56820 FAST TRACKBRIEF January 28, 2009 The IEG report "Improving Effectiveness and Outcomes for the Poor in Health, Nutrition, and Population," was discussed by CODE on January 28, 2009 Improving Effectiveness and Outcomes for the Poor in Health, Nutrition, and Population The Bank Group now funds a smaller share of global support for health, nutrition, and popula- tion (HNP) than it did a decade ago, but its support remains significant--$17 billion in country- level project financing, in addition to policy advice, analytic work, and engagement in global partnerships by the World Bank and $873 million in private health and pharmaceutical investments by IFC from 1997 to mid-2008. The Bank Group continues to play an important role and add value in HNP. About two-thirds of the Bank's HNP projects show satisfactory outcomes. Performance can be substantially improved by reducing project complexity, strengthening risk assessment and mitigation, conducting more up-front institutional analysis, and incorporating more evaluation to promote evi- dence-based decisions. The performance of IFC health investments, mainly hospitals, has improved markedly, but IFC has had limited success at diversifying its health portfolio. The accountability of Bank Group investments for demonstrating results for the poor has been weak. The Bank's investments often have a pro-poor focus, but their objectives need to address the poor explicitly and outcomes among the poor need to be monitored. Importantly, the Bank needs to increase sup- port to reduce high fertility and malnutrition among the poor and ensure discussion of HNP in poverty as- sessments. IFC-financed hospitals mainly benefit the non-poor; IFC needs to support more activities that both make business sense and yield broader benefits for the poor. The Bank Group has an important role in helping countries to improve the efficiency of health sys- tems. The Bank needs to better define efficiency objectives, track efficiency outcomes, and support better information and vigorous evaluation of reforms. IFC needs to enhance support to public-private partner- ships and improve collaboration and joint sector work with the Bank. The potential for improving HNP outcomes through actions by non-health sectors is great, but incentives to deliver them are weak. Adding HNP objectives to Bank projects in other sectors, such as water supply and sanitation, raises the incentive to deliver health benefits. Strengthening the comple- mentarity of investments in HNP and other sectors can also improve outcomes. In IFC, incentives, in- stitutional mechanisms, and an integrated approach to health are needed to improve coordination across units. S ince the late 1990s, when the World Bank Group was the largest source of HNP finance to developing coun- tries, new aid donors and institutions have emerged and global HNP development assistance has more than The Scope of the Evaluation This evaluation aims to inform the implementation of the World Bank's and IFC's most recent HNP strategies to enhance the effectiveness of future support. It covers the doubled, from an annual average of $6.7 billion in 1997/98 to about $16 billion in 2006. The international community period from FY1997 to FY 2008 and is based on desk re- has adopted the Millennium Development Goals (MDGs) views of the portfolio, background studies, and field visits. and other global targets, with a new emphasis on aid effec- The evaluation of World Bank HNP support focuses on tiveness, results orientation, donor harmonization, align- the effectiveness of policy dialogue, analytic work, and ment, and country leadership, reflected in the 2005 Paris Dec- lending at the country level, while that of IFC focuses on the laration on Aid Effectiveness and the 2008 Accra Agenda for Ac- performance of health investments and advisory services tion. The World Bank Group, now one of many large players before and after its 2002 health strategy. The evaluation in international HNP support, is reassessing its comparative does not cover the Bank's global partnership engagements advantage in the context of the new aid architecture. in HNP, some of which have been previously evaluated, except to the extent that they are reflected in country-level World Bank Group support to HNP remains substantial. support. The evaluation benefitted from feedback from an From 1997 through mid-2008, the World Bank (IBRD and Advisory Panel of international experts. IDA) committed nearly $17 billion to 605 HNP projects in more than 120 countries, sponsored analytic work, and Bank Support to the Public Sector for HNP offered policy advice. This support aimed to improve Over the review period, the World Bank directly supported health and nutrition status and reduce high fertility; im- HNP outcomes in countries through lending and nonlending prove the access, quality, efficiency, and equity of the services. Most lending was from projects managed within the health system; reform health systems through changes in Bank's HNP sector units ($11.5 billion, 255 projects). Beyond health finance, support for health insurance, decentraliza- this, about $5 billion in lending for HNP outcomes was ma- tion, engaging the private sector, and other structural naged by other sectors. Since FY00 the Bank also spent $43 changes; and strengthen institutional capacity and sector million of its own budget and trust funds on HNP-related management. In addition, as of 2007 the Bank was engaged economic and sector work (ESW). Professional HNP staff financially in 19 global health partnerships and participating grew by a quarter, as did the share of health specialists. in 15 more. IFC has financed 68 private investment projects in the health and pharmaceutical sectors of devel- The World Bank's Role oping countries--$873 million in total commitments--and offered advisory services on health to the private sector, Although the World Bank finances a smaller share of including support for public-private partnerships. country-level development assistance under the new international aid architecture, it has an important role The World Bank's 2007 strategy, Healthy Development: The and significant potential to add value. The World Bank World Bank Strategy for Health, Nutrition, and Population Results, brings important institutional assets to bear in helping aims among other things, to improve HNP outcomes on countries make health systems work better and ensuring average and among the poor, prevent poverty due to ill- that health benefits reach the poor: long-term, sustained ness, improve health system performance, and enhance engagement in the sector; international experience; a histo- governance, accountability, and transparency in the sector. ry of support for building country capacity to implement It points to several strategic directions or actions for the programs; large-scale, sustained financing; strong links to Bank to achieve the objectives, among them: a renewed finance ministries; and engagement with many sectors other focus on HNP results; efforts to help countries improve the than health with potential to contribute to HNP outcomes. performance of health systems and to ensure synergy with priority However, the Bank's comparative advantage in a country is disease interventions, particularly in low-income countries; and context-specific, depending on health conditions, govern- strengthened Bank capacity to advise countries on intersec- ment priorities and resources, and the activities of other toral approaches to improving HNP results. development partners. To deliver on its comparative ad- vantages, the Bank needs to improve the performance of its The 2002 IFC health strategy defines the sector's goals to country-level support. improve health outcomes, protect the population from impoverishing effects of ill health, and enhance the per- The Evolution and Performance of World Bank formance of health services. The strategy has both business and developmental objectives, including promoting effi- Support ciency and innovation in the health sector, and calls for While the overall level of HNP project approvals increasing the social impact of IFC investments. changed little, the composition of the lending portfolio saw some major shifts. The number of new HNP- managed projects rose slowly, but new financial commit- 2 ments declined. The share of communicable disease supervision needs to be intensified to ensure that civil projects doubled over the decade, reaching about 40 per- works and equipment are delivered as specified, in working cent of approvals in the second half of the period, as did order, and functioning. the share of multisectoral projects, reaching half of all ap- provals. The share of Africa region projects also increased. Accountability of projects for delivering health results to These three trends were due primarily to an increase in the poor has been weak. Studies of the incidence of public multisectoral AIDS projects. Projects supporting sector- expenditure have shown that in most countries, public health wide approaches (SWAps) in health rose to a cumulative spending favors the non-poor; expansion of services cannot total of 28 operations in 22 countries, about 13 percent of be assumed to improve access of the poor relative to the non- the project portfolio. In contrast, the share of lending with poor. While many projects targeted HNP support to geo- objectives to reform the health system dropped by nearly graphic areas with a high incidence of poverty (including rural half. areas) or financed services or addressed problems thought to disproportionately affect the poor, only 6 percent of all HNP Attention to population and malnutrition was low; sup- projects committed to deliver better health or nutrition among port for population nearly disappeared. About 1 in 10 the poor in their statement of objectives, for which they were projects had an objective to reduce malnutrition, which dis- ultimately accountable. A third of projects with objectives to proportionately affects the poor, but the share of projects improve general health status had no targeting mechanism for with nutrition objectives dropped by half over the decade. reaching the poor. Among closed projects with objectives to About two-thirds of nutrition projects were in countries with improve HNP outcomes among the poor, most measured a high levels of child stunting, but Bank nutrition support change in average HNP status in project areas. Very few ac- reached only about a quarter of all developing countries with tually measured whether poor individuals or poor project areas high stunting. Lending to reduce high fertility or improve benefited in relation to the non-poor or in relation to those in access to family planning accounted for only 4 percent of the non-project areas, and even fewer showed that the poor did lending portfolio, dropping by two-thirds over the decade. disproportionately benefit. In some cases, improvements in Population support was directed