69922 ASSESSING THE CAPACITY OF MATERNAL AND REPRODUCTIVE HEALTH PROGRAMS TO ACCESS PUBLIC SECTOR FUNDS IN THE NEW BUDGET SUPPORT ENVIRONMENTS AND THE NEW AID ARTITECHTURE A Synthesis Report of the Dialogue with the Maternal and Reporductive Health Programs of Eritrea, Ethiopia, and Benin Written By: Peter Bachrach BACKGROUND 1. Context. Every year over half million women die in developing countries as a result of complications related to pregnancy and childbirth and millions more more suffer from reproductive health complications.1 In response, many sub-Saharan African countries have historically relied on direct donor support for maternal and reproductive health programs, which have been operated vertically and managed by specialized staff. 2. While these programs have generally benefitted from their earmarked financing and programatic independence, their special status and relative isolation within the Ministry of Health has come at a price: At a time when traditional funding sources and Governments are addressing a range of innovations (including health sector reforms, SWAps, PRSCs, and other budget support measures), the maternal and reproductive health units are somewhat less experienced and therefore disadvantaged in the competition for scarce resources through the medium-term and annual planning and budgeting processes carried out by the ministries of health and finance. 3. Anecdotal evidence from the literature2 suggests that maternal and reproductive health programs have not yet: (i) formulated sufficiently persuasive, evidence-based policies and strategies; (ii) succeeded in establishing and/or maintaining constituences (within and outside the ministry) to actively support their objectives; and (iii) developed the capacity to participate effectively in the national processes for budgeting and disbursing funds. As a result, there is concern (which must be verified on a country-by-country basis) that maternal health programs are not receiving their fair share. 4. Review of the terms of reference. The overall objective of the study is to develop a methodology for possible use by MRH services to assess program status and to examine potential strategies for advancing the objectives of the program through both financial and non-financial means. The specific objectives as outlined in the terms of reference were to: • assess the status and performance of the current MRH program interventions of the country and the future prospects for improving these interventions • examine institutional arrangements within which the MRH program operates to: (i) carry out program planning and budgeting; (ii) implement MRH interventions; and (iii) monitor and evaluate progress on measures of effectiveness, efficiency, equity, quality, etc.; and • analyze past and future financing for these MRH program elements (and particularly with respect to the financing of commodities) and identify potential budgeting strategies for MRH program managers to use in competing for scarce resources. 5. The terms of reference also proposed a list of questions for consideration by this study; a summary of the answers as well as an assessment of the methodology used may be found in Annex . 1 Marge Koblinsky, et al., "Issues in Programming for Safe Motherhood" (Arlington, VA: John Snow Inc., September 2000), p. 7. 2 See the paper, panels discussions, and recommendations in "Making the Link Between Sexual & Reproductive Health and Health Systems Development," Proceedings of an International Conference (Leeds, UK: September 2003). 2 6. Conduct of the study. The initial terms of reference envisoned a number of phases involving: (i) development of a methodology, with the advice from World Bank Reproductive Health technical staff, to be used systematically across countries; (ii) pilot testing of the methodology in one country followed by implementation in two additional countries; (iii) preparation of a synthesis report and a guidance note; and (iv) dissemination of the findings if the consultation is found to be useful. 7. Headquarters literature review. The literature review and initial data collection (carried out in July and August 2006) focused essentially on: (i) studies reviewing the evolution of MRH program components and the achievement of the MDGs, potential indicators for measuring status, and available sources of data on country progress; (ii) data sources on country-level expenditures on health in general (and MRH in particular), including completed or ongoing public expenditure reviews, national health account data, etc.; and (iii) documents examining the effects of health sector reforms on the delivery of MRH services. An analysis of reproductive health status and per capita expenditures (see Annex 2) helped to identify countries of potential interest for the study. 8. During August and September, data collection formats were designed, terms of reference prepared, and proposals to organize the mission’s in-country work program submitted to the selected countries for review by the respective health ministries. 9. In-country consultations. The initial selection of countries was carried out in September and October with TTLs in Washington and MRH program managers in the potential countries. Preliminary discussions with the MRH program managers indicated their lack of interest in yet another review of their programs but a willingness to exchange ideas about how to argue for additional program resources. As a result, the country assessments were organized as consultations with and for the MRH program managers to share potential methods for: (i) rapidly assessing the current status of their program; (ii) accurately identifying the different constraints for improving the various components, and (iii) developing and implementing budgeting strategies as a function of this analaysis. 10. Some ten countries were contacted, and three agreed to participate in the consultation: Eritrea, Ethiopia, and Benin. While these three countries share many common characteristics in terms of their organizational arrangements for financing the health sector, they differ in important respects: • Eritrea represents the traditional case of development assistance where non salary costs for program-related activities are funded by designated partners. • Ethiopia represents the case of a flexible SWAP with some pooled and some dedicated funding provided by the various partners to the Ministry of Health. • Benin represents the case of budget support to the Government budget from both bilateral and multilateral sources. 