71857 A REVIEW OF HEALTH SECTOR AID FINANCING TO SOMALIA (2000-2006) August 2007 Emanuele Capobianco ecapobianco@worldbank.org Veni Naidu veni@worldonline.co.za ACKNOWLEDGEMENTS The authors thank the individual donors, UN agencies and International NGOs who kindly participated in the study sharing data, attending consultative meetings and providing guidance during and after the data collection. Special thanks to the Health Sector Committee (HSC) members of the Coordination of International Support to Somalis (CISS) and in particular to the HSC Chair, Dr. Renato Correggia, the former HSC Coordinator, Dr. Imanol Berokoetxea, and the HSC Programme Associate, Ms Grace Nduma, for providing the researchers with the opportunity to engage with all key stakeholders. The authors are particularly grateful to the following experts who provided valuable comments to improve the first draft of the report: Christopher Lovelace, Son Nam Nguyen, Christopher Walker (World Bank), Ellie Bard (DFID), Francesco Checchi, Egbert Sondorp (LSHTM), Alessandro Colombo (WHO), Ed Cooper (IFRC) Austen Davis (UNICEF), Elena Vuolo (GWU) and Enrico Pavignani. Typesetting of the report was done by Mrs Illona Budiscak, whose excellent contribution is acknowledged. ii ACRONYMS AND ABBREVIATIONS BCG Bacillus Calmette-Guerin CAP Consolidated Appeals Process CCM Country Coordination Mechanism CDC Centers for Disease Control and Prevention CISS Coordination of International Support to Somalis DAC Development Assistance Committee DFID UK Department for International Development DPT3 Diphtheria, Pertussis and Tetanus DRC Democratic Republic of Congo EC European Commission ECHO European Commission Humanitarian Office EPI Expanded Program on Immunization EU European Union FGM Female Genital Mutilation FAO Food and Agriculture Organization of the United Nations FSAU Food Security Analysis Unit FTS Financial Tracking System GAVI Global Alliance for Vaccines and Immunization GDP Gross Domestic Product GFATM The Global Fund to fight AIDS, Tuberculosis, and Malaria HIV Human Immunodeficiency Virus HSC Health Sector Committee ICRC International Committee of the Red Cross IDP Internally Displaced Person IEC Information-Education-Communication IFRC International Federation of the Red Cross INGO International Non-Governmental Organization JNA Joint Needs Assessment M&E Monitoring and Evaluation MDG Millennium Development Goal MDR-TB Multi Drug Resistant Tuberculosis MICS Multiple Indicator Cluster Survey MoH Ministry of Health MSF Medecins Sans Frontieres NGO Non-Governmental Organization OCHA Office for the Coordination of Humanitarian Affairs ODA Official Development Assistance ODI Overseas Development Institute OECD Organization for Economic Cooperation and Development OFDA Office of Foreign Disaster Assistance OVC Orphans and Other Vulnerable Children PEPFAR US President’s Emergency Plan for AIDS Relief PHC Primary Health Care RDP Reconstruction Development Plan iii SACB Somalia Aid Coordination Body SWAp Sector Wide Approach TB Tuberculosis TFG Transitional Federal Government UN United Nations UNAIDS The Joint United Nations Programme on HIV and AIDS UNCAS United Nations Common Air Space UNCT United Nations Country Team UNDP United Nations Development Program UNFPA United Nations Fund for Population Activities UNHCR United Nations High Commissioner for Refugees UNICEF United Nations Children’s Fund UNIFEM United Nations Development Fund for Women UNOSOM United Nations Operation in Somalia USAID United States Agency for International Development WB The World Bank Group WHO World Health Organization iv TABLE OF CONTENTS EXECUTIVE SUMMARY...........................................................................................viii 1. BACKGROUND, AIM AND OBJECTIVES.......................................................... 1 2. CONCEPTUAL FRAMEWORK ............................................................................ 6 Trends in Overall Aid to Developing Countries .......................................................... 6 Trends in Aid to the Health Sector............................................................................... 7 Trends in Aid to Fragile States..................................................................................... 8 Trends in Aid to Somalia ........................................................................................... 10 Donor Harmonization: Principles and Challenges ..................................................... 12 3. METHODOLOGY................................................................................................. 14 Approaches................................................................................................................. 14 Quantitative and Qualitative Data Collection Process ............................................... 14 Types of Quantitative Data Collected ........................................................................ 15 Methodological Limitations and Challenges.............................................................. 17 Usefulness of the Data................................................................................................ 18 4. KEY FINDINGS.................................................................................................... 19 Financial Aid Flows ................................................................................................... 19 Total Health Sector Aid Financing (2000-2006)........................................................ 20 Health Sector Aid by Disease/Program (2000-2006)................................................. 24 Health Sector Aid by Zone (2000-2006).................................................................... 32 Health Sector Aid by Activity (2006 expenditure) .................................................... 33 5. QUALITATIVE FINDINGS ................................................................................. 35 Aid Financing............................................................................................................. 35 Donor Harmonization................................................................................................. 37 6. CONCLUSIONS AND RECOMMENDATIONS ................................................ 38 Conclusions ................................................................................................................ 38 Recommendations ...................................................................................................... 41 v LIST OF TABLES Table 1: Millennium Development Goals relevant to the health sector and Somalia...... 2 Table 2: External aid allocated to health care in fragile states (18) ............................... 10 Table 3: Top ten donors of gross ODA to Somalia (2000-2005 average US$) ............. 11 Table 4: 2006 DAC classification of fragile states ........................................................ 11 Table 5: Percentage difference between data collected from donors and recipient/implementing agencies.................................................................................... 16 Table 6: Total health sector aid financing using current and constant rate of exchange (2000-2006).................................................................................................................... 21 Table 7: Per capita (US$) health sector aid financing.................................................... 24 Table 8: Health sector aid (US$ million) by disease/program (2000-2006) .................. 24 Table 9: Confirmed wild poliovirus cases in Somalia (2000-2006) .............................. 26 Table 10: Nutrition indicators for children under 5 years of age ................................... 29 Table 11: Expenditure by activity (2006) ...................................................................... 33 LIST OF FIGURES Figure 1: Four fragile state contexts and Somalia’s asymmetric situation ...................... 1 Figure 2: Trends in ODA per capita to Somalia (1999-2003).......................................... 4 Figure 3: Actual and projected official development assistance (1990-2010)................. 6 Figure 4: Long term capital flows to sub-Saharan Africa and other developing countries (2003) ............................................................................................................................... 7 Figure 5: Development assistance for health by source (1997-2003).............................. 7 Figure 6: Private and public health expenditure in sub-Saharan Africa (2002)............... 8 Figure 7: Percentage of humanitarian assistance to total aid budget (2002).................... 9 Figure 8: Bilateral ODA to Somalia by sector (2000-2005) .......................................... 12 Figure 9: Explanations for the difference between donor disbursement and recipient/implementing agencies expenditure................................................................ 17 Figure 10: Financial aid flows in the Somalia health sector .......................................... 19 Figure 11: Total health sector aid financing (2000-2006).............................................. 21 Figure 12: Total health sector aid financing by donor category (2000-2006) ............... 22 Figure 13: Percentage contribution of health sector aid financing by donor category, all years (2000-2006) .......................................................................................................... 23 Figure 14: Percentage contribution of health sector aid financing (2000 and 2006) ..... 23 Figure 15: Percentage contribution by programs by year (2000-2006) ........................ 25 Figure 16: Health expenditure: TB, Malaria and HIV (2000-2006) .............................. 25 Figure 17: Health expenditure: Poliomyelitis (2000-2006) ........................................... 26 Figure 18: Health expenditure: Tuberculosis financing versus TB case detection and TB success rate (2000-2006)................................................................................................ 27 Figure 19: Health expenditure: EPI financing versus BCG and measles coverage (2000- 2006)............................................................................................................................... 28 Figure 20: Health expenditure: Reproductive health (2000-2006) ................................ 28 Figure 21: Health expenditure: Nutrition financing versus malnutrition indicators (2000-2006).................................................................................................................... 29 Figure 22: Health expenditure: Non communicable diseases (2000-2006) ................... 30 vi Figure 23: Health expenditure: Health systems support (2000-2006) ........................... 31 Figure 24: Health expenditure: Primary health care (2000-2006) ................................. 31 Figure 25: Health expenditure: Emergency (2000-2006) .............................................. 32 Figure 26: Distribution of health expenditure by zone: all years (2000-2006).............. 32 Figure 27: Distribution of population and health expenditure by zone: all years (2000- 2006)............................................................................................................................... 33 LIST OF ANNEXES Annex 1: Study sample (26 organizations) .................................................................... 44 vii EXECUTIVE SUMMARY Background The study on the 2000-2006 aid flows to the health sector in Somalia is a first attempt to fill a large gap of knowledge in this area. The primary objectives of the study were to assess how levels of donor financing varied over the years; which health interventions were prioritized by donors; and how evenly health sector aid was distributed to the different zones of Somalia. The overall aim of the study was to create evidence for donors, implementers and health specialists involved in allocation of financial resources to the Somali health sector. The study was conducted in close collaboration with the Health Sector Committee (HSC) of the Coordination of International Support to Somalis (CISS). HSC members were consulted prior, during and after the completion of the data analysis. The HSC endorsed the research protocol in February 2007, provided guidance on the sampling frame, reviewed and discussed the preliminary findings and provided feedback on the draft report. The results of the study are based on quantitative and qualitative data collected from donor organizations active in Somalia. The study focused on traditional donors, including bilateral, multilateral and others (GFATM and Red Cross/Crescent Movement). Twenty-six agencies were contacted and a response rate of 96 percent (n=25) was achieved. Of the 25 organizations that responded, 88 percent (n=23) provided relevant health sector data. Key Findings Conventional donor funding for the health sector grew almost three fold in seven years, passing from US$23 million in 2000 to US$ 62 million in 2006. The panorama of health donors in the country changed considerably over the years. The contributions of bilateral donors decreased from 63 percent in 2000 to 35 percent in 2006. Multilaterals donors, and particularly the UN, considerably increased their share. New donors appeared on the scene: GFATM contributions increased from 2004 onwards to reach 22 percent of total donor aid to the health sector in 2006. Per capita aid financing for health grew from US$3 in 2000 to US$7 in 2006. The 2006 figure is conservative and is an approximation based on UNDP population figures. The figure is in line with per capita health aid financing in other fragile states such as South Sudan (US$7) and DRC (US$2-3). Somalia, like South Sudan and DRC, remains an “aid orphan�, despite the increase in funding for health observed in recent years. Health sector financing for health progressively shifted from horizontal to vertical programs. In 2006, malaria, TB, HIV and the polio programs alone accounted for 50 percent of total aid compared to 25 percent in 2000. The polio program has been the largest funded program in