to only about a quarter of HNP status were only measured at the national level. the 35 countries the Bank identified as having fertility rates of more than five children per woman. Analytic work and The Bank delivered several high-profile analytic products staffing to support population and family planning objectives on HNP and poverty in the past decade--notably the nearly disappeared. Substantial analysis of high fertility and Reaching the Poor with Health, Nutrition, and Population Services malnutrition rarely figured in poverty assessments, though project and the World Development Report 2004: Making Servic- both are most acutely felt by the poor. es Work for Poor People. Nevertheless, the share of country poverty assessments with substantial discussion of health Two-thirds of HNP projects had satisfactory out- declined, from 80 percent in FY00-03 to only 58 percent in comes, and the portfolio's performance stalled. Field FY04-06. Only 7 percent of poverty assessments had sub- assessments found that support for reducing malaria in stantial discussion of population, and major discussion of Eritrea and schistosomiasis in Egypt, raising contraceptive nutrition declined by half, from 28 to 12 percent. About a use in pilot areas of Malawi, and reforming the health sys- quarter to a third of Bank HNP analytic work was poverty- tem in the Kyrgyz Republic, for example, showed good related, also a decline. results. However, about a third of the HNP lending portfo- lio did not perform well, a share that has remained steady An increasing share of projects had monitoring indica- while performance in other sectors has improved over the tors and baseline data at appraisal, but overall monitor- decade. Only one in four HNP projects in Africa achieved ing remains weak and evaluation almost nonexistent, satisfactory outcomes. Complex projects--multisectoral presenting a challenge for the HNP strategy's results projects and SWAps--in low-capacity environments were orientation and commitment to better governance. Al- least likely to achieve their objectives. However, health though nearly a third of projects supported pilot interventions reform projects in middle-income countries also performed or programs, or intended to evaluate the impact of a specific less well and are complex and politically volatile. Poor- activity or program, few proposed evaluation designs in ap- performing projects displayed common characteristics: in- praisal documents and even fewer evaluations were actually adequate risk analysis or technical design, inadequate su- conducted. Pilot activities without an evaluation design de- pervision, insufficient political or institutional analysis, lack scribed in the appraisal document were never evaluated. There of baseline data on the basis of which to set realistic targets, were consequences of poor M&E and absence of baseline overly complex designs in relation to local capacity, and data: irrelevant objectives and inappropriate project designs; negligible monitoring and evaluation (M&E). These prob- unrealistic targets--either too high or below the baseline val- lems are similar to those cited in IEG's 1999 evaluation of ue; inability to assess the effectiveness of activities; and lower the HNP sector. The results of the recent Detailed Imple- efficacy and efficiency because of limited opportunities for mentation Review of HNP projects in India suggest that, learning. even among projects that achieve their objectives, field 3 Approaches for Improving HNP Outcomes to prevent the impoverishing impacts of illness. Bank sup- port for health reforms has been mainly to middle-income The evaluation reviewed findings and lessons for a number countries, where health reform projects represent about half of prominent approaches to raising HNP outcomes-- of the portfolio. communicable disease control, health reform, SWAps, and intersectoral action. These approaches have been supported Many lessons have been learned over the past decade about by the Bank and the international community and are not the successes and pitfalls of support for health reform. First, mutually exclusive. the failure to assess fully the political economy of reform and to prepare a proactive plan to address it can considerably Support for communicable disease control can improve diminish prospects for success. Political risks, the interests of the pro-poor focus of health systems, but excessive key stakeholders, and the risk of complexity are often neg- earmarking of foreign aid for communicable diseases lected in risk analysis in project appraisal documents for can distort allocations and reduce capacity in the rest of health reform projects. Second, reforms based on careful the health system. One of the strategic directions of the prior analytic work hold a greater chance of success, but ana- 2007 HNP strategy is to ensure synergy between priority lytic work does not ensure success. Third, the sequencing of disease interventions and strengthening of the health system. reforms can improve political feasibility, reduce complexity, The rationale for investing in infectious diseases is that they ensure that adequate capacity is in place, and facilitate learn- disproportionately affect the poor, their control has large ing. When implementation is flagging, the Bank can help positive externalities, and interventions have been shown to preserve reform momentum with complementary program- be cost-effective in many settings. Dedicated communicable matic