11. The country assessment missions comprised a series of discussions and data collection efforts involving central level staff (both in the Ministry of Health and in other relevant Ministries of Finance, Planning, etc.), with MOH staff (at decentralized levels and within health facilities), and with other relevant stakeholders to analyze: 3 • the current status of selected MRH program elements and the possibilities for expanding and improving these services; • the existing institutional arrangements planning and implementing MRH services and the possibilities for mobilizing additional resources; and • planned and actual expenditures on maternal and reproductive health and the possibilities for increasing the efficiency and effectiveness of future expenditures. 12. The consultants visited Eritrea (October 18-30), Ethiopia (November 26-December 8), and Benin (December 11-20) and shared their preliminary findings with program officials at the end of each trip. Revised individual reports on each of these countries are attached to this summary report. 13. Principal conclusions. In each of the countries, the consultants met with MRH program managers who had common concerns and reached similar conclusions on the issues of concern to the mission. • MRH program managers judge their programs with severity. They recognize that: (i) maternal health indicators are simply not improving rapidly enough (para. 12); (ii) their programs cannot provide either the full complement of interventions or the level of quality for the available services (paras. 14-17); (iii) the time frame for program improvements is, in many instances, likely to take several years at best (para. 18); and (iv) the constraints on improvement are not just technical in nature but involve as well institutional and political constraints (paras.19-20). • MRH program managers are obviously aware of the various public sector reforms in progress, but they are not on the whole persuaded that these have had any especially positive effects on the health sector in general or on maternal health in particular (paras. 21-23). They note in particular a disconnect between: (i) the language of health sector reform (emphasizing system strengthening) and MRH (emphasizing advocacy for RH interventions, packages and services); and (ii) the subordination of specific MRH operational issues to broader sector-wide financing and implementation issues (paras. 24-26). The effects of these ongoing reforms are both felt (changes in planning and budgeting) and unfelt (lack of change for financing various program interventions, for procurement and disbursement, and for management of human resources) (paras. 27-33). Finally, the impact on MRH of issues such as partnership and decentralization remain essentially unresolved (paras. 34-35). • While MRH program managers understand that developing country governments face a range of competing priorities across sectors which limit increases to the health sector, they are as well frustrated by the lack of budgetary priority accorded their programs within the Ministry of Health. On balance, MRH progams have received declining shares of ministry budgets and remain dependent on earmarked external financing, even in countries benefiting from budget support (paras. 36-37). Though faced with these realities, the MRH managers have neither developed adequate budgeting strategies (paras. 38-42) nor marshalled the most appropriate arguments to justify increases and persuade the appropriate authorities (paras. 43-44). 4 PROGRAM COMPONENTS 14. The country assessments attempted to collect data on the current status of certain MRH indicators. The incomplete data, based on official government sources made available, are presented below; a somewhat more extensive table may be found in Annex 3. Key MRH indicators 2000 2001 2002 2003 2004 2005 Antenatal consultations (at least 1 visit) Eritrea 73.0% Benin 88.0% 76.5% 91.2% 92.3% 91.0% Ethiopia 27.0% 42.1% Deliveries assisted by skilled health staff Eritrea 28.0% Benin 65.5% 62.8% 76.0% 76.0% 76.0% Ethiopia 5.6% 9.7% Use of modern family planning methods Eritrea 5.1% Benin 15.6% 13.2% 16.4% 11.0% Ethiopia 25.2% 15. While the information is far too incomplete for discussion of the situation in Eritrea and Ethiopia, the situation in Benin has not improved substantially over the past five years and has even deteriorated with respect to the use of modern family planning methods. 16. On the basis of a number of efforts to develop the components of a maternal health program,3 a summary framework was adapted for use in the field to assess: (i) the inclusion of the program component in the overall maternal health program; (ii) the current quality of service delivery; (iii) estimate the time frame for improving service delivery; and (iv) assess potential constraints on improving the delivery of the service. 17. Analysis of the current status, quality, and time frame for improving program performance. As a first step in the discussions with maternal health program managers, a rapid assessment of the program was carried out. This approach was used at the central level in all three countries visited and at the regional (zoba) level as well in Eritrea. The results are summarized in Annex 4; more detailed results with succinct comments from the program managers may be found in the individual country assessments. Assessment of number and quality of program components Eritrea Ethiopia Benin Maternal health program components Number of components included 38 33 33 Percentage 95% 83% 83% Quality Number of components of sufficient quality 16 5 7 Percentage 42% 15% 21% 3 UN, Program of Action adopted at the ICPD, Cairo 1994. Volume 1, (New York 1995); Allan G. Hill, "The Boundaries of Reproductive Health," Background Paper: EDI Core Course on Adapting to Change: Reproductive Health and Health Sector Reform (Washington, DC 20 September – 8 October 1999; Adam Wagstaff and Mariam Claeson, Rising to the Challenges (Washington, DC: The World Bank, 2004). 5 18. Number of program components. Inclusion of program components was based on the program manager’s judgment; no systematic effort was made to compare the manager’s assessment with the national strategy documents or existing program norms and standards. Of the 40 program elements discussed during the assessment, Eritrea concluded that 38 were actively being addressed while Ethiopia and Benin concluded that 33 were actively being addressed. Eritrea’s program differed from the other two programs essentially on the range of obstetric and gynecological services that were considered to be included in the package. 