesh,there was wide stakeholder participationin defining the Essential Service Package,with the 1993 World Development Report as background.However, the final package was a wishlist that was not prioritized, was not technicallyand financially feasible, and was not based onBangladesh-specific data. 34 The Russia Health ReformPilot project identified two pilot regionsbased on their indicated interests. This strengthenedownership, but each region proposed manydifferent reform activities, increasing complexityand the likelihood of incomplete reforms. Further,the project involved multiple layers of governmentand a large number of separate health and healthpolicy challenges. 35Reforms based on careful prior analyticwork and evidence relevant to the countryin question hold a greater chance of success,but analytic work does not ensure success.The reform projects in Peru 36 and the Kyrgyz Republic37 both benefited from extensive Banksponsoredanalytic work that helped to ensure thatthe reform agenda was technically sound and encouragedownership by the technical stakeholders.In Egypt, analytic work sponsored by the Bankand USAID included stakeholder and institutionalanalyses. However, in the case of Egypt, this couldnot compensate for the absence of a local reformteam with the willingness, technical capacity, andpolitical support to lead the process. 38M&E are critical for implementing andmonitoring health reforms and for demonstratingimpact, but the record of Banksupportedhealth reform projects in actuallyevaluating pilot activities is weak. There wasno evaluation design for the activitiespiloted in the Russia Health Reform Stakeholder participationPilot Project; there was no basis for decisionsfor wider replication. Nor was it is still necessary tocan be important, butthere any evaluation design for the prioritize.health reforms supported in specificregions of Peru. Strong and consistent M&E isimportant—first, for understanding whether theproposed reforms will work, given that they involvechanging the incentives for both providers and patients.Thus, most of the projects first launch reformsin pilot regions. Second, based on these47

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONM&E are critical in healthreform projects.results, successful reforms can be expanded morebroadly. Evaluation of pilot reforms and rapid disseminationof results can also demonstrate toskeptics that the reforms are feasible, weakeningpolitical resistance. The lessons of USAID’s pilotreform activities in the Issyk-Kul region of the KyrgyzRepublic were fully incorporated into the designof the first Health Reform Project, and thedemonstration effect of that region encouragedother regions to accelerate reforms. As discussedin chapter 2, however, the Bank’s recordof ensuring that pilot and reform projectsinclude rigorous evaluation is weak.To summarize, about a third of Bank projects supporthealth reform objectives, mainly in middleincomecountries. The share of projects withhealth reform objectives has been in decline. Theoutcomes of health reform projects and the Bank’sperformance are lower than for other HNP projects.This is an area with a lot of risk—political andtechnical—but nevertheless highly relevant. Manylessons have been learned. Two of the most importantof these, which should affect future performance,are the need to carefully assess thepolitical risks beforehand—including the interestsand incentives of key stakeholders—and to try tominimize complexity.Box 3.6: Genesis of the SectorwideApproach in Health: An InternationalConsensusDuring the 1990s, the concept of a programwide approachwas discussed at a forum on health sector reformchaired by WHO with the active participationof Ghana, Zambia, and other partners. This led to thefirst meeting of several countries and developmentpartners to discuss the approach in 1997.Cohosted by the Danish Ministry of Foreign Affairsand the World Bank, the meeting coined the term“sectorwide approach” and reached a consensuson two follow-up actions: (a) to commission a SWApguide for the health sector and (b) to create an Inter-Agency Group to foster learning and promotion ofSWAps, with WHO as the chair and with the activeparticipation of partners and developing countries.Source: Vaillancourt forthcoming.Sectorwide ApproachesIn 1995, the World Bank defined and promoteda new approach to lending to addresschronic problems in implementing healthprojects. These problems included insufficientlocal ownership and commitment; the lack of anynoticeable trickle-down effect from some projects;low sustainability of benefits after initial implementation;confusion and dissipationThe Sectorwide Approach, of effort caused by the approaches supportedby different donors; excessive ex-or SWAp, was developedbetween 1995 and 1997; patriate technical assistance personnel;the Bank was a major the weakening of government capacityproponent. by the proliferation of donor-financedproject units; and unsatisfactory resultsfrom some adjustment operations in the allocationof public expenditure (Harrold and others 1995).The term sector investment program was coined,encompassing six principles of sound project developmentthat supported a “broad sector approachto lending.” A sector investment programhad to be:• Sectorwide in scope, covering all current andcapital expenditures• Based on a clear sector strategy and policyframework• Run by local stakeholders, including government,direct beneficiaries, and representativesof the private sector• Adopted and financed by all main donors• Based in common implementation arrangementsamong all financiers• Reliant on local capacity, rather than on technicalassistance, for implementation.Two years later this approach was relabeleda “Sectorwide Approach,” or SWAp, at ameeting of donor agencies (box 3.6), but theprinciples remained the same. 39 The anticipatedbenefits from the approach included: greatercountry ownership and leadership in managinghealth support; improved coordination and oversightof the technical and financial inputs of allpartners; strengthened country capacities andsystems for strategic sector management, fiduciaryfunctions, and implementation arrangements;48

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESreduced transaction costs; more efficient use ofdevelopment assistance; more reliable support forthe health sector strategy; and greater sustainabilityof health programs. Ultimately, the approach,through its support of national healthpolicies and programs, was to contribute to improvementsin health sector performance andsustained improvements in people’s health. 40The SWAp concept represented a fundamentalchange in the focus, relationship, and behavior ofdevelopment partners and government—a reformin the relationship between governmentand development partners and among the partners.The specific health policies and programssupported by this approach depend on the contentof the national sector strategy, which variesgreatly across countries. World Bank and otherdonor support for a SWAp can be financed in anumber of different ways—through parallelproject-specific financing, 41 pooled financing, 42general budget support to the finance ministry, ora combination of these.Between fiscal 1997 and 2006, the WorldBank approved 28 HNP projects supportinghealth SWAps in 22 countries. 43 Thus, in thedecade following the launch of the approach,about 13 percent of all approved HNP projects supporteda SWAp. The World Bank project approvalscame in two spurts—immediately after the 1997meeting, followed by a two-year pause, and thenmore projects approved in fiscal years 2003–06 (figure3.5). Two-thirds of the projects that supportedhealth SWAps were in Sub-Saharan Africa. Supportfor health SWAps is mainly found in low-incomecountries, accounting for nearly a quarter of HNPprojects approved in those countries, comparedwith only 6 percent of those in lower-middleincomecountries, and none in upper-middleincomecountries. 44 In 71 percent of the projects,Bank resources were pooled with those of governmentand other donors. 45The assessment of the Bank’s support forhealth SWAps addresses three questions.First, were the benefits of the approach realized,in terms of better donor harmonization and coordination,reduced transaction costs, capacityCommitments (US$ millions)Figure 3.5: After an Initial Spurt, Growth in WorldBank Support for Health SWAps Resumed after 2002400371 376835073002502001501005002900 01997 1998 1999 2000 2001 2002 2003 2004 2005 2006Fiscal year approvedSource: Vaillancourt forthcoming.140Commitmentsbuilding, and so forth? Second, were Over the period fiscalthe objectives of the health strategies 1997–2006, the Banksupported by government and developmentpartners achieved? Third, in health SWAps in 22approved support for 28what ways did channeling support countries.through a SWAp affect the efficacy ofthe Bank’s support, including its ability to conductpolicy dialogue? A desk-based portfolio reviewof all Bank-supported SWAps approved in fiscal1997–2006, including 11 completed operations,and fieldwork in five countries (table 3.2) serveas a basis to answer these questions.Benefits of the ApproachThe objectives, anticipated benefits, and specificmechanisms of the SWAp approach arenot clearly articulated and often not discussedin the Bank’s project appraisaldocuments. These documentsusually deal with the substance of healthprograms and policies. There tend to bea lot of process indicators that implicitlyconvey what the objectives of anticipatedbenefits were, and they aregenerally consistent with the SWAp featuresdescribed in the literature.165260Number of projects2446543210Number of projectsThe objectives,anticipated benefits, andspecific mechanisms ofthe SWAp approach arenot clearly articulatedand often are notdiscussed in appraisaldocuments.49

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable 3.2: HNP Projects Supporting Health SWApsin IEG Country Cases or Project EvaluationsCountryProjectBangladesh Health and Population Program (1998–2005)HNP Sector Program (2005–10)Ghana Health Sector Support (1997–2002)Health Sector Support II (2003–07)Kyrgyz Republic Health and Social Protection (2005–11)Nepal Health Sector Program ( 2004–10)Malawi Multisectoral AIDS (2003–08)Health Sector Support (2004–08)Source: World Bank data.Note: In all five countries, World Bank SWAp support was pooled.Country capacity in the Country capacity has been strengthenedin the areas of sector plan-health sector has beenstrengthened through the ning, budgeting, management, andapproach. fiduciary systems. Resources areused to finance an explicit Program ofWork linked to the national strategy; annualor semi-annual joint review meetings of nationaland external partners are held for planning,programming, budgeting, resource allocation,and monitoring of progress in implementation andachievement of objectives. Governments havebeen enabled to exercise greater leadership in directingthe use of development partner resourcesand conducting and coordinating the dialogue.The performance of SWAps • In the Kyrgyz Republic, the SWApin reducing transaction has resulted in greater delegationcosts cannot be assessed of planning and decision making tobecause they have not staff at the levels of department headbeen monitored. and below within the central ministry,and from central to decentralizedlevels. Fiduciary strengtheningin the health sector has begun toinfluence the overall governance ofthe country. 46• In Bangladesh, line directors prepare annual operationalplans and budgets, and all fundingis on budget, 47 in contrast with the previousarrangement of disparate, multiple projectsmanaged by project directors who would bypassline managers. All national procurement is nowhandled by government, where previously it wasprovided by, among others, UN agencies.• In Nepal, the SWAp has helped to consolidateand coordinate dispersed donor assistance.Budgeting and financial processes within theSWAp account for government, pooled, andnonpooled resources, but attempts to improvefinancial management under the SWAp haveencountered some obstacles that are governmentwideand difficult to address (Shawforthcoming).• Government capacity for procurement andfinancial management in Ghana was developedat central, regional, and district levels, enablingthe consolidated management of the publicbudget and pooled financing. However, spendingis not always consistent with sector prioritiesand the poverty focus articulated in theProgram of Work. Nor is spending routinelymonitored or allocations adjusted to achievegreatest impact.•InMalawi, however, health stewardship andgovernance is weak, aside from the structuredarrangements for cooperation with developmentpartners. The use of government fiduciarysystems in the health SWAp was ruled out;World Bank policies and procedures are followedfor all pooled funding, and the Bankissues no-objections on the use of all pooledfunds. Donors have continued project financingoutside of the pool. 48 In contrast, theHIV/AIDS SWAp enjoys a higher level of confidenceof donors, due in part to the stronger capacityof AIDS institutions, compared with theMinistry of Health—a result of the ability of theHIV/AIDS SWAp to attract the best people, includingthose from the Ministry.Weaknesses persist in the design of M&Eand in the use of country M&E systems. Allof the countries supported have instituted periodicjoint reviews of health-sector performanceusing a common set of indicators, with datagenerated increasingly from the countries’ healthmanagement information systems. However, withthe exception of Kyrgyz, the national strategiesand programs are not underpinned by a welldevelopedM&E framework. For the most part, theresults chain of the strategies and programs is notfully articulated. In Ghana, for example, the M&Eplan, strategy, methodology, roles, and responsi-50

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESbilities were not articulated at the outset; the 20indicators were largely process indicators andhad only a modest overlap with indicators formeasuring health sector performance and outcomes.In Senegal and other countries, baselinedata were outdated or missing. Duplication ofM&E and reporting has been reduced in Nepal andthe health management information system is ofgood quality and reliable, but information andsystems are scattered, making it difficult to assemblea coherent picture of the evolution ofoutputs and how they relate to outcomes. 49The performance of SWAps in reducingtransaction costs—a major anticipated benefitof the approach—cannot be assessedin any country because they have not beenmonitored. In fieldwork, IEG was unable tocompile any data measuring transaction costs beforeand after adoption of a SWAp, for the governmentor for donors. Nor do there appear to beany studies of staff time allocations across tasksin government or by the development partnersbefore and after adoption of a SWAp. Efficiencygains were mentioned in interviews with publicsector officials, citing the reduction in the numberof individual donors, missions, and projects;the reduction in reporting requirements; and theconsolidation of M&E as good steps. However,high transaction costs were mentioned in associationwith the time and expense of preparing forjoint review meetings.The allocation of time of development partnerstaff since the launch of the approach is not regardedas entirely efficient. Some partners continueto micromanage, while in other cases there is a mismatchbetween the partner’s skills (for example,generalists or health experts) and the needs ofthe SWAp (financial management, M&E, amongothers). World Bank supervision costs for theBangladesh SWAp were one-third of pre-SWAp levelsbecause fewer Bank staff and more partnerswent on joint missions. In contrast, in the KyrgyzRepublic, Bank supervision costs are reported tobe very high because of the high transaction costsof collaborating and communicating with otherpartners, oversight of procurement and financialmanagement, and investments in capacity building.HNP Outcomes under Health SWApsThere are countries where progress is being madeboth in achieving the objectives of the SWAp approachand in improving HNP outcomes. TheTanzania health SWAp, supported by the HealthSector Development Project, was favorably evaluatedby a joint external evaluation financed bythe international donor community and had satisfactoryoutcomes (COWI, Gilroy, andEPOS 2007). Health outcomes continueto improve in Nepal, although itis not clear whether they should be attributedto health policy, the SWAp,both, or neither.However, achieving the objectives of theapproach has not always ensured betterhealth outcomes. The implementation andefficacy of the policies supported by theapproach are key. For example, during the 10years of Ghana’s health SWAp, despite the increaseduse of country systems and better harmonizationand coordination of donors, therewere no improvements in health status indicators(figure 3.6), nor were there significant improvementsin health sector performance, such asassisted deliveries, that might have contributedto better outcomes. Inefficienciesin the health sector persist outcomes, however,Achieving better healthbecause of duplication and rivalry depends on thebetween the Ministry of Health and implementation andthe Ghana Health Service, an issue that efficacy of the policiesthe development partners and governmenthave not been able to resolve. approach.supported by theSubstantial investment in the SWApprocess, suboptimal allocation of resources, 50 inadequateflows of funds to the districts, and failureto resolve key impediments to service deliverywere all factors affecting performance in Ghana.Only 4 of the 11 closed projects supporting healthSWAps had satisfactory outcomes in terms ofachieving their relevant program objectives.The approach can have adverse effects onoutcomes in the short run: setting up aSWAp can be disruptive, diverting time andenergy from the achievement of results.There is evidence from Bangladesh, Ethiopia,Ghana, Mali, and Senegal that in the short run,There has been progress inachieving the objectives ofthe SWAp approach andimproving HNP outcomesin several countries.51

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONDeaths per 1,000 live birthsFigure 3.6: Neither Mortality nor Fertility Declinedduring the 10 Years of Ghana’s Health SWAp2001501005001556.477Second Health andPopulation Project1991–971195.566Health SectorSupport Program1998–2002Health SectorSupport Program2003–07108 111 1114.6 4.4 4.4571988 1993 1998 2003 2006Under 5 IMR Fertility6471109876543210Total fertility rateSWAps have often supported overlyambitious programs of work, involvingmany complex reformsand activities that exceed governmentimplementation capacity. Examplesof the types of activities found in the programsof work of countries such as Ethiopia, Mauritania,Senegal, and Tanzania include: delivery ofa basic package of essential health services; reorganizationand decentralization of services; redefiningthe role of the central Ministry of Health;contracting with the nongovernmental sectors;hospital reform; new supply- and demand-sideservice delivery mechanisms; health financing;and cost-sharing. All of these are in addition tothe challenging capacity building and reforms indevelopment partner relationships inherent in aSWAp. Guidance for reasonable sequencing andimplementation that would make such reformsfeasible is often absent. The result is that noteverything gets done, the higher-priority interventionsare neglected, and outcomes are belowexpectations.Sources for 1988–2003: Ghana Demographic and Health Surveys.Sources for 2006: Multiple Indicator Cluster Survey for 2007 (for under 5 and infant mortality); Population ReferenceBureau (for fertility).Note: The mortality rates are averaged over the five years preceding the Ghana Demographic and Health Surveys(GDHS). The fertility rates for GDHS 1988, 1993, and 1998 are averaged over the five years precedingthe surveys, while the fertility rates for GDHS 2003 and the Multiple Indicator Cluster Survey for 2006 areaveraged over the three years preceding the surveys.Setting up a SWAp can the process can take priority over results.In Bangladesh, despite the longdivert time and energiesfrom delivering on history of working with a consortiumsubstance. of donors, there was disruption withthe formal adoption of a SWAp. However,in Kyrgyz there was a longstandingand productive working relationship amongexperts in government and between governmentexperts and the donors’ trust, and the SWAp formalizedwhat was already happening. Beyondthe short-run disruption, it is difficult to tellwhether the improved coordination, ownership,and other reforms in foreign aid introduced by theSWAp have had any long-run impact on the qualityor efficacy of health programs. This remainsa hypothesis.SWAPs often supportedoverly ambitiousprograms thatexceeded the capacity ofgovernment to implement.Impact of the Approach on the Bank’sEffectivenessIn the operations that IEG studied, the Bankhas often been instrumental in getting otherdevelopment partners to pool resources; thedonors who pool generally have a seat at the tablein the policy dialogue (Vaillancourt forthcoming).The Bank has played a crucial role in supportingnational capacities and systems for financial management,procurement, and other fiduciary aspectsof the SWAps within the health sector.The Bank has not withdrawn from policy dialogueunder the SWAp, but there is a riskthat consensus decisions among developmentpartners and government will inhibitstrategic choices and the setting of priorities,areas in which the Bank has often providedvalued support to improve the efficacyof health policy. The SWAp implies ceding leadershipand decision making to government andreducing the bilateral relationship with governmentin favor of collaborative partnerships withdevelopment partners. But the partners collectivelymay not have the will to intervene. In Ghana,for example, the partners were unsuccessful inpressing the government to address duplicationof services between the Ministry of Health andthe Ghana Health Services, although inaction in52

LESSONS FROM THREE APPROACHES TO IMPROVE OUTCOMESreform is likely one reason for stalled health results.They may also merely continue to pressfor their own priorities within the SWAp. Aidesmemoirefor joint review meetings often representan amalgam of the priorities of each of theparticipating development partners.There is often not a clear understandingbeforehand of the decision-making rulesamong individual development partnersand between the partners and the government,in the event that the government reversesor does not act on agreed-upon policy.This leads to ambiguity with respect to how theBank should act. In the case of Bangladesh, thedevelopment partners were split on how to reactto the government’s decision not to merge familyplanning and health services within the Ministryof Health and Family Welfare in the face ofpolitical pressure. The partners were divided onthe appropriate response; the Bank, supported bysome of the partners, temporarily stopped disbursements,trying to force something that wasnot politically feasible, an action that was shortlyreversed. In contrast, in Nepal, a politically drivenpolicy decision was made to provide free basic careto all that would have serious repercussions onhealth finance. In response, the Bank and otherpartners are supporting analytical work to documentthe cost and equity implications of the newpolicy, as a contribution to the Ministry of Healthstrategy on sustainable financing.For most of the SWAps studied by this evaluation,the rules for resolving disagreements are still somewhatambiguous, which does have implications forthe efficacy of the Bank’s support and Participation in SWApsits ability to engage government and has not meant athe partners on critical policy issues withdrawal from policythat have not achieved a consensus. dialogue.Memoranda of understanding remaintoo general and have not provided the guidanceneeded during times of fundamental disagreementon policy.To summarize, SWAps implicitly have two typesof objectives—those of the approach, and thoseof the program supported by the approach.Certain anticipated benefits of the approach supportedby the Bank are being realized—strengthenedcountry sector capacity to plan budgeting,management, and fiduciary systems, and greatercountry leadership in setting the direction of thesector—but weaknesses persist in M&E of theapproach and the programs. One majoranticipated benefit of the approach— But there often is not areduced transaction costs—has not clear understanding ofbeen systematically monitored, but the decision-making rulesfield visits suggest that transaction costs among the partners andin some instances remain high. The between the partners andrules for resolving differences among the government.donors and between donors and governmentare not well articulated and have led tofriction in some cases and a lack of prioritization.To date, there is little evidence that the approachby itself has had additional positive impacts on theeffectiveness of the overall sectoral program andpolicies. Even while adopting these reforms inworking relationships, it is important for the programto support the right things, that it be properlyimplemented, and that the focus on resultsbe maintained.53

Chapter 4Evaluation Highlights• The CAS has helped identify the keysectors that affect HNP, but has notbrought those other sectors to bear onHNP outcomes, nor has it documentedor monitored their contributions.• The efficacy of multisectoral HNPprojects hinges on prioritizing sectorsin the face of limited implementationcapacity and ensuring strong coordinationof activities.• The potential for Bank support forwater supply and sanitation andtransport to contribute to health outcomesand the health MDGs is great,but the incentives to deliver health resultsand a poverty focus are weak.• Results are more likely when thehealth objectives are made explicitand when targets and monitorableindicators are included for retrofittedhealth components.

Girls from Kuje village in Abuja, Nigeria, demonstrate how to wash hands with soap and clean water.Photo courtesy of UNICEF Nigeria/2006/Moses.

The Contribution of OtherSectors to Health, Nutrition,and Population OutcomesRecognizing that it takes more than health services to improve HNP outcomes,the 2007 HNP strategy calls for leveraging interventions inother sectors to deliver results. 1 One of the five strategic directions ofthe strategy is to strengthen Bank capacity to advise countries on an intersectoralapproach to HNP results. The proposed vehicle for coordinating sectors tomaximize impact on health outcomes is the CAS.This chapter seeks to inform implementation ofthe HNP strategy by assessing, first, to what extenthave CASs actually been used as a vehicle forachieving multisectoral synergies to improve HNPoutcomes? Second, what has been the experienceof multisectoral HNP lending operations—that is, HNP projects that engage multiple sectorsin a single operation—in improving HNP outcomes?Finally, what has been the contribution oflending in other sectors—in particular, water supplyand sanitation and transport—in achievinghealth benefits?Intersectoral Approaches inCountry Assistance StrategiesThe CAS is the business plan that guides WorldBank Group activities in a member country engagedin borrowing or is receiving grant funding.Management and the Board use the CAS to reviewand guide the Bank Group’s country programsand to judge the impact of its work. IEG reviewed137 CASs approved between fiscal 1997 and 2006with respect to their use as an instrument to ensurethat actions from other sectors are broughtto bear on HNP outcomes. 2Over the past decade, the CAS has fallenshort of its promise of coordinating andbringing to bear the actions of multiple sectorson HNP outcomes. Three-quarters of CASsacknowledged the contribution of other sectorsto HNP outcomes, and about half proposed amultisectoral lending strategy for HNP (figure4.1). 3 Water supply and sanitation was the sectormost often cited as having an impact on HNPoutcomes, followed by education and the environment.The two main multisectoral lendingstrategies to improve health outcomes involved,first, financing projects in sectors that are complementaryto health (41 percent of CASs), and,second, financing multisectoral HNP projects thattied the actions of many sectors into a singlelending operation (18 percent of CASs). 4 Most ofthe CASs proposing multisectoral HNP projectsplanned a multisectoral HIV/AIDS project (12 percentof CASs). However, almost none of the CASsthat incorporated complementary lending byother sectors proposed any specific HNP targetsor common management arrangements, or explainedwhat form of coordination with the HNPsector was envisaged. 557

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 4.1: Multisectoral Approaches to HNP in Country Assistance Strategies80707560Percent of CASs504030494120100Acknowledegemultisectorallinks to HNPProposemultisectorallending strategyPropose lendingin othersectors18ProposemultisectoralHNP projects12. . . of whichHIV/AIDSprojects2 1CommontargetingacrosssectorsCommonmanagementof paralleloperationsSource: Sinha and Gaubatz 2009.Though three-quarters There was little evidence in theof CASs acknowledged CAS Completion Reports (CASCRs)the contribution of that multisectoral collaborationother sectors to HNP or tracking of outcomes aroseoutcomes, almost none from the multisectoral approaches.included coordinated, Eighteen of the 19 completed CASscomplementary lending for which a completion report hadby other sectors. been written had proposed a multisectoralapproach to HNP outcomes.While in most cases the proposed lending waswholly or partly implemented, there was littlemention in the completion reports of the resultsof health components in the complementarylending, and in most of these the health componentswere dropped. Only in one CAS, forMali (fiscal 2004), were health indicators tracked—in that case to assess the impact of safe wateron cholera incidence. It was unclear in any ofthe CASCRs whether collaboration across sectorsoccurred to achieve HNP outcomes. Even inMali, the CASCR commented that sectoral staffworked independently, which was attributed toinsufficient incentives to work in cross-sectoralteams.This lack of intersectoral coordination inmost of the CASs completed to date does notdetract from the fact that investments inother sectors can have important impacts onHNP outcomes, even without coordination.This was shown in IEG’s evaluation of maternaland child health outcomes in Bangladesh (IEG2005b). That study concluded that improvementsin access to safe water accounted for a quarter ofthe decline in stunting, and that increased levelsof female secondary education supported by theBank had an impact on reducing infant and childmortality that was independent of health programs.The 2007 HNP strategy points to the potentialof better coordination and synergy throughthe CAS, beyond what might have been achievedwithout such coordination.Multisectoral Health, Nutrition, andPopulation LendingOne strategy for bringing to bear the contributionof other sectors to HNP outcomes is through assigningresponsibility for those contributions ina single multisectoral lending operation. This sectionreviews in greater depth the rationale andperformance of the multisectoral part of the HNPmanagedlending portfolio, drawing on the resultsfrom IEG’s in-depth portfolio review of projectsapproved from fiscal 1997 to 2006.58

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESFigure 4.2: Two-Thirds of MultisectoralHNP Projects Involve HIV/AIDSWater supplyand sanitation2%Social insuranceor protectionPopulation or 2%safe motherhood2%Health7%Social sectorexpenditure/reform/restructuring8%Nutrition10%Source: IEG portfolio review.Note: N = 84.Earlychildhooddevelopment2%Othercommunicablediseases2%HIV/AIDS63%More than a third of the HNP-managed projectsapproved between fiscal 1997 and 2006(38 percent) were managed or implementedby more than one ministry or agency in theborrowing country. Nearly two-thirds of multisectoralprojects were HIV/AIDS projects; theremaining third included projects in nutrition,health (multisectoral components), water supplyand sanitation, early childhood developmentand multisectoral social sector lending, (figure4.2). Compared with the rest of the HNP portfolio,multisectoral projects were more likely tohave objectives to improve health or nutritionstatus or to change behavior. They were less likelyto have objectives of improving the access to orquality of health care, reforming or decentralizingthe health system, or to have poverty-targetedobjectives.The share of HNP multisectoral projects doubledin 10 years, from a quarter to nearly halfof all projects. This increase is attributable to agrowing number of multisectoral HIV/AIDS projects.The share of non-HIV multisectoral projectshas remained constant, at 14 percent, while theshare of multisectoral HIV projects tripled, from 11to 35 percent (figure 4.3). Multisectoral projects aremore common in low-income countries (45 percentof all HNP projects) than in middleincomecountries (35 percent). Remanagedprojects wereMore than a third of HNPgionally,more than half of projects in managed or implementedAfrica, a third in East Asia and Latin by more than oneAmerica, and a quarter in South Asia ministry or agency.are multisectoral (figure 4.4). MultisectoralHIV/AIDS projects comprised45 percent of all HNP projects approved in Africa,while about a third of all projects approved in EastAsia were non-AIDS multisectoral projects.The main rationale for involving severalsectors in managing or implementing theseprojects is their complementarity in producinghealth outcomes. The choice of sectorsis based on the perceived or demonstrated comparativeadvantages in relation to the outcome inquestion. Early child development, for example,may involve both learning opportunities andproper health and nutrition. Nutrition objectivesmay require inputs from health, agriculture, oreven industry (in the case of salt iodization).There are many examples of collaboration betweenthe education and health ministries to promotehealth education in the schools. In projectsfeaturing social sector expenditure reform, healthPercentFigure 4.3: Multisectoral HIV/AIDS Lending Accountsfor All of the Increase in Multisectoral HNP Lending1009080706050403020100Source: IEG portfolio review.1411751997–2001(n = 99)Fiscal year of approvalOther multisectoral HNPOther HNP1435512002–06(n = 121)Multisectoral HIV-AIDS59

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 4.4: Multisectoral HNP Projects Are Unevenly Distributed across Regions10080127141014 324244515Percent604088837164 611120440Middle East andNorth Africa(n = 8)Europe andCentral Asia(n = 40)South Asia(n = 28)East Asiaand Pacific(n = 25)Latin Americaand theCaribbean(n = 46)Africa(n = 73)Multisectoral AIDSOther multisectoral HNPOther HNPSource: IEG portfolio review.The share of HNP is bundled with other sectors facingmultisectoral projects similar expenditure problems for reasonsof efficiency. Some health sup-doubled in 10 yearsbecause of the growing port is multisectoral because more thannumber of HIV/AIDS one key ministry is involved in healthprojects. care delivery or reform—the Ministryof Health and the social security agency,for example. In these instances, there may ormay not be any implied synergy; the multisectoraloperation may simply be packaging activities forseparate entities, intending to work in parallel.There are two distinct justifications for multisectoraldesign offered for HIV/AIDS projects. The firstis based on the notion of sectoral expertise andcomparative advantage to address an issue—forexample, enlisting the Ministry of Defense forHIV/AIDS prevention in the military, the Ministryof Justice for HIV/AIDS in prisons, the Ministry ofEducation for HIV/AIDS education in the schools,and the Ministry of Social Welfare to care fororphans. This is similar to the argument for othermultisectoral projects, and achieving the objectivewould require generating working relationshipsand intersectoral collaboration between theseagencies and the Ministry of Health. The secondrationale is essentially mobilization: becauseHIV/AIDS affects all sectors of the economy, all sectorsmust be involved in its prevention and mitigation.6 This rationale leads to the involvement oflarge numbers of ministries and public agencies,irrespective of their sectoral mandate or expertise,with or without the collaboration of the Ministryof Health.Most HIV/AIDS projects were managed by recentlyestablished multisectoral institutionswithout long experience in cross-ministerialcoordination. 7 Nearly two-thirds of multisectoralprojects not pertaining to HIV/AIDS are managedentirely or partly by the Ministry of Health,while two-thirds of multisectoral HIV/AIDS projectsare managed by an entity under the presidentor prime minister, or by the Ministries of Finance,Economy, or Planning (table 4.1). Only 36 percentof multisectoral AIDS projects rely on Ministrymanagement exclusively, or in collaboration withanother ministry. The Ministry is explicitly identi-60

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESTable 4.1: Distribution of Multisectoral HNP Projects by Management and ImplementationArrangementsTotal HIV/AIDS Non-AIDSMultisectoral management or implementation Number Percent Number Percent Number PercentManaged by an entity directly under the president or primeminister and implemented by the Ministry of Health and/orother sectors 31 (37) 30 (57) 1 (3)Managed by the Ministry of Health but implemented byother sectors 29 (35) 16 (30) 13 (42)Managed by the Ministry of Finance, Economy, or Planningbut implemented by the Ministry of Health or other sectors 11 (13) 4 (8) 7 (23)Managed and implemented by more than one ministry(one of which is the Ministry of Health) 9 (11) 3 (6) 6 (19)Managed by another ministry, implemented by the Ministryof Health 4 (5) 0 (0) 4 (13)Total 84 (100) 53 (100) 31 (100)Source: IEG portfolio review.Note: For the types of non-AIDS projects, consult figure 4.2.fied as at least one of the implementing agenciesin project design documents in all of the multisectoralHNP projects and all but eight of the multisectoralHIV/AIDS projects.Multisectoral projects are more demandingof institutions and require a greater degreeof coordination. In 90 percent of themultisectoral projects, at least one implementingagency—and as many as six—was explicitly identified,and in most cases the activities that the agencieswere accountable for were clearly spelledout. 8 The multisectoral HIV/AIDS projects hadfewer assigned implementing agencies (1.3 agencies,on average) than did the other multisectoralHNP projects (2.8 agencies, on average),but this is deceptive.More than half of multisectoral projects also hada demand-driven element at the ministerial level:ministries not explicitly identified beforehandcan submit funding proposals to achieve theproject’s overall objectives. Some of the PADsestimate the total number of ministries or agenciesthat will likely be involved, with a range offrom 1 to 20. For eight HIV/AIDS projects, therewere no assigned implementationagencies; all multisectoral implementationcame from the demand-drivencomponent.Taking into account both the explicitlyidentified implementing agencies andthe anticipated number involved in demand-drivenministerial components,multisectoral HIV/AIDS projects hadtwice as many implementing agencies(about 6, ranging from 0 to 20), as did other multisectoralHNP projects (about 3, ranging from 1to 6; figure 4.5). Yet even this is an understatement.A third of the HIV/ AIDS projects with demanddrivencomponents did not set a maximum numberof implementing agencies; the ultimate numberand their activities are determined duringimplementation.The ability to deal with this complexity wasoften weak. Among the closed projects, ineffectivemultisectoral coordination was cited as acontributing factor to low outcomes in 10 unsatisfactoryprojects, while effective multisectoralcoordination was cited as contributing to 2 satis-Most non-AIDSmultisectoral projectsare managed by theMinistry of Health;most multisectoral AIDSprojects are managed bycoordinating agenciesunder the president orprime minister.61

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONNumber of projectsFigure 4.5: Distribution of Multisectoral HNPProjects by the Number of Assigned andDemand-Driven Implementing Agencies109876543210Source: IEG portfolio review.0 1 2 3 4 5 6 7 8 9 10 >10Total number of assigned and demand-driven implementing agenciesMultisectoral HIVOther multisectoral HNPBox 4.1: Quality-at-Entry forMultisectoral Projects Is WeakIn 2008, the Bank’s Quality Assurance Group (QAG)issued an assessment of quality at entry of the lendingportfolio approved in fiscal 2006–07, highlightingresults for multisectoral projects. QAG’s conclusionswere based on a review of a sample of 10 developmentpolicy operations and 10 investment projectsBank-wide. The report expressed the following concernsabout the outcomes of multisectoral projects:• Excessive complexity and overly ambitiousobjectives• Weak institutional capacity• Lack of readiness of the first year’s program for implementation• Task teams lacking adequate technical expertiseand global experience• Fragmented managerial guidance because of themultisectoral nature of the projects.Source: World Bank 2008b.factory projects (both in nutrition). 9 The Bank’sQuality Assurance Group (QAG) recently expressedconcerns with respect to quality at entryfor multisectoral projects Bank-wide, across all sectors(box 4.1). Evidence from IEG field assessmentsof multisectoral projects in Eritrea andGhana contrasts a project design that encouragedintersectoral collaboration with the Ministryof Health with one that mobilized a large numberof actors with less Ministry oversight and collaboration(box 4.2).The lower performance of Multisectoral projects had lowerHIV/AIDS projects drove outcomes than did other HNP projects,but this is mainly because ofdown the outcomesof multisectoral projects the lower performance of HIV/as a group. AIDS projects. Fewer than half of themultisectoral projects approved andcompleted during fiscal 1997–2006 hadsatisfactory outcomes, compared with about twothirdsof single-sector projects (figure 4.6). 10 Bankand borrower performance were also lower formultisectoral HNP projects, although the differencefor borrower performance is not statisticallysignificant. However, the performance ofnon-AIDS multisectoral projects was similar tothat of single-sector HNP projects, while the multisectoralHIV/AIDS projects performed at a muchlower level. Both types of multisectoral HNP projectshad lower institutional development impact(IDI) than other HNP projects: only 43 percent ofprojects had substantial or high IDI, comparedwith 61 percent for single-sector HNP projects. 11In sum, multisectoral projects are inherently moredemanding, and their share in the HNP portfoliohas doubled, almost wholly due to the increasein HIV/AIDS project approvals. Multisectoral HNPprojects with objectives other than HIV/AIDS performat levels similar to projects in the rest of theHNP portfolio, but multisectoral HIV/AIDS projectsdo not. The large number of sectors involved,the lack of specificity in design documents aboutthe roles and responsibilities of each participatingsector, the relatively new institutions asked tomanage the complex design, and the other factorsthat bring about lower performance in the AfricaRegion likely all contribute to lower outcomesfor these projects.62

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESBox 4.2: Greater Selectivity in Sectoral Participation Can Improve Multisectoral PerformanceExperience from two communicable disease projects suggeststhat the complexity and efficacy of multisectoral projects can beimproved by enlisting a smaller number of priority sectors and bystronger collaboration with the Ministry of Health.In Eritrea, for example, the HIV/AIDS, Malaria, STD, and TB(HAMSET) Control Project engaged a limited group of priorityministries with a direct stake or comparative advantage in preventingand treating the HAMSET diseases—health, education,defense, transport, and labor and human welfare—under theleadership of the Ministry of Health and built on past collaboration.While the project nevertheless challenged the Ministry,by avoiding creation of new institutions, scarce human resourceswere conserved, duplication avoided, and complexityreduced. aIn contrast, the Ghana AIDS Response Project (GARFUND) wasmanaged by a newly formed Ghana AIDS Commission under thepresident and financed at least 16 non-health ministries and publicagencies, in addition to research institutions, regional coordinatingcouncils, district assemblies, parliamentarians, traditionalcouncils, and chiefs. The National AIDS Control Program (NACP)in the Ministry of Health retained responsibility only for implementingactivities falling within a very narrow mandate. As a consequence,“GARFUND subprojects continued to suffer from poortechnical quality and inadequate public health content.” bSources: IEG 2007c, 2009a.a. The project nevertheless remained highly complex, addressing four diseases, the zoba-level line ministries, and a community-driven component.b. IEG 2007c, p. 32.Health in the Lending Portfolios of OtherSectorsSince 1997, the World Bank has committed$5.0 billion in the form of 350 HNP componentsof projects managed by other sectors.In contrast to the part of the portfolio managed bythe HNP sector, which is relatively flat, approval ofprojects in other sectors with HNP components isgrowing, while the size of the components is relativelysmall—amounting to only 30 percent of allHNP commitments since 1997 (figure 4.7). Thispart of the portfolio has grown steadily and steeplysince 1988 and reflects lending for social funds, 12initiated in fiscal 1989, and of poverty reductionsupport credits (PRSCs), begun in fiscal 2001. Until2001, almost all projects with HNP componentswere investment projects; since then, the majorityhave been development policy lending. 13 Theefficacy of these HNP components is not easilyassessed unless they reflect explicit objectives forwhich the projects are accountable. IEG evaluatedsocial funds in 2002, 14 and an ongoing evaluationof PRSCs will look at the adequacy of theseinstruments for achieving HNP results (box 4.3).Figure 4.6: Multisectoral Projects Had LowerPerformance than Other HNP ProjectsPerent of projects rated moderatelysatisfactory or higher1008060402004317Source: IEG portfolio review.63Outcomerating63Bank performanceratingThis section examines in greater depth the extentto which lending in two key sectors—water supplyand sanitation and transport—has been usedto improve HNP outcomes. The selection of thesetwo sectors for evaluation is illustrative, becausethere are many others with large, demonstrated573169 67 6961 6358BorrowerperformanceAll multisectoral (n = 28) Multisectoral AIDS (n = 12)Multisectoral HNP (n = 16) Other HNP (n = 71)63

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 4.7: Approval of HNP Components Managed by Other Sectors Has Grown Steadily Since1988, though the Commitment per Project Is Small2,500 502,000First PRSCapproved, 2001454035Millions of US dollars1,5001,0005000First social fundproject approved,19891970197119721973197419751976197719781979198019811982198319841985Net commitmentsSource: World Bank data.Note: The commitment for the entire project is attributed to the fiscal year of approval.1986198719881989Fiscal year199019911992199319941995Projects1996199719981999200020012002200320042005200620072008302520151050Number of projectsBox 4.3: Poverty Reduction Support Credits: Multisectoral Development PolicyLending in Support of HNPSince 2001 there has been a steady climb in approvalsof Poverty Reduction Support Credits (PRSCs), which areprogrammatic development policy operations intendedto support certain IDA countries as they implement a nationalPoverty Reduction Strategy (PRS). PRSCs mayinclude simultaneous policy actions in several sectorsaligned with the Strategy and do not necessarily seekto exploit cross-sectoral synergies to produce HNPoutcomes. They are typically implemented in a series ofthree to four single-tranche operations.As of the end of fiscal 2008, 87 individual PRSC operationshad been approved, of which 83 percent hada health policy measure. According to an ongoing IEGevaluation of PRSCs to be delivered in 2009, the five topHNP issues addressed were:• Improved access and utilization of health care• Improved efficiency and accountability of resourceuse• Improved health outcomes• Enhanced service delivery• Improved sectoral management and regulation.These are largely similar to the objectives of HNPmanagedinvestment projects. PRSCs are sometimesused to address sectoral policy issues that require attentionand support beyond a specific sector (for example,by the Ministry of Finance or prime minister)and can be used to enhance the dialogue between theline ministry and the Ministry of Finance. While HNP investmentproject approvals are essentially flat, demandfor PRSCs has been growing.The forthcoming PRSC evaluation will assess, amongother issues, the effectiveness of the PRSC in promotingsector dialogue and achieving results in sectors deliveringservices (HNP, education, and water supply) under varyingcountry conditions; the extent to which outputs andoutcomes are actually tracked; and the extent to whichthere have been sectoral investment or technical assistanceprojects working in tandem with PRSCs.64

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESimpacts on HNP outcomes, such as educationand social protection, the latter being key to thesector’s objective of reducing the impoverishingeffects of illness.The evidence base for this section is an in-depthreview of all 117 water supply and sanitation projectsand 229 transport projects approved from fiscal1997 to 2006. The review examined the extentto which the projects cited potential health benefits(or risks, in the case of transport) in projectdesign documents; included explicit objectives toimprove health (or mitigate health risks); proposedand implemented environmental improvementsthat could plausibly provide healthbenefits; and targeted services and health or behavioraloutcomes to the poor. 15 For the closedprojects—26 for water supply and sanitation and105 for transport—it assessed the extent to whichexpected health benefits or objectives have beenmeasured, achieved, and can be attributed to theactivities implemented.Water Supply and SanitationExtensive evidence has emerged supportingthe potential effectiveness of improvementsin water supply and sanitation infrastructureand hygiene behaviors on health outcomesin developing countries, particularly on theincidence of diarrheal and other waterrelateddiseases. 16 Diarrheal diseases accountedfor an estimated 1.6–2.1 million deaths annuallybetween 1990 and 2000 and remain among the topfive preventable killers of children under five indeveloping countries (Keutsch and others 2006).There is a strong correlation between unhygienicconditions in poor households and communitiesand the frequency and severity of diarrhealepisodes. Water supply and sanitation was thesector most frequently cited in CASs as having potentialbenefits for health.While water supply and sanitation interventionscan have an impact on diarrhealand other water-related diseases, the researchliterature shows that these benefitscannot be assumed. The number of highqualitystudies demonstrating impact is small,and meta-evaluations of impact studies of watersupply and sanitation interventions Over the fiscal 1997–2006show high variability in the findings period, the Bank(Overbey 2008; IEG 2008i). For example,one synthesis of findings across 350 HNP components incommitted $5 billion to10 relatively rigorous studies of the impactof improved hygiene found a me-other sectors.projects managed bydian reduction of 33 percent in diarrheaepisodes, but the reduction ranged from 11 to 89percent (Huttly and others 1997; see also studiesreviewed in IEG 2008i). The variability in impactsis brought about by variations in the technology;the extent to which interventions were actuallyimplemented; pre-intervention levels of pathogens,sanitation, water quality and quantity, andhygiene behavior; and the socioeconomic statusand culture of the beneficiaries. In short, the effectivenessof an intervention depends criticallyon contextual factors, local conditions andpathogens, and technology. There islittle conclusive evidence on the extentto which water supply, sanitation,and hygiene interventions are complementsor substitutes in producinghealth benefits; few studies have testedeach separately and in combination inthe same setting.Even when water supply and sanitation interventionsgenerate health benefits on average,the benefits do not necessarily reachthe poor. Few studies measure thebenefits of water supply and sanitationimprovements on the health of thepoorest beneficiaries; most measureaverage impacts. The handful of studiesthat have examined the distributionof health benefits show that onecannot assume that the poor arehelped. An analysis in rural India, for example,found a positive association betweenexpanding piped water and areduction in the prevalence and duration of diarrheain the lowest two quintiles only if there wasa woman with more than a primary education inthe household (Jalan and Ravallion 2003).Recognition by the Bank’s water supply andsanitation sector of the potential contributionof water and sanitation interventions toWater supply andsanitation interventionscan have an impact onwaterborne diseases, butthose benefits cannot beassumed.Even when watersupply and sanitationinterventions dogenerate health benefits,those benefits do notnecessarily reach thepoor.65

