Pro-poor health policies in poverty reduction strategies

HEALTH POLICY AND PLANNING; 18(2): 138 145 Health Policy and Planning 18(2), doi: 10.1093/heapol/czg018 Oxford University Press 2003, all rights reserved. Pro-poor health policies in poverty reduction strategies LEONTIEN LATERVEER 1, LOUIS W NIESSEN 2 AND ABDO S YAZBECK 3 1 ETC Crystal, Leusden, the Netherlands, 2 Institute for Health Care Policy and Management/Institute for Medical Technology Assessment, Erasmus MC, Erasmus University, Rotterdam, the Netherlands and 3 Health and Population Team, The World Bank Institute, Washington, DC, USA Since 1999, the International Monetary Fund and World Bank have required low-income countries soliciting for debt relief and financial support to prepare a Poverty Reduction Strategy Paper (PRSP). The objective of this study is to arrive at a systematic assessment of the extent to which the first batch of interim PRSPs actually addresses the health of the poor and vulnerable. A literature study was used to design and test a semi-quantitative approach to assess the pro-poor focus of health policies in national documents. The approach was applied to the existing interim proposals for 23 Highly Indebted Poor Countries. Results show that a majority of proposals lack country-specific data on the distribution and composition of the burden of disease, a clear identification of health system constraints and an assessment of the impact of health services on the population. More importantly, they make little effort to analyze these issues in relation to the poor. Furthermore, only a small group explicitly includes the interests of the poor in health policy design. Attention to policies aiming at enhancing equity in public health spending is even more limited. Few papers that include expenditure proposals also show pro-poor focused health budgets. We conclude that our systematic assessment of a new international development policy instrument, PRSP, raises strong concerns about the attributed role of health in development and the limited emphasis on the poor, the supposed primary beneficiaries of this instrument. There is a need and an opportunity for the international development community to provide assistance and inputs as poor countries shift their policy thinking from an interim stage to fully developed national policies. This paper presents a menu of analytical and policy options that can be pursued. Key words: poverty reduction, equity, public health expenditure, developing countries, pro-poor, PRSP Introduction In 1999, the International Monetary Fund and World Bank announced a renewed set of conditions for debt relief and financial assistance (IMF/World Bank 1999a; World Bank 2000). Consequent to the new framework, poverty reduction itself has become the main development policy focus. Developing countries are required to formulate a so-called Poverty Reduction Strategy Paper (PRSP) (IMF/World Bank 1999a, b, c; World Bank 2000; IMF/World Bank 2001). A fully developed PRSP should present the country s macroeconomic, structural and social policies for achieving economic growth and poverty reduction, describe the external financing needs and the major sources of finance, and show a participatory approach. Developing countries participating in the Heavily Indebted Poor Countries (HIPC) Initiative are those expected to first present a PRSP. In the future, other low-income countries will follow (World Bank 2001a). Countries unable to present a full PRSP at the time of their HIPC-Initiative decision points or new lending arrangements will receive assistance based on an interim PRSP (iprsp). The iprsp should describe the government s commitment to poverty reduction and a participatory approach. In addition, it should sketch a 3-year macroeconomic framework and a policy matrix including the main elements of the poverty reduction strategy (World Bank 2000). As only a limited number of full PRSPs were completed at the time of the review for this paper (early 2001), the focus of the analysis and the findings is on the iprsps of 23 heavily indebted countries. Moreover, since an iprsp is intended to be a building block for a more consultative and analytical full PRSP, assessing iprsps is an upstream opportunity to influence the development of the full national policy documents. The health impact of the policy shift to PRSPs is under debate (Eurodad 2000; Verheul and Cooper 2001; Verheul and Rowson 2001). The current study is the first to systematically explore and assess possible pro-poor health policies that can be used in the formulation of PRSPs (Laterveer 2001). We focus on the distribution of public health expenditures and evaluate the iprsps against the specific options selected. We have not studied the PRSP-formulation process nor included other national health policy documents. Methods We reviewed the literature to define concepts of distributive equity and equity policy instruments (Laterveer 2001). Based on this review, we have defined equity as distributing public health expenditures according to health care needs (Culyer and Wagstaff 1992; Wagstaff and Van Doorslaer 1993; Claeson et al. 2001a). In this study, equity is interpreted in vertical terms. It is assumed that the poor have a worse health