Somalia, accounting for 20 percent of the total health budget over the period 2000-2006; a polio outbreak in 2005 was the trigger for a further viii increase to polio financing in 2005 and 2006. The large funding for the polio program testifies the political commitment of the donor community, but also raises questions about the opportunity costs of the eradication campaign. Other vertical programs have benefited from increased funding provided by the GFATM since 2004. In some cases, for example TB, the large increase in funding does not seem fully justifiable based on program performance. Within vertical programs, EPI, reproductive health, non communicable diseases and nutrition received inadequate funding. Given the high burden of disease and the dramatically poor indicators on EPI coverage, reproductive health and nutrition, additional financial support would be required to reverse the negative trends in these areas. Support for health system strengthening was relatively large (36 percent of total aid financing over the period 2000-2006), declined in 2006, and was generally fragmented. Partners reported lack of coordination in this area and called for more joint planning. Recent initiatives like the development of the Reconstruction Development Plan (RDP) and the creation of a Health System Working Group in the CISS indicate a positive change towards more coordination among stakeholders. The distribution of aid to the health sector by zones was fairly evenly distributed during the period 2000 to 2006. Sixty-one percent of funding went to South and Central zone (which accounts for 52 percent of the population); 19 percent to Puntland (which accounts for 20 percent of the population); and 20 percent to Somaliland (which accounts for 28 percent of the population). Recommendations • Somalia requires increasing and long term financial support for the health sector to address the needs of its population. Although US$7 per capita aid for the health sector compares favorably with other sub-Saharan African countries, Somalia’s financial needs for the health sector remain high given the exceptionally low health indicators, and the high operational costs linked both to the difficult logistics of the country, and the reliance on international actors mostly located outside Somalia. • More importantly, contributions to the health sector should be made more strategic: funding gaps in key areas (EPI, reproductive health, nutrition and non communicable diseases) should be addressed as a matter of priority. Needs analysis and data on the burden of diseases are required to guide the prioritization process. However, known but neglected causes of high morbidity and mortality should be addressed without further delay. • While supporting thematic interventions, the donor community should invest more in rebuilding the health system. The findings of the JNA and the RDP represent the blueprint for systemic efforts and should guide future planning and implementation efforts. The recently created Health System Working Group should take the opportunity to lead donors and implementers in ix the rebuilding process, possibly drawing on the positive experiences of other fragile states (e.g. Mozambique, Afghanistan, Liberia, Rwanda, etc.). • The mix of relief aid and humanitarian assistance should be sustained until the political situation normalizes. Providing funds for emergencies will serve the immediate needs of the Somali population, while supporting activities that are more developmental in nature will lay the ground for the future health system in the country. A too rapid shift towards development may lead to loss of lives during emergencies. On the other hand, an excessive focus on humanitarian assistance may lead to a very weak health system emerging at the end of the transition phase. • A “one size fits all� strategy may not be appropriate for a fragile state such as Somalia. All three zones operate within differing political contexts. It is therefore essential that partners reflect this diversity in their financial and technical approaches to the health system of the country. • Coordination mechanisms could be strengthened, mostly in terms of priority setting and reducing transaction costs. The current mechanisms have been effective in fund raising and sharing of technical know-how. However, the function of assessing needs and establishing priorities for the health sector has for long lied outside the coordination mechanisms. More efforts could be done to bring together donors, implementers, political authorities, civil society, NGOs and religious leaders to jointly define health priorities and set outcome targets. Donors with limited in-country staff could introduce delegated cooperation arrangements with other donors to monitor aid and activities in the field. To reduce the administrative burden on implementers, a single financial reporting tool may be introduced. • Financial tracking of donor resources to the health sector should become an integral part of the health information system. To be meaningful, financial tracking should be matched with burden of disease and program outcome data. It should be expanded to include tracking of expenditure by key activities and future funding. By doing so, aid financing analysis could become an extremely useful tool for policy planners both at government and donor level. The data collection instrument developed for this study could be adapted to become part of a financial tracking tool for the health sector. • Operational research is needed to integrate the findings of this study and to allow a better understanding of health financing in Somalia. Topics to be studied include health financing by (i) the private sector, (ii) the diaspora through remittances; and (iii) non conventional donors. Studies on household spending on health would complete the picture by providing information on private expenditure. x The report is organized in six sections. Section 1 provides the aims, objectives and benefits of the study and contextualizes the study area, Somalia. Section 2 provides the conceptual framework for the research by (i) looking at recent aid financing trends in developing countries, in the health sector, in fragile states and in Somalia; and (ii) reviewing the literature on harmonization in fragile states. Section 3 describes the methodology, the data collection process, types of data collected and methodological limitations and challenges. Section 4 presents the quantitative findings in terms of (i) total health sector aid financing, and expenditure (ii) by disease, (iii) by zone and (iii) by activity. Section 5 presents the qualitative findings in terms of aid financing and donor harmonization. Section 6 offers conclusions linked to the three primary study objectives and provides recommendations for future funding and better harmonization in the health sector. xi 1. BACKGROUND, AIM AND OBJECTIVES 1.1 Since 1991 Somalia has experienced a prolonged humanitarian crisis due to a civil war still affecting large parts of the country. Somalia has been without a functioning central government since 1991, when the military regime collapsed and civil war among clan militias broke out. During the subsequent decade, there has been fragmentation of political power along clan lines in different parts of the country. In the first half of the 1990s in South-Central Somalia, the civil war destroyed most of the infrastructure, disrupted food supplies and displaced large populations (1). In 2004, a UN-backed Transitional Federal Government (TFG) was formed in South-Central Somalia largely based on clan affiliation. In 2006 the TFG was challenged by the sudden rise of the Islamic Courts, but in December 2006 the TFG regained power thanks to the military support from Ethiopian troops. As of August 2007, the TFG has yet been unable to bring about peace and security. The TFG authority remains challenged by the Islamic Courts and by the self proclaimed state of Somaliland (North Western Zone). Somaliland, twice affected by civil strife (1992 and 1994-1996), in recent years acquired a remarkable level of peace and stability: local elections and presidential elections were held in 2002 and 2003 respectively. The regional state of Puntland (North Eastern Zone) was formed in August 1998 as an autonomous self- governing entity. Puntland’s administrative structures are said to be less mature, infrastructure less developed and its sources of potential revenue more limited (1). 1.2 As a result of the conflict, three operational environments can be identified to contextualize Somalia (see Figure 1). Somaliland is placed in the “gradual improvement� quadrant, which is characterized by the presence of some government reform, but entrenched systems where change is often difficult and slow. Puntland is placed in the “post-conflict transition� quadrant which is characterized by the conclusion of peace and renewed international engagement; while South-Central Somalia is in between the “post-conflict transition� and the “prolonged crisis�. The prolonged crisis is characterized by no consensus between donors and government on development strategies. Figure 1: Four fragile state contexts and Somalia’s asymmetric situation Prolonged Crisis or Impasse Post-conflict Transition South-central Northeast Somalia (Puntland) Deterioration Gradual Improvement Northwest (Somaliland) Source: World Bank, 2007. 1 1.3 The man-made humanitarian crisis has been worsened by a series of natural disasters. In recent years major emergencies included several droughts (in Sool/Sanag and in Bay/Bakool), regular floods (in the Shabelle and Juba regions), and the 2004 tsunami (in Puntland). In 2005/6 failure of the autumn rains caused 1.7 million people to be in dire need, while 400,000 internally displaced people required assistance and protection (2). All the above natural events generated mass movement of people, caused major blows to the local economy and increased the needs of the affected population. 1.4 As a consequence of the prolonged civil strife and natural disasters, the health status of the Somalis (see Table 1) is among the worst in the world1. In 2003, the under-five mortality rate (deaths per 1,000 births) for low income countries was estimated at 123 and for sub-Saharan Africa at 171 (3). The rate for Somalia is estimated at 224 deaths per 1,000 births representing almost twice the rate of other low income countries and a third above the sub-Saharan African average. Similarly, the maternal mortality rate is among the highest in the world. Immunization rates are extremely low: DPT3 coverage in 2002 was 40 percent compared to 60 percent in Mozambique and 58 percent in Zimbabwe (4). Other health concerns include poor nutritional status, and high prevalence of communicable diseases, like TB and malaria, endemic in several parts of the country. HIV prevalence remains below one percent. However, experience from sub-Saharan countries has shown that when the rate of HIV exceeds one percent, it could double or triple in two to three years (5). Non communicable diseases such as mental health are also likely to place a heavy burden on the Somali population. Civil war and trauma have left Somali youth at high risk of developing emotional and psychological disturbances as found in a Canadian study on Somali immigrants (6). There are two additional health problems specific to Somalia. First, it is estimated that about 98 percent of women (5) have undergone some form of female genital mutilation (FGM). This practice carries immediate and long term health risks, including tetanus, hemorrhage, urinary tract infections and obstructed labor. Second, chewing of khat2 is a common practice in Somalia with serious economic, social and mental health consequences. Table 1: Millennium Development Goals relevant to the health sector and Somalia Indicator Somalia Somaliland Puntland South/Central Target for 2015 Under-five mortality rate 224 188 225 231 75 (per 1,000) (1999) (1999) (1999) (1999) Infant mortality rate (per 133 113 133 137 1,000 live births) 44 (1999) (1999) (1999) (1999) One-year-olds immunized 40% -- -- -- -- against measles (%) (2004) 1 Data presented in 1.4 need to be cautiously considered given the paucity and variable quality of data available on the Somali health sector. 2 Khat is an intoxicating plant, classified as an illegal drug in some countries 2 Indicator Somalia Somaliland Puntland South/Central Target for 2015 Children underweight for 26% 21% 26.8% 27.5% 13% age (% of children under 5) (2000) (1999) (1999) (2005) Maternal mortality rates 1,100 1,600 -- -- 275 (per 100,000 live births) (2002) (2005) Births attended by skilled 25% 27% 49.2% 27.6% -- health staff (%) (2002) (1999) (1999) (1999) Halt and HIV Prevalence (% adults 0.9% 1.4% 0.9% 0.5% reverse 15-49) (2004) (2004) (2004) (2004) spread of HIV Halt and Prevalence associated with 118 113 Reverse -- -- Malaria (per 100,000) (2000) (2005) spread of Malaria Halt and Incidence of TB (per 412 460 Reverse -- -- 100,000) (2003) (2000) spread of TB Source: Somalia Joint Needs Assessment (5) , Human Development Report 2006 (7) 1.5 A fragmented health sector provides limited services to the Somali population. The public health care network is small. It mostly relies on national and international NGOs that tend to be concentrated in towns and in secure areas. Direct provision by ministries of health is marginal. Private health care outlets have proliferated throughout the country, and are now estimated to be in the thousands, with large variations in their size, services offered, staff qualifications and performance. 1.6 Various donors have provided assistance throughout the 15 years of conflict in an attempt to alleviate the suffering of the Somalis. A peak of aid was provided between April 1992 and March 1995 during the UN Missions UNOSOM I and UNOSOM II (see Figure 2). The higher levels of aid during that period were linked to the broad mandate of the UN, including support for national reconciliation, rebuilding of the central government, and revival of the economy. According to OECD, from 1995 onwards ODA declined and stabilized between US$10 and US$20 per capita (8). 