19. Quality of the program services. Program managers were then asked to assess each of the program elements considered to be included in the package in terms of whether the quality of the service provided was sufficient or insufficient4. Based on the program manager’s judgment, none of the three programs was judged to be of sufficient quality, though Eritrea evaluated program status more favorably than did Ethiopia and Benin. While there was very little overlap among the components deemed to be of sufficient quality (only immunization for tetanus toxoid), all three program managers noted the need to improve the quality of service delivery for the following components: STI / HIV / AIDS and PMTCT; outreach activities; prompt referral of pregnancy complications (communications and transportation); availability (24/7) of trained surgical, anaesthetic, and midwifery capabilities; management of complications from unsafe abortion; post- partum monitoring; maternal death audits; family planning commodities management; and management of gynecological infections. 20. This analysis of the MRH program components confirmed the results of a recent study from Ghana indicating that the breadth and complexity of the interventions proposed constitute significant problems for many programs.5 Moreoever, the study points out the particular “challenge of harmonizing a comprehensive definition of RH and the reality of selective implementation at the district level,� since “the reality districts face is that they do not have enough capacity to do all that has been defined in the national policy and therefore have to make choices within the institutional arrangements defined in the health sector.�6 This is shown to a limited extent by the analysis done at Zoba level in Eritrea, where the assessment of quality dropped from 42% to 34%. 21. Time frame for component improvement. Managers were asked to estimate a time frame for addressing program improvements: within 1 year, within 2-3 years, and beyond 3 years: Time frame for addressing program improvements Eritrea Ethiopia Benin Estimated time to improve service delivery of various components Components to be addressed in Year 1 9 2 16 Components to be addressed in Years 2-3 9 17 10 Components to be addressed in Years 3+ 4 9 0 22. While the discussion did not address precisely which improvements would be addressed or how they would be implemented, Benin’s program manager was the most optimistic about rapidly improving program services, while Ethiopia’s was the least optimistic. 4 Other efforts to measure program efforts have been far more detailed. See J.A. Ross, et al., "The Maternal and Neonatal Programme Effort Index (MNPI)," Tropical Medicine and International Health, 6 (October 2001), pp. 787-798; and the periodic country reports on line at: www.policyproject.com/pubs/mnpi.cfm. 5 Hall, op cit., p. 2. 6 Harriet Birungi, et al., Priority Setting for Reproductive Health at the District Level in the Context of Health Sector Reforms in Ghana (April 2006), p. vii. 6 23. Assessment of the constraints on improvements. Program managers were subsequently asked to indicate the technical and politico-institutional constraints on implementing program improvements; in the discussions, a distinction was made between: (i) less feasible and more feasible technical interventions; and (ii) greater and lessernterest among potential sources of support for financing these interventions. The results from the individual country consultations are summarized in Annex 5. 24. Based on the interventions above (where a consensus among program managers exists on the need to improve the quality of services) and the results of Annex 4 (where the program managers assessed the technical and political considerations for improving the quality of services), the table below indicates the wide differences among countries. Examination/Treatment for STIs, incl. syphilis (Eth) Examination/Treatment for STIs, incl. syphilis (Eri) PMTCT (Eth) Voluntary HIV counseling and testing (Eri, Ben) PMTCT (Eri) More feasible (H) Extension of coverage via HW outreach (Ben) Extension of coverage via HW outreach (Eth) Referral of complications/Communications (Eri, Eth) Referral of complications/Communications (Ben) Technical-Programmatic considerations Referral of complications/Transport (Eri) Referral of complications/Transport (Ben) Post-partum monitoring (Ben) Management of complications from unsafe abortion (Eri) Maternal death audits (Eth) FP commodities management (Eth) FP commodities management (Eri, Ben) Examination/Treatment for STIs, incl. syphilis (Ben) Voluntary HIV counseling and testing (Eth) PMTCT (Ben) Less feasible (L) Extension of coverage via HW outreach (Eri) Availability (24/7) of trained personnel (Eri, Eth, Ben) Referral of complications/Transport (Eth) Post-partum monitoring (Eri, Eth) Management of complications from unsafe abortion (Eth, Ben) Maternal death audits (Ben) Maternal death audits (Eri) Management of gynecological infections (Eri, Eth, Ben) Lower interest (L) Greater interest (H) Politico-Institutional considerations 25. On only two interventions do the assessments of the program managers in the three countries coincide: on the management of gynecological infections and on the permanent availability of trained personnel. While this may be partly due to differing understandings of the issue, there is clearly no “one-size fits all� approach possible in seeking to help a manager advance his or her program. Rather, except for the case where there is more feasibility and greater interest (and thus may involve the need for additional financial resources), the other cases may or may not require additional financial resources as a first step. Institutional Arrangements 26. To better understand the basis for these assessments of feasibility and interest, the interviews with the MRH (and other) program managers within the Ministry of Health addressed a series of institutional issues, many of which are related to ongoing or proposed health sector reforms. The 7 observations of those interviewed are briefly presented in the relevant annexes of the country reports; the overall conclusions are summarized here in terms of their concerns with: (i) policy formulation and program coordination; (ii) organizational reform and sector resource management; and (iii) decentralization and partnership. 