Pro-poor health policies and PRSPs 139 status than the rich and therefore have more health care needs. Vertical equity in this case implies that more resources should be spent on the poor (Claeson et al. 2001a). Distribution is defined here as the flow of public health expenditures within the health sectors of developing countries, as proposed by the iprsps. We selected 10 health policy options for enhancing distributive equity, which we divided into three main categories (Birdsall and Hecht 1995; Gwatkin 2000a; Claeson et al. 2001b). The first two categories draw heavily on Gwatkin (2000a). The main difference between his approach and ours is that we specifically focused on public health expenditures. The first category comprises those health policies that seek to directly benefit particular individuals that have been identified as poor. Direct or individual targeting is defined here as distributing public health expenditures directly to the poor, for example by means of direct health subsidies, fee waivers and/or exemptions. The second category consists of health policies that seek to target particular groups or programmes, rather than directly targeting poor individuals. Targeting by age is defined as distributing public health expenditures to (health programmes focusing on) poor youth. Targeting by disease applies to the distribution of public health expenditures to the diseases that are most commonly found among the poor, i.e. communicable diseases (Gwatkin and Guillot 1999). Level of care delivery targeting is defined as distributing public health expenditures to the health services that predominantly benefit the poor, such as basic health services. Geographic targeting, excluding urban/rural is defined as distributing public health expenditures to the poorest geographic regions. Similarly, urban/rural targeting applies to the distribution of public health expenditures to urban slums and/or rural areas, i.e. the zones in which the poor tend to cluster. Finally, the third category comprises those health policies that we consider to enhance the impact of the previously mentioned policies. The first option in this category is to distribute public health expenditures so as to increase the number of health personnel working for health services that predominantly benefit the poor. The second option is to distribute public health expenditures so as to improve the geographic distribution of health personnel. The third option is to increase public spending on health. Finally, the last option, closely related to the third, is to mobilize additional public resources for health. The selected targeting types show a clustering and overlapping of the real problems among the poor. However, we listed a number of stand-alone policy options and looked at the iprsps to see if these options were considered at all. We reviewed the iprsps to assess formulated and operationalized health policies. A recording form (checklist) was developed to conduct a quantitative document analysis (Laterveer 2001). Our approach focused on three main issues: (i) the extent to which the iprsps address health (from a pro-poor perspective), (ii) the extent to which the iprsps address (the distribution of) public health expenditures, and (iii) the health policies the iprsps propose for enhancing equity. Table 1 provides an overview of these Table 1. Recording form Main issues Items (i) Extent to which iprsps address: Health from a pro-poor perspective Poverty data Poverty-related health data: - Studies analyzing burden of disease among the poor - Studies analyzing pro-poor health services - Studies analyzing health system constraints for the poor Describing burden of disease among the poor Describing health system constraints for the poor Health sector strategy referring to the poor Consultation of the poor (participation in health) Monitoring and evaluating impact of health strategy on the poor (ii) Extent to which iprsps address: a. Public health expenditures (PHE) PHE-reviews Providing data on current and projected PHE b. Distribution of PHE Providing data on current distribution of PHE Analyses of current distribution of PHE (e.g. benefit incidence analyses) (iii) Proposed health policies for enhancing equity: Selected policy options Direct targeting Targeting by age Targeting by disease Level of care delivery targeting Increasing health personnel working for pro-poor services Geographic targeting, excluding urban/rural Urban/rural targeting Improving geographic distribution of health personnel Increasing public health spending Mobilizing additional public resources for health

140 Leontien Laterveer et al. issues, including the specific items formulated to operationalize them. The study s pro-poor focus is reflected by the issues and items shown in Table 1, but also by the categories corresponding with these items, i.e. the recording form answers. To reflect the intensity and focus of the formulated national health policies, we made a distinction between three categories. The first category applies to explicit statements, i.e. statements in which the iprsps explicitly refer to the poor. The second category applies to implicit statements, i.e. statements that can be considered pro-poor, but in which there is no explicit link to the poor. Finally, the third category applies to the absence of information on the (distributional) health focus. Next, we graded the recording form based on the perceived relevance