3 Figure 2: Trends in ODA per capita to Somalia (1999-2003) Source: OECD, 2006 1.7 Funding of the health services in Somalia has been largely depended on foreign aid. In the past 15 years, the health sector in Somalia has greatly benefited from large contributions from bilateral donors, UN Agencies and, more recently, global programs like the Global Fund to fight AIDS, Tuberculosis and Malaria (GFATM) and, on a smaller scale, the Global Alliance for Vaccines and Immunization (GAVI). Interventions in the health sector have covered the three main zones of Somalia and have been both developmental and humanitarian in nature. 1.8 Aid financing to the health sector in Somalia, however, has never been assessed. Although data on aid for the health sector are available through the reports compiled by the Somalia Aid Coordination Body (SACB) and the Office for the Coordination of Humanitarian Affairs (OCHA’s) Financial Tracking System (FTS), an in depth analysis of aid to the health sector has never been conducted. Quantifying levels of aid and identifying how aid has been prioritized is important to create evidence for health policy specialists/donors involved in planning and allocation of financial resources to the Somali health sector. This is particularly important in a fragile country like Somalia where health needs are almost infinite and resources are scarce. 1.9 In order to fill the knowledge gap, this study aims to quantify and analyze aid financing3 to the health sector in Somalia over the period 2000-2006. This is in line with the conclusion of the High Level Forum on the Health MDGs that “the key to successful post conflict health work is understanding the health sector and the context within which health services are being delivered. The task of analyzing this should start as early as possible, preferably before the start of the transition from war to peace�(9). Furthermore, the study analyzes the context of aid support to the health sector at a time when transition to peace already occurred in some zones such as 3 Aid financing in this study refers to public aid provided by developed nations, members of the Organization for Economic Cooperation and Development (OECD), to less developed nations. It is also referred to as Official Development Assistance (ODA). 4 Somaliland and which is yet to come in larger parts of the country. The timing of the study is considered crucial given the growing interest of the international community to financially support peace-building initiatives in the country. 1.10 The study’s primary objectives are: • To assess how levels of donor financing to the health sector varied throughout the years; • To understand which health interventions were prioritized by policy planners through financial aid allocations; and • To evaluate how evenly aid to the health sector was distributed to the different zones of Somalia. 1.11 In addressing the primary objectives, the benefits of the study are: • To highlight imbalances in aid support to the health sector. More specifically, to provide key information on the prioritization of health interventions based on availability of external aid and on regional differences. The results of the study may help stakeholders to redefine criteria and address imbalances for the allocation of resources to the Somali health sector. The study results could be used both in the scenario of continued conflict and in the event of transition to peace; • To provide a baseline for future research work on health aid financing in the country. An in-depth knowledge of the current resource envelope will facilitate the preparation of resource forecasts, which are central to the development of meaningful strategies in post conflict countries; • To provide health policy planners with evidence-based conclusions to address the main priorities identified by the High Level Forum on the Health MDGs: ensuring longer-term predictability of aid flows, reducing shorter-term aid volatility and promoting coordination, harmonization and alignment; • To assess the impact of global initiatives on the overall health budget and on the level of coordination among partners; and • To increase the scarce literature on Somalia health sector and the literature on health financing in fragile states and in Africa. 5 2. CONCEPTUAL FRAMEWORK Trends in Overall Aid to Developing Countries4 2.1 After a decline in the 90s, official development assistance (ODA) has increased since 1998, reaching $70 billion in 2003 (see Figure 3). In real terms ODA increased by seven percent in 2002 and by 3.9 percent in 2003. As a percentage of gross national income in OECD countries, assistance declined from 0.34 percent in 1992 to 0.22 percent in 2001 before increasing slightly in 2003 to 0.25 percent. The increases reported in the past five years are mostly due to debt relief, emergency and disaster operations, technical cooperation, and administrative overhead. Of the total nominal increase between 2001 and 2003, 66 percent went to debt relief and technical cooperation (10). Figure 3: Actual and projected official development assistance (1990-2010) 2.2 In sub-Saharan Africa, ODA sharply increased in recent years (see Figure 3). Sub-Saharan Africa received 54 percent of the total increase in ODA funding. ODA is the main source of external finance in sub-Saharan Africa, representing more than 55 percent of total external flows (see Figure 4). Foreign direct investment represented another 25 percent of the total long-term flows, remittances 15 percent, and other private flows five percent. In the rest of the developing world, where foreign direct investment and remittances account for the bulk of external financial flows, ODA accounts for only nine percent of such flows (10). 4 The analyses presented on trends in overall aid to developing countries and trends in aid to the health sector (paragraphs 2.1-2.3) mostly rely on Gottret B Schieber G.(10) 6 Figure 4: Long term capital flows to sub-Saharan Africa and other developing countries (2003) Source: World Bank 2005 Trends in Aid to the Health Sector 2.3 ODA for health has risen steadily in recent years (from about US$2 billion in 1990 to more than US$10 billion in 2003). This increase is mostly due to a rise in private philanthropic funding – for example, Bill and Melinda Gates Foundation - and global programs – such as GFATM, GAVI; Roll Back Malaria; and the United States President’s Emergency Plan for AIDS Relief (PEPFAR). In Figure 5 Michaud reflects the increasing importance of global partnerships and private philanthropic funding in development assistance for health. Figure 5: Development assistance for health by source (1997-2003) Despite the considerable ODA, government expenditure on health in sub-Saharan Africa tends to be low, while private spending exceeds public spending. Household out-of-pocket spending accounts for 80 percent of private spending and almost 50 percent of total health spending (11). 7 Figure 6: Private and public health expenditure in sub-Saharan Africa (2002) Source: Bitran (11) Trends in Aid to Fragile States 2.4 Fragile states tend to receive 40 percent less aid per capita than strongly performing countries (12). This is primarily due to disproportionately low aid flows from bilateral donors. Within fragile states, some countries receive substantially higher aid flows than predicted by poverty and policy (“aid darlings�) while a similar number receive substantially lower aid flows (“aid orphans�). In addition, fragile states experience much higher volatility in aid flows than other low income countries, almost double in the decade 1992-2002. 2.5 Supporting fragile states is difficult: it is costly, and is regarded by donors as very high risk (13). As mentioned by Bourguignon, “much aid is lost due to instability and conflict� (14). It is therefore not surprising that donors are reluctant to provide support to fragile states. According to Randel and German in the year 2000 the majority of DAC donors allocated less than 15 percent of their total aid budget to humanitarian assistance (15). 8 Figure 7: Percentage of humanitarian assistance to total aid budget (2002) Source: Randel & German, 2002 2.6 However, the costs of doing nothing or of failing to be effective, in both human and security terms have recently been recognized. Several studies have attempted to quantify the cost of neglect and the cost benefit ratios of investment in conflict prevention. According to McGillivray (16), on average for every GB£1 spent on conflict prevention, GB£24 of savings are generated; on the other hand, having a fragile state as a neighbor may generate losses of 1.6 percent of GDP every year (16). The above results should be read with caution, given their speculative nature but they do point towards substantial pay-off for investment in fragile states. An additional push for higher investment in fragile states comes from the Millennium Development Goal (MDG) agenda. With 14 percent of the world’s population and one third of the world’s poor, the MDGs cannot be achieved without results in fragile states (17). 2.7 With regard to the health sector, investment in post conflict countries are particularly important because they may: i) contribute to alleviate human suffering of many individuals; ii) support the peace process and iii) provide long- term return in terms of equity, efficiency and effectiveness of services provided (18). The case for investing in the health sector is also supported by stark health statistics. According to Cassels, (13) fragile states account for: • 60 percent of global disease epidemics; • One third of global maternal deaths; • One third of people living with HIV in developing countries; 9 • 50 percent of children dying before 5 years of age; • Malaria death rate 13 times higher than other developing countries; and • High malnutrition rates (one third of the population is malnourished). 2.8 Available data on aid to the health sector in fragile states (see Table 2) show that many countries are relatively neglected by the international community, thus falling into the category of “aid orphans�. Bradbury suggests that priorities in aid to the health sector tend to be set on the basis of donors’ political needs rather than on the objective needs of people in distress (19). Data presented in Table 2 seems to confirm that aid allocations may be swayed by geopolitical and media concerns more than by population needs. Table 2: External aid allocated to health care in fragile states (18) Country Year Per Capita Aid Source Cambodia 1994 $2 Lanjow et al., 1999 Mozambique 1995 $5 Ministry of Health 1997 East Timor 2000 $36 Tulloch et al., 2003 Kosovo 2001 $60 Southern Sudan 2003 $7 Health Secretariat of the new Sudan, 2004 DRC 2005 $2-3 World Bank, 2005 Source: Pavignani, 2005 2.9 It should be noted however that the global knowledge on aid financing to the health sector in fragile states remains limited. This is mostly due to the inherent difficulties of tracking financial flows in contexts characterized by high insecurity, frequent natural and man made catastrophes and political changes. Collecting financial data is also made difficult by fragmentation and ambiguities of roles within the donor community, incompleteness of information available, variety of planning cycles and budget formats, as well as resistance to share financial information (20). Trends in Aid to Somalia 2.10 Net ODA to Somalia has increased from US$174 million in 2003 to US$200 million in 2004 and US$236 million in 2005. Aid per capita grew from approximately US$22 in 2003 to US$30 in 2005 (21). The main donors in 2004 and 2005 were the EC, the US and Norway. Other donors are presented in Table 3. 10 Table 3: Top ten donors of gross ODA to Somalia (2000-2005 average US$) Source: OECD 2.11 Based on need, levels of aid, as well as policy and institutional performance of fragile states, in 2006 DAC classified Somalia in the group of “countries with low levels of aid, but relatively weaker policy and institutional quality� (see Table 4). The states belonging to this group suffer from acute fragility, have low resource allocations and high levels of need. In most cases there are indications that increased aid may not be warranted. Within this category Somalia was considered a priority for increased international attention and more coherent engagement (22). Table 4: 2006 DAC classification of fragile states Countries with low levels Countries with relatively of aid, but relatively Marginalized countries higher and/or improving weaker policy and aid levels institutional quality Burundi, Guinea, Nigeria, CAR, Cote d’Ivoire, Cambodia, Chad, Eritrea, Uzbekistan, Yemen, DRC Liberia, Togo, Myanmar, Gambia, Guinea Bissau, Somalia, Sudan, Zimbabwe Níger, Sierra Leone, Tajikistan 2.12 According to OECD, ODA for health in 2004-05 represented approximately five percent of the total aid to Somalia (21). Figure 8 illustrates that in 2004-2005, 67 percent of the contributions were used for emergency assistance. ODA also strongly supported education and other social sectors. 