lending, as it did in Peru and Kyrgyz Republic. Finally, disease projects have dramatically increased as a share of the M&E are critical in health reform projects--to demonstrate lending portfolio over the past decade, and Bank support has the impact of pilot reforms for garnering political support contributed to country capacity in national disease control but also because many reforms cannot work without a well- programs. Support for control of communicable diseases, functioning information system. with the exception of AIDS, has shown better outcomes than the rest of the HNP portfolio. Both equity and cost- SWAps have contributed to greater government leader- effectiveness are particularly important to address in AIDS ship, capacity, coordination, and harmonization within programs, given the large commitments to that disease and the health sector, but not necessarily to improved effi- the fact that HIV does not always disproportionately strike ciency or better health results. Sectorwide approaches the poor. represent a reform in the way that government and interna- tional donors work together (the approach) to support the Since the initial increase in Bank-supported communicable achievement of national health objectives (the program). The disease control in the early 2000s, mainly for AIDS, the in- overwhelming focus of SWAps supported by the Bank has ternational community has also generously expanded funding been on setting up and implementing the approach. Field- through the Global Fund to Fight AIDS, TB, and Malaria work found that country capacity has been strengthened in and the (U.S.) President's Emergency Plan for AIDS Relief the areas of sector planning, budgeting, and fiduciary sys- (PEPFAR), other bilateral contributions, and private founda- tems. However, weaknesses persist in the design and use of tions. In some low-income countries with high HIV preva- country M&E systems; evidence that the approach has im- lence, earmarked AIDS funds from international partners proved efficiency or lowered transaction costs is thin because account for 30 or 40 percent or more of all public health neither has been monitored. funding. Where human resource capacity within the health system is scarce, the allocation of resources across health Adopting the approach does not necessarily lead to better programs and budget lines needs to be balanced, to ensure implementation or efficacy of the government's health pro- that large earmarked funds for specific diseases do not result grams: only a third of Bank projects that supported health in lower efficiencies or reduced care elsewhere in the health SWAps have had satisfactory outcomes. SWAps have often system. There is little evidence that this issue has been consi- supported ambitious programs with many complex reforms dered in recent funding decisions or in risk analysis. and activities that exceed government implementation capac- ity. Important lessons are that programs need to be realistic Health reforms promise to improve efficiency and go- and prioritized and that the process of setting up the SWAp vernance, but they are politically contentious, often should not distract from ensuring the implementation and complex, and relatively risky. About a third of HNP efficacy of the overall health program and a focus on results. projects have supported reform or restructuring of the health SWAps have been most effective in pursuing health program system through changes in health finance, development of objectives when the government is in a leadership position health insurance, decentralization of health systems, and reg- with a strongly owned and prioritized strategy (as in the Kyr- ulation or engagement of the private health sector. These gyz Republic). When this is not the case, there is a risk that objectives affect efficiency and governance, which are valid the health program will be less prioritized reflecting the fa- objectives in their own right. Health insurance reforms aim 4 vored elements of the diverse partners, weakening effective- tor at IFC, involving both the Health and Education and ness (as in Ghana). General Manufacturing departments (for pharmaceuticals). The contribution of other sectors to HNP outcomes has The performance of IFC's health investments, mostly been largely undocumented; the benefits of intersectoral hospitals, improved substantially. Before 1999, four-fifths coordination and multisectoral approaches need to be of all health investments performed poorly, contributing to balanced with their costs in terms of increased complexi- financial losses. The reasons included the impact of financial ty. The contribution of other sectors to HNP outcomes has crises, delays in obtaining regulatory clearances from the been captured through multisectoral HNP projects (projects authorities, and IFC's weaknesses in screening and structur- that engage multiple sectors in a single operation) and parallel ing health sector deals owning to lack of sector-related expe- lending in projects managed by other sectors, in some cases rience. More recent investments have realized good financial with explicit health objectives. Multisectoral HNP operations returns and performed better on achieving intended devel- have risen from a quarter to half of HNP lending, increasing opment outcomes. An evaluative framework for IFC's Advi- the complexity of the portfolio. Most of the increase is due to sory Services was only recently launched, so very few health multisectoral AIDS projects. The