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 4.4: Health Has Been Featured in World Bank Water Supply and SanitationStrategies Since 1993The first World Bank water supply strategy to recognize healthbenefits of water supply and sanitation investments was the 1993policy paper, Water Resources Management: A World Bank PolicyPaper (World Bank 1993b). It emphasized the potential healthbenefits of clean water supply and better hygiene, particularly inreducing the incidence of diarrheal diseases. It also advocatedpublic health education on the safe handling of water to changehygiene behaviors and improve health outcomes, especiallyamong the poor.The 1993 policy paper led to the Bank’s involvement in an increasingnumber of international partnerships such as the GlobalWater Partnership and World Water Council, both formed in1996, and the World Bank–United Nations Development ProgramWater and Sanitation Partnership. However, the comprehensiveprinciples at the heart of the strategy, including those thatparticularly affect health outcomes, initially were not widelyadopted in Bank water-related projects (IEG 2002a). Initiativesspecifically related to sanitation, hygiene, and health becamemore common after 2000, following the World Water Forum andadoption of the MDGs.The World Bank Group’s Program for Water Supply and Sanitationwas adopted in 2004 with objectives and priorities similarto those of the overarching water strategy issued 10 years earlier.It acknowledged the critical relationship between better sanitationand hygiene and improved health outcomes, noting that thehealth benefits from water supply and sanitation investmentsdepend on a “three-pronged strategy: (i) access to sufficientquantities of water; (ii) sanitary disposal of excreta; and (iii) soundhygiene practices.” Improving health outcomes was recognizedas one of five “cross-cutting operational, policy, and institutionalpriorities” requiring both investment in water supply and sanitationinfrastructure and behavioral change. The strategy also advocatedtargeting interventions to the poor as an institutionalpriority. The Sanitation, Hygiene and Wastewater Advisory Servicewas created in 2004, and a health specialist was hired for theprogram in 2005.Source: Overbey 2008.Though half of all watersupply and sanitationprojects cited potentialhealth benefits asjustification for theinvestment, only 10percent had an objectiveto improve health.health outcomes dates back to 1993, whenit introduced its first comprehensive strategyfor water resources (box 4.4). Over thedecade from fiscal 1997 to 2006, the World Bankcommitted more than $7.2 billion in resources to117 new water supply and sanitation projectsworldwide that were managed by the water supplyand sanitation sector.Eighty-nine percent of the 117 water supplyand sanitation projects approved from fiscal1997 to 2006 financed infrastructurethat plausibly could improve health. Theseincluded interventions to improve water supply,sewerage, wastewater treatment, solid wastemanagement and water quality and toconstruct latrines or toilets or handwashingfacilities (figure 4.8). Accordingto project design documents, only28 percent of these projects targeted infrastructureto the poor.While half of all water supply andsanitation projects cited potentialhealth benefits to justify the investment,only 1 in 10 had an objective to improvehealth for which it was accountable. Evenfewer—only 3 percent of all approved projects—had an objective to improve health outcomesamong the poor. Thus, the primary objectivesunderlying the water supply and sanitation lendingportfolio are expanding services, increasingthe efficiency of utilities, and reducing economiccosts and the time it takes to fetch water—not producinghealth benefits or ensuring that thosebenefits reach the poor.Accountability for health outcomes in thewater supply and sanitation lending programappears to be declining. Projects approvedfrom fiscal 2002 to 2006 are less likely tohave been justified by health benefits, to have explicithealth objectives, or to plan to collect healthindicators than projects approved in the precedingfive years (figure 4.9). They are also lesslikely to target behavior change, which is criticalin transforming water supply and sanitation infrastructureimprovements into sustainable health66

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESFigure 4.8: A Large Percentage of Water Supply and Sanitation Projects Invested inEnvironmental Improvements That Could Boost Health OutcomesPercent of projects100908070605040302010089Total28Among thepoor68Watersupply30SewerageEnvironmental Improvements Likely to Provide Health Benefits26Wastewatertreatment22Latrines ortoilets14 14DrainageSolid wastemanagement9Waterquality4HandwashingfacilitiesSource: Overbey 2008.Note: N = 117. Categories are not mutually exclusive, because a project may include more than one intervention.gains. Interviews with water supply and sanitationstaff indicate that since 2000 the sector has focusedlargely on the MDG of improving access tosafe water.The actual health benefits of water supplyand sanitation investments as implementedremain obscure. Among the 26 projects approvedfrom fiscal 1997 to 2006 that have closedand for which there were completion reports,only four documented a change in health outcomes,all of them positive. Those with explicithealth objectives were more likely to collect healthoutcome data than were projects without the objectivesor indicators, though the sample is small. 17Among the projects that showed improvement inhealth outcomes, there was no indication in projectdocuments of specific collaboration withinthe country or the Bank between people in thewater supply and sanitation and HNP sectors.Four water supply and sanitation projectsmeasured positive health outcomes, but attributionof the improvements to water sup-Percent of projectsFigure 4.9: The Share of Water Supply and SanitationProjects with a Health Perspective Has Declined100908070605040302010092Source: Overbey 2008.85Environmentalimprovementswith plausiblehealth benefits5544Health benefitscited asrationale155Explicit healthobjective52Health objectivetargeted tothe poorFiscal years 1997–2001 (n = 62) Fiscal years 2002–06 (n = 55)67

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONThe actual delivery of ply and sanitation interventionshealth benefits by water was weak. The extent to which infrastructureimprovements have been car-supply and sanitationprojects is unclear; only a ried out is generally well documented.few have documented However, poor sanitation and hygienepositive health outcomes. behavior can wipe out any potentialhealth benefits. Few of the projectsmeasured these behavioral indicators, whichwould offer greater confidence in interpretingthe outcomes. Further, the completion reports donot account for the effects of other factors—suchTransport investments canhave both positive andnegative health impacts.as rainfall, better health facilities, or successfulhealth campaigns—on healthoutcomes, a point made by the projectstaff in the case of Madagascar.None of the completed projects measuredhealth outcomes separately for poor andnon-poor project beneficiaries. The NepalRural Water Supply and Sanitation Project hadan explicit health objective and was targeted torural areas, which are apt to be poor (box 4.5).However, the project primarily helped communitiesalong or near main roads, while the reallypoor tend to live in more remote areas. Thefollow-on project is now targeting those remoteareas.TransportTransport investments can produce importanthealth benefits in terms of improved accessto health care, including access toemergency obstetric care, and better distributionof drugs, blood, and other medicalsupplies. Moreover, better access to marketscan lead to greater household income and betternutrition and health. But the impact of better accessto health care depends on context: it will notresult in better health if the health facility offersbad care, for example.Better transport can also have detrimentalimpacts on the environment and on humanhealth. Every year, 1.2 million people die in roadaccidents worldwide (over 3,000 per day), and asmany as 50 million more people are injured, somesuffering permanent disabilities (WHO 2004).Eighty percent of the deaths and 90 percent ofthe losses from road traffic injuries occur in lowandmiddle-income countries. 18 While traffic accidents,injuries, fatalities, and noise have longbeen considered negative externalities, recentevidence suggests that there are direct effects oftransport-induced air pollutants on mortality andrespiratory disease. WHO estimates that sus-Box 4.5: Better Hygiene Behavior through Rural Water Supply and Sanitation in NepalThe Nepal Rural Water Supply and Sanitation (RWSS) Project(1996–2003) had an objective to “deliver sustainable health andhygiene benefits to the rural population through improvement inwater supply and sanitation facilities.”Building on a previous pilot, the project created a semiautonomousRWSS Fund Development Board to finance infrastructure.Support from nongovernmental organizations helpedcommunities implement sustainable schemes. There was apparentlyno formal or informal collaboration between the schemeand the health sector—either in Nepal or within the World Bank.The RWSS project built capacity in rural villages to maintain watersupply and sanitation investments, provided funds for infrastructurethat was cofinanced by communities, and extended hygieneeducation to improve health outcomes. Good sanitation practiceswere reinforced through community grants for latrines sufficientto cover 25 percent of the total beneficiary population at a rate of750 rupees per household. The community then on-lent this to individualhouseholds at an agreed rate of interest for 6 to 12 months.Between 2000 and 2002, the share of residents in project areaswho used tap stands rose from 18 to 91 percent, and the shareusing a latrine rose from 24 to 64 percent. a The share of respondents(mostly women) who reported washing their hands afterdefecation rose from 65 to 88 percent, and the percent washinghands before cooking from 58 to 81 percent. However, there werefewer behavioral improvements in the most remote and poorerareas of western and far-western Nepal, and the impact of betterhygiene on health status could not be confirmed except by areduction in self-reported water-related diseases.Source: IEG forthcoming.a. The project was implemented in four “batches” at different times. These results are from Batch III. Similar results were observed for Batch IV.68

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESBox 4.6: Health and Transport in the World Bank’s Sector StrategiesSustainable Transport: Priorities for Policy Reform (1996) accordeda higher priority to moving people than vehicles, ensuringgreater transport safety, and minimizing adverse health effects.It recommended benchmarked safety programs, adoption ofcleaner fuels, and systematic estimation of the impact of transportprograms on safety and air pollution. It also highlighted threeareas where the Bank could improve its advice and reduce theaccident rate: the separation of motor vehicles from pedestriansand nonmotorized traffic; improvements in driver behavior throughbetter education, regulation, and enforcement; and introductionof geometric road designs that take into account the large numberof vulnerable road users in developing countries.Cities on the Move (2002) advocated the development of a nationalroad accident data collection and analysis capability. Itmaintained that accident frequency and severity can be reducedby improved road design and traffic management. It also recognizedthat poor people are the most vulnerable to the effects ofair pollution, and that little is known about the environmental impactof urban transport. A road safety specialist was appointedto the transport anchor, and collaborative work with WHO waslaunched. In 2004, WHO and the World Bank jointly published theWorld Report on Road Traffic Injury Prevention (WHO and WorldBank 2004), which highlighted the growing public health burdenof road deaths and injuries in low- and middle-income countries.The World Bank contribution was a collaborative effort bystaff in the HNP and transport sectors.HIV/AIDS interventions are a priority for the transport sectorbecause the risk of contracting and spreading HIV at transportconstruction sites is high, and because people engaged in transportare at high risk of spreading HIV along transport corridors.Intensifying Action against HIV/AIDS in Africa (1999) committedto mainstreaming HIV into other sectors, including transport. The2005 Global HIV/AIDS Program of Action committed to includingHIV/ AIDS activities in all Bank-funded construction contracts, inall new transport projects in India and Africa, and in all ongoingAfrica transport projects at mid-term. This approach has been incorporatedinto the recent Transport Business Strategy (2008),which is designed to strengthen transport sector capacity and institutionsengaged with HIV/AIDS prevention strategies. The 2008Business Strategy also commits to make roads safer, includingsupport for governments to develop and implement “strategies,policies, institutions, infrastructure design, vehicle and driverregulations, and enforcement mechanisms.”Sources: Freeman and Mathur 2008; World Bank 2005c, 2008f.pended particulate matter leads to the prematuredeath of over 500,000 people each year (WHO2002). Other potential health risks of transport investmentsinclude water pollution, disease transmission,and reduced physical activity, raising therisks of heart ailments, cancer, and diabetes.Several strategy documents have helpedshape the Bank’s approach to health in relationto the transport sector since 1996(box 4.6). From fiscal 1997 to 2006, the WorldBank committed nearly $28 billion to 229 newprojects managed by the transport sector. Themain health-related transport interventions included:elimination of hazardous locations by rehabilitationor upgrading; improvements in roadtraffic management, bus priority, and risk behavior(such as traffic signals, lane markings,pedestrian crossings, traffic calming, seatbelt andhelmet usage); emissions controls; and publichealth interventions concerning the movement ofpeople along transport corridors ortraveling by air (for HIV/AIDS and othercommunicable disease).About a quarter of transport sectormanagedprojects in the review periodhad an objective to improve health outcomesor mitigate a health risk. Aboutone in five projects had a road safety objective;other objectives having to do with institutional capacityand reducing accidents and “black spots”(hazardous road locations) on roadways were relatedto road safety (figure 4.10). Only a handfulhad objectives related to aviation or waterwaysafety, HIV/AIDS prevention, or improved air quality.19 However, nearly half of the projects had ahealth component (46 percent); almost all of thecomponents were about transport safety (42 percentof all projects), with the next most commoncomponents for HIV/AIDS prevention (8 percent)and air quality (4 percent).About a quarter oftransport projectsincluded an objective toimprove health outcomesor mitigate health risks,but about half had ahealth component.69

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 4.10: The Majority of Health-Related Objectives in Transport Projects Are forRoad Safety302524.0Percent of projects20151019.750Total withhealthobjectiveRoadsafety3.5Institutionalcapacity2.6Accidents/death0.9 0.9 0.4 0.4 0.4BlackspotsAviationsafetyWaterwaysafetyHIV/AIDSAirqualitySource: Freeman and Mathur 2008.While the share of approved transport projectswith health objectives has been stable,the share with health components hasincreased dramatically over time—from36 percent in 1997–2001 to 59 percent in2002–06. Almost all of this increaseThe share of transport can be attributed to transport safetyprojects with health (rising from 33 to 52 percent), and almostall of that to road safety. Thecomponents has increaseddramatically. share of projects with HIV/AIDS componentsat appraisal also rose, from 3to 15 percent. The increase in transport projectswith health components was particularly evidentin middle-income countries, South Asia, Europeand Central Asia, Latin America, and Africa (figure4.11).Projects with healthobjectives are more likelyto plan and collect healthindicators than thosewithout the objectives.One in four transport projects planned tocollect health outcome indicators, but only5 percent planned to collect health outputor behavioral indicators thatwould help to link the projects’activities with health outcomes.The road safety outcome indicators includedthe number of accidents, injuries,and fatalities and accident, injury,and fatality rates. None of the projects proposedto collect health outcome data specifically for thepoor.Among the 105 closed transport projects, aquarter (28 projects) had explicit healthobjectives or financed components with potentialhealth benefits. All of the planned roadsafety, waterway, and port safety componentswere at least partly implemented. However, moreHIV/AIDS components were implemented thanhad been planned at the time that the projectswere designed, which is consistent with the effortsby the health and transport sectors to “retrofit”HIV/AIDS components into ongoing transportprojects.Closed projects with explicit health objectiveswere far more likely to plan and to collecthealth indicators than were projectswith health components but no explicit objective(figure 4.12). Two-thirds or more of transportprojects with explicit health objectivescollected health outcome data, compared withonly a fifth of projects without them. All 15 closedtransport projects that measured health outcomes70

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESFigure 4.11: The Share of Transport Projects with Health Components Has Increased Sharply90Percent of projects807060504030201040526232 33771714294751583160196844740Lower-middleincome (LMIC)Low-income(LIC)Upper-middleincome (UMIC)Middle Eastand NorthAfricaSouth AsiaEast Asiaand PacificEurope andCentral AsiaLatin Americaand theCaribbeanSub-SaharanAfricaCountry incomeRegionsFiscal years 1997–2001 Fiscal years 2002–06Source: Freeman and Mathur 2008.were related to transport safety, and in all butone, the indicators showed an improvement. Reductionsin fatalities per 10,000 vehicles havebeen documented for projects in several countries(figure 4.13).None of the small number of closed transportprojects with HIV/AIDS components collecteddata on health outcomes, so the effectivenessof these activities is unknown. 20 A recent updateon the implementation of the Bank’s GlobalHIV/AIDS Program of Action notes that more thanhalf of the projects in the active transport lendingportfolio include HIV/AIDS activities, most ofthem added after the project was approved. Theexperience in the rest of the transport portfolio suggeststhat health indicators are rarely collected inthe absence of an explicit objective.The link between transport safety improvementsand project outputs was often weak.Outputs were not well documented, other factorsthat might have affected the accident rate were notconsidered, behavioral and other intermediatevariables (such as seatbelt use and accidents byPercentFigure 4.12: Projects with Explicit Health ObjectivesAre More Likely to Measure Health Outcomes80604020067cause) were not measured, and the health outcomedata were often not specific to the areas coveredby project interventions. It is thus difficult to identifythe share of the reductions in fatalities and ac-72Projects with explicit healthobjective (n = 18)Planned to collect health outcomesSource: Freeman and Mathur 2008.20 20Projects without explicit healthobjective (n = 10)Actually collected health outcomes71

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 4.13: Reduction in Fatalities per 10,000 Vehicles in Closed Transport Projects8076Fatalities per 10,000 vehicles60402060523919.98.412.110.5 9.25.87.6 6.601998 2005 1999 2003 1995 2000 1999 2004 1997 2003 1996 1999Bangladesh India Latvia ArmeniaRomaniaLithuaniaCountry and yearSource: Freeman and Mathur 2008, table 5.12.cidents observed in so many of the projects thatarise from traffic safety programs (box 4.7).In summary, CASs invoke the language of synergiesand health benefits to justify the lendingportfolio, but lending activities in complementarysectors generally proceed independently of eachother. While it is likely that investments in othersectors are contributing to HNP outcomes,these benefits are not well doc-None of the projects withHIV/AIDS components umented, and attempts to coordinatecollected data on health or monitor the contribution of Bankoutcomes. support for complementary sectors inproducing health outcomes are rare. 21Multisectoral HNP projects can produce resultsif the degree of complexity can be minimized byprioritizing the sectors, keeping the numbermanageable in relation to country coordinatingcapacity, and ensuring that collaborating agencieshave clearly defined roles and responsibilities.There is great potential for improving HNP outcomesby enlisting both the water supply andsanitation and transport sectors in a more systematicand effective way. However, it is importantto remember that the main objectives of supportin these other sectors are not for health. It wouldmake sense to assess beforehand, case-by-case,the expected costs and benefits at the marginof incorporating explicit health objectives. Onecould argue, however, in the case of transport, thatthere are always potential negative externalitiesof road projects, so road safety objectives mightbe relevant in virtually all cases. In the case of watersupply and sanitation, it is important that the sectorrecognize that its responsibility for achievingthe MDGs is not limited to improved access to safewater, but also includes contributing to the healthand nutrition goals. Incentives to deliver HNPoutcomes are improved in both water supply andsanitation and transport projects if the health objectivesare made explicit. There is virtually no accountabilityfor the results of retrofitted healthcomponents—this is something that needs to beaddressed. Attempts to ensure that the healthbenefits reach the poor are rare.72

THE CONTRIBUTION OF OTHER SECTORS TO HEALTH, NUTRITION, AND POPULATION OUTCOMESBox 4.7: What Accounts for Fewer Road Fatalities in Romania?The Romania Roads II Project had two health objectives—to improveroad safety and reduce lead emissions. Road safety activitiesincluded reducing accident “black spots” and launchinga 10-year Safety Action Plan, which involved public education todeter speeding; wear seat belts; use child restraint devices; andreduce driving under the influence of alcohol, medication, anddrugs. The modernizing, safer national vehicle fleet would alsolikely have contributed to fewer accidents. The accident ratenationally declined, even as the number of registered cars roseby a quarter (see figure, below). The road safety activities reachedtheir height in 2002–03 with the launch of a “Year of Traffic Safety.”Yet, the accident rate was already in decline even before thelaunch of the program. Engineering improvements to the first1,000 kilometers of roads under the previous project, Roads I, isa possible contributory factor; rising fuel prices and a difficult economicclimate might also have led to less driving or driving shorterdistances, even as the vehicle fleet increased. The failure to accountfor factors affecting the accident rate—such as seatbeltuse, sales of child restraining devices, and the number of drugand alcohol-related accidents—makes it difficult to understandthe contribution of road safety activities to the reduction in trafficfatalities.Road Fatalities Began to Decline in Romania Long Before SafetyInterventions Were LaunchedVehicle registration4,500 109.2Year of traffic4,1044,000safety 2002–033,50083,1283,0005.92,5002,000641,5001,0002500001997 1998 1999 2000 2001 2002 2003 2004 2005 2006Vehicle registration (000)Fatalities/1,000 vehiclesFatality rate/1,000 vehiclesSource: IEG 2008g.73

Chapter 5Evaluation Highlights• The health sector is relatively new toIFC and its investments are small andgeographically scattered.• The early projects did not performwell, but performance has improvedand lessons learned led to the 2002strategy.• Support for public-private partnershipshas expanded, especially throughAdvisory Services.• Greater diversity in the portfoliowould improve the social impact ofIFC’s investments.• Health sector activities and responsibilitiesin IFC are segmented andlinked to different departments, implyinga need for internal coordination.• Closer collaboration with the WorldBank on regulatory issues would improvethe climate for expanding IFCinvestments and Advisory Services.

Woman receives a health exam in Beirut, Lebanon. Photo by Alan Gignoux, courtesy of the World Bank Photo Library.

IFC’s Health Strategyand OperationsHealth is a relatively new area to IFC. Until the early 1990s, it had onlya few, sporadic health projects and no health department or specializedhealth staff. Between 1997 and 2007, IFC approved only 54 investmentprojects in the health sector, with total new commitments of about $580 million;health investments represented only 2 percent of total IFC projects andcommitments (appendix D). During the same period, Advisory Services projectson health were approved for a total of $23.67 million.This chapter reviews the evolution of IFC’s approachto investments in the private health sector,examines trends in the content and efficacyof the portfolios of investments and Advisory Servicessince 1997, and assesses the design and implementationof its 2002 health strategy up tothe present. While it is possible to review thecharacteristics of the health investment portfoliosince 2002, most of the investments launchedsince then are not sufficiently mature to evaluatetheir efficacy.Evolution of IFC’s Approach toPrivate Investment in HealthFollowing adoption by the World Bank of its 1997HNP strategy, IFC planted the seeds for developmentof a health strategy with the formationof a Health Care Best Practice Group in February1998 (figure 5.1). The Group was to analyze potentialinvestments in health and to share andleverage knowledge about the health care industrythat was developing across IFC departments(IFC 2002, p. 24).In 1999, the Health Care Best Practice Groupissued “Investing in Private Health Care: ANote on Strategic Directions for IFC,” whichhighlighted the potential contribution ofIFC health investments to improving healthoutcomes for people in developing countries,particularly among the poor (IFC HealthCare Best Practice Group 1999). First, such investmentswere expected to provide private healthservices and finance to low- and middle-incomepatients in the event that public health systemsfailed to reach them with quality care. Second, incountries where private investment not only complementsbut also competes with public sectorhealth services, investment in private health carewas expected to alleviate the burden on overstretchedpublic resources and improve the efficiencyof the health sector more generally. In thelong run, this was expected to enhance health carefor the entire population, including the poor.Third, private facilities were expected to set asidebeds and services for poor clients.IFC subsequently adopted a frontier countrystrategy to steer resources toward underservedsectors in high-risk or low-income countries (IFC1998). Health care, including investments in privateclinics, hospitals, and health care management,was considered a frontier sector, and prioritywas assigned to the social sectors more generally.77

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure 5.1: Timeline of Health Sector-Related Events in IFC (1997–2007)World Bank HNPstrategy (1997)presented to theBoardIFC introducedfrontier countrystrategySocial sectordesignated asone of the prioritysectors for IFCIFC launchedsustainabilityinitiativePrivate sectordevelopmentstrategy of WBGestablishedMajor IFCdepartmentalreorganizationWorld Bank HNPstrategy (2007)presented to theBoardIFC AfricaHealthStrategypresented tothe Board▼1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007▼Health CareBest PracticeGroup formedHealth andEducation UnitestablishedIFC AgainstAIDS initiatedGlobal practicegroup for socialsector formedInvesting inHealth: Noteon StrategicDirections for IFCSource: IEG.Note: WBG = World Bank Group; SME = small and medium enterprises.▼▼▼Health andEducation Unitbecame adepartment▼▼ ▼▼ ▼▼ ▼ ▼▼2002 HealthStrategypresented tothe BoardLife Sciencegroupestablishedwithin the GlobalManufacturingDepartment▼IFC AgainstAIDS integratedwith SMEdepartment▼In April 2000, IFC established the Global PracticeGroup for the Social Sectors, which becameIFC’s Health and Education Department in September2001.The 2002 strategy “Investing in PrivateHealth Care: Strategic Directions for IFC”was the first IFC health strategy presentedto the Board of Directors. The goals for thesector were broadly defined: to improve healthoutcomes, protect the population from the impoverishingeffects of ill health, and enhance performanceof health services (IFC 2002,IFC’s first Boardapprovedstrategy was jectives (see chapter 1, table 1.2). Thep. 3). The strategy had two sets of ob-issued in 2002; it planned business objectives were similar toto continue to invest in those of any other IFC investment: “financingshould be provided in situa-hospitals but to diversifyinto other areas as well. tions where other investors are notprepared to invest; the investmentshould encourage private resourceflows through a demonstration effect; investmentsshould be made only in financially viableoperations; and managerial and financial valueadded should be provided where necessary” (IFC2002, p. 32). The developmental objectives soughtto ensure that IFC investments would contributeto institutional and systemic capacity building andpromote efficiency and innovation within the sector,while improving health security and expandingfinancial protection against the impoverishingeffects of ill health (IFC 2002, p. 4).Under the 2002 strategy, IFC planned tocontinue to invest in the hospital sector butto diversify its portfolio. Hence, it focusedmore on private health insurance to benefit thelower-middle and middle classes in countrieswithout universal risk pooling; support for supplementaryinsurance; investment in pharmaceuticalproduction, medical device manufacture,and biotechnology; and greater investments in theeducation and training of health workers (IFC2002 p. 7). The strategy proposed six follow-up actionsto implement the new investment strategy:develop instruments for small projects; increase78

IFC’S HEALTH STRATEGY AND OPERATIONSefforts to reach the poor; enhance promotionalactivities; strengthen collaborative efforts andleverage existing knowledge; deepen collaborationwith other multilateral financial institutions;and develop an M&E framework for IFC investmentsin the health sector (IFC 2002, p. 37).The 2002 strategy took into account both theinternal and external environments of thehealth sector. Internally, it considered the relativenewness of health investments to IFC.Externally, three issues were considered (IFC2002, pp. 6–19). First, demographics were changing;health expenditures were increasing; therewere advances in medical technology and practice,information, and biotechnology; and changesin consumer behavior and expectations were increasinginequity. Second, global trends pointedto a significant and expanding role for the privatesector, particularly as a partner with the public sectorin the provision of health care. Many governmentswere rethinking the roles of public andprivate agents in the health sector and were beginningto turn to market instruments to enhancethe efficiency and quality of health care provision.Third, the strategy acknowledged the WorldBank Group mission of reducing poverty andpromoting economic growth and the role of theprivate sector for growth and poverty reduction(IFC 2002, p. 21).The strategy subscribes to and supports thegoals of the Bank’s 1997 HNP strategy (IFC2002, p. 20). The IFC strategy identifies the differentroles of the World Bank and IFC, recognizestheir potential complementarity, and advocatescloser collaboration. 1 IFC clearly recognized thatwhile it has the mandate, the growing staff capacity,and the instruments to address directly private sectorfinancing, the staff of the Bank’s Human DevelopmentNetwork have the relevant sectorknowledge and the mandate to inform and leadpolicy dialogue with country authorities and tosupport needed changes through the developmentand financing of public sector operations(IFC 2002, p. 41). Moreover, the strategy includesa specific subobjective that health investmentsshould support the Bank’s health sector objectives.It mentions that discussions with Bank colleaguesthrough the appraisal processare vital and aim at deriving a meaningfullyintegrated approach betweenIFC and the Bank—one that shouldmaximize the benefits to clients and toboth organizations (IFC 2002, p. 23).The strategy did not specify targets, estimatethe resources needed for implementation,or put in place M&E arrangements.Goals for the sector were defined broadly, and astrategic investment approach that would maximizeIFC’s impact and role in the health sector wasnot defined. There was no estimate of the resourcesrequired to implement the strategy. Thestrategy proposed to refine the evaluationframework and to develop a setof baseline indicators that would adequatelymeasure IFC’s health sectorobjectives. However, the strategy itselfprovided no indicators.Since 2004, 2 addressing the regulatory, financing,and implementation constraintsto private sector investment in infrastructure,health, and education has been one ofIFC’s five strategic priorities. For the healthsector, IFC’s corporate strategy aimed to:• Facilitate public-private partnerships (PPPs).• Provide input and advice on regulatory improvements.• Provide innovative financing, advice, and projectdevelopment activities with local privateplayers, such as expanding the use of local currencyand taking new approaches to financingsmall health facilities.• Develop cooperative programs with the WorldBank (IBRD and IDA) to address sectorwideconstraints comprehensively.IFC’s strategy in the health sector is continuingto evolve. The Health and Education Department’sstrategy is moving away from top-tierbricks-and-mortar projects to focus on repeat investmentswith strategic clients with broad reach,South-South investments in large, for-profit providers,and in local clients’ operating networksof hospitals. The Life Sciences strategy is to em-The strategy supports thegoals of the Bank’s 1997HNP strategy.It did not specify targets,estimate resources neededfor implementation,or establish M&Earrangements.79

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONOver fiscal 1997–2007,IFC approved 54investments in the healthsector with $580 millionin commitments.phasize innovation through new products to addressunmet needs and neglected diseases, andbroader access to medicines through productionof generics with high quality standards. IFC is extendingits partnership with private foundationsin the area of health. 3 In 2007, IFC adopted astrategy to improve private financing and provisionof health services in Sub-Saharan Africa, partneringwith other institutions such as the Bill andMelinda Gates Foundation, with the objective ofmobilizing up to $1 billion over five years in investmentand Advisory Services to boost healthcare in the Region. Because this strategy has onlyrecently been launched, it is not a focus of thisanalysis.IFC’s Investment Portfolio in HealthBetween fiscal 1997 and 2007, IFC approved54 investment projects in the health sector,with total new commitments 4 of about$580 million. The average investment in individualhealth projects was therefore small—onlyabout $11 million. Health investments representedonly 2 percent of total IFC projects andcommitments for the period. The majority of theprojects have been in hospitals, pharmaceuticals,and biotechnology (table5.1). IFC had only one project each inmedical equipment manufacture, medicalequipment leasing, medical training,and health payment processing. 5Trends in IFC Health InvestmentsThe health investment portfolio grew unevenlyover the fiscal 1997–2007 period.New investment commitments were stable between1997 and 2002, but after the establishmentof the Health and Education Department in 2001and introduction of the 2002 strategy, the numberof new projects and commitment volume actuallydropped. Investment growth then surged from2005 onward (figure 1.3, chapter 1). This is consistentwith a pattern observed in the dynamics ofother industries in IFC where newly established departmentsfirst go through a consolidation phase,sorting out portfolio issues and developing strategiesand standards for investments, before embarkingon an expansion phase.For the purposes of evaluation, IFC project approvalscan be divided into three periods: (a)committed projects approved before fiscal 2000,prior to the issuance of any strategic documentsabout IFC and health; (b) committed projectsapproved from fiscal 2000 to 2002, when the Departmentof Health and Education was establishedand the Board was presented with a formalstrategy for the sector; and (c) committed projectsapproved from fiscal 2003 onward. The analysisof the first period is only illustrative, since nostrategic directions were in place. The secondperiod is short, but it has been used to assesswhether any of the ideas of the 1999 Strategic NoteTable 5.1: IFC Health Investments by Type of Investment and PeriodProjects by fiscal year of approvalType of investment Total Total net commitmentsHospitals Pharmaceuticals Laboratory Other a projects $US million (Percent)1997–99 8 2 — 1 11 70 (12.1)2000–02 7 4 3 — 14 140 (24.1)2003–07 14 11 1 3 29 370 (63.8)Total 29 17 4 4 54(Percent) (53.7) (31.5) (7.4) (7.4) (100)Net commitmentsTotal 298 197 41 44 580 (100.0)(Percent) (51.3) (34.0) (7.0) (7.7) (100.0)Source: IEG data.a. Medical equipment leasing, medical training, medical equipment manufacture, and provider of medical payment transactions services.80

IFC’S HEALTH STRATEGY AND OPERATIONSTable 5.2: Geographic Distribution of IFC Projects by PeriodEurope and Latin America Middle EastPeriod (fiscal years) Africa Asia Central Asia and Caribbean and North Africa1997–99 0 2 3 6 02000–02 0 5 3 5 12003–07 2 13 8 4 2Total 2 20 14 15 3Source: IEG data.were implemented. It is too early to assess theprojects in the third period.Hospitals have continued to be the mostsignificant type of investment, but the shareof hospital projects has decreased from almost80 percent in the first period to around50 percent in the later periods (table 5.1). 6The notable increase in life science projects in thethird period reflects the efforts of the specializedLife Sciences Group in 2005.Although the number of projects and netcommitments expanded in life sciences,IFC did not expand operations in private insuranceas the strategy intended. 7 To date, IFChas not financed a freestanding insurance project,though it has invested in some prepaymentarrangements. The lack of proper regulatoryframeworks in countries, weak IFC focus and experiencein the subsector, and the lack of institutionalarrangements required to promote theseinvestments (such as regulatory regimes) contributedto the gap.Health projects have been concentrated inthree Regions—Asia, Latin America and theCaribbean, and Europe and Central Asia.These Regions account for 89 percent of the projectsin the sector. This distribution follows thegeneral pattern of IFC’s portfolio as a whole.However, the distribution of commitments indicatesthat Asia has a much larger share in thehealth portfolio.Geographically, IFC health sector operationsdid not experience increased diversificationuntil after fiscal 2002(table 5.2). Before 1999, IFC’s healthprojects were concentrated in LatinAmerica and the Caribbean. The onlyhealth-related operations in Africa werethose in the Africa Enterprise Fund(AEF). 8 The distribution evened outacross Regions from fiscal 2000 to 2002.Then, from fiscal 2003 to 2007, IFCcommitted to health projects in Africafor the first time and substantiallyexpanded its operations in Asia andEurope and Central Asia. The net commitmentvolume followed a similar geographicdistribution pattern.IFC’s health investments are concentratedin middle-income coun-diversify geographicallyIFC investments did nottries. This, too, largely mirrors the until after 2002.pattern in IFC’s overall investmentportfolio. But there is a larger concentration ofhealth investments in low-risk countries and aslightly larger concentration in nonfrontier countries(middle-income countries not at high risk).These results could be explained by IFC’s greaterexperience working in low-risk countries wherethe environment for business is favorable. However,health itself was considered a frontier sectorin 1998 (IFC 1998) because the operationswere new and risky. Data for the corporation asa whole show a relatively more even distribution.Efficacy of IFC’s Investment Operationsin HealthIEG’s evaluation framework for assessing theefficacy of IFC operations focuses on both IFC’sinvestment outcome 9 and project-level develop-Investments areconcentrated in middleincomecountries andhospitals continue toaccount for the largestshare of investment.Although investment inlife sciences expanded, itdid not grow in privateinsurance as the strategyintended.81

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable 5.3: Profitability of IFC Health and Pharmaceutical InvestmentsHealth projects’profitability (percent)Pharmaceutical projects’profitability (percent)Excluding Including Excluding Includingunrealized unrealized unrealized unrealizedPeriod (fiscal years) capital gains capital gains capital gains capital gains1997–99 –51.0 –48.0 –3.4 –2.42000–02 –8.6 –6.7 –13.0 –14.52003–07 –2.2 –0.4 11.6 33.3Source: IEG data, based on IFC’s accounting of operational income and expenditure by project. The data reflect the profit and loss calculation of all projects active during each financialreporting year.ment outcome, based on methodology consistentwith the Multilateral Development Bank’s GoodPractice Standards for Evaluation of Private SectorProjects (MDB, ECG, and WGPSEThe early health2006, p. 2). The development outcomeinvestments, including inof projects is based on four developmentdimensions—project businesspharmaceuticals, werehighly unprofitable.success, economic sustainability, environmentaland social effects, and private sectordevelopment impact—relative to what wouldhave occurred without the project, and comparedagainst established benchmarks as well asstated objectives. 10In fiscal years 1997–99,both investment anddevelopment outcomeswere low.IFC’s health investment results between fiscalyears 1997 and 2002, including pharmaceuticals,were highly unprofitable (table5.3). The financial losses, minus 51 percent inprofitability for health in the first period andminus 13 percent for pharmaceuticals in the secondperiod, indicate that these lines of businesswere a significant cost to IFC, a situation that wasnot sustainable. In the subsequent periods, IFCcontained its losses. More recently, pharmaceuticalbusinesses have started to contribute to IFC’sbottom line, and the profitability of health operationshas been improving.The evaluation assessed the resultsof projects that reached earlyoperating maturity. This occurswhen the project has been substantiallycompleted and has generated atleast 18 months of operating revenuesfor the company, and when IFC has received atleast one set of audited annual financial statementsthat cover at least 12 months of operatingrevenues (MDB, ECG, and WGPSE 2006, p. 2). Thisusually happens about five years after project approval.Thus, only projects in the first and secondperiods meet early operating maturitycriteria and will be fully assessed below,while projects of the third, most recent, periodwill be included selectively, based onthe extent to which particular effects couldbe discerned at this stage.In the first period (fiscal years 1997–99), investmentand development outcomes wereboth low. Only 18 percent of health projects (2out of 11) achieved both a high development ratingand high investment return (figure 5.2), comparedwith 48 percent for all IFC investments.About two-thirds of projects in the first periodwere referred to the Department of Special Operations—adedicated team that handles projectsin distress—for workout, restructuring, rescheduling,and recovery. The reasons for these referralsincluded long delays in completing projects,negative influence of sponsors’ other lines ofbusiness, difficulties in increasing the number ofnew patients served, the impact of financial crisesin certain regions, and delays in obtaining regulatoryclearances. The experience also reflectsIFC’s weakness in screening and structuring healthsector deals owing to a lack of sector-related experienceduring the period. Frequently, IFC providedforeign exchange–denominated loans to82

IFC’S HEALTH STRATEGY AND OPERATIONSclients, but this approach was not appropriate tothe business, which generates revenues mainly inlocal currencies.Figure 5.2: IFC Development and InvestmentOutcomes in Two PeriodsIn many cases these early poor-performing operationsresulted in: complete failure of the businessand bankruptcy of the sponsor company,followed by sale of the project company to a thirdparty; early termination of investment funds, becauseof poor ramp-up of the investment portfolio;abandonment of the project construction; or illiquidity,when a company ran out of cash from operationsin other countries. Other projects withlow development outcomes included hospitalswith large underutilization of facilities and heavyfinancial losses in early years. Although some ofthe businesses are still operational, the financialand economic returns were lower than the benchmarks.Pharmaceutical projects with low developmentoutcomes failed to achieve export marketpenetration, which was the original project objectiveand critical for their business growth. Theyalso encountered policy obstacles (price controls,collection problems with state health funds).Lessons from hospital projects in the first periodare summarized in box 5.1. The projects with lowdevelopment outcomes tended to confirm theperception of the health sector as high risk, andthus had negative demonstration effects.Development RatingLOW HIGHDevelopment RatingLOW HIGH0%High developmentratingLow IFC return64%Low developmentratingLow IFC returnLOWHealth SectorIFC Investment ReturnBased on 11 projects0%High developmentratingLow IFC return21%Low developmentratingLow IFC return18%High developmentratingHigh IFC return18%Low developmentratingHigh IFC returnPeriod 1 (Fiscal 1997–99)HIGH71%High developmentratingHigh IFC return7%Low developmentratingHigh IFC returnDevelopment RatingLOW HIGH10%High developmentratingLow IFC return33%Low developmentratingLow IFC returnLOWPeriod 2 (Fiscal 2000–02)Development RatingLOW HIGHAll IFC48%High developmentratingHigh IFC return8%Low developmentratingHigh IFC returnIFC Investment ReturnBased on 201 projects3%High developmentratingLow IFC return27%Low developmentratingLow IFC returnHIGH58%High developmentratingHigh IFC return12%Low developmentratingHigh IFC returnLOWSource: IEG data.IFC Investment ReturnBased on 14 projectsHIGHIn contrast, nearly three-quarters of projectsapproved in the second period (10 outof 14 projects, fiscal 2000–02) had high developmentoutcomes and a high investmentreturn, substantially better than the rest ofIFC portfolio (figure 5.3). Only 4 projects wererated low in development outcome. One was acomplete business failure, and the company hasbeen liquidated. Two were hospitals that had lowerthan-expectedpatient admittance and struggledto ramp up revenues. Both are operational and expectto be sustainable, though with a lower returnto financiers than anticipated. The fourth, a pharmaceuticalproject, decided to drop investmentsin research and development of new drugs followingpessimistic market views of blockbusterdrugs, 11 and to cut back on employment. Although3 of the 10 projects with high development outcomesshowed business results belowthe benchmark, all met or exceededeconomic benefit benchmarks and contributedpositively to private sectordevelopment. Between the first andsecond periods, performance improvedin all four dimensions of developmentand investment outcomes (figure 5.3).LOWIFC hospital projects offer several potentialbenefits to governments. First, to the ex-IFC Investment ReturnBased on 179 projectsHIGHNearly three-quartersof the projects in thesecond period had highdevelopment outcomesand high investmentreturn—better than therest of the IFC portfolio.83

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 5.1: Early Hospital Investments Provided Important Lessons1. Sponsors of hospital projects may not be focused on thebottom line.Hospital projects sponsored by doctors are unique—the sponsorsinvest in and work in the facilities they build, and trade-offs arealways necessary. Doctors may not expect financial returns (especiallyfrom equities), but they may prioritize access to facilitiesfor more lucrative practices. Professional management in financialcontrol is necessary. IFC must insist that measures be put inplace to control expenditures during the construction phase aswell as during operations.2. Just because people are sick, that does not mean they willgo to a new hospital.Competitively positioning a health care project in the local marketis difficult. To be successful, a project must establish that thequality of service provided justifies its higher cost. If a hospitalcharges too much, potential clients will find other options forcritical care. At the time of appraisal for a new health care project,IFC should carefully review the market plan.3. IFC should expect conflicting interests in health careprojects among stakeholders and plan accordingly.Managing the relationships among doctors, sponsors, managers,contractors, technical partners, and others is complex, and conflictsare bound to arise. IFC should review the procedures for disputeresolution at appraisal and monitor them during implementationand operation. It is particularly important that such proceduresinclude ways to deal with cash shortfalls. IFC has experienced difficultywith different project structures, including 50–50 joint ventures,single sponsors, and multiple sponsors. The strengths andweaknesses of the proposed structure should be reviewed carefullyat appraisal, and weaknesses should be mitigated.4. IFC should be cautious in providing loans denominated inforeign currencies.Because health care services usually generate revenues in localcurrency, providing foreign currency loans increases project risk.In countries vulnerable to capital flight, IFC should consider mechanismsto provide local currency loans. If this is not possible, itshould consider either providing guarantees to local banks thatsupply local currency loans or establishing a hedging mechanismat secure, global banks.Source: IEG analysis of ELRN data.Percent of projects rated highFigure 5.3: IFC Evaluation Results Show SubstantialImprovement in the Second Period9080706050403020100Source: IEG data.18 18Projectbusinesssuccess712285277950 50Economic EnvironmentsustainabilityandsocialDevelopment outcomePrivatesectordevelopment6383Loan0EquityInvestment outcomeFiscal years 1997–99 (n = 11) Fiscal years 2000–02 (n = 14)tent that they attract patients among the nonpoor,private hospitals can reduce the burden onpublic hospitals so that they can focus on theneeds of the poor. Second, the projects may introduceservices not previously available, increasecapacity, or improve hospital management. For example,a project in the Philippines was the firstmajor tertiary hospital built in the capital city in25 years. Third, projects supported by IFC oftenhave state-of-the-art technology and equipmentthat are out of the reach of governments.Managers of client medical institutions in Argentinaand the Philippines reported that stateof-the-artfacilities in the country attracted doctorsthat had established successful careers in medicalinstitutions in Europe and the United States. Fourout of seven IFC projects appeared to attract doctorsback to their native countries to practice.However, some facilities experienced difficultiesin retaining quality nurses. Training for medical em-84