11 Figure 8: Bilateral ODA to Somalia by sector (2000-2005) Source: OECD 2.13 Other countries not represented on the OECD DAC have reportedly provided international aid to Somalia. Precise estimates are not available, though such aid is likely to be substantial given the large presence of Islamic charities mostly financed by the Gulf States. A recent study by ODI found that on a global scale non- DAC contributions constituted between one and 12 percent of total global humanitarian assistance reported on OCHA’s Financial Tracking System in the period 1999-2004 (23) 2.14 Remittances are vital to the Somali economy. It is estimated that remittances to Somalia may be roughly four times official development assistance (24). It is also estimated that between 25 and 40 percent of all families in Somalia receive remittances (25). At least half of remittances are used for direct consumption by the household, including education and health. As demonstrated in the education sector (26), Somalis in the diaspora, mostly emigrants and business people, provide considerable finance for initiatives to support local schools, including staff salaries. Contributions are often channeled through informal solidarity networks. However, finances from the diaspora also support structured organizations such as local NGOs and training institutions. In the health sector, diaspora remittances appear to be used for i) meeting the needs for common and emergency medical treatment; and ii) supporting health facilities, principally hospitals (27). Donor Harmonization: Principles and Challenges 2.15 Donor Harmonization has been central to the development agenda since the 2005 Paris Declaration on Aid Effectiveness (28). Donor harmonization is intended to foster constructive engagement and support existing dialogue and coordination processes in countries with problems of weak governance and conflict or temporary fragility (29). The goals of donor harmonization are to coordinate donor activities internationally so as to keep administrative burdens low, utilize resources more efficiently and achieve a fair distribution of funds among various regions and sectors (30). 12 2.16 International interventions can create societal divisions and worsen corruption and abuse, if they are not based on strong conflict and governance analysis (29). The principles for good engagement in fragile states call for tracking of aid governance trends with the potential to adjust aid as a last resort in serious cases of corruption (29). Recipient partner countries must demonstrate the willingness to implement good governance measures (30) and must ensure ownership and mutual accountability (31). Furthermore, donors should avoid neglecting any geographical region within a country. Real or perceived discrimination in fragile states can lead to increased conflict and service delivery failures (29). 2.17 There should be practical coordination mechanisms between international actors. In the absence of strong government leadership, donors should take on the responsibility for joint assessments, shared strategies and coordination of political engagement (29). Practical initiatives such as joint donor offices, agreed division of labor amongst donors, delegated cooperation arrangements5, multi-donor trust funds and common reporting and financial requirements should be implemented (29). Effective implementation of harmonization depends on vigorous donor leadership in- country. Donors should create top-level advocates as champions of harmonization in their organizations with decentralized decision making to in-country staff (32) to engage constructively with partners and other donors (31). 2.18 Large numbers of actors (government and non-government) involved in donor and recipient countries render the tasks of planning and coordination less efficient (30). Large numbers of projects in a single country may also be an enormous task for the country’s administration in terms of annual reporting and monitoring and evaluation (30). This may be contrary to the aims of harmonization which is intended to reduce transaction costs on all sides. The range of actors in fragile states (humanitarian, diplomatic and military) each with its own objectives and approaches render the development of harmonization strategies more difficult (31). 2.19 Various donor harmonization strategies can be implemented in fragile states. In states that are fragile but with commitment and growing capacity, full alignment may be feasible (31). In cases of concerns about legitimizing a particular government, or a prolonged humanitarian presence, it is important for donors to focus harmonization efforts on the creation of mechanisms that enhance the emergence of country leadership and ownership (31). Donors may also choose to partially align themselves to programs that have sufficient capacity and commitment whether in agencies or regional governments. This is called shadow alignment (31). 5 This is an arrangement in which one or several donors with limited capacity or no country presence transfer management of funds and fiduciary responsibilities to other donors to act on their behalf (34). 13 3. METHODOLOGY Approaches 3.1 The following approaches were used to study aid financing to Somalia over the period 2000 to 2006: • Consultations were held with the Health Sector Committee (HSC) of the Coordination of International Support to Somalis (CISS), formerly known as SACB. - The research protocol was presented in February 2007 and endorsed in March 2007 by the HSC, which represents all donors, UN agencies, NGOs and Somali authorities involved in the health sector. HSC members played an important role in facilitating the data collection process at their agencies and provided critical input during the data analysis phase. • Literature review - Secondary data were collected through a literature search on aid flows in the global context, and more specifically on Somalia and the health sector. The review also focused on donor harmonization in fragile states. Desk reviews of relevant literature on Somalia were carried out. • Quantitative questionnaire - Primary quantitative data collection on financial and in-kind contributions were undertaken via the use of a financial tool developed and pilot tested by the researchers in March 2007. To identify donors (multilateral, bilateral and others), the Somalia Health Sector Donor Reports were reviewed and the list of donors was verified with HSC members. In total 26 organizations were identified. (See Annex 1). All but one agency responded giving a total response rate of 96 percent. Of the 25 agencies surveyed, 22 (88 percent) provided relevant data and three (12 percent) did not have health specific information to be used for the study. Quantitative data were collected between March and May 2007. • Qualitative questionnaire - In-depth interviews of approximately one hour duration were conducted with 14 health program managers at various donor agencies based in Nairobi. Interviews were conducted from 23 March to 23 April 2007. • Consultative workshops – Preliminary findings were presented to HSC members at a meeting held on 11 May 2007. A more detailed presentation and in- depth discussions were held at a consultative workshop with HSC members on 14 May 2007 in Nairobi. Quantitative and Qualitative Data Collection Process 3.2 Data collection: Two types of quantitative data were collected: i) disbursements by donor agencies; and ii) expenditure by agencies that received funds from donors and implemented health sector activities (henceforth referred as recipient/implementing agencies). As a first step, all donors (bilateral, multilateral and 14 other donors) were requested to supply total disbursements6 for the calendar years 2000 to 2006 in support of the health sector7 in Somalia. Commitments or pledges were not considered. As a second step, recipients/implementing agencies were requested to supply actual expenditure on health sector programs for the same period. Data were mainly extracted from computer print-outs provided by recipient/implementing agencies. 3.3 Rate of exchange: Data for the calendar years were collected in the currency of the donor agency. Funds obtained in foreign currency were converted into US Dollars by using the average annual rate of exchange from the Oanda website8. 3.4 Data verification: All disbursed funds were accounted for at the donor agency level. To avoid double counting, approximately 80 percent of all funds disbursed were cross-checked with recipients/implementing agencies or against contracts issued by the donor agency. After compiling data by donors and recipients/implementing agencies, all finance or program managers interviewed were sent a copy of the data sheets for verification. 3.5 In-depth interviews: To complement the results obtained through the quantitative questionnaire, qualitative data were collected via in-depth interviews. The researchers interviewed 14 donors and recipient/implementing agencies health sector program managers, primarily located in Nairobi. 3.6 Confidentiality: All respondents were informed and reassured about the confidentiality of the data collection process. All financial records were securely kept, disclosed only to the research team and will be destroyed after six months from the publication of this report. Types of Quantitative Data Collected 3.7 Four main types of retrospective quantitative data were collected: [A] total aid financing, and health sector expenditure [B] by disease, [C] by zone and [D] by intervention. While A was obtained from the financial data of donor agencies, B, C and D were obtained from financial data of recipient/implementing agencies or from a review of partner contracts at the donor level. • [A] Total donor health sector aid financing refers to disbursement (financial and non financial contributions) made by donors for health sector activities. 6 Disbursements are financial (monetary) and non-financial (in-kind) contributions made by donors to recipient/implementing agencies. In this study a donor had disbursed funds when the contributions had left the agency or the agency’s bank account. 7 In this study, financing for the health sector refers to aid for health only: food aid or support for water and sanitation interventions were not included. 8 Average annual exchange rates are available at: http://www.oanda.com/convert/fxhistory 15 • [B] Health expenditure by disease refers to expenditure by recipient/implementing agencies on various disease programs ranging from vertical programs such as polio, TB, Malaria or HIV (see Figure 15 for the number of diseases reported) to horizontal programs (such as primary health care and health systems support). • [C] Health expenditure by zone refers to expenditure by recipient/implementing agencies incurred for the benefit of beneficiaries in the three zones (Somaliland, Puntland and South-Central zone). A fourth category, “countrywide�, includes expenditure not targeted at any specific zone but benefiting the entire country. • [D] Health expenditure by activity refers to expenditure by recipient/implementing agencies incurred for each of the major activities such as prevention, supplies, monitoring and evaluation, staff costs, etc. 3.8 As expected, the study found differences between the disbursements reported by the donors and the expenditure reported by recipient/implementing agencies (see Table 4). The average difference for all years was 16 percent. Table 5: Percentage difference between data collected from donors and recipient/implementing agencies 2000 2001 2002 2003 2004 2005 2006 Donor Disbursements 100% 100% 100% 100% 100% 100% 100% Expenditure by recipient/ implementing agencies 86% 91% 76% 94% 74% 86% 84% % Difference 14 9 24 6 26 14 16 This difference is explained by four factors (see Figure 9). First, there is a time lag between disbursement of funds by donor agencies and the implementation by recipient/implementing agencies. This means that funds disbursed in one fiscal year by donors are often utilized in other fiscal years by the recipient/implementing agency. The analysis of the data revealed that in 2002 differences were found in two donor organizations providing larger than usual funding which were used in the following years. From 2004 to 2005, disbursements by the GFATM accounted for the majority of the difference as start-up processes delayed recipient/implementing agencies from spending the grants. In 2006, the difference is due to the Italian Cooperation disbursing US$6.5 million for health systems development activities that will be implemented by the recipient/implementing agencies in subsequent years. Second, when funds are transferred from a donor to a recipient/implementing agency, the recipient/implementing agency may take a portion of the funds (generally between five and 15 percent) as overheads. This generally occurs at the headquarters level. In this study the overheads could not be accounted for, as interviewed recipient/implementing agencies in Somalia did not have specific data on overheads. Third, some donors provide un-earmarked funds that are categorized as health sector disbursements in the donor agency. However, at the recipient/implementing agency level the same funds 16 could be utilized for initiatives not strictly linked to the health sector, such as food aid or water and sanitation. Fourth, some reporting errors cannot be ruled out. Figure 9: Explanations for the difference between donor disbursement and recipient/implementing agencies expenditure Donors (100%) Recipients/Implementers Reporting Overheads Non Health & Health & Nutrition Program Error (%?) (5-15%) Nutrition (%?) Expenditures (84%) Year 1 Year 2 Year 3 Time Lag Methodological Limitations and Challenges 3.9 The study has the following limitations: • Scope of the study - The study focuses on health sector aid financing by DAC donors. Due to the study design, remittances from the diaspora, funds from non- DAC donors, private funding of local and international NGOs, and funds from the three Ministries of Health are not included. • Missing data - Data were not obtained from DFID and Norway for 2006 and from ICRC for the period 2000 to 2003. • Agency overheads9 – Overheads that are not part of direct disease program costs for donor and recipient/implementing agencies could not be obtained. For instance, costs related to UNCAS and ECHO flights for transport of personnel and supplies to Somalia were not captured. The cost of security or money transfer was also missed. • The inclusion of all HIV expenditure as health expenditure. Although not all HIV expenditure are strictly health sector-related, it was impossible to disaggregate between health and other multi-sectoral HIV interventions, as most donors categorize HIV expenditure under health. 