of the three main issues for the health policy focus of this study. Furthermore, to reflect the difference between explicit statements, implicit statements and absence of information, we allocated the highest scores to categories applying to explicit statements. Categories applying to implicit statements received lower scores. Categories applying to the absence of information received zero scores. At the start of the document analysis, 28 iprsps had been made available on the World Bank s PRSP-website. We decided to limit our focus to those 23 countries with HIPC status (World Bank 2001b), for two reasons. The first and most important reason is that HIPC countries are actually required to address health issues in their (i)prsps. Secondly, as stated earlier, developing a poverty reduction strategy is most immediately on the agenda of HIPC countries (World Bank 2001a). The 23 countries are: Benin, Bolivia, Cameroon, Central African Republic, Chad, Gambia, Ghana, Guinea, Guinea Bissau, Guyana, Honduras, Kenya, Madagascar, Malawi, Mali, Mozambique, Nicaragua, Niger, Rwanda, São Tomé and Príncipe, Senegal, Tanzania and Zambia. Results The review highlights a number of trends that raise concerns about the process and the potential returns to the health of the poor. We summarize the basic findings for (1) the extent to which health issues are approached from a pro-poor perspective, (2) the attention given to public health expenditures, and (3) the proposed policies for enhancing distributive equity. In addition, we present four typical country profiles based on the main evaluation criteria. Pro-poor health perspective Using poverty data Starting with basic poverty data, all countries recognized the importance of collecting and using such information. Poverty studies, such as household surveys, are mentioned by all but one iprsps. Nearly all countries provide information on the distribution of poverty across the population. While all countries pay attention to the urban/rural distribution of poverty, not all provide data on the geographic distribution of poverty. Using poverty-related health data An area for concern is the effort put into collection and use of poverty-related health data. Although 57% of the documents state that health surveys were conducted, the poor are rarely mentioned. Similarly, little attention is paid to data collection regarding the impact of health services and the prevailing health system constraints. More importantly, the iprsps do not outline how the preparation of the full PRSP will address this weakness. Similarly, none of the iprsps refer to studies analyzing the burden of disease among the poor. None of the papers explicitly identify the burden of disease among the poor, even though about half describe the general population health status using country averages. It should be noted that these iprsps do frequently mention those diseases that are known to be most commonly found among the poor, such as HIV/AIDS, malaria and tuberculosis (Gwatkin and Guillot 1999). However, they do not explicitly identify these diseases as such. Documenting health system constraints Turning next to health system performance, the assessment looked at the extent to which these documents used analytical efforts to identify system constraints impacting the poor. Only a quarter of the papers (26%) explicitly describe why their health systems fail the poor and just one paper based its findings on actual studies examining the health system constraints for the poor. When references were made to health system constraints (43%), regardless of the poverty focus, the most frequently mentioned constraints were the lack of drugs, medical supplies and health personnel. Health sector strategies and the poor Each iprsp reviewed proposed a health sector strategy. Not surprisingly, these strategies show substantial variation in degrees of development and specificity. Consequently, for some countries, the term strategy may actually be too lofty. Although 95% have formulated one or more specific policy measures, only one-third included a detailed time-line for the implementation of the strategy. Up to 70% of the iprsps referred to the poor in their health sector strategy. However, this by no means implies that all of these strategies are actually pro-poor. Some iprsps made vague references to what they call vulnerable population sections, while others explicitly stated a focus on improving the health of the poor. Specific health sector policies The assessment included some specific pro-poor strategy components in the analysis. The most frequently mentioned policy components were disease targets, i.e. efforts taken to reduce the burden of disease (96%). Targeting specific population groups, such as women, children or the rural population, came next (78%). Policies to improve coverage and quality of health services were mentioned by 74%, and improving health indicators and access to health services by 65%. Next came policies for improving the procurement and delivery of drugs and medical supplies (57%), followed by improving the geographic distribution of health services and

