Institutionalization of National Health Accounts: The Experience of Madagascar Paper prepared for the World Bank NHA Initiative March 11, 2009 1
List of Abbreviations CRESAN DEP ETIMCNS INSTAT MoH MTEF NHA PDSSPS PHRD PRSP USAID WHO World Bank Second Health Sector Support Project Directorate of Studies and Planning Inter-Ministerial Technical Team for the National Health Accounts Malagasy National Institute of Statistics Ministry of Health and Family Planning Medium Term Expenditure Framework National Health Accounts National Health Sector Development Plan Policy and Human Resources Development grant Poverty Reduction Strategy Paper United States Agency for International Development World Health Organization 2
I. BACKGROUND AND COUNTRY CONTEXT Country Health Sector Issues 1. Madagascar is one of the World s poorest countries: over two-thirds of the Malagasy population lives below the poverty line. The last decade has witnessed marked improvements in basic social indicators, albeit from a low base. Child mortality rates have declined significantly, from 159 deaths per 1,000 live births in 1997 to 94 in 2003/2004, and immunization rates significantly improved from 53 percent of all children 12-23 months fully immunized in 2003/2004, to 71.5 percent in 2008. Chronic malnutrition rates of children under the age of three decreased from 43 percent in 1997 to 32 percent in 2007. The prevalence of HIV/AIDS in the country remains low, but has rapidly increased among high risk groups and there are an estimated 180,000 persons living with HIV in the country. Madagascar s health indicators are better than other African countries at a similar income level, but they remain low, particularly among the rural population and the urban poor. Thus, although there are encouraging developments, there is still a long way to go given where Madagascar is today relative to the rest of the world. 2. Health is a key pillar of Madagascar s poverty reduction strategy, developed in a National Health Sector Development Plan (PDSSPS), which identifies a number of bottlenecks to increased access and use of health services and improvements in health indicators across the population of Madagascar, including four key areas of weakness: (i) inadequate demand for health services and low levels of utilization, (ii) uneven staffing of health facilities, especially in rural and remote areas: (iii) poorly equipped health centers and low levels of capacity to produce and deliver health services, and lastly (iv) low levels of health financing and inefficiencies in resource allocation. 3. Madagascar spent around US$6 per capita on health care in 2005, significantly lower than the average of US$15.4 per capita for sub-saharan Africa 1. Despite increases of the resource envelope of the Ministry of Health and Family Planning (MoH) since 2005, the budget in 2008 of around US$144 million, or 1.6 percent of GDP, was not sufficient to adequately finance the implementation of this PDSSPS. Moreover, the budget execution rate, although improving, has remained weak with estimates at 73.4 percent in 2007. Furthermore, even when resources are available, they have been allocated in ways that do not necessarily favor the poor, are not sufficiently directed to basic health centers, and the formula for allocating health resources does not take into account demographic or socio-economic differences across the regions. As a result, high impact health interventions, especially those needed to improve child survival are not targeted to where they are most needed. History of the National Health Accounts (NHA) in Madagascar 4. Recognition of these shortcomings in financing and allocative efficiency has spurred the initiative to develop the NHA in Madagascar. Prior to 2003, available health expenditure data was limited to administrative data basically capturing budgetary allocations to the MoH as well as data on execution by budget line. 1 excluding South Africa. 3
5. Thus in November 2001, during a World Bank mission to Madagascar to analyze and discuss the policy framework of the health system in the context of the Poverty Reduction Strategy Paper (PRSP), the idea of carrying out a NHA study was first considered. The NHA was seen as an important exercise in Madagascar for two reasons: i. in 1998 user fees for health services were introduced in all public health facilities to improve the performance of the sector: NHAs were seen as an opportunity to gather the necessary data for a comprehensive evaluation of this program; ii. MoH was about to adopt a new decentralized structure and the NHA would allow to evaluate various alternatives and serve as the base for post-decentralization comparisons. However, due to lack of finances, the study was not launched in 2002 (analyzing data from 2000) as initially hoped for. 6. In Madagascar, the NHA is currently championed jointly by WHO and the World Bank as an important instrument for obtaining much more detailed information on health sector spending beyond public sector spending, as well as on the composition and origins of these resources and the flow of funds at various administrative levels. Incidentally this fitted into the broader policy agenda of the Malagasy government as well as that of the MoH and donors. 