large number of sectors in- Advisory Services projects have been evaluated. However, volved, the lack of specificity in design documents about their the few that have been evaluated have performed lower than roles and responsibilities, the relatively new institutions put in the IFC portfolio overall. charge, and other factors affecting lower performance in Afri- ca all contribute to lower outcomes for multisectoral AIDS IFC has not been able to diversify its health portfolio projects. Multisectoral HNP projects with fewer implementing as envisaged. In 2002, the sector set objectives to diversify agencies have maintained stronger intersectoral collaboration the portfolio beyond hospitals and improve the social impact and better outcomes. of IFC health operations. IFC has continued to finance pri- vate hospitals; the share of pharmaceuticals and other life Both the 2007 HNP strategy and its predecessor foresaw sciences investments has grown, but more slowly than envi- Country Assistance Strategies as the instrument to coordi- sioned in the strategy. IFC has also financed public-private nate intersectoral action for HNP outcomes. However, over partnerships (PPPs) in health and expanded health Advisory the past decade lending in diverse sectors like water supply Services with a focus on Africa. Investment numbers and and sanitation (WSS) and education for the most part has volume have increased since 2005. However, IFC has not been pursued independently, although this does not mean financed any health insurance ventures and has financed only that they have not contributed to health outcomes. one project on medical education. Lending in other sectors may contribute directly to HNP IFC's health interventions have had limited social im- outcomes, in some cases by including health objectives or pact, although efforts are being made to broaden those health components, or indirectly. For example, half of all impacts. IFC's investments in hospitals have targeted WSS projects claim that health benefits will be generated, middle- and upper-income groups. Linkages to public in- and one in 10 has an objective to improve health outcomes. surance schemes will be necessary for IFC-supported hos- However, fewer WSS projects include health objectives to- pitals to meet the health needs of a wider population. Ex- day than was the case 5-10 years ago. Interviews with staff panded support to PPPs jointly with the World Bank, such suggested that the sector has focused primarily on what is as a recent output-based aid project to improve maternal perceived to be "their" MDG, namely increased access to care among some of Yemen's poorest people, and more safe water. Yet research has shown that context matters; strategic deployment of Advisory Services, such as to assist better access to safe water does not necessarily translate into social enterprises in Kenya and India, could help broaden better health. In contrast, the health content of transport the social impact of health investments. projects has greatly increased, particularly in the field of road safety and HIV/AIDS prevention. Trends in accident statis- Recent IFC health projects have had positive results for effi- tics have been relatively well documented for road safety ciency, governance, and affordability. State-of-the-art facili- components; the outputs and results for HIV/AIDS com- ties in some IFC-supported projects have attracted profes- ponents have not. Delivery of health results in non-health sionals with established successful careers in developed sectors has been generally weak except when an explicit countries. Many hospitals supported by IFC have posted fees health objective was identified at project appraisal. and introduced control of doctors' side practices outside of the institutions. Most IFC-supported pharmaceutical projects IFC Support for Development of the Private have resulted in significant declines in the prices of generic drugs, thus enhancing affordability. Health Sector IFC has made support to private health investment one of its The need to collaborate closely with the World Bank's HNP strategic priorities. Health is a relatively small and recent sec- sector is recognized as important in both IFC and World Bank strategies to promote greater efficiency in the health 5 sector through finance of private health care. The evaluation disease activities in countries where other donors are found some interaction, particularly in middle-income coun- contributing large amounts of earmarked disease fund- tries, but there is no real model of how that collaboration ing and additional earmarked funding could result in dis- should occur in a situation where IFC health activities are tortions in allocations and inefficiencies in the rest of the few and very small in relation to the entire World Bank health system; and supporting improved health informa- Group HNP sector in a country. tion systems and more frequent and vigorous evaluation of specific reforms or program innovations to provide Recommendations timely information for improving efficiency and efficacy. For IFC this means supporting PPPs through Advisory · Intensify efforts to improve the performance of the Services to government and industry and through its in- World Bank's health, nutrition, and population sup- vestments; expanding investments in health insurance; port. This calls for matching project design to country and improving collaboration and joint sector work with context and capacity and reducing