27. In each of these areas, the progress achieved in implementing reforms (though in some cases promising) has left the maternal health program (and often other programs as well) in an awkward situation since the reforms have been: neither sufficiently successful to ensure the hoped-for transition (to adequately financed, technically sound, and professionally accountable programs) nor so obviously unsuccessful as to allow for a return to previously-established ways of doing business. In essence, both the new and the old ways of doing business coexist, with a disconnect between the language of the health sector reform literature and the daily experiences of the MRH program managers. 28. MRH is seemingly stuck between: (i) a history of (and in some cases a continuing preference for) independence (or at least earmarked funding) and vertical control (of program interventions, special personnel, etc.); and (ii) an increasing trend toward the integration of program activities and funding. When confronted by the data on ministry funding of MRH, one official responded that: (i) virtually everything the ministry finances is linked to mothers and children in some fashion; and (ii) in any case, the MRH program has ample resources from its traditional partners. 29. Policy formulation and program coordination. The formulation of MRH policies and strategies is at different stages in each of the countries: Eritrea has drafted a broad policy statement; Benin has prepared a strategy with detailed activities and costs; and Ethiopia has completed a strategy that has been published and distributed. What is less clear is the intended use of these documents; the MRH program managers were unable to confirm either that these documents would be considered (and presumably adopted) by the Government, or that they would provide the basis for increased resources within the Ministry of Health. In Benin, documents prepared in March 2006 had received only limited circulation by December 2006 and had not contributed to the budgetary debate for 2007. 30. In addition, as much of the literature has pointed out, MRH presentations have tended to be pre-occupied with advocating for RH interventions, packages and services, while health sector reform discourse has focused mainly on system strengthening interventions, such as financing mechanisms and human resources management.7 The strategy documents of both Eritrea and Benin support this observation, and though the documents do address the different components of service delivery (manpower, equipment, drugs and other consumables and standards and protocols), they tend to address these in parallel fashion (or over different periods of time) with the result that the documents are not effectively operational. The documents list interventions and activities with little regard for priority setting or sequencing of improvements, which would penetrate the concerns of harried government officials or impress parlementarians with their urgency and/or feasibility. 31. Finally, with the increased emphasis on implementing sector-wide issues and financing, previous coordination efforts among agencies concerned with MRH have often been subsumed broader coordinating structures and procedures and not replaced with the kind of operational 7 Birungi, et al., op cit., p. 2. The issue of language is cited in a number of articles, including Hall, op. cit., p. See also Marianne Lubben, Susannah H. Mayhew, Charles Collins, & Andrew Green, "Reproductive health and health sector reform in developing countries: establishing a framework for dialogue," Bulletin of the World Health Organization, 80 (Aug. 2002), pp. 667-674. 8 coordination required or the political weight of combined voices to ensure that maternal health is at the forefront of the national health dialogue. 32. Organizational and resource management reforms. Currently, organizational and resource management reforms have focused on: (i) financing issues, but with more emphasis on planning and budgeting than on procurement and disbursement; and (ii) human resources, but with more emphasis on performance than on manpower planning and management. 33. Financing issues. Though reforms (more modest in Eritrea and Ethiopia than in Benin) are being implemented in planning and budgeting, the overall situation does not seem to have changed dramatically in any of the countries. The medium-term expenditure framework (MTEF) is not fully operational in Benin and has not been adopted in Eritrea; in both countries, government budget ceilings are determined principally on previous allocations (for ministries and among programs within ministries) and negotiations are limited both by the time allotted for budget preparation and by the relatively non-participatory practices of the budgeting units within the Ministries of Health of the two countries. 34. Financing of maternal health continues to be based on a patchwork of different sources without any comprehensive view of the comparative advantage of different sources of financing for different types of interventions and associated expenditures. Program directors in Benin and Eritrea pointed out that much of MRH management involves matching non-Government funding sources with identified needs, and this despite the fact that budget support (as in the case of Benin) was supposed to resolve such problems. Annex 6 presents the current MRH program funding sources and uses in Benin and indicates the continuing complexity of managing funding sources. In Eritrea, the MRH program manager, with fewer partners, has (at least until the advent of the World Bank’s HAMSET II project) used UNFPA funds as the central source of funding and tried to “fill in� with other external and internal financing. 35. Thus far, none of the participating countries has considered an overall financing plan for the different interventions and their related costs, which would take into account: (i) the advantages and disadvantages of the various sources of program financing; and (ii) the possibilities for generating additional resources from the different sources. The following table indicates some of these considerations.8 Administrative Sources of funds Predictability Timeliness Flexibility Local control Requirements Central gov’t budget Yes Variable Some Sometimes Many and long Local gov’t budget No No Some Always Many and long Cost recovery Yes Yes Yes Always Few External assistance Variable Variable Variable Sometimes Variable 36. Negotiating for additional resources will require a more strategic overview of these sources and uses by the MRH program managers at all levels. 