7. Two NHA exercises were carried out in Madagascar. The first NHA study was undertaken on the basis of data from the year 2003 but the results were only made available late 2005/early 2006. Funding for the first NHA study was secured in 2004 via a Policy and Human Resources Development grant (PHRD) to CRESAN, the WB-financed health project. To carryout the NHA, the Maryland-based consulting firm Abt Associates was selected, based on the technical quality of its proposal and significant experience in carrying out NHAs notably in francophone Africa. On the Malagasy side, an inter-ministerial technical team (ETIMCNS) was constituted bringing together representatives of the MoH and specialists from the National Institute of Statistics (INSTAT). The mission of the consulting firm, as defined in detailed TORs, was to not only roll out the first series of NHA but also to transfer the skills and knowledge to the client and build the capacity of the ETIMCNS. The results of the NHA 2003 served as an input to the preparation of the Medium Term Expenditure Framework (MTEF) for 2006-2008. 8. A second NHA study is being carried out for the year 2007, with WHO support and the services of a consultant from Niger with extensive experience in NHA. Its results were expected to be made available in 2009. However, a specific date is not yet available. Financing of the private sector data collection is on hold because of the suspension of USAID activities in Madagascar due to the current political situation. 9. Since the government is still conducting the 2007 NHA study there are no plans yet for a third round of NHA. 10. Initially, the NHA was to be carried out wholly by the MoH staff, with guidance from two international consultants and a Scientific Advisory Committee guaranteeing international standards of the exercise. While there is a consensus among donors and the MoH that institutionalizing the NHA would be desirable and while the potential to inform policy is recognized, this has not happened effectively. Three sets of factors can be outlined: a) Lack of foresight on behalf of the donors 11. When the first NHA was carried out, it was conceived and packaged as a single, self standing exercise. The exercise did seek to transfer knowledge (more on that below) but funding was secured only for one round of data collection, analysis and dissemination. The longer term financing sustainability of the NHA process was not factored in. 4
b) Lack of resources and administrative issues 12. The MoH did welcome the NHA initiative but did not create the environment to make it sustainable. The office of the Directorate of Studies and Planning (DEP) was tasked with following up on the NHA. This is a well defined unit but it did not receive additional resources or staff for NHA institutionalization. In addition, the assignment of this task to the DEP was made purely following the functional distribution of responsibilities within the MoH, notwithstanding the fact that the DEP director himself was most likely not the most qualified (or interested) person within the MoH to carry out this assignment. Moreover, the DEP staff currently lacks the skills required to undertake and analyze NHAs; it does not have a single statistician/economist (let alone one that could be dedicated solely to the NHA on a full-time basis). An alternative would be to create a specific post within the MoH with capacity to coordinate continuous data collection from different agencies (of the MoH, other Ministries and other financial partners), adequate training to identify necessary technical resources, and proper time-availability. c) Lack of broader Government buy-in 13. While the NHA has had significant buy-in from the MoH, no real efforts have been made to advocate the concept more broadly in the Government (specifically with the Ministry of Finance). As a result, the role of NHA to guide policymaking or to serve as input into the budgetary allocation process is undermined. In fact, the application of the NHA is limited to the ability of the MoH to effectively use the data. 14. The results of the NHA have been widely used by donors and the MoH. Key statistics are often quoted and the NHA has fed directly into the elaboration of the MoH s MTEF. This said, the case of Madagascar is probably a good illustration of the limitations inherent in the use of a sophisticated instrument in a broader policy context where most policy decisions and budgetary allocations are not made based on evidence (for a number of reasons that relate to capacity and governance). The key lesson is that the instrument itself and the