IFC’S HEALTH STRATEGY AND OPERATIONSployees was observed in projects with both highand low development outcome. Furthermore,three of them have formal arrangements withuniversities or medical schools to accept internshipsor continuing education for doctors in theirfacilities.Many hospitals supported by IFC have addressedgovernance issues in the privatehealth sector. IFC projects in Eastern Europe introducedthe posting of fees that doctors arecharging, with the objective of reducing informalpayments. Other projects have introducedsome control of doctors’ side practices outside theinstitutions. More generally, IFC requires financialand environmental reporting that prompts clienthospitals to improve their management practices,which in turn can bring forth better decisionmaking.IFC’s Advisory Services in HealthTrends in IFC Advisory Services for HealthIFC’s Advisory Services projects have workedto address the regulatory, financing, andimplementation constraints to private investmentin the health sector. 12 Between fiscalyears 1997 and 2007, 36 advisory projects inhealth were approved for a total of $18.48 million,excluding HIV/AIDS projects. The largest numberof such projects was for studies relatedto investment project preparation andfollow-up (table 5.4). This is consistentwith the need to address IFC’slack of knowledge about the best wayto do business in health.Advisory Services resources forpublic-private partnerships comprisethe largest share of total fundingand have shown the largestincrease over the study period (box5.2). From fiscal 2003 onward, AdvisoryServices have increasingly worked withgovernments to encourage publicprivatepartnerships in the form ofseminars and advice on design and implementationof IFC investment projects(table 5.4). Public-private partnershipscan be effective for expanding IFC’sreach, but they have been difficult to implementbecause it is the governmentthat decides which partnership modelwould be the most appropriate. Thedecision also depends on the government’s capacityto regulate and effectively control the qualityof health care delivery. This is important for thesuccess of the partnership, because after the agreementis signed, there are economic incentives toreduce the provision of services or their quality. Fur-Projects for public-privatepartnerships make up thelargest share of AdvisoryServices funding andproject approvals.Managers of two clienthospitals reportedthat upgraded facilitieshelped attract doctorswith successful careersin Europe and theUnited States.Many hospitals supportedby IFC have addressedgovernance issues.Table 5.4: IFC Advisory Services in Health (1997–2007)Assistance Public- IFCTime period to investment private Against(fiscal years) clients Retail a partnerships Studies AIDS TotalProjects1997–99 5 4 92000–02 3 1 7 112003–07 2 2 9 3 16Total 10 2 10 14 n.a. b 36CommitmentsTotal (US$ million) 1.02 0.45 12.51 4.50 5.19 23.67(Percent) (4) (2) (53) (19) (22) (100)Source: IEG.a. Retail refers to IFC’s direct assistance to nongovernmental organization/grassroots business entities through its Advisory Services window.b. HIV/AIDS activities, carried out by IFC Against AIDS, were not recorded in the centralized database until recently, and historical project-specific data are not able to be mapped outagainst the other Advisory Services operations at this time.85

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 5.2: What Are Public-Private Partnerships inHealth?A public-private partnership in health is any joint program or project thatinvolves collaboration between the public and private sectors, includingcontracting between the public sector (government or developmentagencies) and private providers to offer health services and goodsand/or private financing to support health infrastructure. Governmentsworldwide are developing such partnerships in health to manage increaseddemand for health services and fiscal constraints. Private participationin public hospitals is sought to improve efficiency, reducecosts, and transfer operational risk.Public-private partnerships can take many forms with different degreesof private sector responsibility and risk. They are differentiatedby whether the private sector manages medical services, owns orleases the facility, employs the staff, or finances and manages capitalinvestments.IFC Against AIDS wasinitiated to get businessesinvolved in the fightagainst the diseasethrough risk managementand workplace programs.A new system forevaluating IFC AdvisoryServices has been piloted,but only 10 advisoryoperations in the healthsector have beenevaluated.thermore, public consensus is neededbecause, in practice, opposition canarise over the fairness of privatizationand health care cost increases. The otherchallenge for IFC has been to find privateinvestors willing to participate inthese partnerships. To meet these challenges,IFC emphasized Advisory Serviceoperations to governments.The Regional distribution of Advisory Servicesfunds differs from that of investmentprojects. Africa accounted for 38 percent of thetotal funding, Europe and Central Asia for 28 percent,while Asia accounted for only 6 percent.The significance of IFC funding for AdvisoryServices in Africa is explainedby IFC Against AIDS, as well as by a$3.2 million study (almost 14 percentof total funding) to analyze opportunitiesfor private health investments inAfrica, which also shaped the new IFCBusiness of Health in Africa initiative.IFC Against AIDS, a corporate initiative, was initiatedin 2000 as a response to the identificationof the HIV/AIDS epidemic as a critical obstacle forsustained competitiveness of enterprises (costincreases, productivity drops, and losses of experiencedpersonnel) in regions where the diseaseis more prevalent. The aim of the initiativeis to get businesses involved in the fight againstHIV/AIDS through risk management and implementationof workplace programs in countrieswhere IFC operates. The program is detailed andassessed in box 5.3.Efficacy of Advisory ServicesIFC is rolling out a new system for evaluatingits Advisory Services operations. Under thissystem, Project Completion Reports (PCRs) will beprepared by IFC staff, cleared by IFC management,and reviewed for quality of content by IEG. Thissystem assesses development effectiveness acrossfive dimensions: strategic relevance, efficiency,outputs, outcomes, and impacts. It also rates AdvisoryServices on the role and contribution of IFC,which reflects the extent to which IFC brought additionalityor some special contribution to theproject. The system has been implemented on apilot basis on 293 operations that closed in2004–06. Of these, 10 are in the health sector.Inferences about the performance of AdvisoryServices in health are constrained bythe small sample of evaluated health projectsand considerable data gaps. Among thesmall group of health Advisory Services that wasevaluated, ratings on some of the dimensionswere not available. For example, only 2 PCRs hadvalid judgments on the dimension of impactachievement (in both cases, negative), while theresults of the other 8 projects were either inconclusiveor it was too early to tell at the time ofproject completion (table 5.5). The overall developmenteffectiveness for Advisory Servicehealth projects is based on only 6 projects. 13 Projectsrated high include public-private partnerships(see box 5.4 for a successful example) andstudies for the introduction of health insurance.The six Advisory Services projects in health thatwere rated on outcome achievement performedbelow those in the rest of IFC: only a third hadhigh outcome achievement (meeting expectedoutcomes or better), compared with 71 percentof projects in other IFC sectors. The cost-effectivenessof the health projects was also lower.86

IFC’S HEALTH STRATEGY AND OPERATIONSBox 5.3: IFC Against AIDS: A Preliminary AssessmentIFC initiated its IFC Against AIDS program in 2000. Its mission isto protect people and profitability by being a risk-management partner,HIV/AIDS expert, and catalyst for action. The program hasthree objectives: to prevent new infections, to deal with or manageexisting infections, and to mitigate the effect of HIV on the companyitself. The biggest beneficiaries of the initiative have beencompanies in Africa, but IFC has been trying to expand to otherRegions and countries as well (China and Russia). Except for oneoperation conducted by the Private Enterprise Partnership forAfrica (PEP-Africa), all IFC Against AIDS work is directly relatedto IFC investments.Completion Reports have been issued for IFC Against AIDS programcomponents, with the following findings:• The program objective has been relevant and aligns with thestrategic directions and priorities of IFC. It takes advantage ofIFC’s position with private sector clients and links their corporaterisk management issues to the social agenda. By targetingworkers, IFC Against AIDS is able to work with the adultpopulation, which is difficult to reach by traditional healthawareness campaigns (IFC 2006, p. 27).• The program has developed products and services, tools, andpublications to implement HIV/AIDS policies and training programsin businesses. As of January 2008, the program hadreached 88,000 employees and over 628,000 people in nearbycommunities (IFC 2008). In Africa, the program engaged in activitiesranging from large seminars to assisting individual firms(large companies and small and medium enterprises) as well asnongovernmental organizations. In India, it worked with fourclient companies (covering about 20 sites nationally) to proactivelyaddress HIV/AIDS in the workplace, clinical facilities,and communities at risk. IFC prepared and disseminated studieson Occupational Health and HIV/AIDS Perceptions in Russiaand a briefing book on HIV/AIDS in China.• IFC Against AIDS often relies on other departments/operationsto reach potential clients. Since it is targeted to IFC clients, closecooperation across IFC was critical. However, in a joint projectwith the Africa Project Development Facility (APDF, predecessorof PEP-Africa), the training program lost momentum becauseof IFC organizational changes and shifting strategicpriorities. Similarly, IFC Against AIDS encountered a problemabout the selection criteria of the small and medium enterprises,which were opportunistic and resulted in the inclusionof some businesses that were not linked to IFC activities.• The project in China only delivered a handbook after nearlytwo years of implementation. Although the project seems to havebeen in support of business development aligned with IFC’svalue added on the AIDS agenda, IFC appears to have droppedthe project after the business development mission to the country.IFC did not achieve the two goals—increased involvementof private enterprises in the fight against HIV/AIDS by being arisk-management partner to clients, and being supportive of IFC’sbusiness development by providing value-added services andcontributing to IFC’s brand.Since 2006, IFC Against AIDS has incorporated a broader approachthat also considers tuberculosis, malaria, and maternalhealth. Although it has been in operation for seven years, IFCAgainst AIDS has not been subject to a systematic outcomeand impact assessment of the whole program. In December2007, IFC launched a comprehensive external evaluation of theprogram.Source: IEG review of Project Completion Reports.IFC has also started to assist social enterprisesthat are directly serving rural andpoor people. Among these is the SustainableHealth Enterprise Foundation in Kenya, a microfranchisenetwork of outlets that extend affordablehealth care and medicines to rural areas,which includes community pharmacies andcommunity-based clinics that employ certifiednurses. IFC’s Grassroots Business Initiative helpedthe network to become autonomous by providingcapital loans to partially support the organization’sefforts to improve its operational performanceand franchise management. Anotherexample is GNRC community pharmacy outletsin northeast India. The Advisory Service projectwas to transform community “medi-shops” intoretail pharmacies delivering not just medicines,but also health planning and counseling.This was done with a partnershipbetween IFC and GNRC Hospital,with IFC providing advice on businessoperations.IFC has started to assistsocial enterprises thatdirectly serve rural andpoor people.87

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable 5.5: Percentage of Health and IFC Projects Rated High, PCR Pilots 1 and 2Health sector projectsNumber ofIFC overallMain ratings and dimensions Percent projects (percent)Development effectiveness 33 6 69Strategic relevance 80 10 89Efficiency 33 9 72Output 60 10 84Outcome 33 6 71Impact 0 2 57IFC role and contributions 78 9 87Source: IEG data.Box 5.4: Outpatient Dialysis Services in Romania—A Successful Advisory ServicesPublic-Private Partnership in HealthBetween 2003 and 2005, Romania privatized dialysis services.Previously, inpatient and outpatient dialysis services were providedin about 40 public hospitals. Dialysis supplies were purchasedby the National Health Insurance Fund (NHIF), and theequipment was purchased by the government and allocated to thehospitals. Funding difficulties and increasing demand for serviceshad created a backlog of patients, and the facilities neededto be upgraded and expanded.IFC’s Advisory Services helped the government: (i) revise andupdate national dialysis standards and practices and preparelegislation, including harmonization with European Union clinicalguidelines; (ii) establish regional survey reporting of dialysis costand prices; (iii) conduct cost analysis of dialysis; and (iv) createmodel tender documents.The partnership was structured as a contract for dialysis servicesbetween the government and private service providers,who became owners. The Ministry of Health set prices, and theprivate operators were required to re-equip the facilities andprovide about 200,000 outpatients a year with dialysis treatmentsat quality and service standards comparable to those of the EuropeanUnion. The individual contracts, totaling €20 million peryear, could be extended if the operator constructed a new facilitywithin 18 months. All public patients were to receive free treatment,and the operators were to be reimbursed by the NHIF underspecified fee schedules.The Advisory Services project introduced transparency intogovernment procurement processes. In 2004, NHIF conductedsimultaneous tenders for eight dialysis centers, with strict prequalificationcriteria to ensure participation of experienced providers.Bidders were restricted to winning two centers to limit marketconcentration. Winning bidders were selected based on highest investmentcommitments. The average investment commitment in acenter was $2 million, which was realized within 18 months afterawarding the contracts. NHIF did not have to finance the modernizationfrom public funds and there were significant operational costsavings to NHIF (estimated at about $4 million). The project improvedpatient services at lower cost to the national health system. The qualityof services and patient satisfaction increased at lower cost becauseof the new standards, improved equipment and facilities, andmore efficient organizational structure.Sources: IEG review of Project Completion Report; Nikolic and Maikisch 2006; Maikisch 2007.IFC’s Institutional Arrangementsfor the Health SectorThe activities and responsibilities in IFC relatedto the health sector are fragmentedand linked to different departments, placinga premium on coordination across units.While health facilities and service providers arethe responsibility of the Health and EducationDepartment, the industries related to the healthsector—pharmaceuticals, medical supplies, and88

IFC’S HEALTH STRATEGY AND OPERATIONSmedical equipment—are all the responsibility ofthe Global Manufacturing and Services Department(CGM), under its Life Sciences Group. However,that group also handles investments inproducts, such as biotechnology, cosmetics, andnutritional products. 14 Advisory Services are the resultof the work of different units across IFC: theInfrastructure Advisory Department, Small andMedium Enterprise Department, and numerousRegional facilities, as well as Regional departmentsand the Health and Education Department. Beyondthis, other IFC units implement health-relatedprojects with varying degrees of specialization.IFC’s organizational structure is broadly inline with the traditional organization of thehealth sector, but recent developments inthe sector pose challenges to this structure.Historically, there has been a clear separation inthe industry between the main players—manufacturersand health providers. However, thehealth care market is becoming more integrated.The boundaries between statutory and privatehealth insurance are becoming blurred; payers andproviders are coming together in integrated healthcare provision and e-health; and the industry andhealth insurance companies are developing jointbusiness models. 15 These developments requiremore integrated approaches to business developmentand client service in IFC’s health sector.Many of IFC’s potential clients are operating inboth pharmaceutical and service provision, andthese two are complementary in addressing healthoutcomes. In IFC, groups dealing with health arepart of different departments, and incentives differand have inhibited response to business opportunitiesand implementation of cross-sectoralinitiatives.The Health and Education Department issmall—until recently the smallest of IFC’sinvestment operations departments. TheDepartment averaged 19 full-time/full-year professionalstaff between 2001 and 2005. 16 The numberof staff increased by 42 percent, to 27, from2006 to 2007. Investment officers in the departmentare not specialized and cover both thehealth and education sectors. Between 2001 and2008, 56 percent of the department’s staff time wasspent on health sector activities. Originally,the Life Sciences Group withinthe Global Manufacturing and ServicesDepartment had a sector leader, five investmentofficers, and two analysts. By2007, another investment officer and alife sciences specialist joined the group.IFC’s activities have tended to complementWorld Bank Group operations, but efforts tocollaborate and pursue synergies have notbeen consistent. IFC’s support to expand theproduction of vaccines and to improve access tomedicine for diseases such as malaria in EastAfrica had strong complementarities with WorldBank HNP strategies and operations. In a fewcases, IFC projects followed from the World Bank’ssupport in sector policy reform, with intensivecollaboration between the two IFC’s health activitiesinstitutions. For example, a hospital have tended toinvestment in Argentina was developed complement Bankin close collaboration with the World operations, but efforts toBank and sequenced with World Bank collaborate and pursueprojects. 17 However, efforts to exploit synergies have beensynergies as outlined in the sector inconsistent.strategies have not been consistent. InJuly 2008, for example, IFC’s Health and EducationDepartment pointed out that “one possiblebarrier to working together is a lack of knowledgeand understanding of the strategy and processesof the IBRD and IFC” and called for a dialogue withits colleagues from the World Bank in order to“begin a more focused dialogue.” The institutionalfragmentation of health across several unitsat IFC also inhibits greater collaboration withWorld Bank staff.Given that most of IFC’s operations in thehealth sector are in middle-income countries,coordination with the World Bank inthe health sector faces the same inhibitorsidentified by IEG’s study on middle-incomecountries. These include incompatible timelinesfor projects, differences in organizational culture,and staff concerns that their time can neither beeasily allocated to cooperation nor recognizedin performance assessments (IEG 2007b, p. 63).This is further complicated for health by the asymmetryin the size of IFC interventions relative toHealth sector activitiesand responsibilities in IFCare segmented and linkedto different departments,implying a need forcoordination.89

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBank operations. Although collaboration at the operationallevel turns out to be impractical, thereis recognition that collaboration to develop the appropriateenvironment (policy level) for privateinvestment would be beneficial. World Bank assistancein creating the regulatory environment forprivate sector operations could help to attractprivate investment by establishing appropriate incentivesto invest (eliminating the constraint of lagof regulation in the sector).Collaboration has been Collaboration between the Worldmore situational than Bank and IFC in health has been situationalrather than systematic.systematic; IFC’sinvestments rarely feature Currently, IFC’s investment officers arein joint CASs. communicating with the World Bank’shealth sector experts during project appraisal.Advisory Services for public-private partnerships,which naturally involve governmententities, have active communication within theWorld Bank Group, but there is a limit because ofthe ethical guidelines concerning the actual andperceived conflicts of interest. However, those interactionsdepend on the initiative of particularpeople and are driven by transactions. There is nosystematic avenue of collaboration betweenIFC’s projectsare small in scaleand geographicallyscattered, limiting theoverall impact.the two institutions in developingIFC strategies in health. Because ofIFC’s opportunistic approach in identifyinghealth investment in most countries,its health sector operations rarelyfeature in the Bank-IFC joint CASs. 18Social Impacts of IFC Health InvestmentsThe 2002 strategy did not specify approachesto address poverty, though it recognizedthe need to assess the potential for privateparticipation in health. IFC strategies for healthinvestment have limited direct references topoverty. The 1999 Note on Strategic Directions,for example, highlighted the potential of IFChealth investments to improve health outcomesacross the entire population and among the poor(IFC Health Care Best Practice Group 1999). In theimplementation of the 2002 health strategy, IFCaimed to diversify its portfolio to indirectly promoteaccess to health care services for poor peopleby improving access to and quality of healthservices for the lower-middle and middle class (IFC2002, p. 39) and by working with not-for-profitinstitutions (IFC 2002, p. 5). Then, in a 2007 briefingto the Board of Directors, the Department ofHealth and Education stated that its projects androle were changing, and that it had begun tofocus on clients serving lower-income populations(IFC 2007d, p. 18). The Life Science Group,for its part, has identified three strategic areaswhere it can contribute to the fight against diseasesthat disproportionately affect poor people in developingcountries (IFC 2007c): the productionof generic pharmaceuticals, innovation and researchto help discover technologies or treatmentsfor diseases that affect poor people, andneglected diseases that affect people in developingcountries.IFC’s investments in health services, mainlyhospitals, have benefited primarily upperandmiddle-income people at the “top of thepyramid” (IFC 2007a; IFC and WRI 2007).Among the 12 hospitals in the third period forwhich information is available, 3 were mainly targetedto expatriates and 6 were aimed at high- andmiddle-income populations. Among the evaluatedhospitals, only 1 had a confirmed linkage witha public insurance scheme that paid for treatmentof complicated cases referred to public hospitals.In other cases, links to public insurancefunds helped private diagnostics and laboratoryservice providers to serve people’s needs, irrespectiveof income level, as observed in two evaluatedprojects. Freestanding facilities can besuccessful and provide some desired developmentoutcomes, but linkages to other institutionalarrangements, such as insurance, arenecessary to meet the health needs of a wider population.Among the evaluated projects in the firsttwo periods, only three had such features.All of IFC’s clients in the hospital sectorhave some initiatives that directly targetunderserved populations. In addition, someclients provide free or reduced-price access tohealth care. Others target patients with healthinsurance coverage (often through employers),which allows hospitals to reach a much widerpopulation as well as the middle class. IFC’s clienthospitals and clinics also typically have or con-90

IFC’S HEALTH STRATEGY AND OPERATIONStribute to dedicated foundations for charitableactivities and community programs that targetthe poor and disadvantaged. Contributions normallyaccount for 1 to 5 percent of their income.However, IFC’s additionality in this area is verylimited—many clients were involved in charitableactivities even before IFC investment. In June2008, IFC and the World Bank engaged in a projectthat seeks to help improve maternal careamong some of Yemen’s poorest people. Thisproject uses output-based aid to reach the poor.New opportunities also are arising within the recentIFC Business of Health in Africa initiative.IFC’s life sciences interventions, particularlyin pharmaceuticals, have the potentialfor broader reach and for more direct benefitsto the poor. Household surveys throughoutthe developing world have shown thatpharmaceuticals account for more than half of allhealth spending by people in the lowest incomegroups (Hammond and others 2007). Thus, investmentsby IFC in pharmaceutical projects thatreduce the costs of drugs and improve quality controlhave great potential for helping the poor.The majority of the pharmaceutical projectsfinanced by IFC have resulted in significantdeclines in the price of genericdrugs, enhancing affordability. Four of thesix evaluated projects in pharmaceuticals involveproduction of generic drugs. The introduction ofgeneric drugs in Mexico in 1999 resulted in a 30percent decline in prices, and in Brazil, genericproducts cost an average of 40 percent less thanbrand-name products (Homedes, Ugalde, andForns 2005, p. 695). Some projects, especially inChina and India, are concentrated on the domesticmarkets and use local inputs. Since IFCpromotes Good Manufacturing Practices and otherinternational standards necessary for many exportmarkets, some of the generic drugs are exportedto neighboring countries. Only one projecthad confirmed success in penetrating The IFC strategy did notexport markets with certification from specify approaches tothe regulatory authority. One project address poverty, and itshad an objective to establish a research investments in healthand development facility in a pharmaceuticalcompany. However, as the benefited upper- andservices have primarilysponsor company’s business objectives middle-income people.shifted away from proprietary drug productionbecause of the competitivepressure from international producers as well asstrategic shifts toward generic drug production tothe U.S. market, the company reduced its researchand development functions, which resulted infewer jobs for local scientists and lab technicians.This demonstrates the high-risk nature of researchand development in drug production.In summary, the health sector is still relativelynew to IFC, and the approach to investment hasremained largely a matter of seizing opportunitiesrather than developing them within a clear strategicframework. IFC’s investments in the health sectorare still small and geographically scattered.There were marked improvements in the performanceof the small health investmentportfolio, mainly in hospitals, and IFC’s pharmaceuticalimportant lessons were learned leading investments have theup to the 2002 strategy. The expansion potential for broaderof public-private partnerships supportedthrough Advisory Services has benefit for the poor.reach and more directbeen an important development withpotential to contribute to improved efficiency ofthe health sector of developing countries. Greaterdiversification of the portfolio, particularly towardpharmaceuticals and health insurance, would improvethe social impact of the health investmentportfolio and bring it more in line with the objectivesof the 2002 strategy and the mandate ofthe World Bank Group. Closer collaboration withthe World Bank on private sector regulatory issueswith respect to the health and pharmaceuticalsectors would improve the climate for expandingIFC health investment and Advisory Services.91

Chapter 6

A woman and children pose after taking part in a health survey in Bukoba, Tanzania. Fertility andmaternal mortality remain extremely high in most of Sub-Saharan Africa.Photo courtesy of Martha Ainsworth.

Conclusions andRecommendationsThe global aid architecture in health has changed over the past decadewith the adoption of international goals and a major expansion of thelevels and sources of development assistance, particularly for lowincomecountries. The World Bank Group’s support for health, nutrition,and population has been sustained over the decade, but it is now a smallershare of global HNP assistance.The mandate of the World Bank Group is to promotepoverty reduction and economic growth indeveloping countries. Poor health is both a causeand consequence of poverty and an impedimentto economic growth. This evaluation has pointedto some important accomplishments of the WorldBank’s HNP support to countries, often in difficultsettings, and its contribution in helping build governmentcapacity to manage the sector. The latteris important for improving aid effectiveness moregenerally, given the increasing reliance of the internationalcommunity on government systems.The evaluation has also found improving performancein IFC’s health portfolio and importantlessons learned in this small but expanding sector.The fieldwork for this evaluation also pointsto areas where the World Bank and IFC continueto add value (box 6.1). The results are areminder that the value added by the World BankGroup is not measured solely by the magnitudeof its support, and that its actual comparative advantageor that of any other development partnerdepends not only on its institutional assets, butalso on the country context, health needs, and theactivities of others.Looking forward, the developments in theinternational aid architecture for healthover the past decade and the increased levelsof assistance from other sources presentboth opportunities and challenges for theWorld Bank Group. In the context of the ParisDeclaration on Aid Effectiveness, the Bank hasan opportunity to leverage its experience to helpgovernments strengthen their capacity to manageeffectively their own resources as well as thenew donor funding being channeled throughgovernment. While the Bank’s resources are nowa smaller share of global development assistancefor health, the Bank nevertheless continues tobring key institutional assets to the table: longterm,sustained engagement in the sector; experiencein many countries with similar issues;a history of support for building country capacityfor implementing social sector investments;sustained large-scale financing; strong links to theMinistry of Finance, which can be critically importantboth in leveraging reforms and in improvingthe dialogue between the ministries ofhealth and finance; and a country approach inwhich HNP support is part of a portfolio of activitiesin many sectors, some of them with potentialcomplementary contributions to healthoutcomes. These are assets that the Bank can continueto bring to bear on making health systemswork better and to ensure that health benefitsreach the poor.95

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONBox 6.1: Evolving Value Added of the World Bank Group in HNPField interviews for this evaluation point to major areas in whichthe World Bank’s policy dialogue, analytic work, technical advice,and lending have added value to HNP in countries.In Peru, officials valued highly the Bank’s ability to mobilizefinancial and human resources and connect technical assistancewith financial services; its experience with health reformsin countries with similar conditions; its poverty focus, underscoredby highly effective analytic work providing an “outsider’s”perspective; its long-term relationship, which has helped to maintaina medium-term agenda and dialogue on health in the faceof high turnover in the sector; and its institutional culture, whichhas emphasized outcome-oriented projects, efficiency, and transparencyin the allocation of resources.In the Kyrgyz Republic, the Bank played a leadership role,mediating conflicts among the donors and using its leverage tokeep health reform on the political agenda in the face of strongopposition. It built institutional capacity over time in the Ministryof Health, was able to intervene to work out implementation details,and provided capacity-building inputs for public sector fiduciarymanagement. The country would have implemented healthreforms even without the Bank, but respondents felt that progresswould have been more limited geographically and institutionally,resources would have been used less efficiently, and there wouldhave been less focus on improving access for the poor.In Nepal, the Bank brought value added with its poverty focus,analytic work, and living standards surveys focusing on the healthof the poor. It helped to strengthen national capacity to lead andmanage donor funds and provided technical assistance for monitoringsocial sector expenditures and for financial managementin support of the SWAp. Without the Bank, there would havebeen less financing of civil works and training of village healthworkers, a weaker poverty focus, and less attention to public expendituresand expenditure frameworks. Other donors likely wouldnot have participated in the SWAp without the Bank’s involvement.In Malawi, the Bank’s value added took the form of leadershipon procurement and financial management in support of thehealth SWAp and individual operations; increased knowledgethrough analytic work, exchange of experience, facilitating accessto expertise, and an attempt to maintain the focus on outcomes;and its convening power and strong links with the sectoraland macro level.The proposed health interventions likely would have occurredin Eritrea even without the Bank’s involvement, but not on the samescale. The Bank’s engagement brought credibility and eventualsupport from others in the international community, built the institutionalcapacity of the Ministry of Health, and highlighted theneed to focus both on the supply of health care and on communitydemands, and on both hardware and software.Respondents also expressed views on the “additionality” ofIFC’s health support. Often IFC was the only source of long-termfinance (loan and equity) where private investments in the healthsector were considered risky and local banks did not have experiencein financing the sector. A pharmaceutical business inChina emphasized that IFC’s value added was not so much its financialcontribution, but the value of its “stamp of approval” inenvironmental and social standards, which provides a level ofcomfort to the other investors. Another company said that IFC’sassociation with the World Bank Group, a public entity, facilitatedpublic-private partnerships with local authorities. A third clientvalued IFC’s role as an honest broker between stakeholders, includinggovernment and regulatory agencies, to increase visibilityand bring strategic partners to the project.Sources: IEG 2009a, 2009b, 2008d; Shaw forthcoming; Elmendorf and Nankhuni forthcoming.The evaluation has also revealed substantialchallenges for the future. About a third of theBank’s HNP support is not performing well, andthere has been no sign of improvement. Contributingfactors are the increasing complexity ofHNP operations, particularly in Africa but also insupport of health reform to middle-income countries;inadequate risk assessment and mitigation;and weak M&E. The Bank has underinvested insupport to reduce high fertility and malnutritionamong the poor. The performance of IFC’s healthinvestments has improved markedly, but the programhas been less successful in expanding investmentin activities that both make businesssense and are likely to yield broader benefits forthe poor. Accountability for results has beenweak—both the accountability of Bank- and IFCsupportedprojects for ensuring that results haveactually reached the poor, and the accountabilityof the projects supported by the Bank’s non-96

CONCLUSIONS AND RECOMMENDATIONShealth sectors, such as water supply and sanitationand transport, for demonstrating the healthbenefits of their investments.The expansion of external assistance forhealth also featured a large increase in resourcesearmarked for specific infectiousdiseases or other programs within the healthsector. Many of these previously attracted insufficientattention. To the extent that largeamounts of earmarked support have the potentialto adversely distort resources within the healthsystems of small countries to the detriment ofother critical services for the poor, the Bank needsto consider, on a case-by-case basis, whether additionalearmarked funds of its own for some ofthe same programs are appropriate and costeffectivein improving health, nutrition, and populationoutcomes, given possible alternative investmentselsewhere in the health system.In short, the new international aid architecturehas provided the World Bank Groupwith opportunities to improve health outcomesamong the poor, to prevent povertydue to illness, and to improve the efficiencyand efficacy of health systems. By improvingtheir performance and maintaining focus onthese core objectives, the Bank Group has the potentialto contribute not only to meeting the MillenniumDevelopment Goals but also to ensuringthat the poor benefit and that the benefits will besustained.RecommendationsThe following recommendations for the WorldBank and IFC are offered to help improve the implementationand impact of their respective HNPstrategies and further the mandate of poverty reductionand economic growth in the context ofthe new aid architecture.1. Intensify efforts to improve the performanceof the World Bank’s health,nutrition, and population support.• Match project design to country contextand capacity and reduce the complexity ofprojects in low-capacity settings throughgreater selectivity, prioritization, and sequencingof activities, particularly in Sub-Saharan Africa.• Thoroughly and carefully assess the risksof proposed HNP support and strategies tomitigate them, particularly the political risksand the interests of different stakeholders,and how they will be addressed.• Phase reforms to maximize the probabilityof success.• Undertake thorough institutional analysis, includingan assessment of alternatives, as aninput into more realistic project design.• Support intensified supervision in the fieldby the Bank and the borrower to ensurethat civil works, equipment, and other outputshave been delivered as specified, arefunctioning, and are being maintained.2. Renew the commitment to health, nutrition,and population outcomes amongthe poor.The World Bank should:• Boost population and family planningand other support in the form of analyticwork, policy dialogue, and financing to highfertilitycountries and countries with pocketsof high fertility.• Incorporate the poverty dimension intoproject objectives to increase accountabilityfor health, nutrition, and population outcomesamong the poor.• Increase support to reduce malnutritionamong the poor, whether originating in theHNP sector or other sectors.• Monitor health, nutrition, and populationoutcomes among the poor, howeverdefined.• Bring the health and nutrition of the poorand the links between high fertility, poorhealth, and poverty back into poverty assessmentsin countries where they havebeen neglected.IFC should:• Expand support for innovative approachesand viable business models that demon-97

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONstrate private sector solutions to improve thehealth of the poor, including expansion ofinvestments in low-cost generic drugs andtechnologies that address health problemsof the poor.• Assess the external and internal constraintsto achieving broad social impacts in thesector.3. Strengthen the World Bank Group’s abilityto help countries to improve the efficiencyof health systems. Improving theefficiency of health systems may only indirectlyproduce health benefits, but it is a validobjective in its own right, contributing to moresustainable financing and economic growth.Health reform projects address structural inefficiencies,are inherently politicized, andthus are risky. However, they are neverthelessworth pursuing if deemed relevant and ahigh priority of government.The World Bank should:• Better define the efficiency objectives of itssupport and how efficiency improvementswill be improved and monitored.• Carefully assess decisions to finance additionalearmarked communicable disease activitiesin countries where other donors arecontributing large amounts of earmarkeddisease funding and additional funds couldresult in distortions in allocations and inefficienciesin the rest of the health system.• Support improved health information systemsand more frequent and vigorous evaluationof specific reforms or program innovationsto provide timely information forimproving efficiency and efficacy.IFC should:• Support public-private partnerships throughAdvisory Services to government and industryand through its investments, and expandinvestments in health insurance.• Improve collaboration and joint sector workwith the World Bank, leveraging Bank sectordialogue on regulatory frameworks forhealth to engage new private actors withvalue added to the sector, and more systematicallycoordinate with the Bank’s policyinterventions regarding private sectorparticipation in health.4. Enhance the contribution of supportfrom other sectors to health, nutrition,and population outcomes.The World Bank should:• When the benefits are potentially great in relationto the marginal costs, incorporatehealth objectives into nonhealth projects, forwhich they are accountable.• Improve the complementarity of investmentoperations in health and other sectors toachieve health, nutrition, and populationoutcomes, particularly between health andwater supply and sanitation.• Prioritize sectoral participation in multisectoralHNP projects according to the comparativeadvantages and institutional mandatesto reduce complexity.• Identify new incentives for Bank staff towork cross-sectorally for improving health,nutrition, and population outcomes.• Develop mechanisms to ensure that the implementationand results for small HNPcomponents retrofitted into projects areproperly documented and evaluated.IFC should:• Improve incentives and institutional mechanismsfor an integrated approach to healthissues across units in IFC dealing with health,including the way that IFC is organized.5. Implement the results agenda and improvegovernance by boosting investmentin and incentives for evaluation.Weak M&E is a constraint to achieving the resultsfocus and governance agenda of theBank’s new HNP strategy and inhibits the resultsorientation of IFC.98

CONCLUSIONS AND RECOMMENDATIONSThe World Bank should:• Create new incentives for M&E, for boththe Bank and the borrower, linked to theproject approval process and the midtermreview. This would include requirementsfor baseline data, explicit evaluation designsfor pilot activities in project appraisal documents,and periodic evaluation of mainproject activities as a management tool.IFC should:• Enhance its results orientation by developingclearly specified baseline indicators andan evaluation framework that adequatelymeasure IFC’s health sector objectives andresults.99

Appendixes

An auxiliary nurse midwife speaking to slum dwellers about contraception and other sexual health matters in India. Photo by John Isaac,courtesy of the World Bank Photo Library.

APPENDIX A: WORLD BANK GROUP HNP TIMELINE103

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP TimelineYear World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments1952Economic Survey mission to Jamaica tostudy the country’s development requirementsconsiders the effects of rapidpopulation growth. (March) (1)Concern over the impact of populationgrowth on development is discussed atSeventh Annual Meetings in Mexico City.Chairman of the Board of Governorsargues that the World Bank is well placedto combine sound banking principles withcreative efforts to address populationgrowth issues. (September) (1)1956IFC is established as an institution of theWorld Bank Group to promote sustainableprivate sector investment in developingcountries.1961World Bank begins lending for watersupply and sanitation projects. (2)1964The first IFC investment in pharmaceuticals,“Huhtamaki-Yhtyma Oy” of Finland,is approved.1968Robert McNamara becomes World BankPresident. (April) (1)McNamara calls for governments todevelop strategies to control populationgrowth. He admits that there is noalternative to the World Bank’sinvolvement in “this crisis.” (October) (1)Economics Department’s Special StudiesDivision is reorganized to create aPopulation Studies Division headed byE.K. Hawkins. (3)Population Projects Department isestablished under the Office of theDirector of Projects. (November) (4)K. Kanagaratnam is asked and acceptsthe post as head of the Population ProjectsDepartment; however, he is unable tostart immediately, and in the interimGeorge C. Zaidan becomes the firstdivision chief of the new department. (3)104

APPENDIX A: WORLD BANK GROUP HNP TIMELINE19691970197119721973McNamara calls for emphasis on populationplanning, educational advances, andagricultural growth in his Annual Meetingsaddress. He highlights the need fordevelopment in nutrition, water supply,and literacy. (September) (1)In his Annual Meeting address, McNamaraemphasizes the importance of addressingthe basic problems affecting the dailylives of people in developing countries,including nutrition, employment, andincome distribution, among others. Hedescribes malnutrition as a major barrierto human development. (September) (1)A Bank-wide reorganization creates asenior vice president of operations withfive Regional vice presidents and a vicepresident for project staff. (August) (1)McNamara uses his address at theAnnual Meetings to emphasize the needto incorporate population planning intodevelopment strategies. (September) (1)First population loan is approved for$2 million to support Jamaica’s familyplanning program. (June) (1)As a result of the reorganization, aPopulation and Nutrition Projects (PNP)Department and several others with toofew staff for decentralization are groupedin the Central Operation ProjectsDepartment and provide technicalservices to the Regions. (4)The Board of Executive Directors approvesMcNamara’s proposal for the Bank to takethe lead in mobilizing international fundsfor an onchocersiasis (river blindness)control program. (May) (1)Possible Bank Actions on MalnutritionProblems is released. It is influential incalling attention to the Bank’s role inaddressing malnutrition. (January) (5*)Sectoral Programs and Policies Paperincludes recommendations on populationpolicies. It points to the economic effectsof population growth in developingcountries, describes the Bank’s efforts toassist member countries to reducepopulation growth rates, and outlines itsfuture program in population assistance.(March) (6*)A nutrition policy paper makes the casefor investment in nutrition and proposesthat the Bank “assume a more active anddirect role in nutrition.” (8*)World Bank/WHO Cooperative Program isestablished to address water supply,waste disposal, and storm drainage.(September) (1)World Bank participates in an advisorycapacity in WHO’s Special Program ofResearch Development and Training inHuman Reproduction (HRP). (7)World Bank convenes Meeting ofOnchocersiasis Control Program in Pariswith WHO, the U.N. Food and AgricultureOrganization (FAO), the United NationsDevelopment Program (UNDP). Thepurpose of the meeting is to formulate astrategy to fight river blindness. (June) (1)(Table continues next page)105

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments1974Funds to cover the first year of theOnchocersiasis (river blindness) ControlProgram are mobilized. (March) (1)Population Policies and EconomicDevelopment analyzes the impact ofpopulation growth on the fight againstpoverty. (August) (9*)WHO, FAO, UNDP and the World Bankimplement the Onchocersiasis ControlProgram (OCP), which is endorsed by theseven governments of West Africa, thecountries most affected by the disease.(March) (1)19751975 Health Sector Policy Paper ispublished. As the first formal HNP policystatement, it establishes that lending willbe only for family planning andpopulation. (10*)World Bank cosponsors the TropicalResearch Program along with WHO,UNICEF, and UNDP to coordinate a globaleffort to combat diseases that affect thepoor and disadvantaged through researchand development, and training andstrengthening. (1)1976First loan in nutrition, $19 million toBrazil, is approved. (June) (1)1977World Bank helps to found and becomesa member of the UN Subcommittee onNutrition (SCN). (11)1979The Population, Health, and NutritionDepartment (PHN) is established. TheBank approves a policy to consider fundingfreestanding health projects andhealth components of other projects.(July) (2)John R. Evans appointed PHNDepartment Director. (12)World Bank and UNDP initiate the UNDP-World Bank Water and SanitationProgram (WSP) to analyze cost-effectivestrategies and technologies to bring cleanwater to the poor. (1)1980 WDR 1980: Poverty and Human Developmenthighlights the importance of thehealth sector, education, and social protectionto alleviate poverty. Part of the reportdescribes the role of human developmentprograms, its effects on productivity andpopulation growth. (August) (13*)1980 Health Sector Policy Paper commitsthe Bank to direct lending in the healthsector. The strategy focuses on the needfor basic health services, especially inrural areas, and describes the linksbetween the health sector, povertyalleviation, and family planning. (14*)106

APPENDIX A: WORLD BANK GROUP HNP TIMELINE19811983198419851986WDR 1984: Population and Developmentemphasizes the role of governments toreduce mortality and fertility. (16*)Research Department launches the firstBank-sponsored Living Standards MeasurementSurvey in Côte D’Ivoire. LSMSsare multi-topic household surveys capableof linking the level and distribution of welfareat the household level to health caredecisions, the availability and quality ofhealth services, and HNP outcomes. (17)Barber Conable is appointed as the Bank’s7th President. (July) (1)A Poverty Task Force composed of seniorstaff is established to review the Bank’swork and propose new activities. (19)First loan to expand basic health servicesis made to Tunisia. (15)The first IFC investment in hospitals, theDr. Simo Milosevic Institute located onthe Mediterranean coast of Yugoslavia(now Montenegro), for a medicalrehabilitation facility is approved. (1a)John N. North becomes Director of thePHN Department. (12)Frederick Sai appointed Senior PopulationAdviser. (18)Poverty and Hunger: Issues and Optionsfor Food Security in Developing Countriesargues that food insecurity is causedmainly by poor people’s lack of purchasingpower. It asserts that the role for internationaldonors is to provide assistance todevelop and financing to support improvedpolicies to reduce food insecurity, as wellas addressing international trade factorsthat contribute to food insecurity. (20)World Bank partners with The RockefellerFoundation, UNDP, UNICEF, and WHO toestablish the Task Force for Child Survivaland Development, a campaign to achievethe goal of universal child immunizationby 1990. (1)(Table continues next page)107

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments1987President Conable announces an internalreorganization to be completed bySeptember. (May) (1)PHN becomes a division of the Populationand Human Resources (PHR) Department.Technical departments, including PHNunits, are created within each region, andcountry departments are created withinRegions, combining the functions formerlydivided between programs and projectsdepartments. (21)Ann O. Hamilton is appointed PHRDepartment Director. (12)Dean T. Jamison is appointed ChiefManager of PHN Division. (12)Financing Health Services in DevelopingCountries: An Agenda for Reform arguesthat government expenditures shouldshift toward providing health services forthe poor. The policy study addressesthemes of inefficient public spending onhealth care and recurrent cost financing.(May) (22*)World Bank cosponsors the Safe MotherhoodConference in Nairobi, Kenya. TheBank pledges to take specific steps toaddress issues affecting women, and theSafe Motherhood Initiative is launched.(February) (1)1988First freestanding AIDS project isapproved in Zaire. This is also the firstapproved freestanding Bank project for asingle disease. (21)Anthony Measham becomes PHN ChiefManager. (12)Acquired Immunodeficiency Syndrome(AIDS): The Bank’s Agenda for Action isprepared by the Africa Technical Department.It was not formally adopted bythe Bank management as a strategy butreleased as a working paper. (23*)World Bank becomes a funder of theWHO’s HRP. (24)1989The IDA Debt Reduction Facility isestablished to reduce the stock of debtowed to commercial creditors by IDA-onlycountries. (August) (1)Bank finances the first freestandingnongovernmental organizationimplementedproject for grassrootsdevelopment in Togo. (19)First social fund project is approved. (1)Sub-Saharan Africa: From Crisis to SustainableDevelopment calls for a doublingof expenditure on human resource development:food security, primary education,and health care. (November) (25*)1990The IBRD approves the largest loan atthis point in its history (nominal terms) toMexico to support a debt-reductionprogram, and the Debt-Reduction Facilityfor IDA-only countries undertakes its firstoperation in Bolivia. (19)Steven Sinding becomes SeniorPopulation Adviser. (26)108

APPENDIX A: WORLD BANK GROUP HNP TIMELINE1991199219931994Lewis T. Preston is appointed as the 8thpresident of the World Bank. (September)(21)A report of the Task Force on PortfolioManagement (the “Wapenhans Report”)is transmitted to the Executive Directorsand is a major factor in the Bank’simpetus to redouble its efforts towardeffective implementation of lendingprojects. (1)WDR 1993: Investing in Health evaluatesthe roles of governments and markets inhealth, as well as ownership and financingarrangements to improve health and reachthe poor. It introduces the disabilityadjustedlife year (DALY) to calculate theGlobal Burden of Disease, and argues thatthe international community must committo addressing health issues. (June) (27*)A policy paper, Water Resources Management,proposes a new approach to managingwater resources. The approachadvocates a comprehensive policy frameworkand treatment of water as aneconomic good, along with decentralizedmanagement and delivery structures,greater reliance on pricing, and fuller participationby stakeholder. (29*)Bank issues a statement that abortionis an issue countries themselves mustaddress and denies advocating thelegalization of abortion in Latin America.(March) (1)The first health-related advisory serviceproject is approved by IFC for theThailand Bumrungrad Hospital.AIDS in Asia, the first Regional AIDSsupport unit, is established in the EastAsia and Pacific Region. (21)Janet de Merode becomes Director of thePHN Division. (12)David de Ferranti becomes Director ofPHN Division. (12)Disease Control Priorities in DevelopingCountries provides information on diseasecontrol interventions for the most commondiseases and injuries in developingcountries to help them define essentialhealth service packages. The publicationeventually leads to increased Banklending for disease control. (October) (28*)Better Health in Africa, directed to bothBank and external audiences, argues thatbecause households and communitieshave the capacity to use knowledge andresources to respond to health problems,policy makers should make efforts to createan enabling environments that stimulate“good” decision making. It alsopoints out that health reforms are necessary,that cost-effective packages of servicescan meet needs, and that changesin domestic and international financingfor health are necessary. The publicationwas never approved as an official strategy,but the World Bank supported an independent‘Better Health in Africa’ ExpertPanel that worked to disseminate keymessages to African policy makers. (30*)World Bank joins with UNDP, UNICEF,WHO, and Rotary International to formthe Children’s Vaccine Initiative (CVI).CVI’s goal is to vaccinate every child inthe world against viral and bacterialdiseases. (27)World Bank participates in InternationalConference on Nutrition in Rome.(December) (15)Bank participates in InternationalConference on Population and Development(ICPD) in Cairo and commits to itsplan of action. (31)(Table continues next page)109