37. Furthermore, while planning and budgeting reforms are being implemented, modifications of the procedures for procurement of works, goods, and services and for disbursement of funds have lagged far behind. Thus, in Benin, government funds continue to be disbursed as before: (i) with 8 For additional criteria to assess different types of financing, see among others Logan E. Brenzel, "Planning the Financing of Primary Health Care: Assessing Alternative Methods" (REACH 1987). 9 considerable delay in their initial availability (at times as late as March); (ii) through time- consuming procurement and financial management procedures; and (iii) with strict ceilings for very specific, rigid line items. Decentralization of expenditure responsibilities is effective in Ethiopia and in progress in Eritrea and Benin, but these reforms have been hampered by: (i) conflicting demands on the Ministry of Finance; (ii) insufficient internalization of or authorization to modify existing procedures and practices; and (iii) insufficient availability of funds at the local level to make proposed improvements concrete.9 38. Human resource issues. Though the sector reform process puts increasing emphasis on health systems and personnel performance, discussions with the MRH program managers in the three countries suggest that many of the management issues related to human resources are not within the purview of the Ministry of Health. In Benin, for example, the Ministry of Health controls neither recruitment or conditions of service (Civil Service Ministry), nor pre-service training (Ministry of Higher Education), nor establishment and payment of salaries (Ministry of Finance). In addition, the strength of the existing medical and nursing societies and unions further constrain the allocation and use of health staff particularly in rural communities; this is especially critical in the case of registered midwives, who are overwhelmingly concentrated in urban facilities. In Eritrea, HR management is hindered by a lack of coordination between the HR Department and the Clinical Services Department, both of which have responsibilities for HR management and supervision. 39. Benin has commissioned a study of the HR issue and, among other aspects, the modalities for improving the deployment and remuneration of staff. In Eritrea and Ethiopia, other kinds of HR initiatives are underway: Eritrea is in the process of redefining the nurse midwife criteria and training, and implementing a major effort to provide lifesaving skills (LSS) training to health workers, primarily at health station level. In Ethiopia, over half of the 30,000 Health Extension Workers needed to provide basic services in 15,000 communities have been recruited. However, for the most part, the MRH program managers are not in a position to influence this critical element of service delivery. 40. Partnership and decentralization. Partnership among donor agencies, the private sector, and the government is more developed in Benin and Ethiopia than in Eritrea. Ethiopia has a lengthy history with the Sector-Wide Approaches (SWAps) and pooled funding while Benin has implemented a series of Poverty Reduction Strategy Credits (PRSCs). Similarly, contracting of services with NGOs is being studied and tested in Benin, and NGOs are active partners in Ethiopia, but Eritrea has by and large decided not to utilize this service delivery mechanism. These partnerships could serve an important role in strengthening the delivery of MRH services, but they have not so far done so. 41. While there is much discussion of decentralization, only in Ethiopia (with its federal structure) is decentralization advanced in any meaningful sense. In Eritrea and in Benin, the intentions of government have not yet been matched by the operational measures required to move beyond deconcentration toward decentralization, as indicated in the previous discussion of financial and human resources. 9 See David A. Good, "The Politics of Public Money: Spenders, Guardians, Priority Setters, and Financial Watchdogs Inside Canadian Government," Notes for the Public Sector Financial Leadership Conference (Ottawa 2006), p.2 citing Allen Schick on three basic budget objectives of: (1) maintaining aggregate fiscal discipline, (2) allocating resources in accord with government priorities, and (3) promoting the efficient delivery of services. 10 11 STRATEGIC APPROACHES FOR COMPETING FOR SCARCE RESOURCES 42. Given the expressed need to improve the vast majority of interventions adopted by the MRH strategy and program, the in-country consultation worked with the program manager to: (i) evaluate the prospects for increasing government expenditures for health; (ii) weigh the comparative advantages offered by the different sources of program financing10; and (iii) assess alternative budgeting strategies for increasing financial resources available for improving maternal health interventions. 43. Prospects for increasing government expenditures for health. The mission confirmed in the field the results of WHO’s data on government health expenditures.11 As indicated in the following table, though government financing of health has generally increased in absolute terms, it has not increased over the past half decade as a proportion of overall government expenditures. Country 1998 1999 2000 2001 2002 2003 Benin 13.7% 11.1% 10.0% 9.8% 8.0% 9.8% Ethiopia 10.2% 8.9% 9.3% 10.5% 9.9% 9.6% Eritrea 3.4% 2.9% 4.4% 4.6% 3.9% 4.0% WHO core health indicators: www3.who.int/whosis/core/core 44. In addition, though the bases for calculating the share of financing for maternal and reproductive health care are different for Benin and Eritrea,12 the data in the following table show as well that neither country has increased resources for this program. % of financing for MRH 2002 2003 2004 2005 2006 2007 2008 2009 % of total budget Benin 3.3% 4.8% 2.5% 2.8% 1.8% 1.7% 2.5% % of external funding Eritrea 22.6% 19.4% 11.5% 12.6% 15.4% 14.2% 45. In the absence of sustained increases in financial resources from government funding, overall increases in health sector financing continue to be based on actual (in the case of Eritrea) or projected (in the case of Benin) contributions from external sources, as shown in the following table. % of external financing 2002 2003 2004 2005 2006 2007 2008 2009 Benin 4.0% 4.7% 19.1% 26.7% 31.7% 34.1% Eritrea 42.3% 45.7% 43.8% 55.5% 51.6% 10 The mission did not examine the actual costs proposed by the different strategies; this work was available in Benin but not in Eritrea (although it was planned). The literature seeking to cost the various interventions is voluminous, but was not directly pertinent to the objectives of the mission. 