availability of the data are not sufficient in an environment where the governance framework for using data to inform policy is almost non-existent at decentralized levels. However, the planned World Bank health sector project has an Institutional Strengthening component which aims to provide support to all levels of the MoH for evidence-based planning and use of data for decision-making. II. KEY ELEMENTS OF NHA INSTITUTIONALIZATION Institutional / Governance Structure Capacity and Resources 15. While the initiative to develop the NHA has been endorsed and strongly supported by the MoH, its institutionalization has not been enshrined in the law (such as a law mandating data collection). Instead, as explained above, this was initially a donor-led effort (with MoH buy-in). There was a belief, which was naïve in retrospect, that (i) technical capacity would be transferred as a result of the first study, and (ii) the MoH would eventually make resources available to guarantee the follow-up. This has not materialized. 16. As already explained, the unit responsible functionally for the NHA institutionalization within the MoH is ill-equipped to do this both in terms of human and of financial resources. Interestingly, this is despite the fact that technical capacity does exist in the country (notably at the INSTAT). As a result, the first and second NHA exercises have been almost entirely financed by donors. Even so, this has not been done within a pre-established framework of financing for 5
multiple exercises but largely ad hoc according to donors funding availability and ability to mobilize technical expertise. This has had important consequences for consistency. Consistency and Methodology 17. The MoH has a system in place to extract the relevant public sector data from its financial reporting and monitoring system. However, there is no system in place for tracking financial flows outside of the public sector. Moreover, as a result of the ad hoc planning and funding process, no consistent methodology has been adopted and retained for analyzing the data. In fact each donor or implementing agency has tended to use its own software and programs. For instance Abt Associates, which was recruited as a consultant by the Bank, used software that it has developed exclusively. WHO, which is providing the technical expertise for the second exercise is using a different set of programs. Information and Dissemination 18. Results from the first NHA study have been made widely available and disseminated proactively through workshops funded under the PHRD. As a result, the data has been used and quoted widely. The report itself is available on the WHO s National Health Accounts website for Madagascar (http://www.who.int/nha/country/mdg/en/). However, the complete dataset is not readily available on the internet but can be requested from the MoH. Policy Impact 19. Results from the first NHA study have been used to prepare the MTEF 2006-2008. While public sector data is readily available for the MoH to use in budget planning, the private expenditure component is not adequately documented and readily available. Therefore, a regular updated NHA is important for the proper budgeting and planning process within the MoH. 20. The initial purpose of the NHA was to provide data for the analysis of the user fee policy in Madagascar as well as baseline data for evaluating the effect of decentralization. Since the second round of NHA is not yet available, these studies have not been undertaken. III. OTHER CRITICAL FACTORS AND KEY COUNTRY LESSONS Ensuring Effective Transfer of Knowledge 21. A key assumption about the institutionalization process of the NHA was that capacity would be transferred from the consultant contracted by the Bank to the MoH - during the first NHA study in a learning-by-doing fashion. Several factors can be stressed to explain ex post why this did not happen. This is partly a failure of the consultant to carry out its duties, but also of donors (including the World Bank) to fully appreciate the complexity of such capacity building. 22. In addition, financing considerations and procurement constraints had a direct impact. In 2005, CRESAN called for bids to procure the services of a consulting firm for the first NHA study. The only proposal that was technically acceptable was from Abt. However, Abt s financial proposal was significantly above what CRESAN had budgeted for the work. Since the financing came from a PHRD grant with a limited window of opportunity, the Bank and CRESAN decided to go ahead anyway and requested that Abt should prepare a new leaner proposal with a reduced price tag. The result, although not fully appreciated by CRESAN or the Bank at the 6