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments1995James Wolfensohn is appointed as theninth World Bank president. (June) (1)The Broad Sector Approach to InvestmentLending: Sector Investment Programsdefines sector investment programs (SIP),analyzes experience with the new lendinginstrument and advocates for morelearning and support of SIPs, particularlyin Africa. (32)The Human Development Department isestablished and David de Ferranti servesas Department Director. Richard Feachem(Health), Jorge Barrientos (Implementation),Alan Berg (Nutrition) and ThomasMerrick (Population) are appointed asmanagers/advisers. (July) (4,12)The Bank hosts a conference to launchthe African Program for OnchocersiasisControl, a follow-up to a successfulproject launched in the 1970s. Sponsoredby governments, NGOs, bilateral donorsand international institutions, it implementscommunity-based drug-treatmentprograms in 16 African countries.(December) (1)1995Learning and Leadership Center-HumanDevelopment Network training weekinitiated to provide staff with intensivetraining focused on topical issues in theHNP sector. (15)The Bank participates in the Fourth WorldConference on Women in Beijing (FWCW)and agrees to: reduce the gender gap ineducation and ensure that women haveequitable access and control overeconomic resources. (31)1996World Bank Participation Sourcebooklaunched. Wolfensohn announces thatthe Bank will involve NGOs, the privatesector, community groups, cooperatives,women’s organizations, and the poor anddisadvantaged in decision-makingprocesses. (February) (33)In his Annual Meetings address, Wolfensohndefines the key elements of theStrategic Compact to renew the BankGroup and improve development effectiveness:improving resource mobilization;taking more integrated approaches; buildingpartnerships and sharing knowledge;and restructuring the Bank to be closerto clients through responsive and highqualityproducts. (October) (33)The Bank announces that three newnetworks will be created: Environmentallyand Socially Sustainable Development(ESSD), Finance, Private Sector andInfrastructure (FPSI), and PovertyWorld Bank sponsors tobacco-related andnon-communicable disease conference inWashington, DC. (June) (1)The Flagship Program on Health SectorReform and Sustainable Financing isinitiated by the Economic DevelopmentInstitute (EDI, now World Bank Institute)to provide knowledge and training onoptions for health sector development,including lessons learned and best practicesfrom country experience. Course isoffered at regional and country levels. (1)IFC launches a global study on “PrivateHospital Investment Opportunities” toidentify key success factors forinvestment in hospitals and moregenerally in health. (2a)Special UN Initiative for Africa launched;Bank partners with UN to promote anexpanded program of assistance to Sub-Saharan Africa and improve cooperationbetween the Bank and the UN. Bankcommits to take special responsibility formobilizing resources for basic health andeducation reforms. (March) (1)Wolfensohn announces Bank’s support forthe G-7’s declaration and objective ofproviding an exit strategy for heavilyindebted countries. Bank pledges $500million to a trust fund for debt relief as itsinitial contribution. (June) (33)World Bank cosponsors the Joint UNProgram on HIV/AIDS (UNAIDS) withUNDP, UNESCO, UNFPA, UNICEF, andWHO. (21)110

APPENDIX A: WORLD BANK GROUP HNP TIMELINE1997Reduction and Economic Management(PREM). (December) (1)Poverty Reduction and the World Bank:Progress and Challenges in the 1990s isreleased and vows to redouble Bank’sefforts to ensure success in its mandateto help countries reduce poverty. TheBank says that it will judge itself andstaff by their contributions to achievingthis goal. (June) (1)The Bank and International MonetaryFund launch the Highly Indebted PoorCountry (HIPC) Initiative, creating aframework for creditors to provide debtrelief to the world’s most poor andindebted countries. The HIPC Trust Fundand HIPC Implementation Unit areestablished. (November) (1)Quality Assurance Group (QAG)established with the expressed purposeof improving the quality of the Bank’soperational work within the broad contextof reducing poverty and achievingdevelopment impacts. (34)World Development Indicators 1997, thefirst edition, is published. Wolfensohnpoints to the publication as an example ofthe World Bank’s role in disseminatingknowledge to facilitate decision makingin development. (April) (33)The Strategic Compact period, a threeyearorganization renewal process, islaunched. (April) (1)Bank reorganization leads to the creationof Bank-wide “anchor” units to providequality support to the Regions. Thereorganization was designed to promotebalance between “country focus” and“sectoral excellence.” (21)The Human Development Network (HDN)is formed, along with the HNP SectorBoard, when Bank reorganization groupssector staff into regional sector units ordepartments. Sector staff work withcounty departments in a matrixrelationship. This allows Regionalmanagers working in the HNP sector tocome together. (21)David de Ferranti serves as Vice Presidentand Head of HDN. Richard G.A. Feachemis named HNP Director and serves asChair of the Sector Board. (12)World Bank organizes and hosts anInternational Conference on Innovationsin Health Financing. (36)IFC sponsors a global conference on“Investing in Private Hospitals and OtherThe 1997 Health, Nutrition, and PopulationSector Strategy Paper emphasizesthe importance of institutional and systemicchanges to improve healthoutcomes for the poor, improve healthsystem performance, and achievesustainable financing in the health sector.(September) (15*)Confronting AIDS: Public Priorities in aGlobal Epidemic makes the case for governmentintervention to control AIDS indeveloping countries from epidemiological,public health, and public economicsperspectives. The report advocates thatdonors base their support on evidence ofcountry-specific effectiveness for interventions,and finance key internationalpublic goods. (November) (37*)World Bank becomes a donor to thenewly formed International AIDS VaccineInitiative (IAVI). It is established to ensurethe development of an HIV vaccine foruse around the world. (35)World Bank collaborates with UN EconomicCommission for Africa and UNICEFto organize the Forum on Cost Sharing inthe Social Sectors of Sub-Saharan Africa.Fifteen principles for cost sharing inhealth and education are agreed upon atthe Forum. (38)The World Bank and The Danish Ministryof Foreign Affairs cohost a meeting fordonor agencies in Copenhagen to discusssectorwide approaches. At the meetingthe term SWAp is coined, a SWAp guideis commissioned, and an Inter-AgencyGroup on SWAp is formed. (32)(Table continues next page)111

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitmentsHealth Delivery Systems in DevelopingCountries: Opportunities and Risks,”bringing together IFC and World Bankstaff and other major private health careplayers from developing and developedcountries.1998President Wolfensohn’s address at theAnnual Meetings warns that financialreforms are not sufficient, that humanneeds and social justice must also besought. (1)Assessing Aid: What Works, WhatDoesn’t and Why concludes that there isa role for foreign aid and that properlymanaged aid can contribute to improvingpeople’s lives. It argues that institutionaldevelopment and policy reforms alongwith strong three-way partnership amongrecipient countries, aid agencies, anddonor countries can improve the impactof foreign assistance. (39*)IFC introduces its frontier country strategyto steer resources toward “pioneering” orunderserved sectors in high-risk and/orlow-income countries. (3a)The World Bank launches AIDS VaccineTask Force to speed up deployment ofeffective and affordable AIDS vaccine. Itsupports high-level dialogue withpolicy makers and industry, both “push”and “pull” strategies to generateinvestments in research anddevelopment, and sponsors studies ofpotential demand for a vaccine indeveloping countries. (April) (1)The World Bank Institute develops acourse and learning program titled“Adapting to Change” as a response tothe ICPD. (40)Christopher Lovelace is appointedDirector of the HNP Sector. (12)The Health Care Best Practice Group isformed in IFC to analyze potentialinvestments in health and to share andleverage knowledge about the healthcare industry that was developing acrossIFC departments. Nevertheless, the grouphas no decision-making role. (4a)The World Bank partners with WHO andSmith Kline Beecham to initiate a Programto Eliminate Elephantiasis by distributingdrugs free of charge to governments andcollaborating organizations. (January) (1)The World Bank, WHO, UNDP, and UNICEFlaunch Roll Back Malaria to provide acoordinated global approach to halvemalaria by 2010. (41)1999Wolfensohn calls for development partnersto adopt a Comprehensive DevelopmentFramework, which aims to improvethe effectiveness of development activitiesand move beyond individual projects,promoting national leadership and consensus,and requiring a commitment toThe AIDS Campaign Team for Africa(ACTafrica) unit is created to helpmainstream HIV/AIDS activities in allsectors. (21)Eduardo A. Doryan is appointed HDN VicePresident. (12)Population and the World Bank: Adaptingto Change is shaped largely by its commitmentto the 1994 ICPD and by an emphasison health sector reform in the 1990s. Itsobjective is to address population issueswith a people-centered and multisectoralapproach that improves reproductivehealth through access to information andThe World Bank partners to establishThe Global Alliance for Vaccines andImmunization (GAVI), a public-privatepartnership, to ensure financing to savechildren’s lives and people’s healththrough widespread vaccinations. (46)112

APPENDIX A: WORLD BANK GROUP HNP TIMELINEexpanded partnership, transparency, andaccountability. (January) (33)Bolivia becomes the pilot country for theCDF with two loans for health andinstitutional reform. (June) (1)In preparation for WDR 2000/2001, theBank launches the Voices of the Poorstudy. The study focused on perceptionsof a quality of life; pressing problems andpriorities; the quality of interactions withkey public, market and civil society institutionsin their lives; and changes in genderand social relations. (September) (42*)Wolfensohn appointed for second term asWorld Bank president. (September) (33)Wolfensohn links corruption and povertyat International Anti-CorruptionConference in Durban. He states that theBank will position corruption as a centralissue to development, apply externalpressures for change at the country levelwhile encouraging internal pressures forchange, and create partnerships toaddress corruption issues. (October) (33)The World Bank and InternationalMonetary Fund announce thatconcessionary lending to 81 eligible poorcountries will be based on povertyreduction strategies, initiating the PovertyReduction Support Paper process. (43)Enhanced HIPC launched. HIPC initiativeis modified to provide deeper and broaderrelief, faster relief, and to create a moredirect link between debt relief andpoverty reduction through PovertyReduction Strategy Papers. (1)IEG releases an evaluation of the HNPsector that suggests that the Bankimprove knowledge management,develop more flexible instruments, andsupport increased economic and sectorwork to help countries identify challengesand improve the efficiency, effectiveness,and equity of health reforms. It arguesThe Health and Education Unit isestablished in IFC.services, and recognizes the importanceof contextual factors such as genderequity and human rights. (January) (31*)The Bank’s new strategy to fight HIV/AIDSin Africa in partnership with Africangovernment and Joint UN Program onHIV/AIDS (UNAIDS) approved by RegionalLeadership Team. (May) (21)A Health Sector Strategy for the Europeand Central Asia Region responds tochanges in the health care systems,particularly in transition countries, byproviding a guide to support regionallyappropriate, intersectoral health systemreforms. Key priorities are identified as:(i) promoting wellness and reducing theprevalence of avoidable illness; (ii)creating affordable and sustainabledelivery systems; and (iii) maintainingfunctioning health systems during thereform process. (September) (45*)The document “Investing in PrivateHealth Care: A Note on StrategicDirection for IFC” is prepared by IFC’sHealth Care Best Practice Group. (5a)(Table continues next page)113

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitmentsthat projects had been too complex, hadneglected institutional analysis and thatmonitoring and evaluation was almostnonexistent. It urged that the sector “dobetter, not more,” that is, be moreselective to do a few things better ratherthan too much with poor results. (44)2000World Bank announces a plan to workwith church groups in Africa to fightpoverty and AIDS. (March) (1)Thousands of demonstrators protest atthe Development Committee’s Springmeetings in Washington. The DevelopmentCommittee renews its pledge tospeed up debt relief and to support thefight against AIDS. (March) (1)Wolfensohn addresses the UN SecurityCouncil and calls for increased resourceallocation to fight a “War on AIDS,”noting the epidemic’s devastating effectson the developing world, especiallyAfrica. (January) (33)The first Multicountry AIDS Program(MAP) is approved by the Board andprovides a $500 million envelope forfinancing HIV/AIDS projects in Africa.(September) (21)The overall objective of the World BankStrategy for Health, Nutrition, andPopulation in East Asia and the PacificRegion is to improve the Bank’seffectiveness in health, nutrition, andpopulation in the region. The strategyurges selectivity and flexibility to developnew approaches, as necessary, based onlessons learned and experience in theregion. It prioritizes: improving outcomesfor the poor, enhancing the performanceof health care systems, and securingsustainable financing. (June) (47*)World Bank and WHO issue a publication,Tobacco Control in Developing Countries.It argues that a reduction in tobacco useis essential to improve global health.(August) (48*)Intensifying Action Against AIDS in Africaemphasizes the importance of increasedadvocacy to strengthen political commitmentto fighting HIV/AIDS, mobilizationof resources, and strengthening theknowledge base. It advocates allocationof increased resources and technical supportto assist African partners and theWorld Bank to mainstream HIV/AIDS intoall sectors. (August) (49*)World Bank releases 44 country reportson Socio-Economic Differences in Health,Nutrition and Population. The reportsstress that the poorest sectors of thepopulation must receive adequatehealthcare. (November) (50*)At the World Economic Forum,Wolfensohn urges world leaders tosupport GAVI and its campaign forchildren. (January) (33)At the Second World Water Forum,Wolfensohn pledges the Bank’s supportto ensure that everyone has waterservices for health, food, energy, and theenvironment. The approach he outlinesemphasizes participatory institutions aswell as technological and financialinnovation. (March) (1)At the XIIIth International AIDS Conference,the World Bank pledges $500 million.The Multicountry AIDS Program,developed with UNAIDS, helps countriesto implement national HIV/AIDS programs.(July) (1)The Bank-Netherlands Water PartnershipProgram (BNWPP) is established toimprove water security by promotinginnovative approaches to IntegratedWater Resources Management (IWRM),and thereby contribute to povertyreduction. (51)114

APPENDIX A: WORLD BANK GROUP HNP TIMELINE2001 WDR 2000/2001: Attacking Povertyemphasizes that insecurity, in income orhealth services, is one of manydeprivations suffered by the poor. (52*)The World Bank announces that it willjoin the UN as a full partner to implementthe Millennium Development Goals andto put these goals at the center of thedevelopment agenda. (September) (1)World Bank makes a Declaration ofCommitment at Special Session of theUN General Assembly, reaffirmingpledges made by world leaders to haltand reverse the spread of HIV/AIDS by2015. (June) (33)The Water Supply and Sanitation Program(WSP) Council is created to overseeprogram activities and guide strategicdevelopment in water and sanitation. (53)Board of Executive Directors approves agender and development mainstreamingstrategy. (54)First poverty reduction support credit(PRSC) approved. (1)Bank announces it will build upon currentprograms and follow the CaribbeanRegional Strategic Plan of Action forHIV/AIDS, devoting up to $150 million tothe fight against HIV/AIDS in theCaribbean. (April) (21)Joseph Ritzen appointed HDN VicePresident. (June) (1)Leadership Program on AIDS launched bythe World Bank Institute (WBI) to buildcapacity for accelerated implementationof HIV/AIDS programs. (21)IFC Against AIDS is launched with the ideaof accelerating the involvement of the privatesector in the fight against HIV/ AIDSthrough risk management and implementationof workplace programs. (6a)The Health and Education Unit becomes aDepartment within IFC. (7a)Sub-regional HIV/AIDS strategy for Caribbean.HIV/AIDS in the Caribbean: Issuesand Options released. (January) (55*)The Bank and partners gather in Washington,to further commit to operationalizethe Amsterdam Declaration. The GlobalPlan to Stop TB calls for expansion ofaccess to DOTS and increased financialbacking for the program fromgovernments throughout the world.(October) (56)The Bank’s Water and Sanitation Programforms the Private-Public Partnership forHandwashing with the London School ofHygiene and Tropical Medicine, the Academyfor Educational Development, USAID,UNICEF, the Bank-Netherlands WaterPartnership, and the private sector. (57)The Bank becomes a trustee of the GlobalFund to Fight HIV/AIDS, TB, and Malaria(GFATM), a financing mechanism establishedto foster partnerships between governments,civil society, the private sector,and affected communities to increase resourcesand direct financing toward effortsto fight HIV/AIDS, TB, and malaria. (58)In cooperation with the Gates Foundationand Dutch and Swedish governments, theWorld Bank Health and Poverty ThematicGroup initiates the Reaching the PoorProgram (RPP). RPP is an effort to findbetter ways to ensure that the benefits ofHNP programs flow to disadvantagedpopulation groups through research,policy guidance, and advocacy. (1)The Bank joins the Rockefeller Foundation,Sida/SAREC, and Wellcome Trust tolaunch the INDEPTH Network, an internationalplatform of sentinel demographicsites that provides health and demographicdata and research to enable developingcountries to set evidence-basedhealth priorities and policies. (59)The Bank and USAID cohost the AnnualMeetings of the Global Partnership toEliminate Riverblindness in Washington.The partners pledged to eliminateriverblindness in Africa by 2010. (1)(Table continues next page)115

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments2002Wolfensohn presents a seven-point Post-Monterrey Action Plan to the DevelopmentCommittee on how to boost developmentaid and effectiveness, andtranslate Monterrey commitments intoresults. (April) (33)From this point, Country Assistance Strategies(CASs), the main vehicle for makingstrategic choices about program designand resource allocations for individualcountries, were based on Poverty ReductionSupport Papers in low-income countries.(July) (60)IDA announces that 18–21 percent ofIDA would be in grants and available forspecific activities and for the debtvulnerablepoorest countries. (July) (1)$500 million is approved for the secondstage of its Multicountry HIV/AIDSProgram for Africa (MAP). (February) (1)WBI’s course “Adapting to Change”becomes “Achieving the MDGs: ReproductiveHealth, Poverty Reduction, andHealth Sector Reform.” (40)The HNP Sector Board presents an HNPstrategy update to the Board. The presentationreviews trends in project lendingand objectives, analytic and advisory services,QAG ratings, IFC lending for HNP,and staffing. The update reconfirms thesector’s commitment to the objectives inthe 1997 strategy. It also emphasizes thatgreater country selectivity and diversity inlending instruments will be pursued alongwith efforts to sharpen the focus on qualityand effectiveness, work more closelywith clients and communities, and improvetraining for staff and their allocationto ensure the appropriate skills mix.(March) (61)The 2002 IFC Health Strategy ispresented to the Board of Directors.The Global/HIV AIDS program is createdalong with the Global Monitoring andEvaluation Team (GAMET). GAMET ishoused at the World Bank and supportsefforts with UNAIDS to build countrylevelmonitoring and evaluation capacitiesas well as coordinate technical support.(June) (21)First phase of Bank-Netherlands WaterPartnership-Water Supply and Sanitationinitiated. (51)Global Alliance for Improved Nutrition(GAIN) created at a special UN sessionfor children. The World Bank is a keypartner, mainly managing trust funds andprogram implementation. (62)2003World Bank Annual Report describes theBank’s commitment to meeting the MDGsand emphasizes its commitment to fourpriority sectors including HIV/AIDS, waterand sanitation, health, and education forall. (September) (1)Jean-Louis Sarbib assumes HDN VicePresidency. (July) (12)Board approves first pilots of buy-downmechanism in several polio eradicationprojects in Pakistan and Nigeria. Projectswere financed by Gates Foundation, UNF,Rotary International, and the Centers forDisease Control and Prevention. (63)Romania Dialysis is the first publicprivatepartnership (PPP) project in healthapproved by IFC.Regional AIDS strategy for ECA published:Averting AIDS Crises in Eastern Europeand Central Asia (September) (64*)The Bank and the Pan-American HealthOrganization (PAHO) inaugurate the“Health Partnership for Knowledge Sharingand Learning in the Americas.” Theinitiative promotes the use of technologyto share expertise in order to meet theMDGs across the region. (October) (1)116

APPENDIX A: WORLD BANK GROUP HNP TIMELINE2004 Water Resources Sector Strategy: StrategicDirections for World Bank Engagementis published. The strategy highlightsthe centrality of water resource managementand development to sustainablegrowth and poverty reduction. It arguesthat the World Bank is perceived to have acomparative advantage in the area. It emphasizesthe need to tailor Country WaterAssistance Strategies to be consistentwith country context, CASs, and PovertyReduction Support Papers. (January) (65)Reaching the Poor Program sponsorsglobal conference for researchers todisseminate evidence of how well healthand other social programs reach the poorand to produce policy guidelines basedupon the evidence. (February) (66)The Bank sponsors an event for 35 Africanambassadors, Harmonizing Approaches toHealth in Africa, to intensify efforts toimprove women’s health in Africa andplan follow-up activities. (April) (1)IEG releases an evaluation of the Bank’sapproach to global programs, Addressingthe Challenges of Globalization. The evaluationrecommends that the Bank separateoversight of global programs frommanagement, improve standards of governanceand management of individualprograms, reevaluate selection and exitcriteria, strengthen links between globalprograms and country strategies, andstrengthen evaluations and review ofglobal programs within the Bank. (67*)WDR 2004: Making Services Work forPoor People identifies good governanceand accountability mechanisms as keydeterminants of health systemperformance. (68*)Regional HIV/AIDS strategy for EastAsia and Pacific published AddressingHIV/AIDS in East Asia and the Pacific.(January) (69*)Improving Health, Nutrition, and PopulationOutcomes in Sub-Saharan Africa-—The Role of the World Bank notes thatpositive trends in health indicators haveslowed or reversed in Sub-Saharan Africa.It argues that the Bank must use its comparativeadvantage to work with governmentsand partners to strengthen thecapacity of countries to improve healthoutcomes. Nutrition and population mustremain central issues in development inSub-Saharan Africa and accordingly, thereport presents a regional guide to shapestrategy formulation at the country or subregionallevel. (December) (70*)IFC clarifies five strategic priorities, ofwhich health and education are one. (8a)WHO and the Bank cosponsor the FirstHigh-Level Forum on the Health MDGs.Heads of development agencies, bilateralagencies, global health initiatives, andhealth and finance ministers agree onfour action areas: resources for healthand poverty reduction papers; aideffectiveness and harmonization; humanresources; monitoring performance.(January) (1)(Table continues next page)117

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitments2005 Paul Wolfowitz is approved by the Boardof Executive Directors as the WorldBank’s 10th President. (March) (1)In his speech at the Annual Meetings,Wolfowitz emphasizes the importance ofleadership and accountability, civilsociety and women, and the rule of lawas well as focusing on results. Whenspeaking on the importance of health onthe development agenda, he emphasizesthe World Bank’s commitment to fightmalaria with the same intensity asHIV/AIDS. (September) (71)An IEG evaluation of the Bank’s HIV/AIDSAssistance, Committing to Results:Improving the Effectiveness of HIV/AIDSAssistance, is released. It finds that theBank’s support has raised commitment andaccess to services, but the effect on thespread of HIV and survival is unclear. Itrecommends that the Bank: help governmentsto be strategic and selective, andprioritize high-impact activities and thehighest-risk behaviors; strengthen nationalinstitutions to manage and implementlong-run responses; and improve monitoringand evaluation to strengthen the localevidence base for decision making. (21*)When the Adviser for Population andReproductive, Maternal and Child Health(Elizabeth Lule) is appointed as managerof ACTAfrica, the Adviser position iseliminated. (January) (72)The Life Sciences Group is establishedwithin IFC’s Global ManufacturingDepartment.Rolling Back Malaria: The World BankGlobal Strategy and Booster Programprovides the basis and rationale forinitiating the five-year Booster Programfor Malaria Control. Its objectives are toincrease coverage, improve outcomes,and build capacity. Described as a “newbusiness model,” it prioritizes flexible,country-driven, and results-focusedapproaches. (January) (41*)World Bank partners launch the HealthMetrics Network, a global partnership toimprove the quality, availability, anddissemination of data for decision makingin health. (June) (73)118

APPENDIX A: WORLD BANK GROUP HNP TIMELINE20062007Task Force on Avian Flu for Africaestablished to manage the information,communication, and coordination aspectsof the response to avian influenza. Itsupports country teams to prepareindividual country operations; helpscoordinate the region’s response with theglobal and Bank-wide funding programs,with donors, and mobilize additionalfunding as necessary. (74)Paul Wolfowitz resigns as World BankPresident. (June) (79)Robert Zoellick becomes 11th World BankPresident. (July) (80)Cristian Baeza appointed as Acting HNPDirector (February) (75)Joy Phumaphi becomes Vice President ofthe Human Development Network.(February) (81)Julian Schweitzer becomes HNP SectorDirector. (October) (82)IFC Against AIDS is integrated into theSmall and Medium EnterprisesDepartment.Repositioning Nutrition as Central to Development:A Strategy for Large-Scale Actionaims to position nutrition as a priorityon the development agenda at both thecountry and international levels to bolsterincreased commitments and investment tofight malnutrition. It prioritizes: approachesthat reach the poor and mostvulnerable at strategic stages in their development;scaling-up proven and costeffectiveprograms; reorienting ineffectiveprograms; improving nutrition through deliberateactivities in other sectors; supportingaction research and learning bydoing; and mainstreaming nutrition intodevelopment strategies. (January) (76*)Health Financing Revisited: A Practitioner’sGuide reviews the policy optionsand tools available for health finance inlow- and middle-income countries. Keypriorities include: (i) mobilizing increasedand sustainable government healthspending; (ii) improving governance andregulation to strengthen the capacity ofhealth systems and ensure that investmentsare equitable and efficient; and (iii)coordinating donors to make more flexibleand longer-term commitments thatare aligned with the development goalsof a country. (May) (77*)The objective of the 2007 World BankStrategy for Health, Nutrition, andPopulation Results is to use a selectiveand disciplined framework to redoubleefforts to support client countries to:improve HNP outcomes, especially for thepoor; protect households from illness;ensure sustainable financing; andimprove sector governance and reducecorruption. (April) (63*)World Bank cosponsors the InternationalPledging Conference on Avian and HumanInfluenza in Beijing to assess financingneeds at country, regional and globallevels. (January) (74)World Bank joins the International MonetaryFund and the African DevelopmentBank in implementing the MultilateralDebt Relief Initiative (MDRI), forgiving100 percent of eligible outstanding debtowed to these three institutions by allcountries reaching the completion pointof the HIPC Initiative. The MDRI will effectivelydouble the volume of debt reliefalready expected from the enhanced HIPCInitiative. (78)World Bank signs agreement to join theInternational Health Partnership. ThePartnership aims to improve the work ofdonor and developing countries andinternational agencies to create andimplement plans and services thatimprove health outcomes for the poor.(September) (84)(Table continues next page)119

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix A: World Bank Group HNP Timeline (continued)Year World Bank group events HNP sector events HNP publications and strategies HNP partnerships and commitmentsPopulation Issues in the 21st Century:The Role of the World Bank focuses onlevels and trends in births, deaths,migration and population growth, andrelated challenges. After analyzing globaland regional trends, as well as those forlending for population, the report outlinesthe Bank’s areas of comparative advantage.It concludes that the Bank mustfocus analytical work on populationissues, and collaborate with the privatesector and global partners to develop andmainstream multisectoral populationpolicies appropriate for low-, middle-, andhigh-fertility countries. (April) (83*)The IFC Africa Health Strategy is presentedto the Board of Directors. (9a)Sources: The World Bank part of this timeline was compiled by Mollie Fair, and isextracted from appendix C of “From Population Lending to HNP Results: The Evolutionof the World Bank’s Strategies in Health, Nutrition, and Population,” IEG WorkingPaper, no. 2008/3, February 2008.1. World Bank Group Archives 2005.2. World Bank Web site. “Water Supply and Sanitation Projects the Bank’s Experience:1967–1989.” (http://go.worldbank.org/8LRMSA1520)3. King 2007.4. World Bank Group Archives, “Sector Department Chart.”5. World Bank 1972a.6. World Bank 1972b.7. Golladay and Liese 1980.8. World Bank 1973.9. World Bank 1974.10. World Bank 1975.11. United Nations System Web site. “Standing Committee on Nutrition.”(http://www.unsystem.org/SCN/Publications/html/mandate.html).12. World Bank Group Archives, World Bank Group Staff Directories.13. World Bank 1980b.14. World Bank 1980a.15. World Bank 1997b.16. World Bank 1984.17. Grosh and Muñoz 1996.18. Harvard School of Public Health Web site. (http://www.hsph.harvard.edu/review/fellow.shtml).19. Kapur and others 1997.20. World Bank 1986b.21. IEG 2005a.22. World Bank 1986a.23. World Bank 1988.24. Nassim 1991.25. World Bank 1989.26. People and Planet. Net Web site. (http://www.peopleandplanet.net/doc.php?id=1740).27. World Bank 1993c.28. Jamison and others 1993.29. World Bank. 1993b.30. World Bank 1994a.31. World Bank 1999c.32. Vaillancourt 2009.33. World Bank Group Archives. “James D. Wolfensohn Timeline of MajorDevelopments.”34. World Bank Web site. “Quality Assurance Group.” (http://web.worldbank.org/WBSITE/EXTERNAL/PROJECTS/QAG/0,,contentMDK:20067126~menuPK:114865~pagePK:109617~piPK:109636~theSitePK:109609,00.html).35. IAVI Web Site (http://www.iavi.org/viewpage.cfm?aid=24).36. Schieber 1997.37. World Bank 1997a.38. UNECA, UNICEF, and World Bank. 1998.39. World Bank 1998a.40. White, Merrick, and Yazbeck 2006.41. World Bank 2005b.42. Narayan and Petesch 2002.43. Wagstaff and Claeson 2004.44. IEG 1999.45. World Bank 1999a.46. Walt and Buse 2006.47. World Bank 2000b.48. Jha and Chaloupka 2000.49. World Bank 2000a.50. Gwatkin and others 2000.51. Bank-Netherlands Water Partnership Program Web site (http://www-esd.worldbank.org/bnwpp/).52. World Bank 2001c.53. WSP Web site (http://www.wsp.org).54. World Bank Web site. “Gender and Development.” (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTGENDER/0,,menuPK:336874~pagePK:149018~piPK:149093~theSitePK:336868,00.html).55. World Bank 2001a.56. Stop TB Partnership Web site (http://www.stoptb.org/stop_tb_initiative/).57. Global Public-Private Partnership for Handwashing with Soap Web site (http://www.globalhandwashing.org/).58. Kaiser Family Foundation Web site (www.kff.org/hivaids/timeline).59. INDEPTH Web site (http://www.indepth-network.org/core_documents/vision.htm).60. World Bank Web site. “Strategies.” (http://intranet.worldbank.org/WBSITE/INTRANET/SECTORS/HEALTHNUTRITIONANDPOPULATION/INTHIVAIDS/0,,contentMDK:20120702~menuPK:375837~pagePK:210082~piPK:210098~theSitePK:375799,00.html).61. HNP Sector Board. (draft, February 7, 2002). “Health, Nutrition and PopulationSector Strategy Briefing.”62. GAIN Web site (http://www.gainhealth.org/gain/ch/en-en/index.cfm?page=/gain/home/about_gain/history).120

APPENDIX A: WORLD BANK GROUP HNP TIMELINE63. World Bank 2007a.64. World Bank 2003b.65. World Bank 2004d.66. World Bank Web site. (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTHEALTHNUTRITIONANDPOPULATION/EXTPAH/0,,contentMDK:20744334~pagePK:210058~piPK:210062~theSitePK:400476,00.html?).67. IEG 2004a.68. World Bank 2003b.69. World Bank 2004a.70. World Bank 2004c.71. World Bank Web site. News and Broadcast. “Annual Meetings 2005 OpeningPress Conference with Paul Wolfowitz.” (http://web.worldbank.org/WBSITE/ EXTERNAL/NEWS/0,,contentMDK:20656903~pagePK:64257043~piPK:437376~theSitePK:4607,00.html).72. World Bank Web site. “News and Broadcasts>“ http://web.worldbank.org/WBSITE/EXTERNAL/NEWS/0,,contentMDK:20138122~pagePK:64257043~piPK:437376~theSitePK:4607,00.html73. WHO Web site. “What is HMN?” (http://www.who.int/healthmetrics/about/whatishmn/en/index.html).74. World Bank Web site. “Avian and Pandemic Influenza.” (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTHEALTHNUTRITIONANDPOPULATION/EXTTOPAVIFLU/0,,menuPK:1793605~pagePK:64168427~piPK:64168435~theSitePK:1793593,00.html).75. World Bank Web site. “Acting Assignments in HNP.” (http://intranet.worldbank.org/WBSITE/INTRANET/SECTORS/HEALTHNUTRITIONANDPOPULATION/0,,contentMDK:20131131~pagePK:210082~piPK:210098~theSitePK:281628,00.html)76. World Bank 2006c.77. Gottret and Schieber 2006.78. World Bank Web site. “Debt issues.” (http://web.worldbank.org/WBSITE/EXTERNAL/TOPICS/EXTDEBTDEPT/0,,menuPK:64166739~pagePK:64166681~piPK:64166725~theSitePK:469043,00.html).79. World Bank Web site. “Statements of Executive Director and PresidentWolfowitz.” (http://intranet.worldbank.org/WBSITE/INTRANET/UNITS/INTPRESIDENT2007/INTPASTPRESIDENTS/INTPRESIDENTSTAFCONN/0,,contentMDK:21339650~menuPK:64324835~pagePK:64259040~piPK:64258864~theSitePK:1014519,00.html).80. World Bank Web site. “President’s Staff Connection.” http://intranet.worldbank.org/WBSITE/INTRANET/UNITS/INTPRESIDENT2007/0,,contentMDK:21477815~menuPK:64821535~pagePK:64821348~piPK:64821341~theSitePK:3915045,00.html81. World Bank Web site. January 30, 2007. “Interview with Joy Phumaphi, NewHD Vice President.” (http://intranet.worldbank.org/WBSITE/INTRANET/UNITS/INTHDNETWORK/0,,contentMDK:21199087~menuPK:514396~pagePK:64156298~piPK:64152276~theSitePK:514373,00.html).82. World Bank Web site. “Julian Schweitzer, Sector Director, HNP, Human DevelopmentNetwork.” http://intranet.worldbank.org/WBSITE/INTRANET/ KIOSK/0,,contentMDK:21473063~menuPK:34897~pagePK:37626~piPK:37631~theSitePK:3664,00.html83. World Bank 2007g.84. Department for International Development Web Site. “International Health Partnershiplaunched today.” (http://www.dfid.gov.uk/news/files/ihp/ default.asp).1a IFC 1982.2a IFC 2002, p. 24.3a IFC 1998.4a IFC 2002, p. 24.5a IFC 1999.6a Lutalo 2006.7a IFC 2002, p. 24.8a IFC 2004.9a IFC 2007a.Note: * indicates the publication itself, otherwise, facts are reported in the citedreference.121

A show of hands among Nepalese children who wash their hands after using their new latrine. Photo courtesy of George T. Keith Pitman.

APPENDIX B: DEFINITION OF THE SAMPLES USED FORPORTFOLIO REVIEWS AND WORLD BANK HNP STAFF ANALYSISWorld Bank HNP Project PortfolioProjectsThe review of the portfolio managed by the HNPsector that is presented in chapters 2–4 includedall 220 active and closed projects approved fromfiscal 1997 to 2006 (table B.1). The 220 projectsare listed in appendix C. Approvals of supplementalallocations for active projects are attributedto the original project; they are not consideredseparate operations. The count of projects inother sectors with HNP components includedany project with HNP commitments, as definedbelow.HNP CommitmentsUp to five sector codes are assigned to everyWorld Bank lending operation, and the percentageof the loan to be dedicated to each sector codewas noted in the project design documents. HNPcommitments include the amounts committedunder sector codes JA (health), BK (compulsoryhealth finance), FB (noncompulsory health finance)and other historic codes used for thehealth sector (HB, HC, HE, HH, HP, HR, HT, HY).Total commitments to HNP were calculated by takingthe total amount of each project allocated tothese codes. 1 Because there is often more thanone sector code, even for HNP-managed projects,it means that less than 100 percent of the costof a loan or credit is actually being counted. Note,too, that for multisectoral Development PolicyLoans that are essentially direct budget supportto the government, the allocation across sectorcodes is entirely notional and does not reflectearmarked funds for any sector.World Bank Water Supply and SanitationProject PortfolioThe water supply and sanitation projects reviewedin chapter 4 include all 117 active and closed projectsapproved from fiscal 1997 to 2006 (table B.2)with financial commitments to sector codes WA(sanitation), WC (water supply), WS (sewerage),and WZ (general water, sanitation, and flood protection),and managed by the Water Supply and SanitationSector Board. Projects that are solely aimedat flood protection (WD) and solid waste management(WB) are not included. Supplementalcredits and projects approved under emergencyTable B.1: Projects Managed by the HNP Sector by Fiscal Year of Approval andProject StatusFiscal years1997–2001 2002–06 1997–2006Project status Projects Percent Projects Percent Projects PercentActive 9 9 101 83 110 50Closed 90 91 20 17 110 a 50Total 99 100 121 100 220 100a. Of these, 99 had been reviewed and rated by IEG as of September 30, 2008. In addition, 2 projects were cancelled before they were implemented and thusdid not receive an outcome rating.123

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable B.2: Water Supply and Sanitation Projects Included in the Portfolio ReviewFiscal years1997–2001 2002–06 1997–2006Project status Projects Percent Projects Percent Projects PercentActive 33 53 55 100 88 75Closed a 29 47 0 0 29 25Total 62 100 55 100 117 100a. As of October 31, 2007. Implementation Completion Reports had not been received by IEG for 3 of the 29 projects at the time of the review, reducing to 26the number of completed projects reviewed.Table B.3: Transport Projects Included in the Portfolio ReviewFiscal years1997–2001 2002–06 1997–2006Project status Projects Percent Projects Percent Projects PercentActive 28 22 96 93 124 54Closed a 98 78 7 7 105 46Total 126 100 103 100 229 100a. As of June 25, 2007.procedures were also excluded. Greater detail andthe list of projects can be found in Overbey (2008).World Bank Transport Project PortfolioThe transport projects reviewed in chapter 4 includeall 229 active and closed projects approvedfrom fiscal 1997 to 2006 (table B.3). Only projectsmanaged by the Transport Sector Board and thathad financial commitments under the sectorcodes TA (roads and highways), TP (ports, waterways,and shipping), TV (aviation), TW (railways),and TZ (general transportation, whichincludes urban transport) were included. Supplementalcredits and projects approved underemergency procedures are excluded. Transportprojects managed by other sector boards were excludedbecause the resources needed to trackdown the relatively small number of such projectswere not warranted. 2 Thus, transport projects relatedto improvements of air quality that fell eitherunder the Environment Sector Board or werefunded through the Global Environment Facility(GEF) have not been reviewed.World Bank HNP StaffingThe analysis of HNP staffing in chapter 2 usesfour sets of data on World Bank staff in the HNPsector provided by the Bank’s Human ResourcesDepartment:1. Master dataset of HNP staff. This is a masterdataset of all Bank staff at levels GF andhigher as of the end of every fiscal year (June30), from fiscal 1997 to 2007, who were mappedto the HNP sector. Staff members below levelGF are not systematically mapped to a sector,so are excluded from the analysis. The datasetsincluded the following data for each individual:fiscal year; UPI (staff identification) number; primarymanaging unit (PMU); unit of assignment;entry on duty (EOD); appointment type; level;job title; age; gender; whether Part I or II; yearsin the Bank; years in the PMU; whether postedat headquarters or in the field; duty country;whether a coterminous appointment; 3 programname; whether a manager; and whethera former Young Professional. The information124

APPENDIX B: DEFINITION OF THE SAMPLES USED FOR PORTFOLIO REVIEWS AND WORLD BANK HNP STAFF ANALYSISwas provided separately for each fiscal yearand merged by IEG. This provided the basis fora master panel dataset from which it is possibleto track the movement of staff into and outof the sector, using the UPI number as theidentifier.2. Hub dataset. This is a dataset of all Bank staffand consultants level GE and higher assignedto the “hub” or “anchor,” for the period fiscal1997–2007. The “hub” was defined as includingthe units HDDHE and its successor HDNHE(the HNP hub), and HDNGA (the central unitfor the AIDS program, established in 2004).The data and variables assembled for hub stafflevels GF and higher are identical to those forthe master HNP staff dataset. However, thisdataset also includes staff at level GE and inother categories, such as junior professional associates,junior professional officers, coterminousstaff, and special assignments, who wereworking in those units.3. New hires. A dataset of all new hires of individualsdirectly into the World Bank HNP sector.The variables available included all of thevariables in the master HNP staff dataset, plusthe effective date of the hire.4. Exits. A dataset of all terminations of individualsfrom the World Bank who were mappedto the HNP sector at the time that they left. Thisincludes, for example, resignations, retirements,and deaths of HNP staff. The variables availablefor analysis included all of the variablesin the master HNP staff dataset, plus the effectivedate that of the hire and the reason fortermination.Only a few corrections were made to the originaldata for the analysis. First, the original data includedseveral individuals mapped to units in IFCwho were dropped. No other individuals weredropped, though in a number of instances the unitcodes did not seem to pertain to health (for example,the Board, Staff Association, External Affairs,Commodity Risk Group). Second, in thenew hire and exit datasets there were often duplicatesassociated with conversions from onetype of assignment to another—for example, theperson appears as a new hire, then an exit, thena new hire shortly thereafter in another assignmenttype. In those instances, the first time thatthe individual appears was used for the purposesof counting the number of new hires and their age,and in the exit dataset instances were excludedof those who had been converted to a new assignmenttype. The last observation was used ifthere was more than one appearance and theperson was no longer at the Bank (or at least nolonger in the HNP sector at the Bank).Country Assistance StrategiesThe desk review of CASs included a sample of the211 CASs approved from fiscal 1997 to 2006. Inlight of the large number of countries in three Regions,the study reviewed: (a) all CASs for East Asia,the Middle East and North Africa, and South Asia;and (b) a random sample consisting of roughly halfof all CASs for Europe and Central Asia, LatinAmerica and the Caribbean, and Africa (table B.4).The results reported in chapters 2 and 4 havebeen weighted to take the stratification of thesample into account. A list of the CASs actually reviewedcan be found in Sinha and Gaubatz 2009.IFC Portfolio of Investment Projectsand Advisory ServicesThe portfolio of 52 IFC health projects reviewedin chapter 5 includes 35 active and 17 closed projectsapproved from fiscal 1997 to 2007 (appendixD). Active projects are those for which IFC has financialexposure; closed projects are those withwhich IFC no longer has a financial relationship.Health sector projects included those with thehealth and pharmaceuticals sector code and additionalprojects with business objectives relatedto health (for example, a medical training projectwith an education sector code). Dropped projects,cancellations, rights issues, reschedulings, restructurings,supplementary investments made inthe context of previously approved projects, investmentsthrough the Africa Enterprise Fundand Small Enterprise Fund, and individual investmentsunder agency lines were excluded. 4Chapter 5 assesses the performance of IFC projectsthat reached “early operating maturity.” 5 Theperformance of mature projects was assessedthrough either detailed Expanded Project SupervisionReports (XPSRs) prepared by the investment125

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable B.4: Country Assistance Strategies Issued in Fiscal Years 1997–2006,by Region and Year, and the IEG Sample for ReviewReviewedRegion 1997–2001 2002–06 Total sampleSub-Saharan Africa 29 26 55 31Europe and Central Asia 31 29 60 29Latin America and the Caribbean 20 20 40 21East Asia & Pacific 12 11 23 23Middle East and North Africa 9 10 19 19South Asia 7 7 14 14Total 108 103 211 137departments and validated by IEG, or by IEG’s deskreview of project information for those not coveredby an Expanded Project Supervision Report.For projects that had not yet reached early operatingmaturity, IEG gathered information abouteach project’s characteristics and design, as wellas information on implementation status in the supervisionreports. Field visits supplemented theinformation gathered by the desk review. The objectiveof the field visits was to conduct in-depthvalidation and interviews linked to ongoing andcompleted investments and Advisory Services.The five countries visited were selected to achieveRegional balance, to have more than one currentor past IFC health investment, to include investmentsin different time periods, and to include differenttypes of investments (for example, hospitalsand pharmaceuticals). During field visits, IFCclients, government agency officials, healthrelatedprofessional and business associations,relevant multilateral and/or bilateral developmentorganizations with private sector portfolios, andother stakeholders were interviewed.For Advisory Service projects, the monitoring andevaluation system was still in a pilot phase. IEGreviewed all health projects covered by the previousrounds of project completion report (PCR)pilots, as well as desk review of approval and supervisiondocuments for projects not covered bythe PCR. Interviews were conducted with IFCand World Bank managers, sector specialists, investmentofficers, and project task managers.126