11 See the individual field reports for Eritrea and Benin annexed to this report. 12 See the individual field reports for Eritrea and Benin annexed to this report. 12 46. Alternative budgeting strategies. This study proceeds from the hypothesis that: (i) ministries (and MRH program managers in particular) have essentially four options for increasing resources available to improve MRH service delivery; and (ii) these options are linked to the technical feasibility of and political support for the specific interventions.13 Resources can be increased: (i) through the allocation of additional resources (if available); (ii) through the reallocation of existing resources to other priority interventions, populations, etc; (iii) through regulatory measures (adoption of laws, policies, programs, strategies and standards, etc.; and (iv) through self-regulatory measures providing sufficient information, conferring responsibilities, etc. to individuals, groups and communities to organize their own response to the need for improved maternal health services. 47. Graphically, these four options are linked to considerations of technical feasibility and political support as follows: feasible (L) feasible (H) Technical-Programmatic III IV More considerations Self-Regulation conferring responsibility Allocation / Distribution of for use of resources additional resources I II Less Regulation of the use of current Reallocation / Redistribution of and/or future resources existing resources Lower interest (L) Greater interest (H) Politico-Institutional considerations 48. Linking the analysis of current program interventions to the options for increasing funding allows the MRH program manager to consider: (i) how best to sequence the development of the interventions; and (ii) which budgeting methods are best apt to provide convincing arguments. 49. Sequencing resource allocation decisions. Based on analysis of technical feasiblity and political interest, program managers should be able to identify the most favorable current (budgetary and non-budgetary) options and to envision a sequence for program development over time to add or improve MRH program elements, e.g., moving from less to more feasible, from less to greater interest, etc. Seen from this perspective, the previous table indicates that a Ministry’s reliance on a regulatory response (I) constitutes both a weak and an obligatory response to increase technical feasibility and political interest, which will eventually justify either a future reallocation of resources to new priorities (II) or to confer responsibility (III) for the use of resources to other actors (e.g., private sector, communities, etc.). Following this logic, the Ministry’s decision to seek the allocation of additional resources (IV) is most easily justified if budgeting strategies I, II and III can already be demonstrated to have been used. 50. A comparison of the budgeting strategies in Benin and Eritrea demonstrate quite contrasting styles, as shown in the following table. 13 See for a previous and somewhat different application Robert H. Salisbury and John P. Heinz, "A Theory of Policy Analysis and Some Preliminary Applications," in Ira Sharkansky, ed., Policy Analysis in Political Science (Chicago: Markham Publishing Company, 1970), pp. 39-60. 13 2005 2006 2007 Benin Committed % Budgeted % Budgeted % More feasible (H) / Greater interest (H) 880.0 72% 168.0 44% 424.0 48% More feasible (H) / Less interest (L) 22.1 2% 72.6 19% 97.6 11% Less feasible (L) / Greater interest (H) 237.2 19% 9.0 2% 20.0 2% Less feasible (L) / Less interest (L) 79.2 7% 131.0 34% 347.0 39% Total 1 218.5 100% 380.6 100% 888.6 2006 2007 Eritrea Budget % Budget % More feasible (H) / Greater interest (H) 576 685 55% 319 880 19% More feasible (H) / Less interest (L) 225 838 21% 1 005 295 60% Less feasible (L) / Greater interest (H) 140 742 13% 256 544 15% Less feasible (L) / Less interest (L) 112 200 11% 83 770 5% Total 1 055 465 100% 1 665 489 100% 51. In Benin, committed funds in 2005 overwhelmingly favored interventions where there was greater interest (and generally more feasibility) whereas budget proposals for 2006 and 2007 constitute a dramatic shift toward interventions where there is less interest (and also less feasibility). Eritrea’s budget proposals for 2006 and 2007 show a similar shift from interventions of greater to lesser interest, though with the difference that Eritrea is focusing on interventions considered to be of greater feasibility. 52. Two observations may be in order. First, while in both cases there may be reasons for such budgetary decisions in the long term, it should not be surprising that such decisions do not engender budgetary increases in the near term. Second, though the analysis of feasibility and support is not the only basis for adopting certain budget strategies, program managers should not assume that simply asserting the need for a specific intervention will be sufficient to convince those authorities in a position to influence decisions. 53. Justifying resource allocation decisions. While a number of important principles and techniques have historically been developed to guide decision making in the allocation public sector resources, none “can provide an all-embracing theory of budgeting since the basic budgeting problem is multi-dimensional and has to be tackled simultaneously from various perspectives.�14 This same author goes on to discuss five complementary techniques for determining the allocation of resources: • identifying the underlying rationale for public interventions through an analysis of: (i) the conditions of supply and demand for public and private goods and (ii) the advantage of the state in the economy; • prioritizing alternative applications of public funds by applying the principle of marginal utility using measures of cost-effectiveness; • maximizing utility through an assessment of the net social benefits of public spending using cost benefit analysis; 14 Adrian Fozzard, "The Basic Budgeting Problem: Approaches to Resource Allocation in the Public Sector and their Implications for Pro-Poor Budgeting," Working Paper 147 (London: Overseas Development Institute, 2001), p. 6. 