time, was a proposal that had serious shortcomings and the potential to compromise the quality of the work. 23. Abt reduced the number of missions to Madagascar and came only three times (another time with a very small mission) for short periods of time to work briefly with DEP, the Directorate for Financial Affairs, the Directorate for the Development of Sanitary Districts, and the Directorate for the health of Mother and Child all within MoH. These short missions were not sufficient to ensure the transfer of technical skills or to build a real relationship with the client. Abt used its own software and the analysis was done almost entirely in the Abt US headquarters. The interactions between Abt and CRESAN and the MoH were carried out mainly through directive/prescriptive emails by Abt (typically to request raw data), which were not conducive to establishing a fruitful dialogue with the Malagasy side. Abt produced reports and communications in very poor French. At the end of the exercise, the MoH was not technically able to carry out the work on its own, as had initially been expected. Abt has never received the last payment installment because of this. Ensuring that a National Institutional Framework Exists 24. The NHA has three components: public, private, and household spending. The public spending data is collected by the MoH, while the private and household data is collected by INSTAT, which is the only institution in Madagascar with the ability to carry-out full blown household surveys. INSTAT has a parastatal status which allows its staff to receive consultant fees. Initially INSTAT was supposed to be contracted and paid directly by Abt. However, for a number of reasons this did not happen during the first round and CRESAN had to pay INSTAT directly for their services. This was and continues to be is a source of tension as there is an important discrepancy between the involved MoH staff (receiving no extra payment) and the INSTAT staff involved in the other modules (but paid much higher fees). Given the significant difference in pay between the two parties, it is difficult to motivate MoH staff to be actively involved in this exercise. The DEP raised this as a critical issue and if NHA is ever to be institutionalized, the role of INSTAT and the issue of the differences in fees will need to be addressed. Finally, further complicating the situation is the fact that the MoH is not legally able to hire INSTAT. A clear contracting/hiring/payment mechanism is therefore necessary to correctly and sustainably involve INSTAT in the NHA exercise. Having Realistic Expectations 25. The current events rocking the island of Madagascar 2 provide a fitting illustration of the need for donors to have realistic expectations of the sustainability of their interventions as well as on the ability of the Government in the absence of continued and sustained financial and technical assistance - to institutionalize exercises such as the NHA that are very demanding and taxing in terms of financial and human resources. 26. An additional, more specific point well-illustrated by the case of Madagascar is that the agencies functionally or nominally responsible for carrying out and pushing the 2 In March there was a coup d etat in Madagascar, president Marc Ravalomanana was forced to resign as President of the Republic, and Andry Rajoelina subsequently was installed as the de facto head of state. He subsequently named a new government. This process was condemned by the international community, and technical and financial partners have suspended all non-humanitarian assistance to Madagascar. 7
institutionalization may not be the most able or motivated to in fact do so. Therefore, in some cases there may be a need to create a permanent team within the MoH that is responsible for and has resources to carry out the NHA. General lessons learned 27. In addition to the above mentioned country specific insights, the Malagasy experience offers a few general take-aways: a. Effective knowledge transfer will occur if the exercise is conducted in the country, by a local agency. Building national capacity to use all the tools necessary to independently conduct the exercise takes time and cannot occur during short term missions. b. Any technical assistance offered to the government needs to include a comprehensive long term plan for financing and capacity building. Financing should be secured for all stages of the exercise (data collection and cleaning, analysis, report writing and dissemination). c. Common software should be utilized for all rounds of the NHA in order to ensure comparability among reports, and build upon and strengthen existing in-country capacity. d. Clear roles and responsibilities for data sharing and collecting should be established to ensure that this process is repeatable. Persons interviewed Rozenn Le Mentec health economist with CRESAN supervising the NHA exercise (ex) Dr Josué Lala Andriamanantsoa Director of Planning and Evaluations, MoH (ex) Dr. Remy Rakotomalala National Coordinator for the Project Management Unit, MoH (current; ex: Director of Evaluations for the Project Management Unit) 8