the complexity of the World Bank, leveraging Bank sector dialogue on projects in low-capacity settings, through greater selectivi- health regulatory frameworks for health to engage new ty, prioritization, and sequencing of activities; carefully as- private actors with value added to the sector, and more sessing the risks of proposed HNP support and strategies systematically coordinate with the Bank's policy inter- to mitigate them, particularly the political risks and the in- ventions regarding private sector participation in health. terests of different stakeholders, and how they will be ad- dressed; phasing reforms to maximize the probability of · Enhance the contribution of support from other success; undertaking thorough institutional analysis, in- sectors to health, nutrition, and population out- cluding an assessment of alternatives, as an input into comes. The World Bank needs to incorporate health ob- more realistic project design; and supporting intensified jectives for which they are accountable into non-health supervision in the field by the Bank and the borrower to projects when the benefits are potentially great in rela- ensure that civil works, equipment, and other outputs tion to the marginal costs; improve the complementarity have been delivered as specified, are functioning, and be- of investment operations in health and other sectors to ing maintained. achieve HNP outcomes, particularly between health and WSS; prioritize sectoral participation in multisectoral · Renew the commitment to health, nutrition, and HNP projects according to the comparative advantages population outcomes among the poor. For the World and institutional mandates, to reduce their complexity; Bank, this means boosting population, family planning, identify new incentives for Bank staff to work cross- and other support in the form of analytic work, policy sectorally for improving HNP outcomes; and develop dialogue, and financing to high fertility countries and mechanisms to ensure that the implementation and re- countries with pockets of high fertility; incorporating the sults for small HNP components retrofitted into poverty dimension into project objectives to increase projects are properly documented and evaluated. IFC accountability for health, nutrition, and population out- should improve incentives and institutional mechanisms comes among the poor; increasing support to reduce for an integrated approach to health issues across units malnutrition among the poor, whether originating in the in IFC dealing with health, including the way IFC is or- HNP sector or other sectors; monitoring health, nutri- ganized. tion, and population outcomes among the poor, howev- er defined; and bringing the health and nutrition of the · Implement the results agenda and improve gover- poor and the links between high fertility, poor health, nance by boosting investment in and incentives and poverty back into poverty assessments. For IFC, for evaluation. The World Bank needs to create new this means expanding support for innovative approach- incentives for M&E for both the Bank and the bor- es and viable business models that demonstrate private rower linked to the project approval process and the sector solutions to improve the health of the poor, in- midterm review. This would include requirements for cluding expansion of investments in low-cost generic baseline data, explicit evaluation designs for pilot activ- drugs and technologies that address health problems of ities in project appraisal documents, and periodic eval- the poor; and assessing the external and internal con- uation of main project activities as a management tool. straints in achieving broad social impacts. IFC needs to enhance its results orientation by devel- oping clearly specified baseline indicators and an eval- · Strengthen the World Bank Group's ability to help uation framework that adequately measure its health countries improve the efficiency of health systems. sector objectives and results. For the World Bank this means better defining the effi- ciency objectives of its support and how efficiency will be improved and monitored; carefully assessing deci- sions to finance additional earmarked communicable 6 About Fast Track Briefs Fast Track Briefs help inform the World Bank Group (WBG) managers and staff about new evaluation findings and recommendations. The views expressed here are those of IEG and should not be attributed to the WBG or its affi- liated organizations. Management's Response to IEG is included in the published IEG report. The findings here do not support any general inferences beyond the scope of the evaluation, including any inferences about the WBG's past, current or prospective overall performance. The Fast Track Brief, which summarizes major IEG eval- uations, will be distributed to World Bank Group staff. If you would like to be added to the subscription list, please email us at ieg@worldbank.org, with "FTB subscription" in the subject line and your mail-stop number. If you would like to stop receiving FTBs, please email us at ieg@worldbank.org, with "FTB unsubscribe" in the subject line. Contact IEG: Director-General, Evaluation: Vinod Thomas Directors: Marvin Taylor-Dormond (IEG-IFC) Christine Wallich (IEG-MIGA) Cheryl Gray (IEG-WB) Task Managers: Stoyan Tenev (IEG-IFC) Hiroyuki Hatashima (IEG-IFC) Martha Ainsworth (IEG-WB) Copies of the report are available at: http://www.worldbank.org/ieg/hnp IEG Help Desk: (202) 458-4497 E-mail: ieg@worldbank.org 7