APPENDIX C: WORLD BANK HNP SECTOR PROJECTS APPROVED INFISCAL YEARS 1996–2007Appendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment1997 P006059 Maternal and Child Health and Nutrition II Argentina 95.0 0.0 95.0 100.01997 P043418 AIDS and STD Control Argentina 11.9 0.0 11.9 15.01997 P044522 Essential Hospital Services Bosnia-Herzegovina 0.0 14.4 14.4 15.01997 P004034 Disease Control and Health Development Cambodia 0.0 27.7 27.7 30.41997 P010473 Tuberculosis Control India 0.0 129.6 129.6 142.41997 P010511 Malaria Control India 0.0 159.9 159.9 164.81997 P010531 Reproductive Health India 0.0 223.5 223.5 248.31997 P042540 Iodine Deficiency Control Indonesia 19.1 0.0 19.1 28.51997 P001999 Health Sector Development Program Niger 0.0 37.6 37.6 40.01997 P007927 Maternal Health/Child Development Paraguay 19.2 0.0 19.2 21.81997 P008814 Health Reform Pilot Russian Federation 66.0 0.0 66.0 66.01997 P041567 Endemic Disease Senegal 0.0 13.9 13.9 14.91997 P010526 Health Services Sri Lanka 0.0 17.5 17.5 18.81997 P009095 Primary Health Care Services Turkey 13.3 0.0 13.3 14.51998 P045312 Health Recovery Albania 0.0 13.9 13.9 17.01998 P050140 Health Armenia 0.0 8.4 8.4 10.01998 P037857 Health and Population Program Bangladesh 0.0 242.5 242.5 250.01998 P003566 Basic Health (Health VIII) China 0.0 78.2 78.2 85.01998 P052887 Health Comoros 0.0 6.7 6.7 8.41998 P007015 Provincial Health Services Dominican Republic 28.2 0.0 28.2 30.01998 P039084 Health Services Modernization Ecuador 40.5 0.0 40.5 45.01998 P045175 Health Sector Egypt, Arab Rep. of 0.0 90.0 90.0 90.01998 P043124 Health Eritrea 0.0 17.2 17.2 18.31998 P000825 Participatory HNP Gambia 0.0 17.8 17.8 18.01998 P000949 Health Sector Support Ghana 0.0 33.6 33.6 35.01998 P035688 National Health Development Program Guinea-Bissau 0.0 10.8 10.8 11.71998 P010496 Orissa Health Systems India 0.0 69.5 69.5 76.41998 P049385 Economic Restructuring India 72.3 58.1 130.4 543.21998 P035827 Women and Child Development India 0.0 273.0 273.0 300.01998 P036956 Safe Motherhood Indonesia 41.2 0.0 41.2 42.5(Table continues next page)127

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007 (continued)HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment1998 P001568 Community Nutrition II Madagascar 0.0 19.3 19.3 27.61998 P035689 Health Sector Investment Program Mauritania 0.0 23.8 23.8 24.01998 P007720 Health System Reform - SAL Mexico 700.0 0.0 700.0 700.01998 P055061 Health System Reform TA Mexico 21.5 0.0 21.5 25.01998 P035753 Health Sector II Nicaragua 0.0 18.2 18.2 24.01998 P004566 Early Childhood Development Philippines 17.1 0.0 17.1 19.01998 P002369 Integrated Health Sector Development Senegal 0.0 42.0 42.0 50.01998 P005746 Health Sector Tunisia 48.5 0.0 48.5 50.01998 P040551 Nutrition and Early Childhood Development Uganda 0.0 22.1 22.1 34.0SIL1999 P060392 Health Reform-APL I Bolivia 0.0 20.5 20.5 25.01999 P044523 Basic Health Bosnia-Herzegovina 0.0 8.2 8.2 10.01999 P043874 Disease Surveillance - VIGISUS Brazil 94.0 0.0 94.0 100.01999 P054120 AIDS and STD Control II Brazil 165.0 0.0 165.0 165.01999 P036953 Health IX China 9.4 47.0 56.4 60.01999 P000756 Health Sector Development Ethiopia 0.0 99.0 99.0 100.01999 P052154 Structural Reform Support Georgia 0.0 8.3 8.3 16.51999 P041568 Population and Reproductive Health Guinea 0.0 9.9 9.9 11.31999 P045051 HIV/AIDS II India 0.0 183.4 183.4 191.01999 P050651 Maharashtra Health System India 0.0 123.3 123.3 134.01999 P003967 Health V Indonesia 38.0 0.0 38.0 44.71999 P039749 Health Sector Reform Jordan 34.0 0.0 34.0 35.01999 P046499 Health Restructuring Kazakhstan 39.5 0.0 39.5 42.51999 P058520 Health Latvia 10.8 0.0 10.8 12.01999 P036038 Population/Family Planning Malawi 0.0 5.0 5.0 5.01999 P040652 Health Sector Development Program Mali 0.0 40.0 40.0 40.0(PRODESS)1999 P055003 Nutrition, Food Security and Social Mauritania 0.0 1.9 1.9 4.9Mobilization LIL1999 P005525 Health Management Morocco 64.0 0.0 64.0 66.01999 P040179 Health Pilot Panama 2.2 0.0 2.2 4.31999 P009125 Health Uzbekistan 26.7 0.0 26.7 30.02000 P055482 Public Health Surveillance and Disease Control Argentina 50.4 0.0 50.4 52.52000 P063388 Health Insurance for the Uninsured Argentina 3.3 0.0 3.3 4.92000 P050751 National Nutrition Program Bangladesh 0.0 82.8 82.8 92.02000 P055157 Health Sector Reform Bulgaria 51.3 0.0 51.3 63.32000 P055122 Health Sector Support Chad 0.0 33.6 33.6 41.52000 P051273 Health System Croatia 27.3 0.0 27.3 29.02000 P067330 Immunization Strengthening India 0.0 129.8 129.8 142.62000 P050657 Health Systems Development India 0.0 95.7 95.7 110.0128

APPENDIX C: WORLD BANK HNP SECTOR PROJECTS, FISCAL YEARS 1996–2007Appendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007 (continued)HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment2000 P049545 Provincial Health I Indonesia 0.0 33.8 33.8 38.02000 P059477 Water and Sanitation for Low Income Indonesia 0.0 24.0 24.0 77.4Communities II2000 P069943 Primary Health Care and Nutrition II Iran 82.7 0.0 82.7 87.02000 P053200 Health Sector Reform Lesotho 0.0 4.5 4.5 6.52000 P035780 Health Lithuania 18.7 0.0 18.7 21.22000 P051741 Health Sector Support II Madagascar 0.0 38.4 38.4 40.02000 P062932 Health Reform Program Peru 77.6 0.0 77.6 80.02000 P008797 Health Sector Reform Romania 37.6 0.0 37.6 40.02000 P051418 Health Sector Management Slovenia 5.9 0.0 5.9 9.52000 P058358 Health Sector Development Project Solomon Islands 0.0 3.5 3.5 4.02000 P049894 Primary Health Care Tajikistan 0.0 4.8 4.8 5.42000 P058627 Health Sector Development Program Tanzania 0.0 20.5 20.5 22.02001 P069293 Health Reform LIL Azerbaijan 0.0 4.0 4.0 5.02001 P069933 HIV/AIDS Prevention Bangladesh 0.0 39.2 39.2 40.02001 P075220 HIV/AIDS I Barbados 14.4 0.0 14.4 15.22001 P074212 Health Sector Reform APL II Bolivia 0.0 32.2 32.2 35.02001 P073065 Multisectoral HIV/AIDS Cameroon 0.0 20.0 20.0 50.02001 P071505 HIV/AIDS Prevention & Control Project Dominican Republic 21.8 0.0 21.8 25.02001 P065713 HIV/AIDS, Malaria, STD, and TB Control Eritrea 0.0 33.2 33.2 40.02001 P069886 MAP Ethiopia 0.0 47.8 47.8 59.72001 P060329 HIV/AIDS Rapid Response Gambia 0.0 11.0 11.0 15.02001 P071617 AIDS Response Project (GARFUND) Ghana 0.0 21.3 21.3 25.02001 P067543 Leprosy II India 0.0 27.3 27.3 30.02001 P049539 Provincial Health II Indonesia 58.8 37.2 96.0 103.22001 P070920 HIV/AIDS Disaster Response Kenya 0.0 31.5 31.5 50.02001 P066486 Decentralized Reproductive Health and Kenya 0.0 46.0 46.0 50.0HIV/AIDS2001 P051372 Health II Kyrgyz Republic 0.0 12.5 12.5 15.02001 P066321 Basic Health Care III Mexico 343.0 0.0 343.0 350.02001 P051174 Health Investment Fund Moldova 0.0 9.5 9.5 10.02001 P064926 Health Sector Management Samoa 0.0 3.9 3.9 5.02001 P072482 HIV/AIDS Control Uganda 0.0 36.6 36.6 47.52001 P050495 Caracas Metropolitan Health Venezuela, R. B. de 28.8 0.0 28.8 30.32002 P073118 Multisectoral HIV/AIDS Benin 0.0 13.6 13.6 23.02002 P057665 Family Health Extension Project I Brazil 64.6 0.0 64.6 68.02002 P071433 HIV/AIDS Disaster Response Burkina Faso 0.0 16.3 16.3 22.02002 P071371 Multisectoral HIV/AIDS Control and Orphans Burundi 0.0 10.8 10.8 36.02002 P073525 HIV/AIDS Central African Rep. 0.0 8.0 8.0 17.02002 P074249 HIV/AIDS Cape Verde 0.0 6.5 6.5 9.0(Table continues next page)129

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007 (continued)HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment2002 P072226 Population and AIDS II Chad 0.0 16.5 16.5 24.62002 P071147 Tuberculosis Control China 96.7 0.0 96.7 104.02002 P073892 Health Sector Strengthening and Modernization Costa Rica 16.2 0.0 16.2 17.02002 P071062 Health Sector Development Djibouti 0.0 15.0 15.0 15.02002 P067986 Earthquake Emergency Reconstruction and El Salvador 135.5 0.0 135.5 142.6Health Services Extension2002 P053575 Health System Reform Honduras 0.0 26.6 26.6 27.12002 P074641 HIV/AIDS Prevention and Control II Jamaica 13.5 0.0 13.5 15.02002 P072987 Multisectoral STI/HIV/AIDS Prevention I Madagascar 0.0 15.4 15.4 20.02002 P070290 Health System Development II Nigeria 0.0 111.8 111.8 127.02002 P070291 HIV/AIDS Program Development Nigeria 0.0 82.2 82.2 90.32002 P069916 Social Expenditure Management II Philippines 20.0 0.0 20.0 100.02002 P074059 HIV/AIDS Prevention and Control Senegal 0.0 25.8 25.8 30.02002 P070541 Nutrition Enhancement Program Senegal 0.0 11.3 11.3 14.72002 P073883 HIV/AIDS Response Sierra Leone 0.0 13.1 13.1 15.02002 P073305 Regional Blood Transfusion Centers Vietnam 0.0 38.2 38.2 38.22002 P043254 Health Reform Support Yemen 0.0 25.1 25.1 27.52003 P078324 Health Sector Emergency Rehabilitation Afghanistan 0.0 53.0 53.0 59.62003 P071004 Social Insurance TA Bosnia-Herzegovina 0.0 4.6 4.6 7.02003 P080400 AIDS and STD Control III Brazil 100.0 0.0 100.0 100.02003 P054119 Bahia Development (Health) Brazil 9.0 0.0 9.0 30.02003 P070542 Health Sector Support Cambodia 0.0 24.3 24.3 27.02003 P073603 HIV/AIDS, Malaria and TB Control Djibouti 0.0 3.6 3.6 12.02003 P076802 Health Reform Support Dominican Republic 30.0 0.0 30.0 30.02003 P082395 First Programmatic Human Dev. Reform Ecuador 14.0 0.0 14.0 50.02003 P040555 Primary Health Care Development Georgia 0.0 17.3 17.3 20.32003 P073649 Health Sector Program Support II Ghana 0.0 89.6 89.6 89.62003 P076715 HIV/AIDS Prevention and Control II Grenada 1.5 1.5 3.0 6.02003 P073378 Multi-Sectoral AIDS Guinea 0.0 7.1 7.1 20.32003 P075056 Food and Drugs Capacity Building India 0.0 54.0 54.0 54.02003 P073772 Health Workforce and Services (PHP III) Indonesia 21.8 52.2 73.9 105.62003 P074122 AIDS Control Moldova 0.0 5.5 5.5 5.52003 P078053 HIV/AIDS Response Mozambique 0.0 22.0 22.0 55.02003 P071612 Multisectoral STI/HIV/AIDS II Niger 0.0 10.0 10.0 25.02003 P080295 Polio Eradication Nigeria 0.0 28.7 28.7 28.72003 P074856 HIV/AIDS Prevention Project Pakistan 0.0 37.1 37.1 37.12003 P081909 Partnership For Polio Eradication Pakistan 0.0 20.0 20.0 20.02003 P064237 TB/AIDS Control Russia 150.0 0.0 150.0 150.02003 P046497 Health Reform Implementation Russia 24.0 0.0 24.0 30.02003 P071374 Multisectoral HIV/AIDS Rwanda 0.0 10.7 10.7 30.5130

APPENDIX C: WORLD BANK HNP SECTOR PROJECTS, FISCAL YEARS 1996–2007Appendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007 (continued)HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment2003 P077675 Health Serbia 0.0 20.0 20.0 20.02003 P074128 Health Sector Reconstruction and Development Sierra Leone 0.0 14.0 14.0 20.02003 P074730 National HIV/AIDS Prevention Sri Lanka 0.0 6.3 6.3 12.62003 P076798 HIV/AIDS Prevention and Control Project St . Kitts & Nevis 2.9 0.0 2.9 4.12003 P075528 HIV/AIDS Prevention and Control Trinidad & Tobago 20.0 0.0 20.0 20.02003 P069857 TB/AIDS Control Ukraine 45.0 0.0 45.0 60.02003 P003248 Zanara HIV/AIDS APL Zambia 0.0 16.8 16.8 42.02004 P082613 Regional HIVAIDS Treatment Acceleration Africa 0.0 38.9 38.9 59.82004 P074850 HIV/AIDS Project for Abidjan Lagos Transport Africa 0.0 6.8 6.8 16.6Corridor2004 P071025 Provincial Maternal-Child Health Argentina 115.4 0.0 115.4 135.82004 P072637 Provincial Maternal-Child Health Sector Argentina 675.0 0.0 675.0 750.0Adjustment2004 P073974 Health Systems Modernization Armenia 0.0 17.9 17.9 19.02004 P083169 HIV/AIDS and STI Prevention and Control Bhutan 0.0 2.6 2.6 5.82004 P087841 Social Sector Programmatic Credit Bolivia 0.0 6.3 6.3 25.02004 P083013 Disease Surveillance and Control II Brazil 57.0 0.0 57.0 100.02004 P080721 HIV/AIDS Prevention and Control Caribbean Region 0.0 2.3 2.3 9.02004 P077513 HIV/AIDS & Health Congo, Rep. of 0.0 4.6 4.6 19.02004 P073442 HIV/AIDS Global Mitigation Support Guinea-Bissau 0.0 1.4 1.4 7.02004 P076722 HIV/AIDS Prevention and Control Guyana 0.0 4.7 4.7 10.02004 P050655 Rajasthan Health Systems Development India 0.0 71.2 71.2 89.02004 P086670 Health Sector Management Macedonia, FYR 9.0 0.0 9.0 10.02004 P073821 Multi-Sectoral AIDS Malawi 0.0 3.5 3.5 35.02004 P078368 Multisectoral HIV/AIDS Control Mauritania 0.0 4.2 4.2 21.02004 P082223 Health System (Montenegro) Montenegro 0.0 4.9 4.9 7.02004 P075979 Social Sector Support São Tomé & Principe 0.0 1.7 1.7 6.52004 P082879 Health TA Slovak Republic 4.3 0.0 4.3 12.42004 P065954 Health Reform - SECAL Slovak Republic 50.3 0.0 50.3 62.92004 P050740 Health Sector Development Sri Lanka 0.0 26.4 26.4 60.02004 P082335 Health Sector Development II Tanzania 0.0 58.5 58.5 65.02004 P071014 HIV/AIDS Tanzania 0.0 10.5 10.5 70.02004 P075230 Health Sector Support Tonga 0.0 10.6 10.6 10.92004 P074053 Health Transition Turkey 24.2 0.0 24.2 60.62005 P080406 African Regional Capacity Building Network Africa 0.0 8.5 8.5 10.0for HIV/AIDS Prevention, Treatment, & Care2005 P080413 HIV/AIDS Great Lakes Initiative APL Africa 0.0 11.0 11.0 20.02005 P083180 HAMSET SIL Angola 0.0 1.9 1.9 21.02005 P074841 HNP Sector Program Bangladesh 0.0 120.0 120.0 300.02005 P091365 Social Sector Programmatic Credit II Bolivia 0.0 3.8 3.8 15.0(Table continues next page)131

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix C: World Bank HNP Sector Projects Approved in Fiscal Years 1996–2007 (continued)HNP commitment($US millions)Fiscal year Project Total Totalapproved ID Project name Country IBRD IDA HNP commitment2005 P088663 Health Sector Enhancement Bosnia-Herzegovina 0.0 9.4 9.4 17.02005 P082243 HIV/AIDS Central America 0.0 6.0 6.0 8.02005 P087003 AIDS Control Central Asia 0.0 16.3 16.3 25.02005 P094694 HIV/AIDS/STI/TB/Malaria/Reproductive Eritrea 0.0 12.0 12.0 24.0Health2005 P065126 Health Sector Support Guinea 0.0 17.5 17.5 25.02005 P073651 Disease Surveillance India 0.0 37.4 37.4 68.02005 P075058 Health Systems India 0.0 88.7 88.7 110.82005 P087843 HIV/AIDS Capacity Building TAL Lesotho 0.0 3.0 3.0 5.02005 P083401 Health Sector Support Malawi 0.0 12.0 12.0 15.02005 P040613 Health Sector Program Project Nepal 0.0 43.5 43.5 50.02005 P078991 Health Sector II (APL 2) Nicaragua 0.0 10.2 10.2 11.02005 P079628 Women’s Health and Safe Motherhood II Philippines 12.5 0.0 12.5 16.02005 P078971 Health Sector Reform II Romania 76.8 0.0 76.8 80.02005 P076795 HIV/AIDS Prevention and Control St Lucia 0.8 0.8 1.7 6.42005 P076799 HIV/AIDS Prevention and Control St. Vincent & 1.0 1.0 2.0 7.0the Grenadines2005 P051370 Health II Uzbekistan 0.0 40.0 40.0 40.02005 P082604 HIV/AIDS Prevention Vietnam 0.0 24.5 24.5 35.02006 P082814 Health System Modernization Albania 0.0 13.1 13.1 15.42006 P094220 Health Sector Reform Azerbaijan 0.0 43.0 43.0 50.02006 P096482 Malaria Control Booster Program Benin 0.0 18.6 18.6 31.02006 P093987 Health Sector Support and AIDS Burkina Faso 0.0 35.3 35.3 47.72006 P088751 Health Sector Rehabilitation Support Congo, Dem. Rep. 0.0 135.0 135.0 150.0(Zaire)2006 P088575 Health Insurance Strategy Ecuador 90.0 0.0 90.0 90.02006 P088797 Multisectoral HIV/AIDS Ghana 0.0 6.2 6.2 20.02006 P077756 Maternal and Infant Health and Nutrition Guatemala 31.9 0.0 31.9 49.02006 P085375 Water Supply and Sanitation for Low-Income Indonesia 0.0 6.9 6.9 137.5Communities III2006 P084977 Health and Social Protection Kyrgyz Republic 0.0 10.8 10.8 15.02006 P100081 Avian and Human Influenza Control Lao, PDR 0.0 1.2 1.2 4.02006 P074027 Health Services Improvement Project Lao, PDR 0.0 14.3 14.3 15.02006 P076658 Health Sector Reform Phase II Lesotho 0.0 4.6 4.6 6.52006 P090615 Multisectoral STI/HIV/AIDS Prevention II Madagascar 0.0 10.5 10.5 30.02006 P094278 Health and Nutrition Support Mauritania 0.0 7.3 7.3 10.02006 P083350 Institutional Strengthening & Health Sector Niger 0.0 22.8 22.8 35.0Support Program2006 P097402 Second Partnership For Polio Eradication Pakistan 0.0 46.7 46.7 46.72006 P082056 Mother and Child Basic Health Insurance Paraguay 12.1 0.0 12.1 22.02006 P075464 National Sector Support For Health Reform Philippines 99.0 0.0 99.0 110.02006 P078978 Community and Basic Health Tajikistan 0.0 8.0 8.0 10.02006 P079663 Mekong Regional Health Support Vietnam 0.0 69.3 69.3 70.02006 P096131 Malaria Health Booster Zambia 0.0 5.8 5.8 20.0132

APPENDIX D: IFC HEALTH INVESTMENTS, FISCAL YEARS 1997–2007133

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAppendix D: IFC Health Investments Approved in Fiscal Years 1997–2007Project IFC IFC Total IFCApproval Commitment size loans equity commitmentCountry Project name date date (US$ 000) Primary sector Secondary sector ($ 000) ($ 000) (net, US$ 000)Uzbekistan Core Pharm 30-Jun-97 01-Oct-97 12,200 Chemicals Pharmaceuticals 3,350 500 3,850India Duncan Hospital 30-Jun-97 31-Oct-97 29,300 Health Care Hospitals, Clinics, Laboratories & Other 7,000 87 7,087Latin America Region MSF Holding 20-Nov-97 27-Apr-98 90,000 Finance & Insurance Rental & Leasing Services 15,000 2,000 17,000Argentina Hospital Privado 30-Apr-98 31-Aug-98 21,200 Health Care Hospitals, Clinics, Laboratories & Other 9,600 0 9,600Sri Lanka Apollo Lanka 06-May-98 20-Jun-01 32,000 Health Care Hospitals, Clinics, Laboratories & Other 5,350 1,096 6,447Mexico CIMA Mexico 30-Jun-98 04-Nov-98 11,200 Health Care Hospitals, Clinics, Laboratories & Other 0 4,800 4,800Mexico CIMA Puebla 30-Jun-98 06-Jan-99 32,500 Health Care Hospitals, Clinics, Laboratories & Other 3,500 0 3,500Costa Rica CIMA Costa Rica 30-Jun-98 04-Nov-98 2,800 Health Care Hospitals, Clinics, Laboratories & Other 0 1,200 1,200Brazil Itaberaba 04-Mar-99 24-Feb-00 25,000 Health Care Hospitals, Clinics, Laboratories & Other 0 5,340 5,340Central Europe Region Medicover 15-Apr-99 14-May-99 22,000 Health Care Hospitals, Clinics, Laboratories & Other 7,000 0 7,000Macedonia, FYR Alkaloid 30-Jun-99 20-Jul-99 38,200 Chemicals Pharmaceuticals 4,473 0 4,473Dominican Republic Hospital 16-Jul-99 07-Apr-00 45,900 Health Care Hospitals, Clinics, Laboratories & Other 2,000 0 2,000Brazil Fleury 17-Aug-99 08-May-00 58,000 Health Care Hospitals, Clinics, Laboratories & Other 15,000 0 15,000Philippines Asian Hospital 14-Feb-00 19-Dec-00 88,475 Health Care Hospitals, Clinics, Laboratories & Other 5,000 0 5,000Mexico Hospital ABC 02-May-00 07-Feb-01 81,800 Health Care Hospitals, Clinics, Laboratories & Other 10,294 0 10,294China Wan Jie Hospital 26-Jun-00 28-Nov-01 57,300 Health Care Hospitals, Clinics, Laboratories & Other 15,000 0 15,000Croatia Pliva 09-Nov-00 17-Nov-00 113,000 Chemicals Pharmaceuticals 35,000 0 35,000Russian Federation NMC 27-Dec-00 11-Apr-01 9,000 Health Care Hospitals, Clinics, Laboratories & Other 2,100 0 2,100India Orchid 01-Mar-01 19-May-01 116,000 Chemicals Pharmaceuticals 20,000 0 20,000Philippines AEI 07-Jun-01 20-May-03 4,200 Health Care Hospitals, Clinics, Laboratories & Other 1,000 0 1,000Latin America Region Salutia 12-Jul-01 04-Jan-02 10,000 Health Care Hospitals, Clinics, Laboratories & Other 0 2,500 2,500Eastern Europe Region Euromedic 09-Oct-01 14-Nov-01 33,000 Health Care Hospitals, Clinics, Laboratories & Other 12,976 0 12,976Vietnam F-V Hospital 11-Oct-01 30-Nov-01 40,000 Health Care Hospitals, Clinics, Laboratories & Other 8,000 0 8,000Egypt, Arab Rep. of SEKEM 28-Mar-02 16-Jan-03 13,300 Chemicals Pharmaceuticals 5,000 0 5,000Costa Rica Gutis 24-Jun-02 10-Jan-03 15,800 Health Care Hospitals, Clinics, Laboratories & Other 6,000 0 6,000India Max Healthcare 04-Apr-03 03-Sep-03 84,000 Health Care Hospitals, Clinics, Laboratories & Other 128 0 128134

APPENDIX D: IFC HEALTH INVESTMENTS, FISCAL YEARS 1996–2007Latin America Region Hospital II 09-Apr-03 26-May-03 42,200 Health Care Hospitals, Clinics, Laboratories & Other 12,000 0 12,000Jordan Hikma 19-Jun-03 27-Jun-03 32,000 Chemicals Pharmaceuticals 14,866 0 14,866Turkey MESA Hospital 23-Jun-03 22-Jul-03 45,000 Health Care Hospitals, Clinics, Laboratories & Other 11,000 0 11,000Southern Europe Region Euromedic II 04-Mar-04 12-Mar-04 25,800 Health Care Hospitals, Clinics, Laboratories & Other 12,644 0 12,644Turkey Acibadem 29-Oct-04 11-Nov-04 40,900 Health Care Hospitals, Clinics, Laboratories & Other 20,000 0 20,000India Bharat Biotech 16-Feb-05 26-May-05 4,500 Chemicals Pharmaceuticals 0 4,500 4,500China BioChina 11-Apr-05 10-May-05 25,000 Collective Investment Private Equity Funds 0 4,650 4,650VehiclesMexico CentroMedico PDH 27-Apr-05 10-Jun-05 69,000 Health Care Hospitals, Clinics, Laboratories & Other 14,500 0 14,500Bosnia and Herzegovina Bosnalijek expan 16-May-05 13-Jun-05 24,273 Chemicals Pharmaceuticals 9,403 0 9,403India Dabur Pharma 03-Jun-05 07-Jun-05 69,000 Chemicals Pharmaceuticals 0 15,064 15,064Mexico Centro Espanol 07-Jun-05 27-Apr-06 15,000 Health Care Hospitals, Clinics, Laboratories & Other 5,000 0 5,000India Apollo equity 08-Aug-05 10-Aug-05 NA Health Care Hospitals, Clinics, Laboratories & Other 0 5,085 5,085China United Family 16-Sep-05 13-Oct-05 16,000 Health Care Hospitals, Clinics, Laboratories & Other 8,018 0 8,018China SAC 04-Nov-05 22-Nov-05 13,000 Education Services Education 3,000 1,600 4,600Ukraine Biocon Group 31-Jan-06 15-May-06 7,223 Chemicals Pharmaceuticals 3,500 0 3,500Turkey Acibadem II 24-Feb-06 14-Mar-06 112,894 Health Care Hospitals, Clinics, Laboratories & Other 40,000 0 40,000Kenya Adv Bio-Extracts 23-Jun-06 18-Jan-07 29,300 Chemicals Pharmaceuticals 9,000 0 9,000Egypt, Arab Rep. of Dar Al Fouad 01-Aug-06 29-Apr-07 31,000 Health Care Hospitals, Clinics, Laboratories & Other 0 4,000 4,000Romania MedLife 13-Oct-06 16-Oct-06 30,000 Health Care Hospitals, Clinics, Laboratories & Other 6,348 5,000 11,348China Aier Eye 25-Oct-06 06-Nov-06 29,825 Health Care Hospitals, Clinics, Laboratories & Other 8,112 0 8,112India Ocimum Bio 01-Nov-06 03-Nov-06 19,760 Health Care Hospitals, Clinics, Laboratories & Other 2,600 3,900 6,500China Fosun Pharma 08-Nov-06 08-Nov-06 104,000 Chemicals Pharmaceuticals 40,560 0 40,560Nigeria Hygeia Expansion 25-Jan-07 30-Jan-07 6,230 Health Care Hospitals, Clinics, Laboratories & Other 3,033 0 3,033China Weigao 23-Feb-07 02-Mar-07 20,000 Chemicals Pharmaceuticals 20,000 0 20,000India Max Phase II 24-May-07 29-Jun-07 146,861 Health Care Hospitals, Clinics, Laboratories & Other 0 67,144 67,144India Granules 07-Jun-07 08-Jun-07 31,900 Chemicals Pharmaceuticals 9,000 6,000 15,000135

Indian woman with her children, who help her run her business. Photo by Curt Carnemark, courtesy of the World Bank Photo Library.

APPENDIX E: WORLD BANK SUPPORT FOR POPULATIONThe World Bank’s population strategies andlending over the past decade took place inthe context of a shifting international consensuson approaches to population controland reproductive health and a number ofcompeting agendas. 1 The Bank’s initial involvementin the HNP sector in the 1970s aroseprimarily over concern for the adverse impacts ofrapid population growth and high fertility.However, the 1994 International Conference onPopulation and Development (ICPD) in Cairoshifted the focus of population programs awayfrom fertility reduction and family planning andtoward women’s health, economic and social factorsaffecting the demand for children, and theright to reproductive health care (Robinson andRoss 2007). Shortly thereafter, the internationalcommunity’s attention was drawn to addressingthe rising threat of HIV/AIDS; sectorwide approachesin health; and, following the recommendationsof the World Development Report1993, delivery of a package of basic or essentialhealth services for the poor.The Bank’s 1999 population strategy (World Bank1999c) committed to assist countries to link populationto poverty reduction and human development;advocate for cost-effective policies thatreflect country context; build on analysis and dialogue;provide sustained support; and strengthenskills and partnerships. A year later, the MDGswere adopted with no explicit family planning orreproductive health goal, even though both arekey to achieving many of the other MDGs.The focus of the Bank’s support for reducingfertility and population growth declined overthe decade, and family planning became oneof many components of an essential packageof health or reproductive health services.Over the period fiscal 1997 to 2006, the Bank approvedonly 14 population projects, defined for thepurposes of the review as those with populationin the title and/or including an objective to reducefertility, or with a population or family planning componentor subcomponent. 2 Population projectsthus defined represented about 6 percent of theHNP lending portfolio, declining from 11 to 2 percentof the portfolio over the decade. The populationprojects approved in three-quarters of thecountries were preceded by population projects.Those in Bangladesh, India, Indonesia, and Kenyarepresented the last in a long series stretchingback to the 1970s. However, the series ended duringthe period under review, replaced by operationsfocusing on reproductive health or sectorwideactivities.The Bank’s population support was directedto only about a quarter of the 35 countrieswith high fertility. Among the 13 countries thatreceived Bank support, 8 had a total fertility rateof 5.0 or greater. Only one freestanding populationproject was approved—the Population andFamily Planning Learning and Innovation Loanin Malawi, which implemented community-baseddistribution of family planning in rural areas on apilot basis. Two of the projects packaged populationand AIDS activities, three were linked to maternaland child health or reproductive health, andeight were part of a health or HNP sector project.The main activities financed by these projects included:training health workers; information, education,and communication on family planningand the benefits of smaller families; contraceptives,including social marketing; civil works; communityfunds; policy reform; and economic activities137

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONfor women and youth. Almost all of the supportwas to low-income countries.None of the projects with explicit fertility orpopulation objectives achieved them. Thethree projects with population or familyplanning components but no explicit fertilityobjective 3 at least partially achieved their objectives.Eleven of the population projects approvedfrom fiscal 1997 to 2006 have closed; only3 had satisfactory outcomes. In Guinea, India,Kenya, and Mali, modern contraceptive use andfertility were scarcely affected. 4 In Russia, theabortion rate declined in project areas at thesame rate as in nonproject areas, while moderncontraceptive use was stagnant. In Bangladesh,Gambia, and Senegal, fertility declined somewhat,but there is significant doubt about links tothe support of the Bank, other donors, or publicpolicy more generally. 5 Bangladesh had experienceda spectacular decline in fertility—from 7children per woman in the 1970s to 3.3 by 1999—in large part because of a highly successful familyplanning program supported by the Bank andother donors (IEG 2005b). From 1999 to 2004, fertilitycontinued to decline to 3.0 and the contraceptiveprevalence rate rose from about 50 to 60percent, but it is unlikely that this is primarily attributableto the family planning program supportedby the Health and Population SectorProgram, given the disruption in service deliverycaused by the controversy in attempting (withoutsuccess) to absorb the vertical family planningprogram into other health services.Shortcomings in project preparation contributedto poor outcomes. 6 According to completionreports and IEG fieldwork, project designswere often excessively complex, driven by participatoryor sectorwide approaches. This often resultedin a failure to prioritize activities, whichreduced the project’s feasibility and ultimate impactin the face of low implementation capacity.Also frequently noted was a lack of up-front riskanalysis, risk mitigation actions, and institutionalanalysis. An IEG field evaluation of the VietnamPopulation and Family Health Project (1996–2003)(IEG 2006d), for example, found very little increasein oral contraceptive use, partly because theexisting incentives for two-child families withinthe family planning delivery structure were nottaken into account. This is something that institutionalanalysis in advance of the project shouldhave been able to anticipate.The absorption of population and familyplanning into sectorwide programs—be theySWAps or health reform projects—may havecontributed to the lack of results. There weresignificant improvements in the modern contraceptiveprevalence rate and a reduction in the totalfertility rate during Ghana’s Second Populationand Family Health Project (1991–97). However,under the subsequent Health Sector SupportProject (1998–2002), which supported a sectorwideapproach, there was no progress on eitherof these outcomes (IEG 2005b, 2007d). A similarsituation occurred in Bangladesh, between theFourth Population and Health Project (1991–98)and the subsequent Health and Family PlanningProgram (1998–2005) (IEG 2006b). In both cases,the transition to a SWAp increased the emphasison process, but did not ensure the achievementof health-service performance and output targets,including those for population (IEG 2007d).Field visits in Egypt underscored the findingsof a recent study that family planning and reproductivehealth services are diluted within thebasic package of services delivered through newfamily health facilities supported by the Health ReformProject (1998–present) (IEG 2008b). Clientsreport that there are no longer special rooms infacilities for family planning clients; female physiciansor specialists to discuss the topic are notavailable; and family physicians have less specializedtraining and less time to devote to the clients(Zaky 2007 as reported in IEG 2008b).Nevertheless, there were some importantsuccesses in raising contraceptive use andsupport for fertility decline in high-fertilityenvironments, particularly with respect touse of family planning. The Egypt PopulationProject (1996–2005) contributed to raising contraceptiveuse and lowering fertility in rural UpperEgypt, while the Malawi Population and FamilyPlanning Project was able to raise modern contraceptiveuse in rural areas through community-138

APPENDIX E: WORLD BANK SUPPORT FOR POPULATIONbased distributors. Both of these projects includedimportant demand-generation activities.The Madagascar Second Health Program SupportProject financed training and contraceptives, witha particular focus on a new, long-duration implantablecontraceptive that is easily inserted andespecially convenient in rural and remote areas.The number of locations offering family planningservices increased by 45 percent from 2003 to2007, and the contraceptive prevalence rate rosefrom 9.7 to 24 percent (World Bank 2008e). Unfortunately,the successful investment in pilotfamily planning activities in Malawi was not replicatednationally, and contraceptive use remainslow, with little change (figure E.1). There hasbeen virtually no improvement in Ghana despitesupport by the Bank and other donors for a healthSWAp over the past decade, and there has beenmodest improvement in Bangladesh, despite thedisruption in family planning services by theunpopular and ultimately failed attempt to unifythe delivery of health and family planning services.The problems of high fertility and rapidpopulation growth have regained prominenceinternationally and in the Bank’spopulation strategy. There is greater recognitionthat lowering fertility and population growththrough demand-side activities alone will take avery long time, while simply providing familyplanning as part of a package of basic services isunlikely to raise contraceptive prevalence dramaticallyin high-fertility countries or to lowerfertility among the very poor. Both demand- andsupply-side activities are important. UniversalModern contraception rate in earliest surveyFigure E.1: Trends in Modern Contraceptive UseRates, Case Study Countries706050403020100Eritrea (1995–2002)Nepal (1996–2006)Malawi (2000–2004)Ghana (1998–2003)0 10 20 30 40 50 60 70Modern contraception rate in latest surveyaccess to reproductive health was added as anMDG in 2007. 7 The Bank’s 2007 population strategy(World Bank 2007g) differentiates between reproductive,maternal, and sexual health and theservices that address them and factors that affectdemographic outcomes. It links population, economicgrowth, and poverty reduction and advocatestargeting assistance to the 35 countries withtotal fertility rates exceeding 5.0, many of whichare also among the poorest countries in the world.45 degree lineVietnam (1997–2002)Egypt (1995–2005)Bangladesh (1996–2004)Peru (1996–2004)Source: Nankhuni forthcoming, based on demographic and health survey reports (www.measuredhs.com)Note: Solid circle indicates that the change is statistically significant at 5 or 10 percent level, hollow circleindicates that the change is not statistically significant, and solid triangle indicates that it was not possibleto determine statistical significance. Countries below the 45-degree line experienced improvements in uptakeof modern contraception rates.139

Woman in Burkina Faso cooking a typical meal over a wood fire. Photo by Ray Witlin,courtesy of the World Bank Photo Library.

APPENDIX F: WORLD BANK SUPPORT FOR NUTRITIONThe 2006 World Bank publication, RepositioningNutrition as Central to Development:A Strategy for Large Scale Action(World Bank 2006c), argues that malnutritionis one of the world’s most serious healthproblems and the largest contributor tochild mortality. About a third of children in developingcountries are underweight (low weightfor age) or stunted (low height for age), andabout 30 percent of the population of those countriessuffers from deficiencies in micronutrientssuch as vitamin A, iodine, or iron. 1 Women andchildren are particularly affected, and nutritionaldeficiencies in children while they are still in thewomb and up to age 2 can have lifelong consequences.Malnutrition among children is highestin South Asia and is high and increasing in Sub-Saharan Africa. It affects both the poor and thenon-poor, but is greater among the poor: in 39 outof 46 countries with recent household surveys,more than half of children are stunted, and stuntingis as much as eight times higher among thepoorest wealth quintile than among the richest. 2Tackling malnutrition not only contributesto the MDG of halving the share of peoplewho suffer from hunger, but also to otherMDGs that deal with reducing child mortality,improving maternal health, raisingschool achievement, and reducing incomepoverty.The causes of malnutrition are diverse and includeinadequate breastfeeding, poor child feedingpractices, diarrheal disease and other illness,intestinal parasites, frequent and closely spacedchildbearing, inadequate diet, low access to healthcare, unsafe water, poor sanitation, low purchasingpower, and in some cases inadequate food production.However, the fact that the non-poor alsohave significant levels of malnutrition indicates thatknowledge and behavior are often key.World Bank support is in countries withhigh malnutrition, but coverage of the worstaffectedcountries is low. IEG undertook an indepthdesk review of the 21 projects approvedfrom fiscal 1997 to 2006 with nutrition objectives(10 percent of the HNP lending portfolio), 3 plus6 additional projects with nutrition in the title ornutrition components or subcomponents, for atotal of 27 projects, henceforth called “nutritionprojects.” 4 About half of them are general healthor HNP projects, five are freestanding nutritionprojects, four are mother and/or child health projects,and the remaining five are emergency ormultisectoral programmatic lending. 5 While abouttwo-thirds of the nutrition projects were in countrieswith average child stunting of 30 percent ormore, only about a quarter of countries with suchhigh levels of malnutrition were receiving WorldBank nutrition support. 6 Two-thirds of nutritionprojects were in low-income countries. WhileAfrica had the largest number of nutrition projects(9), South Asia had the highest share of nutritionprojects relative to the rest of the Regional portfolio(29 percent). 7 The share of projects withnutrition objectives declined from 12 to 7 percentbetween the first and second half of the decade.The types of interventions supported by these operationsincluded growth monitoring and nutritionalsurveillance (100 percent), micronutrientsupplements (52 percent), behavior change (nutritioneducation, promotion of growth monitoring,breastfeeding, specific dietary changes,and hygiene, 48 percent), and feeding supplementsor rehabilitation of malnourished children(41 percent). The projects also supported ca-141

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONPercent stunted in earliest surveyFigure F.1: Trends in Stunting in the Rural Areas ofCase Study Countries, 1995–200680706050403020Nepal (1996–2001)Bangladesh (1996–2004)Malawi (2000–2004)Eritrea (1995–2002)Peru (1996–2000)Egypt (1995–2000)Ghana (1998–2003)45 degree line20 30 40 50 60 70 80Percent stunted in latest surveySource: Nankhuni forthcoming, based on demographic and health survey reports.Note: Solid circle indicates statistically significant change at p

APPENDIX F: WORLD BANK SUPPORT FOR NUTRITIONBox F.1: Reductions in Malnutrition in Bangladesh: Lessons from the IntegratedNutrition ProjectThe Bank has supported improved nutrition in Bangladeshthrough two freestanding nutrition projects—the BangladeshIntegrated Nutrition Project (BINP, 1995–2002) andthe National Nutrition Project (2000–07)—and as part oftwo projects supporting sectorwide approaches in HNP.In 2005, IEG evaluated the impact of the BINP, whichwas based on a community-based approach that providednutrition counseling to bring about behavior change andsupplementary feeding for pregnant women and youngchildren.The evaluation found that coverage of the interventionwas high in project areas in general, but that thecausal link between the interventions and nutrition outcomeswas weakened by targeting deficiencies; largeshares of mothers and children receiving supplementalfeeding but no counseling; and the focus of behaviorchange almost exclusively on mothers, who are often notthe main decision makers on nutrition-related practices(both husbands and mothers-in-law have an important influence).Supplementary feeding had some impact amongthe most malnourished, but was a costly part of the programand not sustainable in the long run.The follow-on National Nutrition Program revised thetargeting criteria and attempted to reach out to men withbehavior-change messages. The program was delayedand scaled back, but was able to maintain the achievementsof micronutrient coverage and to promote adoptionof new behaviors. It was unable to demonstratesustainable improvements in birth weights and nutritionstatus of vulnerable groups; however, the baseline wascollected only two years before the end of the project andmonitoring data were not collected.Source: IEG 2005c, World Bank 2007e.the impact was unclear, often due to the failure tocollect data or report on nutrition outcomes.Complexity was cited as contributing to theshortcomings in more than half of the weakperformingprojects. Projects in Bangladesh,Indonesia, Nicaragua, and Sri Lanka had to bescaled back and the efforts prioritized to includefewer activities or a smaller geographic area. TheFood Security and Social Mobilization Project inMauritania suffered from the inexperience andinstitutional weaknesses of the Executive StateSecretariat for Promotion of Women, in additionto high project complexity. The Uganda Nutritionand Child Development Project was designed asa pilot intended for only a few districts; expansionof its geographical coverage without additionalresources greatly increased its complexity.M&E were particularly weak for both projectswith nutrition components and thefreestanding nutrition projects. Nutrition outcomesare affected by many factors beyond theinterventions in these projects; it is thus very importantto attempt to monitor other important factorsthat could be affecting outcomes, a lesson ofthe project in Senegal. However, these projectsfailed even to collect basic data on nutrition outcomes,such as micronutrient consumption (Gambia,Mauritania, Sri Lanka) or the projects’ mainoutputs (Sri Lanka). The Bangladesh project didnot collect baseline data until two years before theend of the project; the Indonesia project producedbaseline and final data from different groupsof people; the India Women and Child DevelopmentProject collected data on children aged 0–3when the target group was aged 0–6. The completionreport for the Nicaragua project reportedno outcomes at all.This experience nevertheless presents someimportant lessons for future nutrition projects.Several projects reported success in theuse of community volunteers to mobilize communitiesor deliver services (Gambia, Senegal,Nicaragua), while one of the reasons for poorperformance in Mauritania was the limited capacityof communities to undertake growth-promotionactivities. The experience also highlightedthe importance of demand generation for nutri-143

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONtion services and behavior change communicationfor success in Senegal: lack of demand was a factorin weak results in India and Mauritania. Multisectoralcoordination and the engagement ofsectors outside of health was lacking in India andMauritania. Two early child development projectspointed to the need for better targeting ofchildren in a more appropriate age range (Uganda)or with more education and counseling (India).Finally, a number of projects cited the need to developsimpler indicators for use by grassrootsgroups, to adopt more realistic and measurabletargets (Bangladesh), and to identify monitoringindicators for improved supervision (Sri Lanka).To summarize, the need to address malnutritionin client countries is great, and the coverage ofBank support for the hardest-hit countries is low.However, the multidimensional determinants ofnutrition tend to lead to complex projects thatinvolve multiple sectors. While there have beendemonstrable results in a few cases, the overall resultsfor the HNP-managed nutrition portfoliowere thin.144