14 • recognizing the primacy of citizens’ expenditure preferences and developing mechanisms of collective decision making so that these can be communicated to decision-makers; and • redistributing resources in order to address social equity and poverty concerns.15 54. Determining the appropriateness of these techniques under the different conditions of feasibility and political support yields the following table. Technical-Programmatic Self-Regulation Allocation / Distribution feasible More considerations Budgeting based on methods of collective decision making Maximisation of utility (cost benefit) Prioritization of public funds (cost effectiveness) Regulation Reallocation / Redistribution feasible Less Budgeting based on comparative advantage of state / Budgeting to address social equity and poverty Supply and demand Lower interest Greater interest Politico-Institutional considerations 55. While program managers may not precisely apply the actual techniques implied by these approaches to the budgeting of individual interventions, they should understand that, for example, arguments based on cost benefit and cost effectiveness are most likely to be effective when they concern interventions for which there is already greater technical feasibility and interest. Where the intervention involves other considerations, other, quite different budgeting concerns may well be required. CONCLUSIONS 56. In view of the targets established in the MDGs and the slow progress being made on maternal health indicators, MRH managers are increasingly concerned about the coverage and quality of program services. At the same time, the technical complexities and the politico- institutional factors of the various components constitute inescapable constraints on rapid improvement of services. Though health sector reform and MRH program development are certainly not contradictory, they have not (in the three countries visited) been shown to be completely compatible: both the inadequate development and sequencing of the MRH program components and the insufficient implementation of the announced reforms should probably share the responsibility for the current situation. 57. In view of these constraints on MRH and the long-term struggle for scarce resources from a wide range of competing priorities across sectors, MRH program managers must first: (i) better assess the feasibility and support for the various interventions; (ii) develop budgeting strategies based on sound sequencing of interventions and adequate budgeting strategies; and (iii) marshal the most persuasive arguments to justify increases to the appropriate authorities. Then (and this will be more difficult given the current state of expenditure procedures) program managers must ensure that the allocated funds are disbursed on the interventions which have been previously identified. 15 Fozzard, op. cit., p. 6. 15 58. Two additional observations should be considered. First, aspects of the preceeding discussion can be summarized in terms of three arguments emphasizing the different dimensions of the debate over the development of maternal and reproductive health services: • The first argument asserts that what matters most is the program, in some instances comprising all of the MRH components of RH and in others focusing more narrowly on commodities (“no product, no program�). • The second argument assumes that what matters most is the array of evolving institutional arrangements and relationships among the key actors in the effort to improve funding and services within the organization specifically, the sector more generally, and the government as a whole. • The third argument presupposes that what matters most is the development of: (i) a consensus concerning the needs of the beneficiary population; (ii) a coherent and coordinated response to these needs; and (iii) a mechanism for ensuring the effective participation of beneficiaries. 59. These perspectives were described in detail more than two decades ago in an article which attempted to examine the elements of management fit.16. These elements of management fit are summarized below and presented graphically on the next page: • From the perspective of institutional arrangements, the key issue for “fit� involves managing the inherent tension between the internal demand to allocate more resources and accelerate procurement and disbursement procedures, and the external demands for transparency, accountability, and coordination of resource use. • From the perspective of program organization, the key issues are first to ensure the fit between objectives (i.e., outputs and service delivery) and beneficiary priorities and second the fit between resources (human, financial, and material) available for implementation and the internal arrangements among collaborating or competing institutions, all with demands for these scarce resources. • From the perspective of the beneficiaries, fit involves both participation in program benefits (i.e., both an individual right to quality services and a collective obligation to promote healthy community behavior) and participation in decision- making (i.e., voice, representation, etc.). • 16 David C. Korten, Community Organization and Rural Development: A Learning Process Approach, Public Administration Review, Vol. 40, No. 5 (Sep. - Oct., 1980), pp. 480-511. 16 60. The terms of reference for this study focused essentially on only two of the three dimensions: program organization and institutional arrangements, a full analysis would have required taking into account as well the perspectives of the beneficiaries. 61. Second, the terms of reference did not explicitly mention accountability, and yet it is implicit in each of the various relationships pictured above.17 If the dialogue with the MRH program managers had used accountability as the starting point of the discussions, the results of the study would probably have been very different. 17 The World Development Report 2004 devotes a large portion of its analysis to the concept of accountability. See World Bank, Making Services Work for Poor People (Washington, DC 2003), especially chapters 3-6. 17 Annex 1 62. Questions. The issues outlined in the terms of reference comprise: (i) an analysis of the institutional and managerial arrangements for budgeting set-up in the country; and (ii) a review of the results of the budgeting and expenditure process. 