Pro-poor health policies and PRSPs 141 the recruitment and training of health personnel (both 48%). Finally, 43% mentioned policies for refocusing public health spending, such as enhancing efficiency or equity. Additional policies mentioned included improving the management of health services, decentralizing the health system, and defining mechanisms for the contracting out of health services to non-governmental actors. Some iprsps either explicitly or implicitly related the above policies to the poor; for example, they proposed policies for reducing the burden of disease among the poor or for improving the delivery of pro-poor health services, such as basic health services and essential drugs. Finally, all countries mentioned some sort of community participation in the poverty reduction process. Only 13%, however, indicated that the poor have been or will be involved in the formulation, implementation and further development of the health sector strategy. Similarly, all countries recognized the importance of monitoring and evaluating the poverty reduction strategy, yet only few paid attention to specifically tracking the effects of the health sector strategy for the poor. Public health expenditures While all iprsps addressed public health expenditures, there was considerable variability in the level of detail provided. Almost three-quarters of the documents addressed the distribution of public health expenditures to some degree. Of the 23 documents, 22% indicated the intention of providing more detailed information in the full PRSP. One paper did not address the issue of distribution in spending at all. Of concern was the finding that none of the documents specifically mentioned quantitative or qualitative studies that have actually analyzed current distribution patterns. Proposed health policies for enhancing distributive equity Of the iprsps that addressed the distribution of public health expenditures, only a quarter explicitly included the intention to make the distribution of expenditures more responsive to the needs of the poor. We assessed the papers on the policy options described earlier. A distinction was made between (1) stating a health policy objective, (2) merely proposing to distribute public health expenditures so as to realize this objective, and (3) showing the amount of expenditures budgeted for this purpose. The results are presented in Table 2. Health policy objectives About one-third of the iprsps explicitly stated to focus on the poor in the context of health, i.e. direct targeting. This equals 50% of the previously mentioned papers that refer to the poor in their health sector strategies. Thirty per cent indicated a pro-poor approach without explicitly stating to target the poor. In terms of explicit policy objectives, direct targeting ranks second. First is level of care delivery targeting: 52% explicitly stated to focus on providing the health services that predominantly benefit the poor. Another 48% included policy objectives mentioning health services that are known to benefit the poor, but did not explicitly link these to the poor. Third was urban/rural targeting (26%). Another 22% proposed to improve the urban/rural distribution of health services, without explicitly stating to target these areas. As can be observed from Table 2, the remaining policy options were mentioned less frequently. Objectives corresponding with disease-based targeting were found in 17% of the iprsps. However, a relatively large proportion (78%) showed a focus on the diseases that are known to be most commonly found among the poor, without explicitly identifying these as such. Increasing the personnel working for propoor services was also mentioned by 17%. About one-third of the papers proposed to increase the number of health personnel, but did not explicitly state to do so for the benefit of pro-poor services. Only 9% explicitly included health policy objectives corresponding with age-based targeting, geographic targeting and Table 2. Proportion of iprsps that propose health policies for enhancing distributive equity (percentages) Health policy options Health policy objectives Distributing public health Costing for enhancing equity (n = 23) expenditures to realize objectives (n = 23) Explicit Implicit No Yes No Yes No Direct targeting 35 30 35 22 78 0 100 Targeting by age 9 78 13 4 96 0 100 Targeting by disease 17 78 4 4 96 100 0 Level of care delivery targeting 52 48 0 30 70 43 57 Increasing personnel for pro-poor 17 35 48 9 91 50 50 services Geographic targeting, excluding 9 26 65 9 91 50 50 urban/rural Urban/rural targeting 26 22 52 17 83 25 75 Improving geographic distribution 9 9 83 9 91 100 0 of personnel

142 Leontien Laterveer et al. improving the geographic distribution of health personnel. But 78% did include policy objectives focusing on (health programmes for) the young, without linking these objectives to the poor. Twenty-six per cent stated to improve the geographic distribution of health services, but did not explicitly refer to the poorest regions. Overall, improving the geographic distribution of health personnel received the least attention. The conclusion is that the number of iprsps explicitly including the interests of the poor in health policy design is rather low. However, the relatively high proportions in the category implicit policy objectives indicate that, in some areas, the iprsps can be considered to favour the poor. Distributing public health expenditures This area comprises the health policy options as defined in the methods section. Similar to health policy