APPENDIX G: WORLD BANK SUPPORT FOR ANALYTIC WORK ON HNPThe Bank’s analytic work in HNP includes both theproducts of the lending part of the institutionand research products and other publications ofBank staff. Official economic and sector work(ESW) financed directly from the Bank budget forcountry, Regional, or global-level HNP analysis istracked in an internal database. 1Between fiscal 2000 and 2008 the Bankspent $43 million of its own budget andtrust funds on a total of 218 ESW tasks thatgenerated reports, policy notes, conferences,workshops, consultations, and country dialogueon HNP. 2 This amounts to about 4 percent ofESW Bank-wide for those years, whether measuredin costs or number of activities.However, the true amount of HNP analyticwork undertaken by the Bank far exceedswhat is in the official database. The trackingsystem excludes the work of the Bank’s ResearchDepartment and major undertakings, such asthe World Development Report 2004: MakingServices Work for the Poor and the multiyear researchon Reaching the Poor with Health, Nutrition,and Population Services (World Bank2005a). 3 An exhaustive search for all individual articles,working papers, studies, toolkits, reports,and research published by the Bank or Bank staffon HNP topics counted a total of 1,457 pieces issuedover fiscal 1997–2006—six times more thanimplied by the official ESW database. 4ESW on HNP tripled over fiscal 2001–05,then dropped to half its previous levels infiscal 2006–08 (figure G.1). About two-thirds ofthese activities were conducted at the countrylevel, about a quarter at the Regional level, and 7percent at the global level. 5 The largest shares ofthese ESW activities were country-level HNP sectorstudies (42 percent), studies of health finance(33 percent), HIV/AIDS studies (31 percent), orhealth strategies or policies (29 percent). The increasein ESW in fiscal 2002 and 2004–05 is mostlyattributable to an increase in country-level sectorstudies or reviews in those years; there followeda large drop in fiscal 2007–08.Prominent among the country-level studies wasa series of more than a dozen HNP Country StatusReports launched by the Africa Region in 2003,with the purpose of informing the HNP discussionin Poverty Reduction Strategies. Most used the analyticframework proposed in the HNP chapter ofthe Sourcebook for Poverty Reduction Strategies(Claeson and others 2002), linking health outcomes,household and community characteristicsand behaviors, health care delivery and financing,and other sectors affecting health. All focused onthe relationship between health and poverty to differingdegrees and included a chapter on publicexpenditures in health and health financing.Nearly two-thirds of all ESW tasks since2002 that stated an objective were intendedto inform government policy, while onlyhalf aimed to inform lending. 6 Other main objectivesof the HNP ESW portfolio were to informpublic debate (42 percent), inform the developmentcommunity (32 percent), and build capacity(25 percent). Not surprisingly, country andRegional ESW were more likely intended to informgovernment policy, and country-level ESW was themost likely to have an objective to inform lending(55 percent), while all—or nearly all—globalHNP ESW intended to inform public debate or the145

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONNumber of ESW activities deliveredFigure G.1: Trend in Official HNP ESW Tasks,Fiscal Years 2000–084540353025201510501413292435421921 21FY00 FY01 FY02 FY03 FY04 FY05 FY06 FY07 FY08Source: World Bank data.All ESW Country-level Regional GlobalTable G.1: Coverage of Topics in an Inventory ofHNP Analytic Work, Fiscal 1997–2006 (n =1,457)Topic Number PercentHealth system performance 597 41.0HIV/AIDS and STIs 321 22.0Child health 152 10.4Communicable diseases other than AIDS 109 7.5Injuries and noncommunicable diseases 109 7.5Nutrition and food security 99 6.8Population and reproductive health a 74 5.1Other HNP and human development 367 25.2Source: IEG inventory of HNP analytic work.Note: Categories are not mutually exclusive; percentages add to more than 100 percent.a. Of which only 11 deal exclusively with population/family planning.development community. IEG’s recent evaluationof Bank-wide ESW found that it led to higherproject quality at entry (IEG 2008h).The broader inventory of analytic work conductedfor this evaluation found that health system performancewas the most common topic, treated in41 percent of all analytic work (table G.1). HIV/AIDSwas the second-most common topic, while only1 in 10 publications addressed child health.Health is also frequently analyzed in publicexpenditure reviews (PERs), a categoryof formal ESW usually conducted by staff inother sectors, and thus not included in thestatistics for the HNP sector. 7 PERs are oftenthe basis for discussion with the Ministry ofFinance about sectoral budget allocations, andabout allocations within sectors. Thus, they are animportant input into discussions of efficiency andsustainability of finance. However, about a thirdof PERs delivered from fiscal 2000 to 2007 hadno chapter or subchapter on health. The healthfocus of PERs has declined over time: between fiscal2000–03 and 2004–07, the share with healthchapters or subchapters declined from 71 to 59percent. Only 3 percent of PERs had a chapter orsubchapter on nutrition or on population, fertility,or family planning. Of the five PERs witha population chapter or subchapter, only one(Ethiopia) was in a high-fertility country. Given theclose relationship between rapid populationgrowth, the dependency ratio, and sustainabilityof public expenditure, it is particularly surprisingthat population is not more widely discussed.146

APPENDIX H: ADDITIONAL FIGURES ON WORLD BANK HNPLENDING, ANALYTIC WORK, AND STAFFING147

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONFigure H.1: IEG Ratings for Closed HNP Projects Approved in Fiscal 1997–2006, byYear of Approval100Percent of projects rated moderatelysatisfactory or higher806040205659647166736663 64700OutcomeQuality at entryQuality of supervision Bank performance Borrower performanceRating categoryFiscal years 1997–99 (n = 55) Fiscal years 2000–06 (n = 44)Source: World Bank data.Figure H.2: IEG Ratings for Closed HNP Projects Approved in Fiscal 1997–2006,by Year of Exit100Percent of projects rated moderatelysatisfactory or higher80604020625469657167646367670OutcomeQuality at entryQuality of supervision Bank performance Borrower performanceRating categoryFiscal years 2001–05 (n = 47) Fiscal years 2006–08 (n = 52)Source: World Bank data.148

APPENDIX H: ADDITIONAL FIGURES ON WORLD BANK HNPFigure H.3: IEG Bank Performance Ratings,by Sector Board and Fiscal Year of Exit100Figure H.4: IEG Borrower PerformanceRatings, by Sector Board and Fiscal Yearof Exit100Percent of projects rated moderatelysatisfactory or higher806040200Source: World Bank data.1987–91 1992–96 1997–2001 2002–06Fiscal year project closedOther sectorsHNP sector8068Percent of projects rated moderatelysatisfactory or higher806040200Source: World Bank data.76711987–91 1992–96 1997–2001 2002–06Fiscal year project closedOther sectors HNP sectorPercent of projects rated likely or highFigure H.5: IEG Sustainability Ratings, HNPand Other Sectors, by Fiscal Year of Exit100806040200Source: World Bank data.(n = 14)78741987–91 1992–96 1997–2001 2002–06Fiscal year project closedOther sectors HNP sectorThousands of US dollarsFigure H.6: Mean and Median HNP ProjectPreparation Costs, by Fiscal Year ofApproval (nominal dollars)450400350300250200150100500Note: N = 220 projects.1997 1998 1999 2000 2001 2002 2003 2004 2005 2006Fiscal year of approvalMeanMedian149

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONMonthsFigure H.7: Duration of HNP ProjectPreparation, by Fiscal Year of Approval2520151050Note: N = 220 projects.1997 1998 1999 2000 2001 2002 2003 2004 2005 2006MeanFiscal year of approvalMedianConstant 2006 US$ thousandsFigure H.8: Average Annual SupervisionCosts for Investment Projects Managed byHNP and Other Sectors120100806040200Source: World Bank data.2000 2001 2002 2003 2004 2005 2006 2007 2008Fiscal yearHealth Sector BoardOther sectorsFigure H.9: Trends in HNP and Hub-MappedStaff of Level GF+ (excluding coterms),Fiscal 1997–2007200183 185184Figure H.10: HNP Operational Staff, byRegion and Fiscal Year6050Number of HNP staff150100500136145158178 17416516915 18 18 19 18 2212 1317 161691997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Fiscal yearHNP mapped staffHDNHE & HDNGA(hub) mappedSource: Nankhuni and Modi 2008.18Number of HNP staff4030201001997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Fiscal yearAfrica South Asia Europe and Central AsiaLatin America and the Caribbean East Asia and PacificMiddle East and North AfricaSource: Nankhuni and Modi 2008.150

APPENDIX H: ADDITIONAL FIGURES ON WORLD BANK HNPFigure H.11: Trend in Population andNutrition Staff25Number of population or nutrition staff201510501997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007Fiscal yearTotal staff with population in titleTotal staff with nutrition in titleSource: Nankhuni and Modi 2008.Note: If staff have both population and nutrition in their title, they are counted in bothcategories.Table H.1: Frequency of Bank Performance Issues in Closed HNP Projects Approved inFiscal 1997–2006, by Project OutcomeProjects withProjects withunsatisfactorysatisfactoryoutcomeoutcomeBank performance issue Number Percent Number PercentInadequate risk assessment 17 (40) 2 (4) 10:1Inadequate technical design 17 (40) 3 (5) 8:1Inadequate supervision 18 (43) 5 (9) 5:1Inadequate political or institutional analysis 17 (40) 8 (14) 3:1Inadequate baseline data or unrealistic targets 17 (40) 8 (14) 3:1Inadequate M&E framework, poor data quality 36 (86) 26 (46) 2:1Overly complex design 12 (29) 8 (14) 2:1Inadequate partner financing or coordination 5 (12) 4 (7) 2:1Implementation disrupted by a crisis 7 (17) 8 (14) 1:1Inadequate prior analytic work 5 (12) 0 (0) —Number of projects 42 57Source: IEG review of ICRs.Ratio of percentunsatisfactoryto percentsatisfactoryoutcome151

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONTable H.2: Probit Regressions on the Determinants of Project Outcome Ratings(n = 94 investment projects approved from fiscal 1997–2006)(1) (2) (3) (4) (5) (6) (7) (8)Variable dP/dx T dP/dx T dP/dx T dP/dx T dP/dx T dP/dx T dP/dx T dP/dx TSatisfactory Bank quality at entry 0.309 1.43 0.325 1.36 0.308 1.44 0.537 3.02 0.580 3.61 0.549 3.12Satisfactory Bank supervision 0.187 0.97 0.225 0.98 0.209 1.02 0.341 1.89 0.339 2.02 0.337 1.89Satisfactory borrower performance 0.800 4.04 0.822 4.06 0.799 4.10IDA –0.097 –0.05 0.055 0.23 –0.054 –0.28 0.224 0.86 0.056 0.29 0.219 0.84 0.089 0.63 0.369 1.90Multisectoral 0.050 0.29 0.040 0.23 0.112 0.68 0.227 0.68 0.234 0.71 0.323 1.45 0.424 1.46IDA* multisectoral –0.554 –2.69 –0.627 –2.96 –0.456 –1.90 –0.531 –1.46 –0.505 –1.34 –0.689 –3.14 –0.700 –2.47Europe and Central Asia 0.518 2.93 0.495 2.88 0.510 2.75 0.442 2.63Middle East and North Africa 0.479 2.58 0.486 2.52 0.481 2.54 0.442 2.07East Asia and Pacific 0.288 1.32 0.306 1.44 0.266 1.16 0.345 1.74Latin America and Caribbean 0.417 1.84 0.355 1.52 0.402 1.68 0.500 2.82South Asia 0.580 3.65 0.578 3.08 0.572 3.38 0.468 3.01$10–50 million –0.027 –0.17 –0.004 –0.02 0.052 0.33 0.258 1.05 0.197 0.83 0.276 1.11 0.082 0.42 0.099 0.42$50–100 million 0.034 0.22 0.070 0.38 0.101 0.66 0.193 0.74 0.120 0.45 0.203 0.78 0.089 0.44 0.030 0.12> $100 million 0.379 1.33 0.399 1.29 0.365 0.22 –0.118 –0.24 –0.194 –0.42 –0.093 –0.19 0.113 0.31 –0.220 –0.50Approved fiscal 2000–06 0.018 0.12 0.009 0.06 0.142 0.91 0.171 1.16 0.139 0.95 0.192 1.23 0.133 1.10 0.204 1.64Support for SWAp –0.450 –2.30 –0.339 –1.98 –0.096 –0.47Africa MAP (HIV) project –0.45 –2.10 –0.303 –1.20 –0.150 0.52Pseudo R 2 0.6433 0.6736 0.6615 0.4551 0.4372 0.4568 0.1443 0.2515Joint tests (p value)Quality at entry, Bank supervision,borrower performance 0.0010 0.0008 0.0007 0.0000 0.0000 0.0000Regions 0.0008 0.0069 0.0028 0.0157Loan size 0.6180 0.6304 0.6103 0.6073 0.6408 0.5800 0.9746 0.7956IDA, multisectoral, interaction 0.0032 0.0006 0.1505 0.2701 0.4436 0.0017 0.0333IDA, multisectoral, Africa MAP 0.0028 0.3892a. Figures in bold are significant at p < .01; figures underlined are significant at p < .05; figures in italics are significant at p < .10.b. The comparison groups are: Africa Region; projects of less than $10 million; approved in fiscal years 1997–99.152

APPENDIX I: EVALUATION OF WORLD BANK PARTICIPATIONIN TWO GLOBAL HNP PARTNERSHIPSThe Medicines for Malaria Venture (MMV)funds and manages the discovery, development,and registration of new medicines for the treatmentof malaria in disease-endemic countries inresponse to the increasing incidence of and mortalityfrom malaria, the declining efficiency offirst- and second-line treatments, and the limitedresponse of the pharmaceutical industry to discoverand develop new antimalarial drugs. TheMMV has been highly successful at achieving itsinitial objectives—to establish and manage aportfolio of antimalarial drug candidates—andthe public sector target price of a full courseof treatment of a dollar or less appears withinreach.The MMV’s mandate has been expanded to includeimproving access and delivery of antimalarialdrugs. It is less clear whether MMV has the organizationalarrangement and institutional relationships(notably with countries) to deliver onthe highly demanding downstream access anddelivery activities and whether it will be able to reconcileits private sector entrepreneurial style withthe public sector requirements for resolution ofpolicy and institutional issues in access and delivery(IEG 2007e).The Population and Reproductive HealthCapacity Building Program (PRHCBP), establishedin 1999, is a merger of three programs:Population and Reproductive Health, Safe Motherhood,and the Program to Reduce the Practice ofFemale Genital Mutilation and Improve AdolescentHealth. Its objective is to build the capacity ofcivil society organizations to develop and implementculturally appropriate interventions in populationand reproductive health. It does this byproviding grants to international intermediaries,which then make grants to grassroots groups, andsupporting operations research and technologyand information transfer. It is financed entirelyby the Bank’s Development Grant Facility, managedwithin the Bank structure, and with no steeringcommittee or other structure that wouldenable actors external to the Bank to participatein decision making and oversight. To date, theBank has allocated $18.3 million of DevelopmentGrant Facility funding to the PRHCBP. The evaluationfound that the objectives of the programwere highly relevant, but that both efficacy and efficiencywere difficult to assess because there wasno systematic measurement of the achievementof the program’s stated objectives. Grant decisionmaking by the Review Committee was a veryinformal process. It was only in 2006 that publicsolicitation of proposals became practice, and in2007 that criteria for evaluating proposals wereestablished. For almost a quarter of all grantsawarded through fiscal 2007, there was no writtenrecord of review decisions. Under the operationsresearch component, PRHCBP has supportedthe International Partnership for Microbicides todevelop new technologies to prevent HIV/AIDSand unwanted pregnancy. The consolidation of thethree programs did not lead to clear objectives andthe links to country-level Bank operations wereweak (IEG 2008e).153

This mobile health education van covers rural areas in Sri Lanka. Photo by Dominic Sansoni, courtesy of the World Bank Photo Library.

APPENDIX J: MANAGEMENT RESPONSEManagement highly values IEG’s evaluation ofWorld Bank Group support in the health, nutritionand population sector, an important assessmentafter 10 years of implementation of the1997 health, nutrition and population (HNP) strategy.1 The evaluation is helpful in articulating someof the contextual difficulties the Bank Group facesas a key partner in the international health environment.Management has some general commentson the changing context for its support,followed by comments on the evaluation’s mainfindings and recommendations. Lastly, the Responsecites the International Finance Corporation’s(IFC) evolving role in the sector. TheManagement Action Record (attached to the ManagementResponse Summary at the front of thisvolume) provides a Bank Group response to IEG’srecommendations. As noted below, the 2007Health Strategy (World Bank 2007a) found issuessimilar to those raised by IEG, not surprising becausethere was regular interaction between HNPstaff and the IEG team in the process of preparingthe new strategy. The recent health StrategyProgress Report (World Bank 2009) lays out theBank’s actions to strengthen its support, includingtaking into account all of the major IEG analysisand recommendations. Annex 2 of the ProgressReport, the Management Action Plan, summarizeshow the actions being taken correspond toIEG’s recommendations, including assigning responsibilitiesand setting benchmarks for measuringprogress. For reference it is appended tothis response.Changing Context for WBG SupportThe context in which the Bank Group providessupport in this sector has changed dramaticallyduring the last decade. That context affects howthe World Bank Group works.A Changing Global Environment RequiresMore from the Bank. The last 10 years haveseen the creation of new international health institutionsand foundations, as well as the emergenceof many different innovative health fundingmechanisms. 2 The welcome result has beensharply increased global funding for HNP. TheBank’s role has undergone a similar paradigmshift. Although lending investment remains significant,we now play a much more nuanced role,working with and through partners and new internationalinstitutions. 3 This changing role, anchoredin the Paris and Accra Declarations, isdesirable from a development perspective, but italso imposes significant additional challenges inattributing development outcomes to Bank financeor technical contributions. The sector isconstantly adapting to this rapidly evolving environment,using the full spectrum of the Bank’scomparative advantages beyond traditional lendingand economic and sector work (ESW). Thebenefits are clear, as are the risks of working in acomplex development area, investing in complexhealth systems while respecting the multiplicityof synergies necessary for a continuum of care andservice delivery, essential for public health and diseasecontrol, and in some instances relying onpartners to deliver results.A Mandate to Achieve Health Outcomesthrough Health in All Policies. The internationalhealth community and the World BankGroup have recognized that, in addition to standaloneinterventions and sector-specific policies, wemust focus on more comprehensive “health in allpolicies.” Given how the Bank Group is organized,this is our comparative advantage—and the HNPsector has been moving in this direction over thepast years. The 2007 HNP strategy (World Bank155

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION2007a) included a critical self-assessment of thesector’s performance since 1997. The IEG evaluationreconfirms much of this self-assessment,and many of IEG’s recommendations are helpfulin achieving greater effectiveness and impact.Main Findings and RecommendationsManagement agrees with many of the findings, andappreciates the recommendations. Managementhas taken them into account in the ProgressReport and the Plan of Action (see the attachedAction Plan). While not detracting from the importancethat it gives to the evaluation and its usefulnessfor the Strategy Progress Report, managementhas a set of observations on some of thefindings.Health Status Has More than One DevelopmentDimension. Management considersbetter health a development outcome in and ofitself, irrespective of its contributions to othergoals. In addition to improving health status, theHNP sector aims to cushion the financial shocksof health costs, which can be substantial in manyclient countries, where out-of-pocket paymentsdominate. Thus, while we agree that investmentsshould focus on health outcomes for the poor, itis also vital to protect those above the poverty linefrom financial shocks from poor health that pushfamilies into poverty.The World Bank Group’s Global HNP EngagementsGo Beyond Projects and ESW.The evaluation emphasizes lending, with someattention to policy dialogue and analytic work.But the HNP sector has expanded the paradigmover the past decade. The World Bank Group usesa range of engagement instruments, such as analyticand advisory activities (including IFC AdvisoryServices), 4 just-in-time policy advice, policy-basedlending led by other sectors, subnational lendingwithout a sovereign guarantee (through IFC), andworking through international networks and partnerships,such as the Global Alliance for VaccinesInitiative (GAVI), the Global Fund to Fight AIDS,Tuberculosis and Malaria (GFATM), and the EuropeanUnion (EU) Observatory. Working with partnersthrough pooled funding, country systemsand joint strategies and supervision (as opposedto ring-fenced Bank operations) is also anchoredin international commitments and agreementssuch as the Paris and Accra Declarations. The successof the Joint United Nations Program onHIV/AIDS (UNAIDS), GFATM, GAVI, Roll BackMalaria, EU Observatory and other major internationalpartnerships is also the shared success ofthe Bank Group’s HNP work, as we exercise substantialtechnical and financial influence in thesenetworks and partnerships. Over the past decade,we have also enhanced internal collaborations toimprove the Bank Group’s impact on global health:Examples include work across units (notably withConcessional Finance and Global Partnerships,Operations Policy and Country Services, and Treasuryfor innovative financing mechanisms—theInternational Finance Facility for Immunization, AdvanceMarket Commitment, and Treasury services)and partnerships with GFATM, GAVI, UNAIDS,and others. Harnessing the Bank’s broader developmentexpertise has significantly impactedthe structure of the global health architecture, inaddition to saving millions of lives.The Evaluation Framework Does Not Assessthe World Bank Group’s Non-TraditionalContributions to Global Health and ClientSupport. The evaluation’s review of four approachesmisses much of the work carried out insupport of global public health, including keyanalytical pieces, the global work on core publichealth functions, water and sanitation, surveillanceand vital statistics, indoor air pollution,avian and human influenza, and so forth. Theevaluation does not evaluate performance-basedapproaches, which we believe have delivered impressiveresults. Regarding SWAPs, we note thatbefore reaching any conclusions on their effectivenesswe need to take into account elementsbeyond the evaluation that reflect the Bank’s supportfor greater donor coordination and the useof country systems as enshrined in the Paris andAccra Declarations.Coverage of Analytic Work. Much of the nontraditionalanalytical work has a potentially largeor larger impact than the Bank’s lending, especiallysince Bank financing is usually a small share ofoverall health spending in middle-income coun-156

APPENDIX J: MANAGEMENT RESPONSEtries. 5 Knowledge-product tasks are often preferredto ESW to get findings out more quickly orprovide just-in-time advice to clients and partners.The evaluation does not sufficiently accountfor the role of analytic and advisory activities inengaging clients and advancing policy dialogue,either as a stand-alone tool to support the client(in particular in middle-income countries) or asa parallel track dialogue to investment and policybasedlending. The evaluation could also havetaken greater account of other knowledge products,such as a large portfolio of Japan Policy andHuman Resources Development Fund (PHRD)grant-financed analytic and advisory activities insupport of project preparation, as well as nonformalESW and technical assistance.Improving Poverty Targeting. Managementagrees with the need to ensure that project designresponds to the priorities and needs of the poor,and to measure the full impact of improved healthservices for the poor. Indeed, the 2007 HNP strategyexplicitly recognizes the need to focus notonly on levels of HNP outcomes but also on theirdistribution, especially among the poor. This focusin the strategy drew heavily on HNP’s pathbreakingReaching the Poor Program, active since2001. Reaching the Poor has delivered global leadershipin the measurement of disparities in HNPhealth-service coverage and outcome indicatorsamong the poor versus the non-poor, as well as ofthe financial burden on households from seekingcare. In 2005, Reaching the Poor published a reviewof interventions and programs that had beensuccessful in reaching the poor (Gwatkin, Wagstaff,and Yazbeck 2005). A new report (Yazbeck 2009)has been produced in the period since the adoptionof the 2007 HNP strategy and was launchedin January 2009. This volume lays out a policymenu emphasizing pro-poor policy reform alongsix dimensions, and a list of the analytical tools tobetter understand the constraints to pro-poor targetingof public health investments. 6 Managementwill use these findings to ensure a better pro-poorfocus in future lending operations.Disease Control Programs and Targeting. Asopposed to specific income groups, disease controlprograms must focus on the prevailing epidemiology.For example, an AIDS program mustfocus on high-risk groups—irrespective of income.A malaria program focused solely on the poorwould fail to eliminate malaria. Polio could only beeradicated from the Western Hemisphere by focusingon large, inclusive campaigns targeting allincome groups. Such investments in public healthand control of communicable diseases are globalpublic goods, generating positive externalities forsociety, irrespective of income status.Investments in Health Systems. The 2007 HNPstrategy underscores the need to focus on healthsystems for delivering improved HNP results, particularlyfor the poorest and the most vulnerable,and the Bank has emerged as a strategic leader inadvancing health systems strengthening for improvedHNP results. Over the past two years, projectswith a primary focus on health systems haveincreased twofold. In line with the strategy, 67percent of Bank programs approved since fiscalyear 2007 that focus on priority disease areas alsoinclude strong components on health systemsstrengthening. A new Health Systems for theHealth MDGs initiative was launched in 2008 tobring together the resources and efforts to fundand implement coherent, country-led health sectorprograms in Africa and selected countries inAsia. 7 This program will improve the Bank’s abilityto rapidly assist and advise HNP operations onthe ground, particularly in the areas of health financeand risk pooling mechanisms, human resourcesfor health, governance, supply chainmanagement, as well as infrastructure planning.HNP in Poverty Assessments and CASs. Managementnotes the substantive improvementsover the past years in quantity and quality of HNPinvolvement in the Poverty Reduction and EconomicManagement Network–led analytical work,and agrees that HNP must be fully included in allPoverty Assessments and fully examined in thepreparation of CASs.Cross-Sector Work. Management welcomes thesuggestion to expand cooperation and crosssectorwork with the Transport, Water and Sanitationsectors. More will be done to harness crosssectoralresults along the notion of “health in all157

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONpolicies.” The 2007 HNP strategy is a good conduitto foster improved collaboration across sectorsin support of health outcomes. We note therecommendation that “when the benefits are potentiallygreat in relation to the marginal costs,”health objectives should be incorporated intonon-health projects, but also note that clients undertakesuch operations largely for non-healthobjectives and it would be unrealistic within currentresources to burden all such projects with potentiallysmall (albeit cost effective) health impactobjectives, or to demonstrate and document empiricallysuch impact and outcomes in each case.The Quality of the Bank HNP Portfolio. Managementappreciates the concern for the qualityof the Bank HNP portfolio—in particular in Africa.The current data on riskiness of the HNP portfolioconfirms the fact that the problems aremost acute in the Africa Region, and that HNP’sperformance across most other Regions is approachingother sectors’ performance. An additionalarea of concern is the underperformanceof projects that have a significant HIV/AIDS component,a high percentage of which have receivedunsatisfactory ratings from IEG 8,9 (with a relativelyhigh disconnect—as management ratedthe outcomes of several of these operations as satisfactory)and which also constitute a disproportionatenumber of projects at risk in the currentportfolio.The HNP Action Plan for the Portfolio. As outlinedin the Strategy Progress Report, the HNP sectoris making major efforts to improve the qualityof the existing portfolio and ensure the quality ofnew operations entering the portfolio. These effortsinclude increasing candor in reporting toclose the realism gap and improving other portfolioindicators. The Africa Region has implementedseveral changes aimed at improving HNP’sportfolio in the Region. In March 2008, the QualityAssurance Group (QAG) was asked to reviewthe performance of HNP projects that were categorizedas being at risk. The QAG panel indicatedimportant areas for urgent attention: strengtheningsector management oversight; reviewingcurrent resources for preparation/supervision;and addressing two key weaknesses: monitoringand evaluation (M&E) and institutional analysis.A detailed Portfolio Improvement Action Planthat includes all at-risk projects, as well as projectsneeding additional management oversight toavoid falling into at-risk status, has been developedand is being monitored on a quarterly basis by theHNP Sector Board. 10 The Portfolio ImprovementAction Plan addresses some of the key concernsraised by the various reviews, namely, intensifyingmanagement oversight, targeting of resourcesto projects most in need, strengthening projectimplementation to focus on monitoring and evaluation,and matching project interventions to thecountry’s institutional capacity. A comprehensiveapproach has also been adopted to improve thequality of HIV/AIDS projects, and this includedan umbrella restructuring package of 11 Multi-Country HIV/AIDS Programs (MAP) for Africaprojects in fiscal year 2007. Additional technicalsupport is being provided to improve implementation,develop impact evaluation capacity,and strengthen governance and accountabilitywithin national AIDS programs.Doing More in Population and ReproductiveHealth. The Strategy Progress Report highlightsplans for strengthening support for populationand reproductive health in a health systemsapproach. That strengthening is critical to improvingmaternal and child survival rates.Investing in Nutrition Support. We also agreewith the evaluation’s findings that nutrition hasplayed a less prominent role within HNP overthe past 10 years. The need for action is evenmore important today in the context of the aftermathof the crises in fuel, food, and fertilizers,as well as the escalating effects of the financial crisis.Management is therefore investing significantresources in the next few years to ramp up theBank’s analytical and investment work and leverageresources from other donors. The agendafor scaling-up nutrition is being catalyzed withadditional budget resources starting in 2009, andcontinuing for three years. The increased allocationsare being utilized principally in Africa andSouth Asia, two Regions where the malnutritionburden is highest. These funds will be complementedby additional trust fund resources from158

APPENDIX J: MANAGEMENT RESPONSEJapan, and possibly from other donors that are currentlyengaged in discussions on this issue.Needed Improvement in Monitoring andEvaluation. Management agrees that M&E needsto be improved, and it is an important part ofStrategy implementation, as noted in the StrategyProgress Report (including the work on retrofittingprojects and improving the design of newprojects). The evaluation proposes a focus onstrengthening health information systems. However,in some cases, routine health monitoringsystems (including surveillance, facility reporting,vital registration, census data, resource tracking,and household surveys) first need to be strengthenedto provide the data and indicators that areneeded. The Bank has deepened its collaborationwith the Health Metrics Network, a global partnershipaimed at building statistical capacity incountries to improve the collection and use ofhealth information. The Bank has produced atoolkit on measuring health system governance tobetter monitor accountability in the sector, and thisis being piloted in five countries with support ofthe Governance and Anti-Corruption Trust Fund.One of the largest impact evaluation trust funds atthe Bank, the Spanish Trust Fund for Impact Evaluation,is housed in the human development sector.The trust fund, which was initiated in 2007 andwill continue until 2010, finances rigorous impactevaluations of interventions aimed at enhancinghuman development as well as learning and disseminationactivities to help promote knowledgeand awareness of “what works” in the human developmentSector. The Bank is also working withpartners, such as WHO, to develop better ways tomonitor the health MDGs, including the estimationof trends in child and maternal mortality forwhich updates have recently been issued.The Problem of Attribution. The evaluation arguesfor a better attribution of health outcomesto inputs financed by the World Bank Group.Health outcomes are hard to evaluate within thetimeframe of a project and are influenced by multipledeterminants. Also, the Paris and Accra Declarationsemphasize the greater developmentimpact of pooled funding, use of country systems,and country-based M&E. The Bank’s policyon M&E (OP 13.60) likewise emphasizes the useof country-level M&E systems. Strictly speaking,attribution is only possible in a tightly designedrandomized trial, which will rarely be feasible inBank-supported investment projects or sectorsupport. The standard should be that sufficient evidenceon outputs, intermediate outcomes, andoutcomes should be collected to establish a credibleresults chain regarding the link betweenBank-financed investments and sector progress.Cost of M&E. Borrowing countries have many developmentand poverty-reduction priorities andworry about the opportunity cost of complexM&E systems, especially those that are separatefrom country systems. They see that large-scaleevaluations have an important global public goodaspect, justifying external (grant) financing inmost cases. This may delay the establishment ofappropriate baseline data and results frameworksprior to project approval, notably with regard toimpact evaluations.Operational Complexity. Management agreesthat we need to strive to reduce complexity inBank-financed HNP operations. However, HNPoperations can rarely be institutionally or technicallysimple, since the desired outcome usually dependson a complex and interacting set of social,cultural, and institutional factors. Investing in simpleprograms would not necessarily provide for lastingimpact. Management acknowledges, however,that complexity can be at least partially addressedby some of the recommendations of the evaluation,such as thorough technical preparation, includingsolid analytical underpinning, political mapping,high quality at entry including a good results framework,and finally in-depth supervision and parallelpolicy dialogue with client and partners.The 2007 Health Sector Strategy and IEGRecommendations. The 2007 strategy actuallycited findings that were similar to the IEG findings,as would be expected since IEG staff worked withthe strategy team to share preliminary results oftheir work. Annex 2 of the Progress Report summarizeshow the overall actions being taken tostrengthen the Bank’s HNP support correspondto IEG’s recommendations (see attached).159

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONIFC’s HNP FootprintIFC has considerably increased its footprint inHNP over the past decade and is prepared to intensifycollaboration within the World Bank Group.During the period under review, much has occurred,both in IFC’s health activities and in theprivate health sector in developing countries.There is a growing acknowledgement of the roleof the private sector in health care in developingcountries. Indeed, in many low-income countries,the private sector pays for a far larger partof health care than the public one, and in manymore, it is at least of equal size. Furthermore,joint research by the World Bank, IFC, and the U.S.Agency for International Development (USAID)conducted in Africa showed that very poor peopleoften obtain health care in the private sectorand that the public sector often subsidizes healthcare for the rich. These findings have led to theBank and its partners designing innovativeconsumer-focused approaches to addressingmajor health financing challenges; such as theAMFm (Affordable Medicines Facility for Malaria),which will reduce the price of Artemisinin-basedCombination Therapy (ACT), and therefore outof-pocketpayments.In a world of growing, aging populations in developingcountries that are likely to “get old beforethey get rich,” both public and private sectorresources are needed to tackle the health needsof the population, as neither has the resources byitself. Indeed, all countries’ health systems are financedby both sectors; it is only the proportionsthat vary.IFC therefore has taken up the challenge to growits work to support the private health sector indeveloping countries to complement the larger,more established work done with the public sectorby the World Bank. The period under reviewhas seen a marked increase in IFC’s activity inhealth, the creation of a dedicated Health andEducation Department, and a specific focus onpharmaceutical and life sciences activities withinIFC’s Global Manufacturing and Services Department.During this time, many lessons havebeen learned, specialist knowledge has deepened,and performance has improved by anymeasure applied. As in other sectors, IFC continuesto strive for greater development impact,and we therefore welcome all input that couldhelp us to do better.160

APPENDIX J: MANAGEMENT RESPONSEWorld Bank Management Action PlanIEG Recommendation Actions to Be Taken How Much & by When By WhomI. Intensify efforts to improve the performance of the World Bank’s health, nutrition, and population support.(a) Match project design to countrycontext and capacity and reducethe complexity of projects in lowcapacitysettings through greaterselectivity, prioritization, and sequencingof activities, particularlyin Sub-Saharan Africa.– Carry out Quality EnhancementReviews focusing on technicalpreparation, M&E, and institutionaland risk assessmentsand mitigation measures.– Reviews of HNP portfolio.– 75% of all new HNP projects havean intensive Quality EnhancementReview focusing on technical preparation,M&E, and institutional andrisk assessments and mitigationmeasures, starting fiscal year 2010.– Quarterly reviews of HNP portfolioby HNP Sector Board ongoing.HNP Sector Board, HNPHub, and Regional qualityteams.(b) Thoroughly and carefully assessthe risks of proposed HNP supportand strategies to mitigate them,particularly the political risks andthe interests of different stakeholders,and how they will beaddressed.– Concentrate on risk managementand mitigation duringQuality Enhancement Reviews.– Expand learning on HNP sectorrisk assessments and mitigationstrategies.– Of the Quality Enhancement Reviewsconducted, 100% includefocus on risk, starting fiscal year2010.– Develop and roll out course on projectrisk analysis for HNP teams, anddisseminate best practices and lessonslearned, starting fiscal 2010.HNP Sector Board, HNPHub, World Bank Institute.(c) Phase reforms to maximize theprobability of success.– Increase analytical work focusingon reform for those HNP projectsfocusing on health systemreform.– 100% of new HNP projects focusingon health system strengthening orbroadly on health reform to bebased on analytical work, includingpolitical risks and the interests ofdifferent stakeholders, startingfiscal 2010.HNP Sector Board, withtechnical support fromHNP Hub, as needed.(d) Undertake thorough institutionalanalysis, including an assessmentof alternatives, as an input intomore realistic project design.– Increase analytic and advisoryactivities for institutionalanalysis, with increasedattention through better policydialogue and analytical work,adapted to country context.– Learning program focusing onHNP sector institutional andstakeholder analysis.– 80% of new HNP projects to bebased on institutional analysis,starting fiscal 2010.– Training program designed anddelivered, starting fiscal 2010.HNP Sector Board,HNP Hub.(e) Support intensified supervision inthe field by the Bank and the borrowerto ensure that civil works,equipment, and other outputshave been delivered as specified,are functioning, and are beingmaintained.– Project design to specifyborrower responsibilities forcivil works and equipmentmaintenance.– Project design to ensure adequaterecurrent cost budgetingfor civil works and equipmentmaintenance.– All new HNP projects startingpreparation in fiscal 2010.– All new HNP projects startingpreparation in fiscal 2010.HNP Sector Board,HNP Hub.(Table continues on next page)161

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONWorld Bank Management Action Plan (continued)IEG Recommendation Actions to Be Taken How Much & by When By Whom– Develop guidelines and standard– By fiscal 2010.specifications for civilworks and equipment andother health inputs, and theirmaintenance.II. Renew the commitment to health, nutrition, and population outcomes among the poor.(a) Boost population and family planningsupport in the form of analyticwork, policy dialogue, and financingto high-fertility countries andcountries with pockets of highfertility.– Analytic and advisory activitypolicy note on reproductivehealth, including familyplanning.– Incorporate family planninginto health-system strengtheningprojects.– In high-fertility countries, incorporatepopulation and familyplanning issue into CAS.– By fiscal 2010.– 2 health system strengthening projectsin high-fertility countries includestrengthening of familyplanning delivery, by fiscal 2010.– 50% of CASs for high-fertility countries,starting fiscal 2010.HNP Sector Board, HNPHub.(b) Incorporate the poverty dimensioninto project objectives to increaseaccountability for health, nutrition,and population outcomes amongthe poor.(c) Increase support to reducemalnutrition among the poor,whether originating in the HNPsector or other sectors.– Ensure adequate attention isgiven to poverty dimensions inproject design and supervision,particularly project developmentobjectives and key performanceindicators.– Scale-up the Bank’s analyticaland investment work andleverage resources from otherdonors.– 80% of all new HNP projects incorporatethe poverty dimension, whereappropriate, starting fiscal 2010.– President’s Regional ReprioritizationFund to hire 6 additional Bank staff(US$4 million committed for fiscalyears 2009–11); Japan TF (US$2 millionwith potential for additionalUS$20 million); possibly funds fromother donors that are currently engagedin discussions on this issue.– Global Action Plan designed andagreed with key partners, by fiscal2010.– Six to 8 analytic and advisoryactivitiess or new investment innutrition by fiscal 2011.HNP Sector Board, withtechnical support fromHNP Hub, as needed.HNP Sector Board,HNP Hub.(d) Monitoring health, nutrition, andpopulation outcomes among thepoor, however defined.– Track health outcomes andintervention coverage amongthe poor.– Publish report on health indicatorsof poor people.– Annual review of health indicatorsamong the poor, starting fiscal 2010.– Annual report, starting fiscal 2010.HNP Hub162

APPENDIX J: MANAGEMENT RESPONSEWorld Bank Management Action Plan (continued)IEG Recommendation Actions to Be Taken How Much & by When By Whom(e) Bring the health and nutrition ofthe poor and the links betweenhigh fertility, poor health, andpoverty back into poverty assessmentsin countries where this hasbeen neglected.– Increase inclusion of HNP inpoverty assessments.– 90% of all poverty assessments andat least 40% of all CASs shouldassess the health status of the poor,starting fiscal 2010.III. Strengthen the World Bank Group’s ability to help countries to improve the efficiency of their health systems.HNP Sector Board, withtechnical support fromHNP Hub and PREM, asneeded.(a) Better define the efficiencyobjectives of its support and howefficiency improvements will beimproved and monitored.– Expanded PAD definition anddiscussion of efficiency objectives,measures and monitoringframework to be expandedand more explicit.– Analytic and advisory activitiesto analyze and review experiencein improving healthsystem efficiency.– 70% of HNP projects to includedefinition and analysis of improvingHNP sector efficiency, includingdiscussion of efficiency-equitytrade-off, starting fiscal 2010.– Start fiscal 2010.HNP Sector Board,HNP Hub.(b) Carefully assess decisions tofinance additional earmarkedcommunicable disease activitiesin countries where other donorsare contributing large amounts ofearmarked disease funding andadditional funds could result indistortion in allocations and inefficienciesin the rest of the healthsystem.– Closely coordinate proposalsfor Bank support for newdisease-specific programswith other partners.– 100% of HNP projects with significantpriority-disease components tomap contributions from other donorsand ensure strengthening of healthsystems, starting fiscal 2010.HNP Sector Board, withtechnical support fromHNP Hub, as needed.(c) Support improved health informationsystems and more frequentand vigorous evaluation of specificreforms or program innovations toprovide timely information for improvingefficiency and efficacy.– Build statistical capacity forclient countries on priority HNPoutcome indicators directlythrough Bank operationsand/or supporting globalpartner’s country support(e.g., MDGs).– Continue support of the InternationalHealth Partnership’s(IHP+) efforts to strengthenmonitoring and evaluation andhealth information systems incountries.– Conduct country assessmentsof health information systems.– 50% of new HNP projects includestrengthening of country M&Esystems, by fiscal 2010.– Strategy for global monitoringarrangement designed (in collaborationwith global partners), by fiscal2009.– Country assessments in 10 countriesin fiscal 2010.HNP Sector Board,HNP Hub.(Table continues on next page)163

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONWorld Bank Management Action Plan (continued)IEG Recommendation Actions to Be Taken How Much & by When By WhomIV. Enhance the contribution of support from other sectors to health, nutrition, and population outcomes.(a) When the benefits are potentiallygreat in relation to the marginalcosts, incorporate health objectivesinto non-health projects, forwhich they are accountable.– Provide incentives to non-HNPtask team leaders to incorporatehealth objectives into nonhealthprojects.– Intersectoral coordination thematicgroup for HNP results established toidentify constraints and incentives,by fiscal 2010.HNP Sector Board, HNPHub, Country Directors.(b) Improve the complementarity of investmentoperations in health andother sectors to achieve health, nutrition,and population outcomes,particularly between health andwater supply and sanitation.– Develop, implement, and managean intersectoral coordinationthematic group for HNPresults.– Group Functioning by fiscal 2010.HNP Sector Board, HNPHub.(c) Prioritize sectoral participation inmultisectoral HNP projects accordingto the comparative advantagesand institutional mandates, toreduce complexity.– Invite other sectors’participation to HNP projectdesign reviews (e.g., QualityEnhancement Reviews) whereappropriate.– 100% of all HNP projects, startingfiscal 2010.HNP Sector Board.(d Identify new incentives for Bankstaff to work cross-sectorally forimproving HNP outcomes.– See response to IV (a) above.– See response to IV (a) above.– See response to IV (a)above.(e) Develop mechanisms to ensurethat the implementation and resultsfor small health componentsretrofitted into projects are properlydocumented and evaluated.– Strengthen HNP portfoliomonitoring, including non-HNPprojects, to document healthresults achieved through non-HNP sectors.– 50% of all HNP and non-HNP SectorBoard operations tracked, startingfiscal 2010.HNP Sector Board,HNP Hub.V. Implement the results agenda and improve governance by boosting investment in and incentives for evaluation.(a) Create new incentives for monitoringand evaluation for both theBank and the borrower linked tothe project approval process andthe mid-term review. This wouldinclude requirements for baselinedata, explicit evaluation designsfor pilot activities in Project AppraisalDocuments, and periodicevaluation of main project activitiesas a management tool.– Implement US$2.8 millionSpanish Trust Fund (SIEF), supportingimpact evaluations.– Pilot and evaluate impact ofoutput- and performancebasedfinancing for HNPrelatedprojects/programs.– Introduce Results Frameworkstargeting HNP outcomes, output,and system performance,including baseline data andoutput targets and programs.– 15 HNP projects, fiscal 2011.– 16 active projects with most loanproceeds allocated to output-basedfinancing, fiscal 2010.– At least 70% of new projects/programs approved by the Board,starting fiscal 2009.HNP Sector Board,HNP Hub.164