9 Overheads are generally described as the ongoing administrative expenses of a business which cannot be attributed to any specific business activity, but are still necessary for the business to function. Examples include rent, utilities, insurance, etc. (www.investorwords.com/3547/overhead.html Accessed May 23, 2007) 17 • Differing financial reporting systems provided challenges in disaggregating expenditure by zone. Expenditure by zone were either obtained from organizations financial reporting systems or disaggregated by program managers. Bias in the data for the allocation by zone is introduced due to two major programs; HIV and Polio. As advised by program managers, HIV expenditure was equally divided by each zone. Polio expenditure was apportioned according to the number of children vaccinated in each zone. • Differing financial reporting systems provided challenges in disaggregating expenditure by activity. Costs were allocated differently in different agencies. For example, travel for training might be allocated to capacity building in one agency and under travel in another. In addition some agencies financial systems did not disaggregate expenditure by specific activities. In a few instances costs had to be re-allocated per activity at the discretion of the researcher and the program or finance managers. 3.10 Data collection provided the following methodological challenges: • The process was lengthy and labor intensive: Data collection and verification of quantitative data took on average five contacts per agency. Follow-up contacts were made via telephone, email or personal visits. • Retrieving financial data from agencies not based in Nairobi was particularly difficult. In two bilateral donor agencies, disbursements and financial accountability were the responsibility of headquarters or regional offices based outside of Nairobi. Contacts had therefore to be established with headquarters/regional offices. • The institutional memory of donors and recipient/implementing agencies was not always strong. Collecting data from the year 2000 proved challenging since in two agencies archived data were difficult to retrieve. In one agency, the annual reports were used to extract data. • Obtaining actual expenditure from recipient/implanting agencies was more difficult than expected. Some agencies required more time to access actual expenditure while information on budgets were easier to access. Usefulness of the Data 3.11 Despite the limitations and methodological challenges experienced, the study offers some useful insights, as illustrated in the findings and in the conclusions/recommendations sections. The study is the first exercise of this kind conducted in Somalia. It utilized a resource tool that can be adopted by partners for future data collection on aid tracking for Somalia. The study provides important baseline data, raises levels of awareness and generates questions about past and future aid financing to the health sector. Finally, it highlights gaps in knowledge that may be filled by future studies. 18 4. KEY FINDINGS Financial Aid Flows 4.1 In Somalia aid to the health sector flows through an intricate network, characterized by three groups of financiers and many intermediaries (see Figure 10): Figure 10: Financial aid flows in the Somalia health sector10 Bilateral, Multilateral, Non DAC Donors/ Diaspora Other Donors Private Donations Remittances UN Somalia Other UN MOH Red Cross/Crescent International Local NGOs/ Agencies Movement NGOs Institutions Beneficiaries i) The so called traditional donors (bilateral11, multilateral12 and others13) are the object of this study. These donors provide direct and indirect funding to several agencies and institutions operating in Somalia. Funding is directly channeled to international NGOs, the Red Cross/Red Crescent Movement or, more frequently, to the UN Somalia Family. The UN play a crucial role 10 Non DAC donors are donors from countries not represented in the OECD Development Assistance Committee (DAC) 11 These include the DAC donors 12 These include EC, ECHO, Development Banks and the UN 13 These include the GFATM, ICRC and IFRC 19 in further channeling funds to a series of implementers that include local NGOs/institutions, international NGOs, ministries of health, and the Red Cross/Red Crescent Movement. UN agencies that receive funds from traditional donors may also directly implement or channel funds to other UN agencies; ii) Non DAC donors14 and private financiers15 usually sponsor international NGOs or national NGOs/institutions. Non DAC donors, mostly Arab countries supporting Islamic charities, and private donors are not included in this study. However, their contributions to the health sector are significant (see 2.11 and 4.3). iii) As discussed in 2.12, Somalis living abroad are reported to significantly contribute to the economy of the country. Funding for health reaches the beneficiaries to meet the costs of medical treatment. In other cases, remittances are used to support local institutions or NGOs. The analysis of Somali remittances to the health sector is not part of this study. Total Health Sector Aid Financing (2000-2006) 4.2 Total health sector aid financing to Somalia grew from US$23 million in 2000 to US$62 million in 2006, almost a three fold increase in six years. The data show a steady increase with peaks reported for the year 2002 and from 2004 onwards (see Figure 11). Even in real terms, keeping the rate of exchange constant using 2000 as the base year, the increase in financing between 2000 and 2006 is almost three fold (see Table 6). 14 These include, for example, Arab Countries. 15 These include private donations that are made to charities, local or international NGOs. 20 Figure 11: Total health sector aid financing (2000-2006) 70 Millions 60 170% increase 50 34% 40 US $ 14% 30 49% -10% 20 1% 30% 10 0 2000 2001 2002 2003 2004 2005 2006 4.3 The totals reported (see Table 6) are conservative estimates of the total aid provided to the health sector in Somalia. DFID contributions for the year 2006 and ICRC disbursements for the period 2000-2003 could not be obtained and are therefore not included. In addition, as explained in the methodology section (see 3.9), the total reported disbursements are limited to DAC donors, multilateral donors and other donors like GFATM and the Red Cross and Red Crescent Movement. The total amount excludes remittances from the diaspora, funds from non-DAC donors, and private funding for local and international NGOs. There is little information available on the amount provided through remittances and by non-DAC donors. However, more is known about international NGOs that operate in Somalia and finance themselves mostly through private donations. The international NGO, MSF operates large humanitarian programs in 12 locations in central and south Somalia (33). In 2006 MSF had a country budget of US$24 million (34), which is 100 percent funded through own resources. This amount is considerable and should be taken into consideration when reviewing the data presented in Table 6. Table 6 shows the annual increase in current US Dollars as well as constant US Dollars using the 2000 US Dollar exchange rate16. Table 6: Total health sector aid financing using current and constant rate of exchange (2000-2006) 2000 2001 2002 2003 2004 2005 2006 Reported Disbursements US $ (Millions) current 23,0 23,3 30,3 27,1 40,5 46,3 62,2 Annual increase current 1% 30% -10% 49% 14% 34% Reported Disbursements US $ (Millions) constant 23,0 23,5 30,0 26,3 36,1 42,7 59,0 Annual increase constant 1% 28% -12% 37% 18% 38% 16 The US inflation over this period was reported at 2.78 percent. 21 4.4 The large increase in financing to Somalia over the period 2000-2006 can be explained by an increase in financing from all three categories of donors analyzed (bilateral, multilateral and others). However, the largest absolute and relative increases are seen among the multilaterals and “others� (see Figure 12). The emergence of the GFATM as main donor to the health sector in Somalia explains the surge observed in the “others� category from 2004 onwards. The increase of multilateral funding seems also to be larger from 2004 onwards and appears to be mostly driven by the UN. On the other hand the level of bilateral support has remained constant over the years. Figure 12: Total health sector aid financing by donor category (2000-2006)17 2005 2004 Bilateral 2003 Multilateral 2002 Others 2001 2000 0 10 20 30 40 50 60 70 Millions US $ 4.5 Considering the aggregate data from 2000-2006 (see Figure 13), bilateral donors were the largest contributors to the health sector in Somalia (48 percent), followed by multilaterals (31 percent) and “others� (21 percent). Within the bilateral donors, the EU member states were the largest donor. If the multilateral funding provided by EC and the European Commission Humanitarian Office (ECHO) is added to the bilateral contributions from EU member states, European countries financed almost 50 percent of health activities in Somalia. Within the “others� category the impact of the GFATM is evident (12 percent of total aid), especially considering that the GFATM funded health activities only for three of the seven years considered in this analysis. Funding for polio activities sustained by CDC and Rotary International (six percent) is also significant. 17 ICRC 2000 to 2003 is missing from “Other� donor category 22 Figure 13: Percentage contribution of health sector aid financing by donor category, all years (2000-2006) IFRC ICRC 2% Global Fund 1% Rotary 12% 3% EU Member States CDC 32% 3% UN 16% Non EU Member EC/ECHO States 15% 16% 4.6 The comparison between the relative contributions for health in 2000 and 2006 shows important changes in the panorama of donors to the health sector in Somalia (see Figure 14). Though increasing their overall funding, bilateral donors have comparatively become less relevant financiers. On the other hand UN contributions have grown both in absolute and relative terms to fund a quarter of the overall health sector in 2006. The GFATM, a relatively new player in Somalia, provided 22 percent of total aid financing in 2006. Figure 14: Percentage contribution of health sector aid financing (2000 and 2006) 2000 2006 ICRC Rotary 3% IFRC 4% Rotary 6% EU Member 1% UN Global Fund EU Member States 22% 17% States 38% 26% EC/ECHO Non EU Member 10% CDC States 4% 9% Non EU Member UN EC/ECHO States 25% 25% 10% 4.7 The per capita aid for the health sector more than doubled in seven years, rising to US$7.3 in 2006 from US$3.2 in 2000 (see Table 7). The 2006 estimate is conservative. Three factors need to be considered while reviewing the data in Table 7. First, the numerator (i.e. total aid financing) lacks some data for DFID in 2006 and 23 ICRC in 2000-2003. Second, the denominator (i.e. the population) is not exact, as population estimates remain highly contentious in Somalia. However, the population estimates used for the analysis were obtained from the Human Development Report 2006 (7) published by the UNDP and are considered by the authors as a good approximation of the real population of Somalia. Third, the per capita aid excludes remittances from the diaspora, funds from non-DAC donors, and private funding for local and international NGOs. As discussed in 4.3, if the funding for humanitarian assistance provided by MSF was included, the per capita aid for 2006 would increase to US$10.2. Table 7: Per capita (US$) health sector aid financing18 200019 2001 2002 2003 2004 2005 2006 Total $ 23.0 23.3 30.3 27.1 40.5 46.3 62.2 Population m. 7.1 7.3 7.5 7.8 8.0 8.2 8.5 Per Capita $ 3.23 3.18 4.01 3.49 5.06 5.62 7.34 Health Sector Aid by Disease/Program (2000-2006) The reader is reminded that the analysis of aid by diseases, by zone and by activity is based on the expenditure by recipient/implementing agencies (see 3.7). These are lower than the total donor contributions, which are used in the analysis for the total health sector aid financing to Somalia (see Table 7). Table 8: Health sector aid (US$ million) by disease/program (2000-2006) 2000 2001 2002 2003 2004 2005 2006 A] Donor Contributions 23.0 23,3 30,3 27,1 40,5 46,3 62,2 B] Expenditure by Recipients/Implementers 20,1 21,3 24,4 25,5 32,0 40,7 53,6 % Difference 14 9 24 6 26 14 16 4.8 During the period 2000-2006, health sector financing in Somalia progressively shifted from horizontal to vertical programs. In 2000, primary health care, and health system strengthening activities constituted 37 percent of the total aid financing; in 2006 funds allocated to these activities decreased to 23 percent of total aid. Conversely, in 2006 malaria, TB, HIV and the polio programs accounted for 50 percent of aid to the health sector compared to 25 percent in 2000. 18 According to the Human Development Report 2006, the population of Somalia was estimated at eight million people in 2004 with a growth rate of 2.9 percent. 