objectives, level of care delivery targeting is most frequently mentioned: 30% of the iprsps explicitly propose to distribute public health expenditures to the health services that predominantly benefit the poor. Direct targeting is next (22%), followed by urban/rural targeting (17%). Little attention was paid to the other expenditure policy options. Only 9% proposed to distribute expenditures so as to increase personnel working for pro-poor services and improve the geographic distribution of personnel. Geographic targeting was also mentioned by only 9%. Finally, age-based targeting and targeting by disease were mentioned by 4%. Costing Few papers that included health expenditure proposals also paid attention to costing these proposals. Only those papers that proposed disease-based targeting and improving the geographic distribution of personnel showed budgeted expenditure amounts. The remaining two policy options mentioned in the methods section, increasing public health spending and mobilizing additional resources, were explicitly mentioned by, respectively, 43 and 61%. As for the latter option, half of the iprsps mentioned national or social health insurance (also including pre-payments schemes) as a means, while costrecovery mechanisms (e.g. user fees) were mentioned by 29%. The documents paid very little attention to impact analyses: only one paper explicitly stated the intention to analyze the effects of the proposed expenditure policies on the poor and only three explicitly mentioned measures for protecting the poor from possible adverse impacts from these policies. Finally, using the recording form, a subset of items was created, which was clustered under the headings Data and Diagnostics and Policy Options. This allows for identifying general trends, and strengths and weaknesses of the iprsps (Figure 1). The internal line in the figure represents the average scores for the 23 countries on each of the selected questions. Starting at the top and going clockwise, the Data and Diagnostics elements, with the exception of the consultations dimension, seem to be non-existent. However, when turning to the left side of the graph, the iprsps show more aggressive attempts at some forms of targeting, failing, however, to increase resources. In summary, Figure 1 shows that while the overall performance of the iprsps is rather weak, the real problems are that there are no attempts at analytical and diagnostic work to explain the problems for the poor, identify their needs or explain why the system is failing them. On the other hand, and in the absence of data and diagnostic work, Burden of disease for the poor Increasing spending on health Pro-poor services analysis Urban/rural targeting Health system constraints for the poor Geographic targeting Consultation of the poor Level of care targeting Disease targeting Age-based targeting Monitoring and evaluating impact on the poor Public Expenditure Analysis (PEA) PEA with distribution Figure 1. Strengths and weaknesses of iprsps (internal line represents average scores for the 23 countries)

Pro-poor health policies and PRSPs 143 the iprsps attempt, on average, to improve targeting of resources in several ways, but fail to take more than basic targeting approaches. It should be mentioned here that for Policy Options, we have mainly looked at the policy position on these issues. Still, the actual costing continues to be lacking even for the targeting tools identified. Country profiles Another way to present the information collected in our review is to organize the country papers along four profiles of achievement. This facilitates identifying documents requiring additional work, as well as good practice examples. Table 3 organizes the 23 countries along four quadrants, using the Data and Diagnostics and Policy Options headings from Figure 1. Quadrant I represents countries with above average iprsps in terms of data and diagnostic work as well as pro-poor policy options. The second quadrant lists the countries that were above average on the diagnostic dimension, but below average in terms of pro-poor policies. Quadrant III is the opposite of quadrant II and the fourth quadrant lists countries below average on both dimensions. Discussion and conclusions While the formulation of health policies is most definitely on the poverty reduction agenda of the countries reviewed, there is still a lot of work to do in order to properly address the health of the poor. Of concern is the lack of countryspecific data on the distribution of disease, the composition of the burden of disease, the prevailing health system constraints and the impact of health services in the majority of iprsps reviewed. More importantly, little or no effort was made to analyze these issues in relation to the intended beneficiaries. As for distributive equity, only a small group of iprsps documented efforts to explicitly include the interests of the poor in health policy design. The majority did not take an explicit pro-poor approach. The attention given to making the distribution of public health expenditures more responsive to the needs of the poor was even more limited. Most frequently mentioned were proposals to distribute expenditures to propoor health services, to the poor themselves, and to urban slums and/or rural areas. Few of the papers that included expenditure proposals also