ENDNOTESChapter 11. Michaud 2003. The share of IDA averaged 11.1 percentof the total for 1997–99.2. IEG calculation based on an annual average of $1billion of World Bank health assistance, of a total calculatedby C. Michaud of $16.665 billion in 2006 (WorldBank 2008a).3. http://www.unicef.org/publications/ index_23557.html4. World Bank 2008a. About 536,000 women dieworldwide from complications of pregnancy and childbirth;the mortality rate in Africa is about 900 maternaldeaths per 100,000 live births.5. The World Bank’s commitments include loansfrom the IBRD and the credits and grants from IDA.6. This excludes 17 IFC health projects fundedthrough the Africa Enterprise Fund (AEF) and Small EnterpriseFund (SEF), totaling $14 million, which are notevaluated here.7. In many countries, access to basic health care isconsidered a basic human right.8. World Bank 1993b, p. 5. An example of a publicgood is information—one person’s consumption doesnot reduce the information available to others. Externalitiesoccur when consumption of a good or serviceby one individual affects others—for example, childrenimmunized against a disease are not only lesslikely to get the disease themselves, but so are othersthey come into contact with (a positive externality). Atthe same time, industries and automobiles generate airpollution, which conveys a negative health externalityfor the surrounding population. Insurance companiesseek to avoid adverse selection—in which peoplewho are the most likely to be sick enroll—by targetingpeople with the lowest health risks, making it difficultfor people with pre-existing conditions to obtain insurance.Moral hazard occurs when people who areinsured have less of an incentive to avoid health risksor demand more expensive or unnecessary care becausethey are insured.9. Before 2006, IEG-World Bank was known as theOperations Evaluation Department (OED).10. A forthcoming IEG evaluation will evaluate supportfor HNP outcomes through multisectoral developmentpolicy lending in the form of Poverty ReductionSupport Credits (PRSC). That support is not assessedhere.11. Improving the selectivity and strategic engagementof the Bank in global programs is one of the fivestrategic directions in the 2007 HNP strategy.12. This was also maintained in the preceding strategy,from 1997.Chapter 21. The trends in formal economic and sector work(ESW) are described in appendix G, as well as the resultsof an inventory of HNP analytic work undertakenby IEG.2. Investment operations provide funding in theform of IBRD loans or IDA credits and grants to governmentsto cover specific expenditures related toeconomic and social development projects. DevelopmentPolicy Loans provide untied, direct budget supportto governments for policy and institutional reformsaimed at achieving a set of specific development results.They may be freestanding operations or, more often,part of a programmatic series of operations. Of the220 HNP-managed projects approved over fiscal years1997–2006, only 6 were Development Policy Loans.3. The number of IDA projects increased by 46 percent,while the number of IBRD projects declined by20 percent.4. The results, findings, and lessons from lending forcommunicable diseases and SWAps are discussed inchapter 3; results for multisectoral HNP projects aretaken up in chapter 4.5. The absolute number of projects with these typesof formal objectives declined from 44 in fiscal 1997–2001to 29 in 2002–06. There may be other projects in theportfolio that include financing, health insurance, andactivities with the private sector for which these are notexpressed as objectives. They are not included in thetable.6. Staff at level GF+ or higher mapped to the HNPsector (Nankhuni and Modi 2008).165

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION7. As explained in appendix B, the analysis of staffspecialties is based on job titles only; there was noknown reliable data source that would have allowedtracking by actual staff expertise over the period. Theextent to which the job titles accurately reflect staff expertiseis unclear.8. IEG evaluates projects by assessing their resultsin relation to their stated objectives. The outcome ratingis based on: (a) the relevance of the project’s objectivesand design in relation to country needs andinstitutional priorities; (b) efficacy, which is the extentto which the project’s objectives have been (or are expectedto be) achieved; and (c) efficiency, which is theextent to which the objectives have been (or are expectedto be) achieved without using more resourcesthan necessary. Outcome is rated on a six-point scale:highly satisfactory, satisfactory, moderately satisfactory,moderately unsatisfactory, unsatisfactory, and highlyunsatisfactory. For ease of comparison, this report usesthe percentage of ratings that are moderately satisfactoryor higher in making comparisons across projectgroupings.9. The share of project exits in other sectors thatwere Development Policy Loans remained a constant15–16 percent between the two periods.10. The share of Africa Region projects with satisfactoryoutcomes rose from 60 to 67 percent amongprojects exiting in fiscal 1998–2001 and fiscal 2003–07,compared with an increase from 73 to 78 percent Bankwide.Outcome ratings for African HNP projects are thusconsiderably lower than the Regional average across allsectors.11. However, it is important to note that there wasa large increase in HNP project approvals in the AfricaRegion in fiscal 2002–06 and they will constitute amuch larger share of exits in the HNP portfolio over thenext five years.12. While Africa Region projects performed lesswell on average in other sectors than did the other Regionscombined (64 percent satisfactory in Africa, 79 percentin other Regions over the 10-year period), theoutcome ratings of Africa Region projects increased ata faster rate (from 58 to 72 percent satisfactory over thetwo periods). When Africa Region projects are excluded,the ratings for other Regions and sectors combinedare higher, but the amount of increase is lower(rising only from 78 to 81 percent satisfactory, as opposedto 72 to 79 percent satisfactory if Africa isincluded).13. Ratings on institutional development impact ofHNP-managed projects (a rating that was discontinuedin fiscal 2006) are as low as those for other sectors.14. Of the 220 projects approved from fiscal 1997to 2006 that were subjected to in-depth review, 110 hadclosed and 99 had been rated by IEG as of September30, 2008. Fifty-eight percent of the HNP projectsapproved from 1997 to 2006 that had closed had satisfactoryoutcomes, while about two-thirds had satisfactoryBank and borrower performance. The outcomeratings for the sample in the portfolio review (58 percentsatisfactory) are different from those in figure2.4 (68 percent satisfactory) because they apply to a differentgroup of projects. Figure 2.4 shows the outcomesfor all projects that closed in a given fiscal year,irrespective of when they were approved, while the IEGsample is of 220 projects approved from fiscal 1997 to2006, only half of which have closed to date. Most ofthe projects that have closed were approved in fiscal1997–2001, but a few approved in fiscal 2002–06 (forexample, Development Policy Loans) are also included.The performance of most of the projects approvedmore recently is unknown.15. While there are several possible explanations forthe overall increase in supervision costs across all sectors,examination of the supervision cost data by Regionreveals no pattern that would explain why averageHNP supervision costs for three years were so muchgreater than for the rest of the Bank.16. The ratio of the percent of unsatisfactory projectswith this characteristic relative to the percent ofsatisfactory projects with the characteristic.17. Projects in Africa are also largely responsible forlower outcomes for IDA projects. Half of IDA-financedHNP projects had satisfactory outcomes, compared toabout three-quarters of IBRD projects. However, thisrises to 77 percent if IDA projects in Africa are excluded.18. For a fuller definition and discussion of sectorwideapproaches (SWAps), see chapter 3.19. See appendix F for more on the performance ofthese projects.20. World Bank Group 2007. The Detailed ImplementationReview (DIR) was conducted by the Departmentof Institutional Integrity.21. World Bank Group 2007, p. 11. The DIR team visitedall 15 food and drug laboratories and 2 offices thatreceived equipment from the project (p. 67).22. World Bank Group 2007, p. 12. The DIR teamvisited 55 hospitals in 23 districts, representing 35 per-166

ENDNOTEScent of project sites. They were selected based on diversityin size and geographic areas, including remoteareas and sites visited and not visited previously byBank supervision missions (p. 191).23. World Bank Group 2007, p. 22. For example, inthe case of the Orissa project, it noted that it was notuntil after 6 years of implementation and two projectextensions that the Bank began to “extensively” visit andreport on project sites (p. 186). In response, the SouthAsia Region developed an action plan that includedstronger supervision structured around the projectsupply chain and vulnerable points, use of multipleinformation sources, more use of community and thirdpartymechanisms, and dedicated consultants to conductrandom visits.24. The average length of the M&E section of thePAD has also increased between the two years, from 445to 1,272 words.25. This reflects results for HNP projects reviewedthrough October 31, 2008. These 45 projects were approvedroughly between 1998 and 2000. There havebeen substantial additional efforts to improve the monitoringof results since then, partly by incorporatinglogframes or results frameworks that try to measure allof the different links in the results chain. However,these improvements cannot yet be evaluated.26. The HNP impact evaluations addressed the followingtypes of interventions: nutrition (20 percent),HIV/AIDS (17 percent), malaria (13 percent), health insurance(9 percent), other health financing and performance-basedcontracting (9 percent), de-worming(6 percent), PROGRESA conditional cash transfers (5percent), health reform (2 percent), and other (21percent).27. This is probably an understatement of the truenumber of intended evaluations of pilot activities or impactevaluations, as some may not be mentioned in thePAD but rather in the detailed project implementationplans. Impact evaluations that were retrofitted intoprojects after they were approved are also not included.28. In a similar vein, Loevinsohn and Pande (2006)found that a typical HNP project in South Asia averagesabout two innovations, neither of which are typicallyevaluated.29. One additional project—the Bangladesh NationalNutrition Program (fiscal 2000)—had an impactevaluation conducted by an external entity, as reportedin the completion report.30. TB, malaria, schistosomiasis, leprosy, and polio.31. For the purpose of this analysis, projects wereflagged if they had any mention of “the poor” or “poorregions” in their objectives. To the extent that the projectsselected poor regions of the country for implementationand this is not reflected in the objectives, thepoverty focus of the portfolio is understated by the objectives.IEG reviewed the design documents for the 47projects with objectives to improve general health status,to identify projects that targeted poor areas, evenif this was not explicit in the objectives. If one were toinclude projects that had poverty targeting in their designbut not in the objectives, the overall share of projectstargeting health outcomes among the poor woulddouble, from 8 to 16 percent.32. A third or more of projects expressed a nonspecificobjective to improve equity in health servicesor in the health system, without specifying whether theintent was access to or use of services, distribution ofresources, or health status of the population.33. Two of the 14 projects were cancelled before theywere implemented, so it is not possible to assesswhether they achieved their objectives. Both of the cancelledoperations involved extending health insuranceto the poor.34. Six major studies, all in the first five years coveredby the inventory, generated a large number ofcountry reports linked to a single program and accountfor a total of 184 pieces of analytic work. Theywere all related to poverty and equity.Chapter 31. The Zaire National AIDS Control Project and theBrazil Amazon Basin Malaria Control Project were bothapproved in fiscal 1989.2. For freestanding projects, the full project cost isincluded, for projects with components, only the costof the communicable disease component is included(Martin 2009). This is likely an understatement, sincemany HNP projects finance childhood immunization asan activity, even when it does not appear as a formalcomponent. A search of “immunization” and “vaccine”in project appraisal documents identified 13 additionalprojects with either immunization subcomponents ormention of immunization as an activity. However, thereis often insufficient information available in projectcompletion documents to assess the efficacy of subcomponentsin a desk study.3. Seven regional AIDS projects were approved, coveringparts of Africa, the Caribbean, Central America, and167

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONCentral Asia. The two main rationales for regional projects,according to design documents, are: (a) to addressexternalities that individual country programs maynot be able to address and (b) to achieve economies ofscale and where investments by individual countries willlikely result in duplication. The first rationale wouldcover the projects that dealt with truckers, migrants, andother mobile populations that cross international borders,like the Abidjan-Lagos Transport Corridor Project.The second rationale would cover projects like the CentralAsia AIDS Control Project, which aims to strengthenregional laboratory services and disease surveillance.However, several are operating in countries where thereis also a freestanding AIDS project. IEG has yet to conductan in-depth field evaluation of these projects; onlyone has closed to date.4. The Bank approved the first of two umbrellacommitments for the Africa Multicountry AIDS Programin 2001, together totaling $1 billion.5. “G8 Summit Communiqué,” Okinawa, Japan,July 23, 2000.6. The Africa Multicountry HIV/AIDS Program,or MAP.7. GFATM Web site. Commitments are as of November2007.8. The World Bank serves on the governing boardsfor Stop TB and Roll Back Malaria. It also serves on theboard of GAVI, leads the Financing Task Force, and isthe treasury manager for the International Finance Facilityfor Immunization.9. Bangladesh Fourth Population and Health; CambodiaDisease Control and Health Development; EgyptSchistosomiasis Control; Eritrea HIV/AIDS, Malaria,STDs and TB (HAMSET) Control; Eritrea Health; GhanaAIDS Response (GARFUND); Kyrgyz Health Reform I.10. Communicable disease projects here are definedas projects with a communicable disease objective,title, or component.11. Only half of projects with communicable diseasecomponents had satisfactory outcomes, although theratings apply to the overall objectives of the projects,which may include more than communicable diseasecontrol.12. Neither of the two completed Africa MAP projectsthat received satisfactory outcome ratings arecountry-level freestanding AIDS projects. The EritreaHAMSET Project was a multiple disease project (AIDS,malaria, TB, STDs) that was rated moderately satisfactorybased on the strong performance of the malariaactivities, highlighted in box 3.1. The Abidjan-LagosTransport Corridor Project aimed at high-risk groupsand risk areas along border crossings; it was rated satisfactory.There is also a group of Caribbean MAPprojects, but none had been rated by IEG as of September30, 2008.13. Martin 2009. The M&E rating is a rating launchedon an informal and experimental basis by IEG in 2006.The rating measures M&E design, implementation ofM&E, and use of the data.14. Problems were noted in the India AIDS II Projectby the Detailed Implementation Report of healthprojects in India prepared by the Bank’s Integrity Department.Among the African MAP projects, issues concerningcorruption were noted in project completionreports for Kenya and Cameroon. In Sierra Leone, theNational AIDS Commission forwarded irregularities toa local investigative body that then published the namesof illicit nongovernmental organizations and barredthem from participation.15. Malawi (fiscal 2007) and Mozambique (fiscal 2003).16. The 2007 strategy refers to synergy 32 times, inthe context of synergy between disease-specific interventionsand health-systems strengthening, and theneed for synergy underpins 1 of 5 strategic directionsarticulated in the HNP strategy. However, nowhere inthe strategy is synergy defined.17. The countries are Malawi (43.5 percent), Mozambique(33.2 percent), Rwanda (45.0 percent), andUganda (42.0 percent) (Zaky and others 2008).18. Of course, the severity of any distortion is relatedto what is actually being financed.19. IEG defined health-reform projects based onstated objectives and the project title. However, therewere a number of projects with neither characteristicthat had components that clearly indicated reform activities—forexample, the Bangladesh Health and PopulationControl Program, the Panama Health Pilot, theMoldova Health Investment Fund, the Ghana HealthSector Program II, and the Nepal Health Sector Program.Many of these projects adopted overall objectives to improvehealth outcomes or to improve efficiency, withoutmention of reform.20. Overall, 28 percent of the projects approvedfrom fiscal 1997 to 2006 had objectives to improve efficiency,and nearly half of health reform projects as definedabove (48 percent) had an explicit objective toimprove efficiency or cost-effectiveness. Most projectswith efficiency objectives that are not in health reform168

ENDNOTEShave titles that indicate that they are health-systemprojects (15/24 projects with efficiency objectives butnot health reform). The share of projects with efficiencyobjectives has also been in decline—from 37 percentof projects approved in fiscal 1997–2001 to 20percent of projects approved in fiscal 2002–06.21. Even when additional projects known to havehealth reform content but without explicit objectivesor a title are included, the share with health reform declinesfrom 48 to 28 percent over the period.22. In the case of low-income Africa, this declinereflects a large increase in the share of communicabledisease projects. See the earlier discussion of communicabledisease projects.23. China Health VIII; Peru Health Reform; SamoaHealth Sector Management; Argentina ProvincialMaternal and Child Health Sector Adjustment; KyrgyzHealth Reform.24. Moldova Health Investment Fund.25. China Health VIII, Kyrgyz Health Reform I andII, Moldova Health Investment Fund; Samoa HealthSector Management; Mali Health Sector Development;Argentina Provincial Maternal and Child Health SectorAdjustment.26. Kyrgyz Health Reform II (Good GovernanceStructural Adjustment Credit), Peru Health Reform II(Programmatic Social Reform Loans I-IV).27. Lithuania Health; Venezuela Caracas MetropolitanHealth; Kenya DARE; Chad Health Sector Support.28. There were probably also efficiency gains inPeru since there was increased utilization of public facilitiesnationwide and no significant expansion inhuman resources. However, these efficiencies are notwell documented.29. Although initially the funding for insurance wasto be financed from the World Bank and the Inter-American Development Bank, the government of Perudecided to fully finance that component. However, insuranceremained part of the PARSALUD project andthe Bank provided continuous technical and analyticalsupport.30. ESSALUD is funded by employers, who paymonthly the equivalent of 9 percent of a worker’ssalary.31. The Peru Health Reform Project was launchedduring the administration of President Fujimori, wholater resigned. The new leadership distrusted programsfrom the previous administration and was not interestedin pursuing comprehensive reforms that somestakeholders would object to. Yet the principle ofextending insurance endured, based on previousanalytic work. The part of the reforms that extendedcoverage was politically popular; the new leadershipconsolidated the existing insurance programs into theSIS and expanded it nationwide, ahead of schedule, evenwhile resistance to more ambitious reforms of otherpublic sector insurers persists. The Bank addressedneeded expansion of community-managed health facilitynetworks, which the new administration was lukewarmto, through conditions in the four Programmatic SocialReform Loans, particularly the first two.32. Only 3 of the 37 were given a risk rating of high,and one of these projects was so risky that it never becameeffective (Argentina Health Insurance for thePoor).33. In the Kyrgyz Republic and Russia, capacity buildingin primary care through family medicine reformswas a precondition for hospital rationalization and establisheda quality alternative to hospital care. Financingand service delivery reforms made possible thelater benefits package and copayment schemes; changesin revenue collection and pooling were prerequisitesto introduction of new purchasing arrangements. However,in Bangladesh, too many reforms were packed intoa single project: unification of health and family planning,decentralization, a new tier of service delivery atthe community level, in addition to the launching of asectorwide approach. Community clinics were launchedbefore staff were available.34. “[T]his resulted in considerable inertia in thetypes of service being provided, but did provide a platformto attempt to shift the focus from costly curativecare.” IEG 2006b, p. 8.35. The team tried to reduce the complexity by reducingthe number of regions from three to two, buteach still entailed multiple reforms and links to a nationalcomponent.36. Peru: Poverty Assessment and Social Policiesand Programs for the Poor (1993, before the BasicHealth and Nutrition Project); Peru: Improving HealthCare for the Poor (1999, before PARSALUD); A New SocialContract in Peru and An Opportunity for a DifferentPeru (2006, before the current election cycle). The 1999piece recommended: (i) allocating funds with a sharperfocus on the needs of the poor; (ii) reinforce pro-poorfocus among Ministry of Health providers by targeting,increase access of the poor to hospitals, more communityparticipation; (iii) create new insurance mecha-169

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONnisms to finance health care for the poor; (iv) improveMinistry of Health’s information and management systems;(v) refine human resource skill requirements andincentives to better serve poverty-oriented programs.These were all issues addressed in PARSALUD.37. Analytic work prior to the first Health ReformProject included analyses of: the sexually transmittedinfection (STI) program (1994); the TB situation (1995);maternal and perinatal health (1995); acute respiratoryinfection and diarrhea control; and health spending(1995). A detailed burden of disease analysis wasused to identify areas where health status could bepromoted through ambulatory care in the most costeffectivemanner. Significant analytic work was undertakenfor the second project as well: health expendituresin the Kyrgyz Republic (1999); Who Is Paying for HealthCare in Eastern Europe and Central Asia? (Lewis2000); “Analysis of Kyrgyz Health Expenditures” (Kutzin1999); Informal Health Payments in Eastern Europeand Central Asia (Lewis 2002); Review of the SanitaryEpidemiological Services in Kyrgyz Republic (KyrgyzMinistry of Health and others 2000); and a social expenditurereview (World Bank 2001b).38. Its technical strength, comprehensiveness, andconsistency were not enough to ensure ownership(Gonzalez-Rosetti forthcoming, p. 37).39. The official World Bank definition of a SWAp is“an approach to support a locally-owned program fora coherent sector in a comprehensive and coordinatedmanner, moving toward the use of country systems”(World Bank Operations Policy and Country Services).40. The broad international consensus that countryledmanagement of aid was more likely to promote developmentwas evident by the endorsement of over 100countries and donor organizations of the Paris Declarationon Aid Effectiveness in 2005.41. Parallel funding is earmarked for project financingthat supports implementation of the national healthplan.42. Pooled funding is the pooling of Bank supporttogether with the financial resources of other developmentpartners for the purpose of supporting the implementationof national health plans and programs.These funds, allocated and managed by the government,can be channeled through the budget or in a separateaccount that is a supplement to the health budget.In all of the examples analyzed for this review, thepooled funds remained separate from the rest of thegovernment budget, as a supplement to the healthsector.43. These 28 projects were defined as having thefollowing characteristics in their design documents:(a) explicitly reference support of a sectorwide approach;(b) appear to support a program or sectorwideapproach, even without explicit reference to a SWAp;or (c) provide for the pooling and joint managementof donor funding. Among that list of projects, those wereretained that had: (d) mechanisms for coordination betweenthe government and donors, and among donors;and (e) a common M&E framework for measuring programperformance used by most donors and governmentand a mechanism for joint reviews of programperformance. Support of a sector policy and programis another fundamental characteristic of a SWAp, but isalso a common feature of non-SWAp Bank support, sowas not used as a defining criterion.44. The portfolio review understates the extent ofWorld Bank support for health SWAps to the extent thatoperations outside of the HNP sector that also supporthealth SWAps are excluded. Two examples are thehealth SWAps in Rwanda and Uganda, which are supportedby Poverty Reduction Support Credits (PRSC).45. The countries in which Bank support was notpooled included: Mauritania (2 projects), Ethiopia,Senegal, Guinea-Bissau, Mali, Sri Lanka, and Lesotho.The Bank’s decision not to pool appears to be linkedto weak country systems and capacities for the allocationand management of pooled funds and, as a consequence,a decision to undertake a more gradualapproach both to build country capacity and to nurtureand encourage other partners to move in this direction.46. Other sectors are discussing adopting internationallyaccepted procedures for procurement, financialmanagement, and audit. The Ministry of Finance hasnoted its appreciation for the fiduciary rigor of the SWApprocedures, as it insulates the sector from corruption.47. This is in contrast to pre-SWAp arrangements,in which donor financing did not appear routinely inthe government health budget and thus could not beused for purposes of government planning.48. Malawi also lacks a sound public expenditureframework.49. Use of country M&E systems still does not absolvethe responsibility of the development partners andgovernment to physically verify that implementation istaking place.170

ENDNOTES50. For example, the underfunding of malaria controlin Ghana; this was also the case with the SWAp in Malawi.Chapter 41. The 2007 HNP strategy proposes a new line of ESWcalled the Multisectoral Constraints Assessment, tohelp country teams identify the investments and sectorwork most likely to result in improved health outcomes,especially among the poor, in the preparationof CASs. Such analysis might result in either multisectoraloperations or strategically complementary sectorallending. The specifics of this new analytic instrumenthave yet to be elaborated. There has been no systematicreview of the extent to which the Bank’s CASsover the past decade have already addressed health ina multisectoral way, and the lessons learned.2. The sample of CASs reviewed included all thoseapproved from fiscal 1997 to 2006 from South Asia, EastAsia and the Pacific, and the Middle East and NorthAfrica, and a 50 percent random sample of the CASsfrom the other three larger Regions (Latin Americaand Caribbean, Europe and Central Asia, and Africa).This amounted to 65 percent of all CASs issued over theperiod. In addition, all 30 CAS Completion Report Reviewsto date were reviewed to assess the extent towhich planned multisectoral action actually occurred.Unless otherwise noted, the results cited in the text areweighted to take into account the stratification of thesample. See Sinha and Gaubatz 2008.3. Only 18 percent of CASs acknowledged the impactof HNP outcomes on other sectors. The most commonlycited linkage was environmental—the impactof population growth on natural resources (6 percent).All of these were in low-income countries and almostall were in Africa. Only 5 percent of CASs acknowledgedthe impact of HNP status on poverty reduction.4. This includes 9 percent that had both multisectoraland complementary lending.5. There was one example in each of three regionswhere there were some linkages between the HNP sectorand sectors proposed for complementary lending—child and maternal health objectives in a water supplyand sanitation project in Ghana, indicator for increasednumber of trips to health clinics for transport projectin Laos, and health care centers as an indicator for infrastructurein Sri Lanka. The lack of coordination acrosssectoral lending programs does not imply that each ofthe sectoral programs individually has not produced benefitsfor HNP outcomes indirectly, but rather that thecloser coordination and synergies envisioned in thestrategy have not really been attempted.6. In fact, implementing a multisectoral response wasitself an objective in many of these projects. A case inpoint is the Ghana AIDS Response Project (GARFUND),for which the first objective was “to intensify multisectoralactivities designed to combat the spread ofHIV/AIDS.”7. The main challenge of these projects has to dowith the number of actors and the need to generateintersectoral collaboration, not the number of components.Multisectoral projects had a mean of 3.7 components(3.9 for HIV/AIDS projects and 3.3 for othermultisectoral HNP projects), compared with 3.6 componentsfor HNP projects that were not multisectoral.8. The most frequently cited implementing agencieswere the ministries of health (91 percent); education(25 percent); labor, social security, or employment (12percent); social welfare, social affairs, or gender (11 percent);agriculture (7 percent); justice or prisons (6 percent);economy, finance, or planning (5 percent); andinformation (5 percent).9. Indonesia Iodine Deficiency Control, SenegalNutrition Enhancement Program.10. Outcome ratings can also be compared for multisectoralprojects according to the executing agencyarrangements: managed by the Ministry of Health (9projects, 44 percent satisfactory); managed by unitunder the office of the president or prime minister (8projects, none satisfactory); managed by finance orother ministries (8 projects, 75 percent satisfactory);managed by multiple ministries (3 projects, 67 percentsatisfactory). However, all of the projects managedby multisectoral commissions had other commondesign features that could have influenced outcomesindependently of the management structure. The borrower’sperformance for that subset of multisectoralprojects was 63 percent satisfactory, but the Bank’sperformance was only 35 percent satisfactory.11. IDI and sustainability ratings have been discontinued,so not all closed projects have them. Fourteenmultisectoral projects had closed that had IDIratings, of which 4 were HIV/AIDS projects and 10were other multisectoral HNP projects; 41 single-sectorHNP projects had closed and received IDI ratings. Thepercent of projects with IDI ratings of substantial orhigher was: all multisectoral (43 percent), multisectoral171

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONAIDS (50 percent), other multisectoral HNP (40 percent),and single-sector HNP projects (61 percent).There were no significant differences in the sustainabilityratings across these four categories—in all cases,70 percent or more of the projects got ratings of likelyor higher on sustainability.12. Social funds are multisectoral programs thatprovide financing (usually grants) for small-scale publicinvestments targeted at meeting the needs of thepoor and vulnerable communities, and at contributingto social capital and development at the local level.13. In fact, although in the Bank’s internal financialsystem certain percentages of Development PolicyLoans (including PRSCs) are notionally attributed to thesector, these funds are not earmarked to the health sector,nor can they even be linked to higher spending inthe health sector. (See note 2 in chapter 2 for a definitionof DPL.)14. Social Funds: Assessing Effectiveness found thatsocial fund projects, many of which have invested inhealth infrastructure, have been highly effective in deliveringsmall-scale infrastructure, but less so in achievingconsistently positive and significant improvementsin outcomes and welfare impacts. While they have deliveredslightly more than proportional benefits to thepoor and the poorest, there have also been a significantnumber of non-poor beneficiaries. Most social fundbeneficiaries are satisfied with the subprojects, but thegreatest community problems have not necessarilybeen addressed and there is no assurance that the selectedsubprojects ensure the highest net benefits tothe community. While social fund facilities are generallyoperating and equipped, they have not been immuneto staffing and equipment shortages. Social fundagencies have developed capacity as effective and innovativeorganizations; wider impacts on existing institutionshave been much more limited.15. For the full portfolio reviews, see Overbey 2008for water supply and sanitation and Freeman andMathur 2008 for transport. The transport review includedonly projects linked to the Transport SectorBoard, with commitments under sector codes TA (roadsand highways), TP (ports, waterways, and shipping), TV(aviation), TW (railways), and TZ (general transportation,including urban transport).16. Aziz and others 1990, Curtis and Cairncross2003, Daniels and others 1990, Esrey and Habicht 1985,Esrey and others 1991, Fewtrell and others 2005, Huttlyand others 1997, Strina and others 2003, VanDersliceand Briscoe 1995, Zwane and Kremer 2007. Waterrelateddiseases include waterborne diseases, waterwashed(or water scarce) diseases that are transmittedbecause of a lack of water for hygiene, water-relatedvector-borne diseases, transmitted by insects that breedin water or bite near water, like dengue, malaria, andtrypanosomiasis; and water-based diseases that transmitvia an aquatic intermediate host, such as schistosomiasisand guinea worm.17. Three of the closed projects had explicit healthobjectives and planned to collect health outcome data,but only two actually did so—the Morocco and NepalRural Water Supply and Sanitation Projects. Among the23 closed water supply and sanitation projects withouta health objective, 3 nevertheless planned to collecthealth data, but only the Kazakhstan Atryau Pilot WaterSupply and Sanitation Project did so, documenting adecline in dysentery, typhoid, intestinal infections, andhepatitis A from 1999 to 2002. A fourth project withouthealth objectives and that did not plan to collect healthdata nevertheless did—the Madagascar Rural WaterSupply and Sanitation Pilot, documenting a decline inwater-borne diseases (cholera, bilharzias, and diarrhea)from 2002 to 2004.18. WHO 2004. Losses were measured in terms ofdisability-adjusted life years.19. The review included only projects under the responsibilityof the Transport Sector Board; projectswith air quality objectives and that fell either under theEnvironment Sector Board or that were funded throughthe Global Environment Facility (GEF) are therefore notin the sample.20. The lack of data on outcomes of HIV/AIDS componentsof transport projects mirrors the experienceof transport components in HIV/AIDS projects. Amongfive recently completed HIV/AIDS projects managed bythe health sector with activities aimed at transportworkers or Ministry of Transport civil service, none reportedbaseline or end-point data on risk behavior orcondom use among transport workers. None of the ICRsprovided information on the impact of the transportsector interventions.21. This is not meant to imply that coordinationacross the complementary sectors is necessary to contributeto HNP outcomes; it is possible that investmentsin complementary sectors, even if not coordinated, aresufficient. However, the outputs and outcomes of in-172

ENDNOTESvestments in complementary sectors are integral to theresults chain contributing to HNP outcomes and wouldwarrant monitoring.Chapter 51. “[O]ne of our key objectives is to complement theWorld Bank’s work in areas of policy advice and dialogue,lending, and analysis in the health sector. Thus,we believe it is essential that the process feature the closestcollaboration possible with the HNP anchor and regionalstaff in the Bank.” IFC 2002, p. 3.2. IFC Strategic Directions, March 23, 2004,pp. 5–16. IFC produces a Strategic Directions paperevery year. This refers 2004 corporate strategy.3. In September 2006, IFC and the Bill and MelindaGates Foundation partnered on a research project thatinvestigated the best private health care business modelsin Africa, conducted by McKinsey and Companyand published as “The Business of Health in Africa: Partneringwith the Private Sector to Improve People’sLives” (IFC 2007a). Based on that report, IFC establishedits Africa Health Initiatives with the objective of mobilizingup to $1 billion over five years to boost health carein the Region. The Gates Foundation also committedto cofinance technical assistance activities related to theimplementation of the said initiative.4. The commitment amount (net of cancellations)represents IFC’s legally binding potential financialexposure to the client. It is the standard measurementof IFC’s financial exposure in its Annual Report andfinancial statements, as well as the standard referenceof multilateral development bank activities in the “GoodPractice Standards for Evaluation of Private Sector InvestmentOperations” (MDB/ ECG/WGPSE 2006, p. 2).5. IFC invested in small hospitals and clinics throughspecial promotional facilities such as Africa EnterpriseFund (AEF) and Small Enterprise Fund (SEF). These initiativesare for IFC to make direct investments in smalland medium-size enterprises where local intermediarieswere not providing sufficient financing to viable projects.IEG evaluated the AEF program as a whole andfound mixed results. While most projects had positivefeatures, only about half were successful. The programhas been costly and required a $5 million fund injectionfor operating costs. The hospital investments sharethis feature. IFC committed 15 projects (average investmentamount was $0.9 million) from 1997 until2001, when the programs were restructured and effectivelylimited its operations. More than half of theprojects failed to generate sufficient revenue to repayloans or provide any returns on equity. Half of the projectsexperienced some write-off of IFC investment;IFC investments were fully written off for a third. Theyare not featured in this study because, in line with themultilateral development banks’ best practice on privatesector evaluation, they were promotional activitieswith their own investment criteria outside of conventionalIFC operations, and only used selectively.6. The hospital projects approved after fiscal 2002 arefocused on modern urban hospitals, but they havestarted to incorporate new features to enhance their potentialdevelopment impact. For example, one projectin Sub-Saharan Africa has a link with output-based aidthat provides health insurance to the poor, enabling thehospital to provide services to underserved segmentsof the population. This is consistent with the approachproposed in the 2002 strategy for the Bank Group.Moreover, IFC emphasizes repeat investment with thesame clients and investment that moves downmarket,that is, into local markets and out of those dominatedby foreigners. Still, it is too soon to evaluate the outreachto local patients. Since 2005, IFC has invested in newareas, such as distance medical education in China.The client company provides distance education fornurses and other medical professions through publicprivatepartnership between local medial universities,local public hospitals, and doctors’ associations.7. Life science projects are also increasing in variety,as they relate not only to generic drug production (fiveof them) but also to vaccine production, drug distribution,and research and development. All of the projectshave focused on research and development. Outof 11 life science projects in the third period, 3 haveactivities related to antimalaria drugs/vaccines; 1 hasfocused its research and development on countryspecificdiseases.8. The AEF was established in 1988 to make directinvestment in small- and medium-size enterprises inSub-Saharan Africa. It was established because of ashortage of suitable private sector intermediaries in theRegion willing and capable of providing long-term financeto such enterprises.9. IFC’s investment outcome rating is an assessmentof the gross profit contribution quality of an IFC loanand/or equity investment, i.e., without taking into accounttransaction costs or the cost of IFC equity capi-173

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONtal. Loans are rated satisfactory provided they are expectedto be repaid in full with interest and fees asscheduled (or are prepaid or rescheduled withoutloss). Equities are rated satisfactory if they yield an appropriatepremium on the return on a loan to thesame company (a nominal U.S. dollar internal rate ofreturn greater than or equal to the fixed loan interestrate plus 5 percent).10. For real sector projects, like those in health,project business success is assessed according towhether they generated a project financial rate of returnat least equal to the company’s cost of capital(with a 350 basis-point spread to its equity investorsover its lenders’ nominal yield). Economic sustainabilityis assessed according to whether the project generatedan economic rate of return of at least 10 percent(in cases in which it can be assessed). Environmentaland social effects are measured by whether environmentalperformance meets IFC’s requirements and bythe projects’ net beneficial impact in terms of pollutionloads, conservation of biodiversity and natural resources,social, cultural and community health aspects,as well as labor and working conditions and workers’health and safety. Private sector development impactsmeasure impacts beyond the project company, demonstratingeffects in creating a sustainable enterprise capableof attracting finance, increasing competition,and establishing linkages with other firms.11. A blockbuster drug is a drug generating morethan $1 billion of revenue for its owner each year.12. Advisory Services specifically refers to a serviceproduct category in IFC. These services cover a broadrange of activities including technical assistance to IFCclient companies, fee-based advice to governments onprivatization, and feasibility and market studies. SomeAdvisory Services projects are funded by trust funds andothers by IFC’s own account.13. According to the rating methodology for AdvisoryServices, development effectiveness can be ratedfor projects even if the impact dimension cannot be determined(because of lack of information, for example).This explains why six projects could be rated overall ondevelopment effectiveness, even though only two hada rating on the impact dimension.14. The 2002 health strategy mentions that an entrystrategy was under development to identify the mostpromising investment opportunities in the biotechnologyindustry. It envisioned that IFC would initiallyemphasize opportunities with significant downstreampotential (product development and commercialization),rather than upstream basic research. IFC 2002,p. 38.15. For example, see Roland Berger’s research athttp://www.rolandberger.com/expertise/industries/healthcare.16. Full-time/full-year staff refers to staff that workedfor more than 2,040 hours in a given year in the Healthand Education Department (CHE).17. The Health Insurance Reform Project (fiscal1996, opening competition among the social insurancefunds), the Provincial Health Sector DevelopmentProject (fiscal 1996), the Second Provincial ReformProject (fiscal 1998) and the Second Maternal and ChildHealth and Nutrition Project (fiscal 1997, basic healthservices). Based on IEG’s field visit, this collaborationbetween the World Bank and IFC was effective until theArgentina financial crisis in 2001, after which communicationand collaboration were reduced as the Bankand IFC took wider actions to avoid actual and perceivedconflicts of interest within the World Bank Group.These actions resulted in increased barriers in sharinginformation.18. As of end of fiscal 2008, about one-third of activeCASs were done jointly by the World Bank and IFC.IFC’s engagement in the World Bank Group CAS hasbeen ad hoc and uneven. For much detailed assessmentof IFC’s involvement in the CAS in general, see IEG’s“Biennial Report on Operations Evaluation in IFC 2008”(August 2008).Appendix B1. For the two projects managed by the HNP sectorthat had no HNP sector code—the HIV/AIDS GreatLakes Initiative (fiscal 2005) and the Indonesia WSSLICIII (fiscal 2006)—design documents were reviewedand the projects were manually coded to accurately reflectthe projects’ true allocation to health.2. According to the IEG review of the transport sector(IEG 2007a), only 16 percent of all transport projectswere managed under sector boards other thantransport.3. A coterminous appointment is a term (maximumof 4 years) or open-ended appointment of indefiniteduration funded 100 percent from sources other thanthe Bank’s budget.4. A rights issue is the additional purchase of a company’sequities in the event of a new issue and exercisingshareholders’ rights to avoid dilution of its position.174

ENDNOTESRescheduling of a project often involves a new projectID, but there is no new investment; for evaluation purposes,it is bundled with the original project. These exclusionsare consistent with the Expanded ProjectSupervision Report sampling methodology, as well asthe Good Practice Standards.5. The projects meet early operating maturity andare subject of performance assessment when: (a) theproject has been substantially completed; (b) it hasgenerated at least 18 months of operating revenuesfor the company; and (c) IFC has received at least oneset of audited annual financial statements covering atleast 12 months of operating revenues generated by theproject.Appendix E1. This appendix draws on the background paper byFair (forthcoming).2. These projects were studied in greater depth becausethey had a population or family planning focus.They understate the extent of Bank support for familyplanning, which is often financed as part of maternaland child health, safe motherhood, reproductive health,or general health investments. Eight of the projects hadexplicit objectives to reduce fertility or populationgrowth, three had objectives to improve the distributionor quality of family planning, two aimed to improvereproductive health outcomes, and one supportedbetter access to an essential health package that includedfamily planning. This discussion is based onthe background paper by Fair (forthcoming).3. Malawi Population and Family Planning, IndonesiaSafe Motherhood, and Madagascar Health SectorSupport Program II.4. According to demographic and health surveydata, the total fertility rate in Guinea increased from5.5 to 5.7 over 1999–2005 and modern contraceptiveuse rose only from 4.9 to 6.8.5. The total fertility rate declined in Gambia fromabout 6.0 to 5.1, which could be attributable in part tohigher female education and delays in marriage ofyounger cohorts of women or to social marketing andgreater availability of family planning services supportedunder the project. The total fertility rate inSenegal declined from 5.7 to 5.3 over the period1997–2004. However, the modern contraceptive prevalencerose only from 9 to 10.3 percent, and only halfof women who wanted more children had access tocontraception.6. The quality at entry was unsatisfactory for six ofthe eight projects with unsatisfactory outcome ratings;for all eight projects with unsatisfactory outcome ratings,the quality of Bank supervision was also in the unsatisfactoryrange.7. Target 5b is to achieve, by 2015, universal accessto reproductive health. The main indicators are the contraceptiveprevalence rate, the adolescent birth rate, antenatalcare coverage (at least one visit and at leastfour visits), and unmet need for family planning.Appendix F1. Shekar, Heaver, and Lee 2006, pp. 3–4. Overnutrition,or obesity, is an increasing problem contributingto morbidity and mortality in developingcountries, and is often found in the same householdsas undernutrition.2. Ergo, Shekar, and Gwatkin 2008, pp. 7–8. Thegreatest inequality in stunting is in Latin America andthe Caribbean, while the greatest average levels arefound in South Asia (more than 50 percent of childrenare stunted in all four countries), followed by Sub-Saharan Africa (average of 41 percent stunted across 26countries with recent household surveys).3. Three-quarters of these projects had an objectiveof improving nutritional status, including child growthand reducing nutrient deficiencies, and about one infive aimed to increase the access to or quality of nutritionservices. One project aimed to mitigate the impactof malnutrition.4. This understates the Bank’s support for nutritionto the extent that nutrition activities are often includedin the basic health care packages, yet not mentionedin the PAD. In addition, it is likely that projects in othersectors make major contributions to nutritional status(such as agriculture).5. Seven of the projects were sectorwide approachesin health (Bangladesh, Burkina Faso, Cambodia, Mauritania,Niger, Sri Lanka) that had either an explicit nutritionobjective or component.6. The 27 projects were approved in 20 countries;12 of these countries had child stunting (height for age)of at least 30 percent, compared with a total of 47 developingcountries overall with this average level ofstunting (Shekar, Heaver, and Lee 2006, appendix 5.6).7. Only 3 percent of projects in Europe and CentralAsia were nutrition projects. For the remaining Regions,nutrition projects accounted for 12–14 percentof the HNP portfolio.175

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION8. Bangladesh Health and Population, India Womenand Child Development, Afghanistan Health SectorEmergency, Burkina Faso Health Sector Support andAIDS.9. See, in particular, chapter 4 for a discussion of thehealth effects of water supply and sanitation interventionsin Nepal.10. The Bank’s performance was rated satisfactoryfor 57 percent of the closed projects, and the borrower’sperformance only slightly better (64 percent).11. World Bank 2004b. Note, however, that thesefigures are based on a comparison of a baseline totalgoiter rate among both schoolchildren and pregnantwomen and a follow-up measurement of schoolchildrenonly. This is an example of the very weak M&E inmost of the nutrition portfolio. A 2001 study by IEGcompared the performance of this project with twoother iodine deficiency projects that promoted production,distribution, and consumption of iodized saltin China and Madagascar (Goh 2001). The study pointedto the extreme difficulty of getting results in Indonesiabecause of the large number of small salt producersthat are difficult to regulate (70 percent of saltproduction comes from small salt farmers). In contrast,salt production and distribution is centrally controlledin China, while about six large producers inMadagascar produce 80 percent of the salt. That studyfound four key factors affecting the success of projectspromoting iodized salt to combat iodine-deficiencydisorders (IDD): (a) information and behavior changecommunication concerning IDD to raise the demandfor iodized salt; (b) easy access; (c) industry compliance,through persuasion and alignment of incentives withthe self-interests of the salt industry and consumers;and (d) quality control, since consumers cannot tell theextent of iodization.12. In the Sri Lanka Health Services DevelopmentProject, low birthweight declined marginally, and anevaluation of one year of work by the nongovernmentalorganizations showed minimal impact; other indicatorslike the consumption of iodized salt were notmeasured. The Mauritania Nutrition, Food Securityand Social Mobilization Project—a Learning and InnovationLoan that was supposed to test two differentapproaches to improved nutrition—did not do so; malnutritionwas reduced 16 percent in two urban areas(though the data being compared are from a lean monthand a normal season, so this may be exaggerated) andincreased in rural areas by 6–8 percent; anemia amongpregnant women and consumption of iodized sale werenot measured. (Mauritania also experienced a majordrought and locust infestation during project implementation.)The Bangladesh National Nutrition Programdid not collect baseline data until two years beforethe project closed, and found that some of the targetsthat had been set were already achieved at baseline; projectmonitoring data suggest a decline in severe proteinenergymalnutrition (PEM) from 24 to 16 percent from2004 to 2006, and moderate PEM from 30 to 25 percent.There was substantial evidence of increased knowledgeand change in some behaviors, however. Finally, therewas no discernable improvement in malnutrition indicatorsin Andhra Pradesh, India, during the EconomicRestructuring Project. (Source: Implementation CompletionReports.)Appendix G1. ESW is an activity that “(a) involves analytic effort;(b) is undertaken with the intent of influencing an externalclient’s policies and/or programs; and (c) representsthe views of the Bank (that is, not attributed toindividual authors).” (IEG 2008h, p. 1)2. Of this amount, $27.7 million was from the Bankbudget and $15.1 million was financed from trust funds.Fiscal 2000 is the starting point for this review, becausebefore that date ESW was not systematically tracked inthe Bank’s internal information systems.3. The Reaching the Poor with HNP Services Projectwas partly financed by the Bank’s research supportbudget, partly as knowledge management through theWorld Bank Institute and the HNP anchor, in additionto the Bill and Melinda Gates Foundation and the governmentsof Netherlands and Sweden. The only partof the exercise captured as official ESW were a few disseminationactivities.4. This total excludes obvious duplicates (for example,when a research working paper was also publishedas a book chapter). See the discussion in appendix H.One reason for the difference is that ESW for fiscal1997–99 is not in the official database. However, becauseof the elapsed time, analytic work from those years wasalso more difficult to locate, despite an exhaustivesearch conducted by IEG.5. The distribution of HNP ESW is slightly moreskewed toward regional ESW and less to the countrylevel, compared to the distribution for all sectors combined(country-level, 78 percent; regional, 17 percent;global, 5 percent) (IEG 2008h).176

ENDNOTES6. The results tracking framework for ESW was introducedin 2004; all ESW with a budget > $50,000 hadto include a statement of objectives as of fiscal 2004,and in 2005, this was expanded to include all ESW.While the share of HNP ESW with an objective of informinglending may seem low (51 percent), it is actuallygreater than for ESW Bank-wide (41 percent with an objectiveof informing lending) (IEG 2008h).7. While most PERs are managed outside the HNPsector, in some cases PERs have been sponsored by theHNP sector.Appendix J1. The 1997 HNP Strategy, Health, Nutrition, andPopulation Sector Strategy Paper (World Bank 1997b),July 3, 1997, called for (a) improving health, nutrition,and population outcomes for the poor; (b) enhancingthe performance of health care systems; and (c) securingsustainable health care financing. This was to beachieved through (i) sharpening strategic directions;(ii) achieving greater impact; (iii) empowering HNP staff;and (iv) building partnerships.2. For example, the Global Fund to Fight AIDS, Tuberculosisand Malaria (GFATM), the Global Alliance forVaccines and Immunization (GAVI), the Bill and MelindaGates Foundation, and the U.S. President’s EmergencyPlan for AIDS Relief (PEPFAR).3. For example, the International Health Partnership,for which the Bank, jointly with the World HealthOrganization (WHO), provides the secretariat.4. See, for example, the investigative study cited inendnote 6.5. Even in low-income countries, total official developmentassistance for health is less than 30 percentof total expenditures.6. An example of a country application of theReaching the Poor work is the investigative study of utilizationof health services and patient satisfaction forthe poor in selected states in India, which has led to arange of reforms including management training, newstaffing and service patterns, provision of essentialdrugs, and repair of equipment and facilities. This programresulted in increased utilization of all types ofhealth facilities (improving absolute levels of utilizationamong the poorest 40 percent of the population) andimproved patient satisfaction at lower-level project facilities(as opposed to hospitals) for the poor.7. The establishment of two regional hubs in Africa(in Dakar and Nairobi) and recruitment of 10 highlevelexperts with a strong health systems focus will becompleted by the end of fiscal year 2009.8. Management notes, however, that during a 2005Board discussion of the report from the Committee onDevelopment Effectiveness, Committing to Results:Improving the Effectiveness of HIV/AIDS Assistanceand Draft Management Response, questions wereraised about the IEG’s [formerly Operations and EvaluationDepartment’s] evaluation methodology of theBank’s HIV/AIDS assistance, and concerns were raisedabout IEG’s tone.9. IEG would like to clarify that the recent results forHIV/AIDS projects highlighted in the HNP evaluationpost-date the 2005 HIV/AIDS evaluation, which coveredresults only through mid-2004 and did not include anycompleted projects from the Multi-Country HIV/AIDSProgram. Further, IEG and management use the sameharmonized criteria for rating projects in all sectors, includingHIV/AIDS.10. The Portfolio Improvement Action Plan includesan analysis of the critical obstacles to project improvement,changes needed to upgrade projects, andmeasures to restructure or cancel if needed. Additionalfinancial and/or technical support to achieve these improvementsis an explicit part of the plan.177

Koh Toch Health Center in Kampot Province, Cambodia. The Cambodia Disease Control and Health Department Project supported investmentsin improved health infrastructure and in control of major communicable disease killers—malaria, TB, and HIV/AIDS.Photo courtesy of Martha Ainsworth.