19 The population for 2000-2003 was adjusted using the 2.9 percent figure also for the years before 2004 24 Figure 15: Percentage contribution by programs by year (2000-2006) 100% 90% 80% 70% Poliomyelitis 60% TB,HIV,Malaria Reproductive health 50% EPI/Nutrition 40% Emergency 30% PHC,Health Systems Support 20% 10% 0% 2000 2001 2002 2003 2004 2005 2006 4.9 The shift towards financing of vertical programs became more noticeable from 2004 onwards due to the presence of a new donor, the GFATM (see Figure 16). In the six calls for proposals issued between 2002 and 2006, the GFATM approved four proposals (two for malaria, one for HIV, one for TB) submitted by the HSC, as Country Coordination Mechanism. The total amount approved for Somalia was US$50 million, though only US$29 million was disbursed as at December 2006. Funds for HIV and malaria are managed by UNICEF, while funds for TB are managed by World Vision. Implementers include NGOs, UN agencies and civil society. Figure 16: Health expenditure: TB, Malaria and HIV (2000-2006) 7 Millions 6 5 4 TB US $ Malaria 3 HIV 2 1 0 2000 2001 2002 2003 2004 2005 2006 4.10 The peak in overall health financing to vertical programs observed in 2005 and 2006 can also be explained by a sharp increase in funding for the Polio program. The Polio program has by far been the largest funded program in Somalia health sector over the period 2000-2006. In seven years it received a total of US$43 million, equal to 20 percent of the total aid to the health sector in Somalia. The program, sponsored by several international donors, is jointly implemented by WHO 25 and UNICEF. It achieved a substantial reduction of cases from 2000 to 2004, but in 2005 was hit by a major outbreak in several regions of the country (See Figure 17). Table 9: Confirmed wild poliovirus cases in Somalia (2000-2006) 2000 2001 2002 2003 2004 2005 2006 46 7 3 0 0 185 36 The response to the outbreak involved doubling the immunization campaigns, and increasing surveillance. These factors largely explain the funding surge experienced by the program and by the overall aid to the health sector in the years 2005/2006. Figure 17: Health expenditure: Poliomyelitis (2000-2006) 14 200 Millions 180 12 160 10 140 8 120 US $ Polio Funding 100 6 Confirmed Cases 80 4 60 40 2 20 0 0 2000 2001 2002 2003 2004 2005 2006 4.11 Due to the lack of precise data on the burden of disease and on program outcomes, it is difficult to argue about the financial needs of some vertical programs (e.g. HIV and malaria). However, based on good quality data available, generous extra funding for the TB program does not seem fully justifiable (see Figure 18). The HIV program is still in its infancy with regard to collecting epidemiological data and tracking outcome indicators: the HIV prevalence rate among pregnant women was reported as 0.9 percent in 2004 (35), but little is known about variations in prevalence over time or about the burden of disease across vulnerable groups. Several activities are currently implemented in Somalia to prevent the spread of HIV but assessing their impact proves challenging. For the reasons above, it is impossible to draw any conclusion on the effectiveness of the large amount of funding provided to the HIV program. Similar challenges are experienced by the malaria program, though a clear move towards evidence-based policy making was initiated in recent years, partially thanks to the catalyst role played by the GFATM. Differently from the “young� HIV and malaria programs, the TB program has been running in Somalia for a long time. The well established TB program has progressed remarkably to achieve the WHO targets of 85 percent treatment success rate and 70 percent case detection rate in 2004 (36;37). These results were obtained before the GFATM brought a six fold increase in the annual TB budget for Somalia. Although funds will be needed to sustain the results reached in 2004 and to address new challenges like MDR-TB and TB/HIV co-infection, the large increase in TB funding may be questionable. 26 Figure 18: Health expenditure: Tuberculosis financing versus TB case detection and TB success rate (2000-2006) 100 3.5 Millions 90 3.0 80 70 2.5 TB Funding % Coverage 60 2.0 Case Detection US $ 50 Rate (%) 40 1.5 Success Rate (%) 30 1.0 20 0.5 10 0 0.0 2000 2001 2002 2003 2004 2005 2006 4.12 Among vertical programs, a few (EPI, reproductive health, nutrition and non communicable diseases) seem to have been neglected by donors (see details in 4.13 to 4.15) 4.13 EPI funding in 2000-2006 was stable around US$2 million/year. During this period vaccine coverage for some antigens remained stable at very low levels and for other antigens worsened considerably. Measles coverage in the years 2000- 2006 remained between 30 and 40 percent; BCG coverage decreased from 70 percent in 1999 to 30 percent in 2006 (see Figure 19) (38;39). Financing for EPI activities was irregular, with a decline between 2002 and 2005 and an increase in 2006 linked to measles campaigns. It is striking that despite tackling six killer diseases, EPI received only US$2.9 million in 2006. This amount is only 25 percent of the total received by the polio program (US$11.4 million) in the same year. It would be erroneous to derive a causality link between funding and performance, as many factors related to access, utilization and quality of services could combine to keep EPI coverage low. However, it is likely that more financial resources for EPI could help to tackle key bottlenecks for service delivery. As EPI relies heavily on the existing health network, the program’s performance may also benefit from increased financing to the overall health system and primary health care. 27 Figure 19: Health expenditure: EPI financing versus BCG and measles coverage (2000-2006) Millions 4 100 3.5 90 80 3 70 % Coverage EPI Funding 2.5 60 BCG Coverage US $ 2 50 Measles Coverage 1.5 40 30 1 20 0.5 10 0 0 1999 2000 2001 2002 2003 2004 2005 2006 4.14 Reproductive health was also neglected for many years and only recently started receiving more attention by donors. Reproductive health indicators have remained abysmally low over the period 2000-2006. Data from the 2006 MICS reveal little improvements from 1999: only 15 percent of women use contraceptive methods; the fertility rate remains high at 6.5 children/woman; institutional deliveries are as low as nine percent and antenatal care attendance is only 26 percent. Given the high maternal mortality in the country (1,100/100,000 live births) and the above indicators, reproductive health will require additional sustained investments in future years. Figure 20: Health expenditure: Reproductive health (2000-2006) 3.0 Millions 2.5 2.0 US $ 1.5 Reproductive health 1.0 0.5 0.0 2000 2001 2002 2003 2004 2005 2006 4.15 Between 2000 and 2006 nutrition accounted for 11 percent of the total aid financing to Somalia. However, data on malnutrition seem to have worsened over the years (see Figure 21). ECHO, DFID and USAID were the main donors, while UNICEF was the biggest implementer. Activities included selective feeding programs for malnourished children, administration of micronutrients, breastfeeding promotion, 28 and regular/emergency nutritional surveys across the country. MICS data suggest that the nutrition situation in Somalia worsened between 1999 and 2006 (see Table 10). Table 10: Nutrition indicators for children under 5 years of age Malnutrition Indicators for MICS 1999 MICS 200620 Children Under 5 Underweight (Weight/Age) 26% 35% Stunting (Height/Age) 23% 38% Wasting (Weight/Height) 17% 11% The increase in funds allocated to nutrition in 2006 is a good step in the direction of addressing a top public health challenge for Somalia. Funding should be sustained and possibly increased until program targets are reached. Figure 21: Health expenditure: Nutrition financing versus malnutrition indicators (2000-2006) 8 40 Millions 7 35 6 30 5 25 Nutrition Funding US $ 4 20 Underweight 3 15 Stunting 2 10 1 5 0 0 2000 2001 2002 2003 2004 2005 2006 4.16 Despite the expected heavy burden of non communicable diseases in Somalia, practically no funding was allocated for that purpose in the period 2000- 2006. Non communicable diseases and injuries represent 27 percent of the total burden of disease in Africa (40) and this percentage may be higher in Somalia, given the high incidence of neuropsychiatry disorders linked to 15 years of war. Yet, the annual average expenditure for non communicable diseases over the study period was less than US$200,000. Although the difficulties of targeting chronic diseases in a fragile state are recognized, the absence of funding for diseases with such a high burden on the population is a matter of concern. 20 NB: As of July 2007, MICS 2006 data had not yet been officially endorsed by UNICEF. 29 Figure 22: Health expenditure: Non communicable diseases (2000-2006) 0.6 Millions 0.5 0.4 US $ Mental health 0.3 People with disabilities 0.2 0.1 0.0 2000 2001 2002 2003 2004 2005 2006 4.17 Between 2000 and 2006 funding for horizontal programs like health system strengthening and primary health care21 increased in absolute terms (see Figures 23 and 24) but decreased as a proportion of overall aid to the health sector (see Figure 15). Between 2000 and 2006 different donors provided funding for human resource development, health care infrastructure, NGO support to primary and tertiary health care, etc. The total financing for health system and primary health care fluctuates between US$13 and 17 million over the period 2004-2006. However, the level of coordination among implementers, especially for health system strengthening, was reported as weak. Funding for the health system was often channeled to individual NGOs or managed by a single UN agency without a coordinated and structured approach. During the qualitative interviews, health specialists in various organizations called for better coordination in this area, particularly in view of new funding for health system strengthening that the GFATM, the EU and DFID recently committed and pledged. The creation of a Health System Working Group under the HSC is a first step to improving donor harmonization in this critical area. This is in line with the conclusions of the JNA and the RDP. 21 Health systems support includes expenditure on staff and equipment, training and capacity building, health management information system and governance (e.g. development of policies, plans and strategies). PHC includes expenditure related to a variety of interventions offered at primary health care clinics. 30 Figure 23: Health expenditure: Health systems support (2000-2006) 12 Millions 10 8 US $ 6 Health Systems 4 2 0 2000 2001 2002 2003 2004 2005 2006 Figure 24: Health expenditure: Primary health care (2000-2006) 8 Millions 7 6 5 US $ 4 PHC 3 2 1 0 2000 2001 2002 2003 2004 2005 2006 4.18 Between 2000 and 2006, emergency funding accounted for six percent of the total aid financing to Somalia. Aid for emergencies appears low, given that Somalia was affected by both natural disasters and continued conflict over the period 2000-2006. Continuous fighting, with high casualties and massive displacement, continued to occur, especially in Central and Southern Somalia. The low levels of funding are probably confounded by two factors. First, funding lines, not specifically allocated to emergencies, were utilized by implementers to cover acute needs of the population. Second, the impact from MSF is not taken into account from 2005 onwards, i.e. from the year in which the organization started operating solely on private funding. If the MSF budget for 2006 was added to the emergency pot, the portion devoted to humanitarian assistance in 2006 would increase to 36 percent of the total aid financing to Somalia. 4.19 The trend in emergency aid reflects the hopes for political stabilization emerged in 2005 (see Figure 25). Unfortunately, the eruption of violence in 2006, coupled with a severe drought, explains the increase in emergency funding in 2006. 31 Figure 25: Health expenditure: Emergency (2000-2006) 3.0 Millions 2.5 2.0 US $ 1.5 Emergency 1.0 0.5 0.0 2000 2001 2002 2003 2004 2005 2006 Health Sector Aid by Zone (2000-2006) 4.20 The distribution of health expenditure by zone over the period 2000-2006 shows that the South-Central zone received 45 percent of the overall funding, while Somaliland and Puntland received each 14 percent. An additional 27 percent was allocated to activities in support of all zones (see Section 3.9). Figure 26: Distribution of health expenditure by zone: all years (2000-2006) Countrywide 27% South-Central 45% Puntland Somaliland 14% 14% 4.21 There is a close relationship between population distribution and health expenditure by region over the period 2000-2006 (see Figure 27). If countrywide expenditure are omitted or proportionately distributed across zone, the allocations are remarkably close to the proportion of the population in each region, as estimated by UNDP (5). Expenditure appears to have been proportionally slightly higher in South/Central Somalia, a not surprising finding, given the gravity of the humanitarian situation in that part of the country. 32 Figure 27: Distribution of population and health expenditure by zone: all years (2000-2006) 70% 61% 60% 52% 50% Percentage 40% % Population 28% % Health Expenditure 30% 20% 20% 19% 20% 10% 0% Somililand Puntland South-Central Health Sector Aid by Activity (2006 expenditure) The following analysis was carried out for the year 2006 in an attempt to disaggregate expenditure by activities. It proved a difficult exercise as agencies tend to categorize expenditure in different and sometimes generic ways. This made the analysis by category particularly challenging (see Section 3.9). Table 10 provides the results of the analysis, but the information should be treated with caution given the limitations of the data. Table 11: Expenditure by activity (2006) Actual Expenditure Percent Activity Purpose US$ Expenditure (millions) Supplies 15.0 27.9% EPI, Malaria, Polio, and other essential supplies Staff & Op Costs 12.2 22.8% Agencies’ transport and staff costs 22 NGO Support to Primary health care activities/essential services. Primary Health 8.8 16.5% It includes salaries paid to clinic staff Care Health System Rehabilitation of health facilities, equipment 5.2 9.6% Support staff costs. Prevention 4.4 8.1% HIV, Reproductive Health and EPI Capacity-building Training of health personnel, meetings and 3.7 6.9% & Training workshops M&E 3.1 5.7% Monitoring and evaluation of programs Media initiatives and social mobilization IEC/Awareness 0.5 0.9% programs 22 Technical support staff and international staff breakdown of costs were not obtained for some agencies 33 Actual Expenditure Percent Activity Purpose US$ Expenditure (millions) NGO Support to 0.3 0.6% Support to hospitals, including staff costs Tertiary Hospitals Support for Ministries of 0.2 0.4% Incentives paid to governments Health Governance 0.2 0.3% Policies and planning Research activities (OVC assessment, Research 0.1 0.2% behavioral research on HIV etc.) Total 53.6 4.22 Supplies and agencies’ operating costs in 2006 constituted more than half of the health sector expenditure. Supplies represented 29 percent of overall expenditure in the health sector. Operating costs of agencies, including staff costs, travel into Somalia and transport accounted for 23 percent of total expenditure. The operating costs (US$12.2 million) are an indication of the high transaction costs to operate in Somalia. It should be noted that operation costs do not include the costs for transporting supplies through the ECHO financed flight services. 