63. Institutional and managerial arrangements for budgeting. In each country, the MRH units were headed by qualified program managers with lengthy experience in the public sector and knowledge about government budgets. That these managers did always possess adequate budgeting skills or extensive knowledge of the budget process is not surprising, what is surprising is the extent to which these managers tended to be informed of rather than active participants in the ministerial decisions concerning allocation of the prospective budget. In Benin (with the DPP) and in Eritrea (with the DOAF), budgeting tended to be the domain of a limited group within the Ministry. 64. While Benin’s MRH program had attempted to estimate the costs of its national strategy and Eritrea’s MRH program was planning to do so as well, neither exercise was sufficiently structured to produce the hoped-for results: • Neither Benin nor Eritrea started with any evidence-based studies linking funding with results to initiate negotiations for additional MRH funding. Data from the Benin’s health management information system would have permitted the program to make the argument that lack of funding has stymied progress, but it did not; and neither Eritrea’s nor Ethiopia’s program produced the necessary data. • Neither Benin nor Eritrea has a realistic medium-term budget ceiling for MRH, without which the notion of presenting the comprehensive program financing needs simply gets lost in long lists of interventions and activities. • Though the programs have estimated (or at least intend to estimate) costs, none has formulated a financing plan, that is, mapped out how the various services and their associated categories of expenditures will be funded by the different sources of financing in a manner which takes builds on the advantages and disadvantages of these sources. 65. Though there is much consideration of advocacy in the strategic orientations of both Benin and Eritrea, there are few details about how this advocacy is to be done, whether from inside with personnel and facilities making the case (for improved training, better equipment, etc.) or from outside through the use of civil society groups to pressure local and national authorities on behalf of mothers and their children. 66. Review of the budgeting and expenditure process. In no country was the information available to fully anser the questions raised in the terms of reference. • Information was available in Benin and Eritrea on total health sector expenditures, but assessing the adequacy of these amounts and of the MRH program’s share is difficult because: (i) MRH related expenditures are scattered throughout the budget (even in Benin where MRH constitutes one of the five major programs); and (ii) the determination of adequacy, particularly within the context of “other competing priorities� is too vague. As a factual matter, what the assessments found was that, by using various measures to compare allocations and expenditures over time, the 18 proportion of actual and projected spending on MRH was not increasing in any of the countries. • In none of the countries was it possible to determine either the financial gap between the amount requested and the amount received or the activities that were not funded as a result of budget reductions. In theory, such a comparison would be possible, since initial program budgets are prepared as part of the Ministry’s total budget request and then prepared again based on the overall allocation from the Ministry of Finance and the internal allocations within the Ministry of Health. In practice, the task would be very time consuming and probably not as relevant as determining the gap between program allocations and program expenditures, where the country assessments found that budget allocations were always greater than expenditures. • In Benin and Eritrea, data on the major types of expenditures (particularly personnel and drugs) were available and were funded somewhat similarly: (i) budget ceilings for salary costs and expenditures were established and executed by the Ministry of Finance; and (ii) drugs and consumables (including contraceptive materials) comprised a patchwork of funding sources, procurement methods, and distribution modalities. It was not possible to determine the adequacy of the funding, although in both Benin and Eritrea MRH (and other) program managers spoke of insufficient staff and of periodic stockouts of drugs and consumables. • Shortfalls in financial resources are expected by MRH program managers, and their responses are consistent, given their constraints on the use of existing resources: (i) use government funds which have already been earmarked for specific activites because modifications are virtually impossible within a fiscal year; and (ii) either convince the usual partners to assume the costs of any unforeseen priority activity or postpone the activity until a later date. 67. Methodologies. Based on research undertaken in selected countries, the study had somewhat mixed results in developing the intended methodologies: • The country assessments did develop approaches for analyzing the existing technical, financial and managerial arrangements for delivering MRH services but did not really succeed in determining the extent to which they might be combined to contribute to increase the availability of resources needed to improve reproductive health results. • The missions did propose measures for discussion among the selected countries to improve institutional relationships, increase programmatic control over the amount and allocation of financial resources, and the report proposes several methods for improving strategic budgeting in a way that would provide more control (though not necessarily increase) the amount and allocation of financial resources. The further application of this approach in a specific country during the budgeting process would be an interesting follow-up exercise. • This report suggests that while there is indeed a need to strengthen the links between program planning and budgeting on the one hand and the continuous availability of necessary inputs at the health facility level, the means to do this are not really within the current possibilities of the MRH program (and perhaps not even the Ministry of Health). 19 68. In short, the major methodological contribution of this study concerns the assessment of current program interventions and the analysis of future program orientations and the budgeting of adequate allocations to implement these orientations. 20