REFERENCESAdeyi, O., E. Baris, S. Chakraborty, T. Novotny, and R.Pavis. 2003. Averting AIDS Crises in Eastern Europeand Central Asia—A Regional Support Strategy.Washington, DC: World Bank.Ainsworth, Martha, and Manisha Modi. 2009. “Overviewof the Health, Nutrition, and Population Portfolio,FY1970–2006.” IEG Working Paper. IEG, Washington,DC.Akin, John S., Nancy Birdsall, and David M. De Ferranti.1987. Financing Health Services in DevelopingCountries: An Agenda for Reform. World BankPolicy Study. Washington, DC: World Bank.Armecin, Graeme, Jere R. Behrman, Paulita Duazo,Sharon Ghuman, Socorro Gultiano, Elizabeth M.King, and Nanette Lee. 2006. “Early Childhood Developmentthrough an Integrated Program: Evidencefrom the Philippines.” World Bank PolicyResearch Working Paper No. 3922. World Bank,Washington, DC.Aziz, K., B. Hoque, K. Hasan, M. Patwary, S. Huttly, M.Rahaman, and R. Feachem. 1990. “Reduction in DiarrhoealDiseases in Children in Rural Bangladeshby Environmental and Behavioural Modifications.”Transactions of the Royal Society of Tropical Medicineand Hygiene 84: 433–38.Bustreo, Flavia, A. Harding, and H. Axelsson. 2003.“Can Developing Countries Achieve Adequate Improvementsin Child Health Outcomes Without Engagingthe Private Sector?” Bulletin of the WorldHealth Organization 81(12): 886–94.Bryce, Jennifer, Cesar G. Victora, Jean-Pierre Habicht,Robert E. Black, and Robert W. Scherpbier on behalfof the MCE-IMCI Technical Advisers. 2005. “ProgrammaticPathways to Child Survival: Results of aMulti-Country Evaluation of Integrated Managementof Childhood Illness.” Health Policy and Planning,December 20, Supplement 1: i5–i17.Cassels, Andrew. 1997. A Guide to Sector-Wide Approachesfor Health Development: Concepts, Issuesand Working Arrangements. WHO, DanishInternational Development Agency (DANIDA), Departmentfor International Development (U.K.),European Commission. Geneva: WHO.Castro-Leal, F., J. Dayton, L. Demery, and K. Mehra.1999. “Public Social Spending in Africa: Do the PoorBenefit?” World Bank Research Observer 14(1):49–72.Charmarbagwala, Rubiana, Martin Ranger, Hugh Waddington,and Howard Nial White. 2004. “The Determinantsof Child Health and Nutrition: A Metaanalysis.”IEG Working Paper. IEG, Washington, DC.Claeson, Mariam, Charles C. Griffin, Timothy A. Johnston,Mildred McLachlan, Agnes L.B. Soucat, AdamWagstaff, and Abdo Yazbeck. 2002. “Health, Nutrition,and Population.” In A Sourcebook for Poverty ReductionStrategies, ed. Jeni Klugman. Washington,DC: World Bank.Cotlear, Daniel. 2000. “Peru: Reforming Health Care forthe Poor.” World Bank Human Development Department,LCSHD Paper Series No. 57. World Bank,Washington, DC.COWI, Goss Gilroy, Inc., and EPOS. 2007. Joint ExternalEvaluation of the Health Sector in Tanzania,1999–2006. Copenhagen: Ministry of Foreign Affairsof Denmark.Curtis, Val, and Sandy Cairncross. 2003. “Effect of WashingHands with Soap on Diarrhea Risk in the Community:A Systematic Review.” Lancet InfectiousDiseases 3: 275–81.Daniels, D.L., S.N. Cousens, L.N. Makoae, and R.G.Feachem. 1990. “A Case-Control Study of the Impactof Improved Sanitation on Diarrhea Morbidity inLesotho.” Bulletin of the World Health Organization68(4): 455–63.Dye, Christopher, Catherine Watt, and others. 2004.“The Effect of Tuberculosis Control in China.” TheLancet 364: 417–22.Elmendorf, A. Edward, and Flora Nankhuni. Forthcoming.“Evaluation of the World Bank’s Support forHealth, Nutrition and Population: Malawi CaseStudy.” IEG Working Paper. IEG, Washington, DC.Ergo, Alex, Meera Shekar, and Davidson R. Gwatkin.2008. “Inequalities in Malnutrition in Low- andMiddle-Income Countries: Updated and ExpandedEstimates.” Country Reports on HNP and Poverty.World Bank, Washington, DC.179

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONEsrey, S.A., and J-P. Habicht. 1985. The Impact of ImprovedWater Supplies and Excreta DisposalFacilities on Diarrheal Morbidity, Growth, andMortality among Children. Cornell InternationalNutrition Monograph Series 15. Ithaca, NY: Divisionof Nutritional Sciences, Cornell University.Esrey, S.A., J.B. Potash, L. Roberts, and C. Shiff. 1991.“Effects of Improved Water Supply and Sanitation onAscariasis, Diarrhea, Dracunculiasis, Hookworm Infection,Schistosomiasis, and Trachoma.” Bulletin ofthe World Health Organization 69(5): 609–21.Fair, Mollie. Forthcoming. “Review of the Portfolio ofPopulation Lending, 1997–2006.” IEG Working Paper.IEG, Washington, DC.———. 2008. “From Population Lending to HNP Results:The Evolution of the World Bank’s Strategiesin Health, Nutrition, and Population.” IEG WorkingPaper No. 2008/3. IEG, Washington, DC.Fewtrell, Lorna, Rachel B. Kaufmann, David Kay, WayneEnanoria, Laurence Haller, and John M. Colford, Jr.2005. “Water, Sanitation and Hygiene Interventionsto Reduce Diarrhea in Less Developed Countries:A Systematic Review and Meta-analysis.” Lancet InfectiousDiseases 5: 42–52.Filmer, Deon. 2003. “The Incidence of Public Expendituresin Health and Education.” Background Noteto World Development Report 2004: Making ServicesWork for Poor People. World Bank, Washington,DC.Freeman, Peter, and Kavita Mathur. 2008. “The HealthBenefits of Transport Projects: A Review of the WorldBank Transport Sector Lending Portfolio.” IEG WorkingPaper No. 2008/2. IEG, Washington, DC.G8. 2000. “G8 Summit Communiqué.” Okinawa, Japan,July 23.Gakidou, Emmanuela, Rafael Lozano, Eduardo Gonzales-Pier,Jesse Abbott-Klafter, Jeremy T. Barofsky,Chloe Bryson-Cahn, Dennis M. Feehan, Diana K. Lee,Hector Hernandez-Llamas, and Christopher J.L.Murray. 2006. “Assessing the Effect of the 2001–06Mexican Health Reform: An Interim Report Card.”The Lancet 368: 1920–35.Giugale, Marcelo, Vincente Fretes-Chibils, and John L.Newman. 2006. An Opportunity for a DifferentPeru: Prosperous, Equitable, and Governable.Washington, DC: World Bank.Goh, Chorching. 2001. “An Analysis of Combating IodineDeficiency: Case Studies of China, Indonesia,and Madagascar.” IEG Working Paper No. 18. IEG,Washington, DC.Gonzalez-Rosetti, Alejandra. Forthcoming. “Evaluationof the World Bank’s Support for Health, Nutrition,and Population: Egypt Case Study.” IEG WorkingPaper. IEG, Washington, DC.Gottret, P., and G. Schieber. 2006. Health FinancingRevisited: A Practitioner’s Guide. Washington, DC:World Bank.Grosh, M., and J. Muñoz. 1996. “A Manual for Planningand Implementing the Living Standards MeasurementStudy.” Survey Living Standards MeasurementStudy Working Paper No. 126. World Bank, Washington,DC.Gwatkin, Davidson, and M. Guillot. 2000. The Burdenof Disease among the Global Poor. Washington,DC: World Bank.Gwatkin, Davidson, Adam Wagstaff, and Abdo Yazbeck.2005. Reaching the Poor with Health Services: WhatWorks, What Doesn’t and Why. Washington, DC:World Bank.Hammond, Allen L., William J. Kramer, Robert S. Katz,Julia T. Tran, and Courtland Walker. 2007. The Next4 Billion: Market Size and Business Strategy at theBase of the Pyramid. Washington, DC: World ResourcesInstitute and IFC.Harrold, Peter, and others. 1995. “The Broad Sector Approachto Investment Lending: Sector InvestmentPrograms,” World Bank Discussion Paper No. 302,Africa Technical Department Series. World Bank,Washington, DC.Homedes, Núria, Antonio Ugalde, and Joan RoviraForns. 2005. “The World Bank, PharmaceuticalPolicies, and Health Reforms in Latin America.” InternationalJournal of Health Services 35(4):691–717.Huttly, S.R.A., S.S. Morris, and V. Pisani. 1997. “Preventionof Diarrhea in Young Children in DevelopingCountries.” Bulletin of the World Health Organization75(2): 163–74.IEG (Independent Evaluation Group). Forthcoming.“Nepal: Rural Water Supply and Sanitation Project.”Project Performance Assessment Report. IEG, WorldBank, Washington, DC.———. 2009a. “Eritrea: Health Project and HIV/AIDS,Malaria, STD and TB Control Project (HAMSET).”Project Performance Assessment Report. IEG, WorldBank, Washington, DC.180

REFERENCES———. 2009b. “Peru: Basic Health and Nutrition Projectand Health Reform Project.” Project PerformanceAssessment Report. IEG, World Bank,Washington, DC.———. 2008a. Annual Review of Development Effectiveness2008: Shared Global Challenges. IEG StudySeries. Washington, DC: World Bank.———. 2008b. “Egypt: Population Project.” ProjectPerformance Assessment Report. IEG, World Bank,Washington, DC.———. 2008c. “Egypt: Schistosomiasis Control Project.”Project Performance Assessment Report. IEG, WorldBank, Washington, DC.———. 2008d. “Kyrgyz Republic: Health Reform I andII.” Project Performance Assessment Report. IEG,World Bank, Washington, DC.———. 2008e. “The Population and ReproductiveHealth Capacity Building Program (PRHCBP).”Global Program Review. IEG, Washington, DC.———. 2008f. Public Sector Reform: What Works andWhy? IEG Study Series. Washington, DC: World Bank.———. 2008g. “Romania: Roads II Project.” ProjectPerformance Assessment Report. IEG, World Bank,Washington, DC.———. 2008h. Using Knowledge to Improve DevelopmentEffectiveness: An Evaluation of World BankEconomic and Sector Work and Technical Assistance,2000–2006. IEG Study Series. Washington, DC:World Bank.———. 2008i. What Works in Water Supply and Sanitation?Lessons from Impact Evaluation. IEG StudySeries. Washington, DC: World Bank.———. 2007a. A Decade of Action in Transport: AnEvaluation of World Bank Assistance to the TransportSector, 1995–2005. IEG Study Series. Washington,DC: World Bank.———. 2007b. Development Results in Middle-IncomeCountries. IEG Study Series. Washington, DC: WorldBank.———. 2007c. “Ghana: AIDS Response Project (GAR-FUND).” Project Performance Assessment Report.IEG, World Bank, Washington, DC.———. 2007d. “Ghana: Second Health and PopulationProject and Health Sector Support Project.”Project Performance Assessment Report. IEG, Washington,DC.———. 2007e. “Medicines for Malaria Venture.” GlobalProgram Review. IEG, Washington, DC.———. 2007f. “Russian Federation: Health ReformPilot Project.” Project Performance Assessment Report.IEG, Washington, DC.———. 2007g. Unlocking the Development Potentialof Regional Programs. IEG Study Series. Washington,DC: World Bank.———. 2006a. Annual Review of Development Effectiveness2006: Getting Results. IEG Study Series.Washington, DC: World Bank.———. 2006b. “Bangladesh: Fourth Population andHealth Project and Health and Population ProgramProject.” Project Performance Assessment Report.IEG, Washington, DC.———. 2006c. Engaging with Fragile States: An IEG Reviewof World Bank Support to Low-Income CountriesUnder Stress. IEG Study Series. Washington, DC:World Bank.———. 2006d. “Vietnam: Population and Family HealthProject.” Project Performance Assessment Report.IEG, Washington, DC.———. 2005a. Committing to Results: Improving theEffectiveness of HIV/AIDS Assistance. An OED Evaluationof the World Bank’s Assistance for HIV/AIDSControl. IEG Study Series. Washington, DC: WorldBank.———. 2005b. Maintaining Momentum to 2015? AnImpact Evaluation of Interventions to ImproveMaternal and Child Health and Nutrition inBangladesh. IEG Study Series. Washington, DC:World Bank.———. 2005c. The Effectiveness of World Bank Supportfor Community-Based and -Driven Development.IEG Study Series. Washington, DC: WorldBank.———. 2004a. Addressing the Challenges of Globalization.An Independent Evaluation of the WorldBank’s Approach to Global Programs. IEG Study Series.Washington, DC: World Bank.———. 2004b. “Cambodia: Disease Control and HealthDevelopment Project.” Project Performance AssessmentReport. IEG, Washington, DC.———. 2002a. Bridging Troubled Waters: Assessingthe World Bank Water Resources Strategy. IEG StudySeries. Washington, DC: World Bank.———. 2002b. Social Funds: Assessing Effectiveness.IEG Study Series. Washington, DC: World Bank.———. 1999. Investing in Health: Development Effectivenessin the Health, Nutrition, and Population181

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONSector. IEG Study Series. Washington, DC: WorldBank.———. 1998. Nongovernmental Organizations inBank-Supported Projects. IEG Study Series. Washington,DC: World Bank.———. 1992. Population and the World Bank: Implicationsfrom Eight Case Studies. Washington,DC: World Bank.IEG-IFC. 2008. Biennial Report on Operations Evaluationin IFC 2008. Washington, DC: World Bank.IFC. 2007a. The Business of Health in Africa: Partneringwith the Private Sector to Improve People’sLives. Washington, DC: IFC, The World BankGroup.———. 2007b. Annual Report. Washington, DC: IFC,The World Bank Group.———. 2007c. “IFC: Innovation for Better Health in DevelopingCountries—Investing in Life Sciences inEmerging Markets.” Brochure. IFC, Washington,DC.———. 2007. “Technical Briefing to the Board of Directors.”Health and Education Department. IFC,The World Bank Group, Washington, DC.———. 2006. “Smart Lessons from Africa,” Johannesburg,South Africa.———. 2004. “Strategic Directions.” IFC, WashingtonDC.———. 2002. “Investing in Private Health Care: StrategicDirections for IFC.” Health and Education Department.IFC, Washington, DC.———. 1998. Beyond 2000. Washington, DC: WorldBank.———. 1982. Annual Report. Washington, DC: WorldBank.IFC Health Care Best Practice Group. 1999. “Investingin Private Health Care: A Note on Strategic Directionsfor IFC.” IFC, Washington, DC. Photocopy.Institute for Health Sector Development. 2004. “PrivateSector Participation in Health.” London.Jalan, Jyotsna, and Martin Ravallion. 2003. “Does PipedWater Reduce Diarrhea for Children in Rural India?”Journal of Econometrics 112: 153–73.Jamison, D.T. 2006. “Investing in Health.” In DiseaseControl Priorities in Developing Countries, 2d ed.Washington, DC: Oxford University Press for theWorld Bank.Jamison, D.T., J.G. Breman, A.R. Measham, G. Alleyne,M. Claeson, D.B. Evans, P. Jha, A. Mills, and P. Musgrove.2006a. Priorities in Health. Washington, DC:World Bank.Jamison, Dean T., Joel G. Breman, Anthony R. Measham,George Alleyne, Mariam Claeson, David B. Evans,Prabhat Jha, Anne Mills and Philip Musgrove, eds.2006b. Disease Control Priorities in DevelopingCountries, 2d ed. Washington, DC: The World Bank.www.dcp2.org/pubs/DCP.Jamison, Dean T., W.H. Mosley, A. R. Measham, and J.L.Bobadilla. Disease Control Priorities in DevelopingCountries. 1993. New York and Washington, DC: OxfordUniversity Press for the World Bank.Jha, P., and F. Chaloupka. 2000. Tobacco Control in DevelopingCountries. New York: Oxford UniversityPress.Johnston, Timothy. 2002. “Supporting a Healthy Transition:Lessons from Early World Bank Experiencein Eastern Europe.” IEG Working Paper Series. IEG,Washington, DC.Keutsch, G.T., O. Fontaine, A. Ghargava, C. Boschi-Pinto, Z.A. Bhutta, E. Gotuzzo, J. Rivera, J. Chow, A.Shhid-Salles, and R. Laxminarayan. 2006. “DiarrhealDiseases.” In Disease Control Priorities in DevelopingCountries, 2d ed., ed. D. Jamison and others.New York: Oxford University Press.Klugman, Jeni, ed. 2002. A Sourcebook for Poverty ReductionStrategies. Washington, DC: World Bank.Kutzin, Joseph. 1999. “Analysis of Kyrgyz Health Expenditures.”World Bank. Photocopy.Kyrgyz Ministry of Health, WHO Regional Office forEurope, World Bank, DFID, and USAID/ZdravPlusProgram. 2000. “Review of Sanitary EpidemiologicalServices in Kyrgyz Republic.” Report of the InteragencyMission to the Kyrgyz Republic.Lele, Uma, Naveen Sarna, Ramesh Govindaraj, andYianni Konstantopoulos. 2004. “Global Health Programs,Millennium Development Goals, and theWorld Bank’s Role.” IEG Working Paper Series. Casestudy for Addressing the Challenges of Globalization.IEG, Washington, DC.Levine, Ruth, and the What Works Working Group,with Molly Kinder. 2004. Millions Saved: ProvenSuccesses in Global Health. Washington, DC: Centerfor Global Development.Lewis, Maureen. 2002. “Informal Health Payments inEastern Europe and the Former Soviet Union: Issues,Trends, and Policy Explanations.” In Funding HealthCare: Options for Europe, ed. E. Mossialos, A. Dixon,182

REFERENCESJ. Figueras, and J. Kutzin. Buckingham, U.K.: OpenUniversity Press.———. 2000. Who Is Paying for Health Care in EasternEurope and Central Asia? Human DevelopmentSector Unit, Europe and Central Asia Region.Washington, DC: World Bank.Loevinsohn, Benjamin, and Aakanksha Pande. 2006.“Measuring Results: A Review of Monitoring andEvaluation in HNP Operations in South Asia andSome Practical Suggestions for Implementation.”South Asia Human Development Unit, World Bank,Washington, DC.Lopez, Alan D., Colin D. Mathers, Majid Ezzati, Dean T.Jamison, and Christopher J. L. Murray. 2006. “Globaland Regional Burden of Disease and Risk Factors,2001: Systematic Analysis of Population Health Data.”The Lancet 367: 1747–57.Lutalo, Martin. 2006. “HIV/AIDS Getting Results: IFCAgainst Aids—Protecting People and Profitability.”IFC, Washington, DC.Mahal, A., J. Sing, F. Afridi, V. Lamba, A. Gumber, and V.Selvaraju. 2000. “Who Benefits from Public HealthSpending in India?” National Council for Applied EconomicResearch, New Delhi.Maikisch, Harald. 2007. “Lessons from Practice: PublicPrivate Collaboration Workshop in Podgorica, Montenegro.”Presented at the Workshop, Sept. 17–18.Martin, Gayle. 2009. “Portfolio Review of World BankLending for Communicable Disease Control.” IEGWorking Paper Series. IEG, Washington, DC.Michaud, Catherine. 2003 (October 29–30). “DevelopmentAssistance for Health (DAH): Recent Trendsand Resource Allocation.” Paper prepared for theSecond Consultation, Commission on Macroeconomicsand Health. WHO, Geneva.Mills, Anne, R. Brugha, K. Hanson, and B. McPake.2002. “What Can Be Done about the Private HealthSector in Low-Income Countries?” Bulletin of theWorld Health Organization 80(4): 325–30.MDB (Multilateral Development Banks), ECG (EvaluationCooperation Group), and WGPSE (WorkingGroup on Private Sector Evaluation). 2006. “MDB-ECG Good-Practice Standards for Evaluation of PrivateSector Investment Operations,” 3d ed. ECG-WGPSE.Nankhuni, Flora. Forthcoming. “Health Outcomes amongthe Poor in Countries Receiving World Bank Support.”IEG Working Paper, IEG, Washington, DC.Nankhuni, Flora, and Manisha Modi. 2008. “Analysis ofTrends in Staffing of the World Bank’s HNP Sector.”HNP Evaluation Note. IEG, Washington, DC.Narayan, Deepa, and Patti Petesch. 2002. Voices of thePoor: From Many Lands. New York: Oxford UniversityPress for the World Bank.Nikolic, Irina A., and Harald Maikisch. 2006. Public-Private Partnerships and Collaboration in theHealth Sector: An Overview with Case Studies fromRecent European Experience. World Bank Health,Nutrition and Population (HNP) Discussion Paper.Washington, DC: World Bank.O’Donnell, O., E. van Doorslaer, R. Rannan-Eliya, A.Somanathan, S. Adhikari, D. Harbianto, C. Garg, P.Hanvoravongchai, M. Huq, A. Karan, G. Leung, C.Wan Ng, B. Raj Pande, K. Tin, K. TIsayaticom, L.Trisnantoro, Y. Zhang, and Y. Zhao. 2007. “TheIncidence of Public Spending on Healthcare: ComparativeEvidence from Asia.” World Bank EconomicReview 21(1): 93–123.Over, Mead, Elliot Marseille, Julian Gold, Peter Heywood,Subhash Hira, and Indrani Gupta. 2004. HIV/AIDSTreatment and Prevention in India: Modeling theCosts and Consequences. Washington, DC: WorldBank.Overbey, Lisa. 2008. “The Health Benefits of Water Supplyand Sanitation Projects: A Review of the WorldBank’s Lending Portfolio.” IEG Working Paper No.2008/1. IEG, Washington, DC.PAHO (Pan-American Health Organization). 2007.Health in the Americas 2007. Vol. I. Scientific andTechnical Publication #622. Washington, DC: PAHO.Revenga, Ana, Mead Over, Wiwat Peerapatanapokin,Sombat Thanprasertsuk, Emiko Masaki, Viroj Tangcharoensathien,and Julian Gold. 2006. The Economicsof Effective AIDS Treatment: EvaluatingPolicy Options for Thailand. Washington, DC: WorldBank.Robinson, Warren C., and John A. Ross, eds. 2007. TheGlobal Family Planning Revolution: Three Decadesof Population Policies and Programs. Washington,DC: World Bank.Roll Back Malaria Partnership Secretariat. 2005. GlobalStrategic Plan 2005–2015. Geneva: Roll Back MalariaPartnership Secretariat.Sahn, D., and S. Younger. 2000. “Expenditure Incidencein Africa: Microeconomic Evidence.” Fiscal Studies21(3): 321–48.183

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATIONShaw, R. Paul. Forthcoming. “Evaluation of the WorldBank’s Support for Health, Nutrition, and Population:Nepal Case Study.” IEG Working Paper. IEG,Washington, DC.Shekar, Meera, Richard Heaver, and Yi-Koung Lee.2006. Repositioning Nutrition as Central to Development:A Strategy for Large-Scale Action. Directionsin Development Series. Washington, DC: WorldBank.Sinha, Shampa, and Judith Gaubatz. 2009. “Have CountryAssistance Strategies Capitalized on IntersectoralLinkages to Improve Health Outcomes among thePoor?” IEG Working Paper. IEG, Washington, DC.Strina, A., S. Carincross, M.L. Barretto, C. Larrea, andM.S. Prado. 2003. “Childhood Diarrhea and ObservedHygiene Behavior in Salvador, Brazil.” AmericanJournal of Epidemiology 157(11): 1032–38.Subramanian, Savitha, David Peters, and Jeffrey Willis.2006. How Are Health Services, Financing, andStatus Evaluated? An Analysis of ImplementationCompletion Reports of World Bank Assistance inHealth. HNP Discussion Paper. Washington, DC:World Bank.UNAIDS (Joint United Nations Progam on AIDS). 2008.Report on the Global AIDS Epidemic 2008. Geneva:UNAIDS.UNECA (United Nations Economic Commission forAfrica), UNICEF (United Nations Children’s Fund),and World Bank. 1998. “Addis Ababa Consensus onPrinciples of Cost Sharing in Education and Health.”New York: UNICEF.UNICEF. (United Nations Children’s Fund). 2006. Stateof the World’s Children. New York: UNICEF.Vaillancourt, Denise. 2009. “World Bank Support for SectorwideApproaches in Health: A Comparative Studyof IEG Evaluations of Health SWAps in Five Countries.”IEG Working Paper. IEG, Washington, DC.VanDerslice, James, and John Briscoe. 1995. “EnvironmentalInterventions in Developing Countries:Interactions and Their Implications.” AmericanJournal of Epidemiology 141(2): 135–44.Van de Walle, D. 1995. “The Distribution of Subsidiesthrough Public Health Services in Indonesia.” InPublic Spending and the Poor, ed. Dominique VanDe Walle and Kimberley Neads. Washington, DC:World Bank.Villar Uribe, Manuela. Forthcoming. “Pilot Projects, ImpactEvaluation, and the Quality of Monitoring andEvaluation in the Health, Nutrition, and PopulationPortfolio.” IEG Working Paper. IEG, Washington, DC.Wagstaff, Adam, and Mariam Claeson. 2004. The MillenniumDevelopment Goals for Health: Rising tothe Challenges. Washington, DC: World Bank.Wagstaff, Adam, and Shengchao Yu. 2007. “Do HealthSector Reforms Have Their Intended Impacts?The World Bank’s Health VIII Project in Gansu Province,China.” Journal of Health Economics 26(3):505–35.White, Arlette Campbell, Thomas W. Merrick, and AbdoS. Yazbeck. 2006. Reproductive Health: The MissingMillennium Development Goal. Washington, DC:World Bank.Wood, Bernard, Dorte Kabell, Nansozi Muwanga, andFrancisco Sagasti. 2008. Synthesis Report on theFirst Phase of the Evaluation of the Implementationof the Paris Declaration. Copenhagen: Kabell KonsultingApS.World Bank. 2009. “Implementation of the World Bank’sStrategy for Health, Nutrition, and Population (HNP)Results: Achievements, Challenges, and the WayForward—Progress Report.” SecM2009-0147. WorldBank, Washington, DC.———. 2008a. Global Monitoring Report. Washington,DC: World Bank.———. 2008b. “Implementation Completion and ResultsReport on a Credit in the Amount of SDR 41.3Million (US$60 Million Equivalent) to the Republicof Madagascar for a Second Health Sector SupportProject.” World Bank, Washington, DC.———. 2008c. “Implementation Completion and ResultsReport on a Proposed Credit in the Amount ofSDR 39.2 Million (US$50.0 Million Equivalent) to theRepublic of Cameroon for a Multi-Sectoral HIV/AIDSProject.” World Bank, Washington, DC.———. 2008d. “Implementation Completion and ResultsReport on a Credit in the Amount of SDR 38.8Million (US$50 Million Equivalent) to the Republicof Kenya for a Decentralized Reproductive Healthand HIV/AIDS Project.” World Bank, Washington, DC.———. 2008e. “Implementation Completion and ResultsReport on a Credit in the Amount Of SDR 41.3Million (US$60 Million Equivalent) to the Republicof Madagascar for a Second Health Sector SupportProject.” World Bank, Washington, DC.———. 2008f. Safe, Clean, and Affordable: Transportfor Development. The World Bank Group’s Trans-184

REFERENCESport Business Strategy for 2008–2012. Washington,DC: World Bank.———. 2007a. Healthy Development: The World BankStrategy for Health, Nutrition, and Population Results.Washington, DC World Bank.———. 2007b. “Implementation Completion Report ona Credit in the Amount of US$50 Million to the NationalAIDS Control Council, Government of Kenyafor a HIV/AIDS Disaster Response Project.” WorldBank, Washington, DC.———. 2007c. “Implementation Completion and ResultsReport on a Credit in the Amount of SDR 11.8 Million(US$14.7 million equivalent) to the Republic ofSenegal in Support of the First Phase Nutrition EnhancementProgram.” World Bank, Washington, DC.———. 2007d. “Implementation Completion and ResultsReport on a Credit in the Amount of SDR 68.50Million (US$92.0 Million Equivalent) to the People’sRepublic of Bangladesh for a National NutritionProject.” World Bank, Washington, DC.———. 2007e. “Implementation Completion and ResultsReport on a Loan in the Amount of US$30.0 Millionto the Bolivarian Republic of Venezuela for aCaracas Metropolitan Health Services Project.” WorldBank, Washington, DC.———. 2007f. Malawi Public Expenditures Review.Washington, DC: World Bank.———. 2007g. Population Issues in the 21st Century:The Role of the World Bank. HNP Discussion Paper.Washington, DC: World Bank.———. 2006a. “Implementation Completion Report ona Credit in the Amount of US$96.7 million to Indiafor a Tuberculosis Control Project.” Report No.34692. World Bank, Washington, DC.———. 2006b. A New Social Contract for Peru. WorldBank Country Study. Washington, DC: World Bank.———. 2006c. Repositioning Nutrition as Central toDevelopment: A Strategy for Large-Scale Action.Washington, DC: World Bank.———. 2005a. Reaching the Poor with Health, Nutrition,and Population Services—What Works, WhatDoesn’t, and Why, ed. D. R. Gwatkin, A. Wagstaff, andA.S. Yazbeck. Washington, DC: World Bank.———. 2005b. Rolling Back Malaria: The World BankGlobal Strategy & Booster Program. Washington,DC: World Bank.———. 2005c. The World Bank’s Global HIV/AIDSProgram of Action. Washington, DC: World Bank.———. 2004a. Addressing HIV/AIDS in East Asia andthe Pacific. Washington, DC: World Bank.———. 2004b. “Implementation Completion Report ona Loan in the Amount of US$18.8 million to the Republicof Indonesia for the Intensified Iodine DeficiencyControl Project.” World Bank, Washington, DC———. 2004c. Improving Health, Nutrition, and PopulationOutcomes in Sub-Saharan Africa—The Roleof the World Bank. Washington, DC: World Bank.———. 2004d.Water Resources Sector Strategy: StrategicDirections for World Bank Engagement.Washington, DC: World Bank.———. 2004e. The World Bank Group’s Program forWater Supply and Sanitation. Water Supply andSanitation Sector Board. Washington, DC: WorldBank.———. 2003a. Averting AIDS Crises in Eastern Europeand Central Asia. Washington, DC: WorldBank.———. 2003b. World Development Report 2004: MakingServices Work for Poor People. Washington,DC: World Bank.———. 2002. Cities on the Move: A World Bank UrbanTransport Strategy Review. Washington, DC: WorldBank.———. 2001a. HIV/AIDS in the Caribbean: Issues andOptions. Washington, DC: World Bank.———. 2001b. Kyrgyz Republic: Review of Social Policyand Expenditures. Vol. 1: Overview of Health,Education, and Social Protection Issues. Report No.22354-KZ. Washington, DC: World Bank.———. 2001c. World Development Report 2000/2001:Attacking Poverty. Washington, DC: World Bank.———. 2000a. Intensifying Action against AIDS inAfrica. Washington, DC: World Bank.———. 2000b. World Bank Strategy for Health, Nutrition,and Population in East Asia and the PacificRegion. Washington, DC: World Bank,———. 1999a. A Health Sector Strategy for the Europeand Central Asia Region. Washington, DC:World Bank.———. 1999b. Intensifying Action against HIV/AIDSin Africa. Washington, DC: World Bank.———. 1999c. Population and the World Bank: Adaptingto Change. Washington, DC: World Bank.———. 1998a. Assessing Aid: What Works, WhatDoesn’t and Why. World Bank Policy Research Report.Washington, DC: World Bank.185

IMPROVING EFFECTIVENESS AND OUTCOMES FOR THE POOR IN HEALTH, NUTRITION, AND POPULATION———. 1997a. Confronting AIDS: Public Priorities ina Global Epidemic. Washington, DC: World Bank.———. 1997b. Health, Nutrition, and Population SectorStrategy Paper. Health, Nutrition, and PopulationFamily, Human Development Network. Washington,DC: World Bank.———. 1996a. Poverty Reduction and the WorldBank: Progress and Challenges in the 1990s. Washington,DC: World Bank.———. 1996b. Sustainable Transport: Priorities forPolicy Reform. Washington, DC: World Bank.———. 1996c. World Bank Participation Sourcebook.Washington, DC: World Bank.———. 1994a. Better Health in Africa. Washington, DC:World Bank.———. 1994b. The Broad Sector Approach to InvestmentLending: Sector Investment Programs. Washington,DC: World Bank.———. 1993a. “Peru: Poverty Assessment and SocialPolicies and Programs for the Poor.” World BankReport No. 11191-PE. World Bank, Washington, DC.———. 1993b. Water Resources Management: A WorldBank Policy Paper. Washington, DC: World Bank.———. 1993c. World Development Report 1993: Investingin Health. Washington, DC: World Bank.———. 1989. Sub-Saharan Africa: from Crisis to SustainableDevelopment. Washington, DC: WorldBank.———. 1988. Acquired Immunodeficiency Syndrome(AIDS): The Bank’s Agenda for Action. Washington,DC: World Bank.———. 1987. Financing Health Services in DevelopingCountries: An Agenda for Reform. Washington,DC: World Bank.———. 1986. Poverty and Hunger: Issues and Optionsfor Food Security in Developing Countries.Washington, DC: World Bank.———. 1984. World Development Report 1984: Populationand Development. Washington, DC: WorldBank.———. 1980a. Health Sector Policy Paper. Washington,DC: World Bank.———. 1980b. World Development Report 1980:Poverty and Human Development. Washington,DC: World Bank.———. 1975. Health Sector Policy Paper. Washington,DC: World Bank.———. 1974. Population Policies and Economic Development.Washington DC: World Bank.———. 1973. Nutrition Policy: Policy Guidelines forBank Nutrition Activities. Sector Program Paper.Washington, DC: World Bank.———. 1972a. Possible Bank Actions on MalnutritionProblems. Washington DC: World Bank.———. 1972b. Sectoral Programs and Policies Paper.Washington DC: World Bank.World Bank Group. 2007. Detailed Implementation Review:India Health Sector, 2006–2007. Volume I.Washington, DC: World Bank.———. 2002. “Private Sector Development Strategy:Directions for the World Bank Group.” Washington,DCWHO. 2008. World Malaria Report 2008. WHO: Geneva.———. 2004. World Report on Road Traffic InjuryPrevention. WHO: Geneva.———. 2003. The World Health Report 2003—Shapingthe Future. WHO: Geneva.———. 2002. World Health Report: Attributable Mortalityby Risk Factor, Level of Development andSex, Statistical Annex. WHO: Geneva.———. 2002. The World Health Report 2002.—ReducingRisks, Promoting Healthy Life. WHO:Geneva.———. 2000. The World Health Report 2000: HealthSystems: Improving Performance. WHO: Geneva.WHO/Department of Control of Neglected TropicalDiseases. 2006. “Neglected Tropical Diseases, HiddenSuccesses, Emerging Opportunities.” .WHO and World Bank. 2004. World Report on RoadTraffic Injury Prevention. Geneva: WHO.Xianyi, Chen, Wang Liying, Cai Jiming, Zhou Xiaonong,Zheng Jiang, Guo Jiagang, Wu Xiaohua, D. Engels,and Chen Minggang. 2005. “Schistosomiasis Controlin China: The Impact of a 10-Year World Bank LoanProject (1992–2001).” Bulletin of the World HealthOrganization 83(1): 43–48.Yazbeck, Abdo S. 2009. Attacking Inequality in theHealth Sector: A Synthesis of Evidence and Tools.Washington, DC: World Bank.Zaky, Hassan H.M., Sherine Shawky, Faten Abdel Fattah,and Eman El-Hadary. 2007. “Evaluation of the Impactof the Provider Incentive Payments on Reproductive186

REFERENCESHealth Services: Egypt’s Health Sector Reform Programme.”Report written for the Ministry of Healthand Population, Central Administration for TechnicalSupport and Projects. Funded by the ReproductiveHealth and Research Department, WHO,Geneva.Zimmerman, Michale B., Pieter L. Jooste, and ChandrakantS. Pandav. 2008. “Iodine Deficiency Disorders.”The Lancet 372: 1251–62.Zwane, Alix P., and Michael Kremer. 2007. “What Worksin Fighting Diarrheal Diseases in Developing Countries?A Critical Review.” Center for InternationalDevelopment (CID) Working Paper, No. 140. Centerfor International Development, Harvard University,Cambridge, Massachusetts.187

RECENT IEG PUBLICATIONSAnnual Review of Development Effectiveness 2008: Global ChallengesAddressing the Challenges of Globalization: An Independent Evaluation of the World Bank’s Approach to Global ProgramsAssessing World Bank Support for Trade, 1987–2004: An IEG EvaluationBooks, Building, and Learning Outcomes: An Impact Evaluation of World Bank Support to Basic Education in GhanaBridging Troubled Waters: Assessing the World Bank Water Resources StrategyChina: An Evaluation of World Bank AssistanceThe CGIAR at 31: An Independent Meta-Evaluation of the Consultative Group on International Agricultural ResearchDebt Relief for the Poorest: An Evaluation Update of the HIPC InitiativeA Decade of Action in Transport: An Evaluation of World Bank Assistance to the Transport Sector, 1995–2005The Development Potential of Regional Programs: An Evaluation of World Bank Support of Multicountry OperationsDevelopment Results in Middle-Income Countries: An Evaluation of World Bank SupportDoing Business: An Independent Evaluation—Taking the Measure of the World Bank–IFC Doing Business IndicatorsEngaging with Fragile States: An IEG Review of World Bank Support to Low-Income Countries Under StressEnvironmental Sustainability: An Evaluation of World Bank Group SupportEvaluating a Decade of World Bank Gender Policy: 1990–99Evaluation of World Bank Assistance to Pacific Member Countries, 1992–2002Extractive Industries and Sustainable Development: An Evaluation of World Bank Group ExperienceFinancial Sector Assessment Program: IEG Review of the Joint World Bank and IMF InitiativeFrom Schooling Access to Learning Outcomes: An Unfinished Agenda—An Evaluation of World Bank Support to PrimaryEducationHazards of Nature, Risks to Development: An IEG Evaluation of World Bank Assistance for Natural DisastersHow to Build M&E Systems to Support Better GovernmentIEG Review of World Bank Assistance for Financial Sector ReformAn Impact Evaluation of India’s Second and Third Andhra Pradesh Irrigation Projects: A Case of Poverty Reduction with LowEconomic ReturnsImproving Investment Climates: An Evaluation of World Bank Group AssistanceImproving the Lives of the Poor through Investment in CitiesImproving the World Bank’s Development Assistance: What Does Evaluation Show?Maintaining Momentum to 2015: An Impact Evaluation of Interventions to Improve Maternal and Child Health and NutritionOutcomes in BangladeshNew Renewable Energy: A Review of the World Bank’s AssistancePakistan: An Evaluation of the World Bank’s AssistancePension Reform and the Development of Pension Systems: An Evaluation of World Bank AssistancePoland Country Assistance Review: Partnership in a Transition EconomyThe Poverty Reduction Strategy Initiative: An Independent Evaluation of the World Bank’s Support Through 2003The Poverty Reduction Strategy Initiative: Findings from 10 Country Case Studies of World Bank and IMF SupportPower for Development: A Review of the World Bank Group’s Experience with Private Participation in the Electricity SectorPublic Sector Reform: What Works and Why? An IEG Evaluation of World Bank SupportSmall States: Making the Most of Development Assistance—A Synthesis of World Bank FindingsSocial Funds: Assessing EffectivenessSourcebook for Evaluating Global and Regional Partnership ProgramsUsing Knowledge to Improve Development Effectiveness: An Evaluation of World Bank Economic and Sector Work andTechnical Assistance, 2000–2006Using Training to Build Capacity for Development: An Evaluation of the World Bank’s Project-Based and WBI TrainingWater Management in Agriculture: Ten Years of World Bank Assistance, 1994–2004The Welfare Impact of Rural Electrification: A Reassessment of the Costs and Benefits—An IEG Impact EvaluationWorld Bank Assistance to Agriculture in Sub-Saharan Africa: An IEG ReviewWorld Bank Assistance to the Financial Sector: A Synthesis of IEG EvaluationsThe World Bank in Turkey: 1993–2004—An IEG Country Assistance EvaluationWorld Bank Lending for Lines of Credit: An IEG EvaluationAll IEG evaluations are available, in whole or in part, in languages other than English. For our multilingual section, please visithttp://www.worldbank.org/ieg.

ISBN: 978-0-8213-7542-6