34 5. QUALITATIVE FINDINGS This section describes the findings from the interviews held with 14 program managers, based in Nairobi. The qualitative findings are discussed under two themes: (i) aid financing and (ii) donor harmonization. Aid Financing 5.1 From the donor perspective three factors drive health sector funding priorities: humanitarian crises, the status of the health system and politics. Respondents mentioned that priorities are determined by states of emergency caused by natural or man made disasters, and by the recognition of the poor status of infrastructure and management of health services in Somalia. There are also historical and political factors that influence the size of financial aid. For many governments, the size is determined by the availability of funds from their Treasury. In other cases, it is a negotiated process between in-country health sector donor staff and political counterparts from headquarters. 5.2 Donors identified three internal challenges for their support to Somalia: small number of Nairobi/Somalia based staff, size of financial aid and lengthy approval processes for the release of funds. Some donors have only one staff member in-country who focuses on all sectors. Having no staff with a health sector focus poses challenges in deciding which health sector programs to fund. Some donors felt that decisions about the size of the aid are driven by political factors at headquarters resulting in smaller financial aid than would be justified in the country. A UN agency expressed concern about the lack of funding for gender-related programs because gender is regarded as a cross cutting issue with little or no budget allocations within vertical programs. The flow of funds from the donor headquarters to Nairobi as illustrated in Figure 10 is a lengthy process causing funding delays and sometimes interrupting service delivery. 5.3 Donors identified two external challenges for their support to Somalia: uncertain political/institutional situation and limited NGO capacity to deliver services. Donors felt that there is no functioning government that can be held accountable. Without a banking system and limited access due to security, the cost of providing aid increases and the opportunity to monitor activities and finances decreases. According to donors, there are only a few effective international and local NGOs providing care and support in the health sector. These NGOs appear overburdened, and their absorptive capacity seems questionable. Donors also mentioned the lack of capacity in Somalia to manage programs and to provide the necessary management and financial reporting information. Some donors also expressed concerns about the hiring and firing practices that are challenging for NGOs operating in Somalia. 35 5.4 Donors stated that the frequency of reporting and length of contracts especially short contracts creates an administrative burden on both recipients and donors. While some donors require reports (financial and narrative) on an annual basis, others require reports on a quarterly or bi-annual basis. This places an administrative burden on recipients who may have to deliver reports for the same program to different donors with different reporting requirements. In almost all cases, the disbursement of funds follow the arrears system, in other words, once funds are accounted for with receipts, release of further funds take place. Some donors who have capacity, mentioned that they conduct field visits which are dependent on security access, to verify expenditure. Donors with limited staff at the country level mentioned that they do not undertake field visits and rely on reporting from international or local NGOs. To reduce the administrative burden, donors mentioned that they are moving towards longer-term contracts of up to three years. However, they stated that funding contracts for emergency relief continue to remain short (one to six months). 5.5 Among donors, the lack of knowledge on the overall aid financing for the health sector was remarkable. Only one donor knew the proportion of its aid versus the total aid financing to the health sector in Somalia. Some interviewees at the country office did not even know their agency’s own aid contribution and had to retrieve the information from their headquarters. This gap in knowledge is partially due to staff turnover at the country office or to the centralized management of funds with limited information sharing at country level. 5.6 Despite the internal and external challenges experienced, most donors (n=10) reported that they would increase their funding in the future. Some of the reasons mentioned to justify the increase include the need to foster development and particularly to improve the health system, the need to fill funding gaps and to scale up activities. Other donors stated they are optimistic about the future political situation in Somalia. 5.7 Two donors reported that funding levels will decrease and two remain unsure. The decrease in funding is due to lack of staff within one donor agency to focus on the health sector. In another donor agency, an NGO supported for many years by the donor, refused funding. Given the limited number of effective NGOs in the health sector, the donor decided to shift funding to another sector where an effective NGO had been identified. 5.8 All except two donors stated that future funding channels will remain the same. Donors mentioned that they will continue to channel funds through existing structures such as UN agencies, international and local NGOs and the Red Cross and Red Crescent Movement. One donor mentioned that future funding is expected to be channeled through the Country Coordinating Mechanism (CCM) via the three AIDS Commissions. Another donor mentioned that if the TFG emerges as an effective counterpart, funding may shift towards the TFG with specific focus on capacity building. 36 Donor Harmonization 5.9 Donors mentioned that the absence of a formal government called for stronger donor coordination. Donors stated that a strong HSC has emerged as a result of weak existing structures in Somalia. The SACB (currently CISS), established in December 1993, has provided a framework and a forum for developing a common approach to the health sector in Somalia. Several working groups (nutrition, disease outbreaks, EPI, HIV, FGM, reproductive health, malaria and TB) have provided technical guidance and guaranteed a good level of coordination among actors. The SACB was also able to raise funds by emerging as CCM for both the GFATM and GAVI. Donors praised the work of the HSC in raising funds (especially from the GFATM) and in providing technical oversight. 5.10 Donors felt that the HSC could be strengthened to play a greater advocacy role on anti-corruption practices and promotion of financial accountability. They recognized that the committee members are volunteers and this limits the coordinated effort to what can be achieved in voluntary time. They saw the benefits of a strong cooperation such as having a shared vision, participating in a sector wide approach, pooling of financial and human resources necessary for scaling up, tracking aid, identifying funding gaps and moving towards a single reporting system. They stated that donors with limited staff capacity may wish to transfer some of their responsibilities to the group. It is acknowledged by donors that for this to happen, trust is required. 5.11 Donors recognized that matching technical needs with political imperatives is necessary but challenging in the absence of a formal government in Somalia. Donors felt that it is crucial to understand what the needs of the Somalis are and to match them with their political objectives. Some donors stated that if they were based in the different zones of Somalia, there might be closer cooperation with the districts resulting in greater understanding of the needs of the Somalis. It was emphasized that since funding is driven more by the political than the technical agenda, pooling funds may cause internal conflicts among the donor group due to donors’ varying political agendas. Conflict may also arise from resistance of how the pooled resources could be utilized. Furthermore, some donors felt that they needed the flexibility to respond to emergencies and feared that they may lose this flexibility if funds were pooled. 37 6. CONCLUSIONS AND RECOMMENDATIONS In this section, the conclusions of the study are presented according to the three primary objectives of the study (see 1.9). Based on the conclusions, a series of recommendations are provided. Conclusions QUESTION 1: How did levels of donor financing to the health sector vary throughout the years? 6.1 Between 2000 and 2006 there has been a 170 percent increase in funding for the health sector in Somalia. The total aid provided by bilateral DAC donors, multilateral (UN, EC/ECHO and World Bank) and other donors (GFATM and Red Cross/Crescent Movement) has steadily increased from US$23 million (2000) to US$62 million (2006). This amount does not include funds provided by non-DAC donors (e.g. Arab countries), private funding of international NGOs (e.g. MSF), domestic funding on health and remittances. The US$62 million is therefore a conservative estimate: the total envelope of aid to the health sector is expected to be significantly higher if all other contributors to the health sector were included. 6.2 Between 2000 and 2006 the aid financing for health per capita more than doubled, moving from US$3 to US$7. As mentioned previously, this amount is likely to be an underestimation of the total health aid per capita. However, even considering other donor funding, the Somalia health sector is likely to remain grossly under-funded. The US$7 per capita health aid financing is in line with other fragile states: South Sudan received US$7 in 2003 (See Table 2) and DRC received US$2-3 in 2005 (see Table 2). Somalia, like South Sudan and the DRC, remains an “aid orphan�, especially when compared to other “darling� fragile states, like East Timor and Kosovo (see Table 2). The data suggests that fragile states tend to receive less aid per capita than strong performing ones (see Section 2.4). The data also seem to suggest that priorities in aid to the health sector tend to be set on the basis of donors’ political needs rather than the objective needs of the country (see Section 2.8 and 5.5). The findings of the qualitative component of this study appear to confirm the political nature of aid allocations to Somalia. 6.3 Between 2000 and 2006 aid financing to the health sector was not volatile in absolute terms, contrary to the findings in other fragile states by other authors (see Section 2.4). However the mix of the health donors in Somalia substantially changed. Within the steady increase of funding over the seven years, the sources of funding have diversified. The percentage contribution of bilateral donors has decreased from 63 percent in 2000 to 35 percent in 2006. Multilateral donors, and especially the UN, considerably grew to account for 35 percent of aid financing in 2006. The GFATM emerged in 2004 and became a major financier in the following years (22 percent of the total health budget in 2006). 38 QUESTION 2: Which health interventions were prioritized by policy planners through financial aid allocations? 6.4 Aid financing to the health sector, especially from 2004 onwards, favored vertical programs: Polio, TB, HIV and Malaria accounted for 50 percent of total aid in 2006. • The polio program is the number one priority in the Somalia health sector. From 2000 to 2006 it received 20% of total aid financing to the health sector, with an average of US$6.1 million/year. The above figures show the political commitment of the donor community but also raise questions about the opportunity costs of the eradication campaign. The polio program achieved zero polio cases in 2003 and 2004. However, faced with a major outbreak in 2005, it mobilized additional financial resources to double the immunization campaigns and strengthen the surveillance system. Given the nature of the eradication program and its global target, a large amount of funding will continue to be needed until the virus is eradicated not only from Somalia but globally. • The TB program, one of the most successful programs in Somalia, achieved the WHO global targets for treatment success and case detection rates in 2004. This was accomplished with limited resources (less than US$0.5 million/year for the period 2000 to 2004). In 2005 and 2006 a six fold increase in funding (US$3 million/year) was allocated mostly through GFATM contributions. Though the TB program is facing several challenges (TB/HIV co-infection, multi-drug resistance, etc.), such a sharp increase in financing may be questionable. • The HIV and malaria programs received approximately US$6 million each during 2006. However, it is difficult to draw conclusions on the adequacy of funding for the two programs. This is mostly due to i) lack of precise estimates on the burden of disease; and ii) lack of reliable outcome indicators that would allow for tracking the effectiveness of activities implemented. Both programs are currently working to rectify the above. 