showed health budgets. Approximately half of the iprsps mentioned policies aiming at increasing public health budgets and mobilizing additional resources. However, if the distributional aspect of health spending is left unaddressed, these policies cannot succeed in making spending more responsive to the health care needs of the poorest groups (World Health Organization 2000). In this study, we adopted a vertical equity definition: if the poor have more health care needs than the rich, more should be spent on the poor. The selected policy options emphasize health strategies for directly or indirectly targeting the poor. However, we acknowledge that equity is more than targeting the poor. In this context, Leon et al. (2001) note that, Although targeting the poor is clearly important, inequalities and inequities in health are not only about the plight of the most deprived in each society. In many low income countries, over half the population may be living in poverty and those who are not will still be living in circumstances that contribute to the poor health of the country as a whole. Moreover, improvements in the health of the poor may also be achieved through other policies than the approach adopted in this study (Gwatkin 2000b). However, it is important to note that the World Bank advocates a pro-poor approach in the PRSP process, which makes it an important line of policy formulation for developing countries. As previously stated, in the targeting types there is a clustering and overlapping of the real problems among the poor. A rational implementation strategy might select one or more options to be implemented based on the actual perception of the problems. Based on the results of our study, we make two recommendations. First of all, the preparation of PRSPs by countries needs to be more analytical and evidence-based (Niessen et al. 2000). This implies investment in the collection of data on the distribution of poverty, the burden of disease, the constraints of the health system, public health expenditures and the impact of public spending on health, particularly in relation to the poor. Secondly, countries should be more explicit and specific in the formulation of national (pro-poor) health policies. This includes an explanation of the adopted health approach, setting long-term and medium-term health objectives, a detailed time-line, measurable indicators and the costing of proposed policies. Such policies promote actual implementation and facilitate impact analysis. The results summarized in this paper should also motivate the international development community, multi-laterals and bi-laterals, to assist low-income countries. While some efforts have been made, such as the dissemination of the World Table 3. Country profiles Above average pro-poor policies Below average pro-poor policies Above average diagnostic Ghana, Guinea, Madagascar, Central African Republic, Chad, work and data Nicaragua, Niger Gambia, Honduras, Malawi, Rwanda Below average diagnostic Bolivia, Guinea Bissau, Guyana, Benin, Cameroon, Mali, Mozambique, work and data Kenya, Zambia São Tomé and Príncipe, Senegal, Tanzania

144 Leontien Laterveer et al. Bank s PRSP Sourcebook for health (Claeson et al. 2001a, b), much more is needed. The results of the iprsp review presented here point to some important activities that can be undertaken by the international development community: (1) Strengthen local capacity for collection and analysis of data, and better use of existing data, relating to the needs of the poor and the functioning of the health sector, especially for the poor. (2) Facilitate consultations with the poor and socially vulnerable to better understand the constraints faced by marginal groups and to better structure effective programmes. (3) Finance South South learning mechanisms and multicountry policy dialogue forums to disseminate experience with (the formulation of) pro-poor policy options to arrive at more explicit pro-poor policies, including explicit targeting and explicit costing. (4) Support the development of monitoring and evaluation mechanisms that capture equity performance and allow countries to learn and adapt. Such mechanisms should ensure that the voices of the vulnerable are heard. (5) Provide financial and technical support for strengthening the implementation capacity in low-income countries. Other recent studies with comparable but less systematic approaches also arrive at some of these conclusions and recommendations (World Health Organization 2001). Considering the countries accomplishments since the launch of the PRSPs in 1999, there should be hope to expect that future papers will show progress in poverty reduction strategies. Our hope is that ultimately, this can also be said for the health of the poor. References Birdsall N, Hecht R. 1995. Swimming against the tide: strategies for improving equity in health. [http://www.worldbank.org/html/ extdr/hnp/hddflash/workp/wp_00055.html] Claeson M, Griffin C, Johnston T et al. 2001a. Health, Nutrition and Population. Technical Notes. [http://www.worldbank.org/ poverty/strategies/chapters/health/hltn0611.pdf] Claeson M, Griffin C, Johnston T et al. 2001b. Health, Nutrition and Population. Draft for comments. [http://poverty.worldbank.org/files/4978_hlth0627.pdf] Culyer AJ, Wagstaff A. 1992. Need, equity and equality in health and health care. Discussion Paper 95. York: Centre for Health Economics, University of York. Eurodad. 2000. An independent guide to PRSP. [http://www. eurodad.org/2poverty/indexpoverty1.htm] Gwatkin DR. 2000a. The current state of knowledge about targeting health programs to reach the poor. Preliminary Discussion Draft. [http://www.worldbank.org/poverty/health/ library/targeting.pdf] Gwatkin DR. 2000b. Health inequalities and the health of the poor. What do we know? What can we do? Bulletin of the World Health Organization 78: 3 18. Gwatkin DR, Guillot M. 1999. The burden of disease among the global poor. Current situation, future trends, and implications for strategy. [http://www.worldbank.org/poverty/health/library/ burden.pdf] International Monetary Fund (IMF)/Word Bank. 1999a. Poverty Reduction Strategy Papers Operational Issues. Draft for discussion. [http:// poverty.worldbank.org/files/operational_ issues.pdf] International Monetary Fund (IMF)/Word Bank. 1999b. Building Poverty Reduction Strategies in developing countries. [http://poverty.worldbank.org/files/building_english.pdf] International Monetary Fund (IMF)/Word Bank. 1999c. Heavily Indebted Poor Countries (HIPC) Initiative Strengthening the link between debt relief and poverty reduction. [http://poverty.worldbank.org/files/strengthening_link_english. pdf] International Monetary Fund (IMF)/World Bank. 2001. Poverty Reduction Strategy Papers Progress in Implementation. [http://poverty.worldbank.org/files/prspprogess.pdf] Laterveer TL. 2001. Progress in poverty reduction strategies. A health policy study of 23 developing countries. Background paper. Rotterdam: Institute for Health Care Policy and Management, Erasmus University Rotterdam. Leon DA, Walt G, Gilson L. 2001. Recent advances. International perspectives on health inequalities and policy. British Medical Journal 322: 591 4. Niessen LW, Grijseels EWM, Rutten FFH. 2000. The evidencebased approach in health policy and health care delivery. Social Science and Medicine 51: 859 69. Verheul E, Cooper E. 2001. Poverty Reduction Strategy Papers (PRSP). What is at stake for health? Briefing paper. Amsterdam: Wemos in collaboration with the Network on PRSPs and Health. Verheul E, Rowson E. 2001. Poverty reduction strategy papers. British Medical Journal 323: 120 1. Wagstaff A, Van Doorslaer E. 1993. Equity in the finance and delivery of health care: concepts and definitions. In: Van Doorslaer E, Wagstaff A, Rutten F (eds). Equity in the finance and delivery of health care: an international perspective. Oxford: Oxford University Press, pp. 7 19. World Bank. 2000. Poverty Reduction Strategy Papers Internal guidance note. [http://www.worldbank.org/poverty/strategies/ intguid.pdf] World Bank. 2001a. Poverty Reduction Strategy Papers Sourcebook. Preface. Draft for comments. [http://www.worldbank. org/poverty/strategies/chapters/preface/pref0618.pdf] World Bank. 2001b. Grouping of HIPCs under the enhanced HIPC initiative: status as of December 2001. [http://www.worldbank. org/hipc/progress-to-date/grouping-_december.pdf] World Health Organization. 2000. World Health Report 2000. Geneva: WHO. [http://www.who.int/whr] World Health Organization. 2001. Health in PRSPs. WHO Submission to World Bank/IMF. Review of PRSPs. WHO, Department of Health and Development. [http://www.worldbank.org/ poverty/strategies/review/who1.pdf] Acknowledgements For her research at the World Bank, Leontien Laterveer received a grant from the AA van Beek Fund, Rotterdam. The opinions formulated in this paper are those of the authors and do not necessarily reflect the official policy of the involved institutions. Biographies Leontien Laterveer is a health scientist who wrote her MSc thesis on PRSPs and health at the World Bank. During her studies, she worked for the Dutch Ministry of Foreign Affairs and Wemos, a Dutch NGO. For the latter, she wrote a review of 20/20 Initiative country proposals for the People s Health Assembly in Bangladesh. Presently, she is junior consultant with ETC Crystal. Louis W Niessen is a medical epidemiologist and a senior staff member at the Institute for Health Care Policy and Management/Institute for Medical Technology Assessment, Erasmus University. He is involved in national and international

Pro-poor health policies and PRSPs 145 projects on health policy, disease burden and economic evaluation in European, African and Asian settings. He has worked in Tanzania, Peru and Nepal. Address: Institute for Health Care Policy and Management/Institute for Medical Technology Assessment, Erasmus University, PO Box 1738, 3000 DR, Rotterdam, the Netherlands. Email: niessen@bmg.eur.nl (also for correspondence). Abdo S Yazbeck is a lead health economist at the World Bank Institute, the training arm of the World Bank. He received his Ph.D. in Economics from Rice University in 1991 and lectured in economics at Rice University and Texas A&M University. He has worked on health sector and development issues in the Former Soviet Union, the Middle East, North Africa, South Asia and Sub-Saharan Africa. Address: Health, Nutrition and Population/Poverty Thematic Group, The World Bank, 1818 H Street, N.W. Washington, DC 20433, USA. Email: ayazbeck@worldbank.org Correspondence: Leontien Laterveer, ETC Crystal, Kastanjelaan 5, PO Box 64, 3038 AB, Leusden, the Netherlands. Email: l.laterveer@ etcnl.nl