6.5 Within vertical programs, EPI, reproductive health, non communicable diseases and nutrition seem to have been neglected. • EPI received a relatively constant amount of financing over the years (US$2.2 million/year). The declining trend of EPI coverage, demonstrated by both the MICS 1999/2006 and the 2000-2005 annual WHO/UNICEF immunization reports, seems to suggest that the amount of funding for EPI may have not even been sufficient to sustain the low levels of EPI coverage (DPT3 and BCG were 35 and 50 percent respectively in 2005). The expenditure on EPI indicates the need for substantial increases in funding, if adequate coverage levels are to be achieved. • MICS 2006 data for reproductive health show little gains from 1999. In a country with maternal mortality rate of 1,044 per 100,000 live births, low percentage of births performed by skilled staff and high rates of FGM, during the period 2000-2006 the average yearly expenditure on reproductive health per woman of child bearing age was US$0.55 cents. Only in 2006 was this amount increased to 39 US$1.40. Increased funding may be required to improve the current low outcomes in antenatal care, institutional deliveries, contraceptive prevalence, fertility rate and ultimately maternal mortality. • Non communicable diseases have been generally neglected by the donor community. The small number of implementing partners and the high cost for service provision may have been limiting factors for resource mobilization. Although there are no official data on the burden of mental illness and disabilities in Somalia, partners were unanimous in recognizing the extent and severity of these diseases within Somali communities. As these needs are largely unmet, investments in non-communicable diseases are needed in the future. • Data on nutrition indicators calls for more resources to be devoted to nutritional programs. As mentioned by Bradbury (19), high levels of malnutrition are treated as an “accepted� norm for communities in Somalia rather than a crisis that requires a humanitarian response. 6.6 Funding for horizontal programs (health system strengthening and primary health care) has been large (36 percent of total aid financing over the period 2000-2006), but sharply declined in 2006. Funding for horizontal programs was often fragmented among NGOs and UN agencies and partners lacked a coordinated approach (see Section 4.17). This is currently changing. The JNA and RDP provide a new framework to address the perceived priorities for the health system. The recent creation of a Health System Working Group under the HSC indicates that donors and implementers are prioritizing health system development and are trying to better coordinate among themselves. 6.7 Funding for emergency activities represented only six percent of the total aid financing over the period 2000-2006, a surprisingly low figure given the number of man made and natural emergencies that occurred during the period of the study. However, other funding allocations such as primary health care, which are not specifically allocated to emergencies, were utilized by implementers to cover acute needs of the population. In addition, if the MSF budget was added to the total funding, the portion devoted to emergencies would increase to almost 36 percent of the total aid financing to Somalia. On the ground, humanitarian assistance and development aid tend to coexist, with donors and implementers often switching from emergency activities to developmental interventions. Donors and implementers work “in dual mode�, quickly responding to emergencies caused by conflicts and natural disasters; while at the same time trying to promote long-term activities to strengthen the health system. This modus operandi, not ideal but inescapable given the circumstances, is likely to be needed until the political situation in Somalia stabilizes. QUESTION 3: How evenly was aid to the health sector distributed to the different regions of Somalia? 6.8 In the period 2000-2006 aid to the health sector was distributed fairly evenly across zones. The answer to this question, however, is limited by two factors: first, population estimates remain highly contentious in Somalia and the proportion of 40 citizens in each zone is only a best estimate based on UNDP Somalia data. Second, 27 percent of aid financing for health is allocated centrally and not regionally. Agencies allocate the purchase of supplies and staff and other operational costs to “countrywide.� A further methodological problem is that two key vertical programs (Polio and HIV) provided only estimates of the regional distribution of their funds. With the above limitations in mind, regional aid financing to Somalia over the period 2000 to 2006 appear to have been distributed evenly according to population estimates. Sixty-one percent of aid for the health sector went to the South-Central zone (which accounts for 52% percent of the population), 19 percent went to Puntland (which accounts for 20 percent of the population), while 20 percent went to Somaliland (which accounts for 28 percent of the population). 6.9 The level of coordination in the health sector is mixed. Through various working groups, the CISS has played an important role in coordinating donors and implementers on technical matters. It also succeeded in uniting partners, including Somali authorities, into a CCM that has been legitimized by the GFATM. It however lacks Somali participation, possibly due to its location in Nairobi. Its effectiveness in favoring joint planning and pooling resources has also been limited. This is due to lack of capacity and the fact that all members of the health sector committee are volunteers coupled with the fact that there are low numbers of in-country staff (see Section 5.2). Other instruments, e.g. the CAP, have proven useful in raising money but less in bringing partners together. Recent efforts have been made to jointly assess the health sector needs through the JNA and to plan for the reconstruction of the health sector through the RDP. The process has been fraught with delays and difficulties; however the JNA and RDP could be important tools to group donors around a similar vision for the health sector in Somalia. 6.10 The analysis of aid contribution by activity is limited by the way donors allocate and record their expenditure. Recommendations 6.11 Somalia requires increasing and long term financial support for the health sector to address the needs of its population. Although US$7 per capita aid for the health sector does not compare badly with other sub-Saharan African countries (see Table 6), Somalia’s financial needs for the health sector remain high given the exceptionally low health indicators, and the high operational costs linked to both the difficult logistics of the country, and the reliance on international actors mostly located outside Somalia. 6.12 More importantly, contributions to the health sector should be made more strategic: funding gaps in key areas (EPI, reproductive health, nutrition and non communicable diseases) should be addressed as a matter of priority. Needs analysis and data on the burden of diseases are required to guide the prioritization process. However, the acute needs in certain areas (EPI, reproductive health, nutrition and non communicable diseases such as mental health and disabilities) should be addressed without further delay. 41 6.13 While supporting thematic interventions, the donor community should invest more in rebuilding the health system. The findings of the JNA and the RDP represent the blueprint for systemic efforts and should guide future planning and implementation efforts. The recently created Health System Working Group should take the opportunity to lead both donors and implementers in the rebuilding process, possibly drawing on the positive experiences of other fragile states (e.g. Mozambique, Afghanistan, Liberia, Rwanda etc.). 6.14 The mix of relief aid and humanitarian assistance should be sustained until the political situation normalizes. Providing funds for emergencies will serve the immediate needs of the Somali population, while supporting activities, that are more developmental in nature, will lay the ground for the future health system in the country. A too rapid shift towards development may lead to loss of lives during emergencies. On the other hand, an excessive focus on humanitarian may lead to a very weak health system emerging at the end of the transition phase. 6.15 A “one size fits all� strategy may not be appropriate for a fragile state such as Somalia. All three zones operate within differing political contexts (see Section 1.2). It is therefore essential that partners reflect this diversity into their financial and technical approaches to the health system of the country. Constant dialogue will build trust. This will promote willingness and accountability from Somali authorities. 6.16 Financial tracking of donor resources to the health sector should become an integral part of the health information system. The tool developed for the study could be adopted, and improved by interested parties. The Health System Working Group should assume the responsibility for financial tracking annually. The exercise should be expanded to include all international NGOs and Ministries of Health. It should also be expanded to include tracking of expenditure by key activities. Financial tracking should be matched with burden of disease and program outcome data. Information could include future funding levels which would allow partners to better plan their programs. By doing so, aid financing analysis could become an extremely useful tool for policy planners both at government and donor level. The instrument developed for this study could be adapted to become part of a financial tracking tool for the health sector. 6.17 Coordination mechanisms could be strengthened, mostly in terms of priority setting and reducing transaction costs. The current mechanisms (CISS, CAP, UNCT) have been effective in fund raising for special programs and in providing technical coordination on certain health subjects. However, the function of assessing needs and establishing priorities for the health sector seem to rely outside the coordination mechanisms, either in individual donor agencies or in UN agencies. More efforts could be done to bring together political authorities, civil society, NGOs and religious leaders to define health priorities and set outcome targets including health systems development through joint planning. Full-time “Champions� could be created to strengthen donor coordination mechanisms. Donors with limited in-country staff could introduce delegated cooperation arrangements with other donors to monitor aid 42 and activities in the field. To reduce the administrative burden on implementers a single financial reporting tool could be introduced. 6.18 Operational research is needed to integrate the findings of this study and to allow a better understanding of health financing in Somalia. Topics to be studied include health financing by (i) the private sector, (ii) the diaspora through remittances; and (iii) non conventional donors. Studies on household spending on health would complete the picture by providing information on private expenditure. 43 Annexes Annex 1: Study sample (26 organizations) Bilateral Multilateral Other Denmark EC GFATM Finland ECHO ICRC France World Bank IFRC Germany Islamic Italy Development Bank Netherlands WHO Sweden UNICEF DFID UNAIDS USAID UNDP Japan UNFPA Norway UNHCR UNIFEM FAO/FSAU 44 Reference List (1) The World Bank. Somalia from resilience towards recovery and development. USA: World Bank; 2006. (2) United Nations Security Council. Report of the Secretary-General on the situation in Somalia. USA: United Nations Security Council; 2006 Oct 23. (3) Schieber G. Sustainable Health Sector Financing. Tokyo, Japan: High Level Forum on the Health MDGs in Asia-Pacific; 2005. (4) UNICEF. One year old immunization - DPT3. www.nationmaster.com/graph/hea_imm_1_yea_old_chi_dpt-1-year-old- children-dpt3.; 2002. (5) UN and The World Bank. Draft Report - Somalia Joint Needs Assessment. USA: UN and World Bank; 2006 Sep 14. (6) Reitsma, K. Needs assessment: Somali adolescents in the process of adjustment. Canada: Children’s Aid Society of Toronto; 2001. (7) UNDP. Human Development Report, 2006. http:hdr.undp.org/hdr2006/statistics/countries/data_sheets/cty_ds_SOM.html .; 2006. (8) OECD. Monitoring resource flows to fragile states, 2005 Report. USA: OECD; 2006. (9) High Level Forum on the Health MDGs. Health Service Delivery in Post Conflict States. Paris: High Level Forum on the Health MDGs; 2005. (10) Gottret, B. and Schieber, G. Health financing revisited. A practioner’s guide. The World Bank; 2006. (11) Bitran, R. Forthcoming. User fees for health care in developing countries: A review of current issues and experiences. Santiago: Bitrans & Associates; 2007. (12) Levin, V. and Dollar, D. The forgotten states: Aid volumes and volatility in difficult partnership countries (1992-2002). For DAC learning and advisory process on difficult partnership; 2005 Jan. (13) Cassels A. Health in Fragile States - An Overview Note. Paris: High Level Forum on the Health MDG’s; 2005. (14) Bourguignon, F. and Sundberg, M. Aid effectiveness - Opening the black box. USA: The World Bank; 2007. 45 (15) Randel, J. and German, T. Trends in the financing of humanitarian assistance. In Macrae J The New Humanitarianism: A review of trends in global humanitarian action. ODI HPG Report. 2002. (16) McGillivray, M. Aid allocation and fragile states. Background paper for the senior level forum on development effectiveness in fragile states. 2005 Jan. (17) Christiansen K. Fragile states, external interventions and the MDGs.: Presentation at the ODI Meeting Series: The MDGs: the 2005 agenda; 2005. (18) Pavignani E. Health service delivery in post-conflict states. High level forum on the MDGs 2005. (19) Bradbury, M. et al. Measuring humanitarian needs. Needs assessment and resource allocation in Southern Sudan and Somalia. ODI, HPG Background Paper. 2003. (20) Pavignani, E. and Colombo, A. Analyzing health sector financing and expenditure. Module 6 "Analyzing disrupted health sectors. Manual". WHO; 2006. (21) OECD. OECD Aid Statistics. http://www.oecd.org/dataoecd/0/38/1878989.gif.; 2007 May 8. (22) Development Assistance Committee. Monitoring resource flows to fragile states: 2006 Report. 2006 Nov. (23) Harmer A, Cotterrel L. Diversity in donorship: The changing landscape of official humanitarian aid.: ODI Humanitarian Policy Group Research Briefing; 2005. (24) Kulaksiz, S. and Purdekova, A. Somali remittance sector: a macroeconomic perspective. in Remittances and economic development in Somalia. An overview. The World Bank; 2006. (25) KPMG/UNDP/EC. Feasibility study on financial services in Somalia -Final Report. KPMG/UNDP/EC; 2003. (26) Lindley, A. The influence of migration, remittances and diaspora donations on education in Somali society in Remittances and Economic Development in Somalia. An Overview. USA: The World Bank; 2006. (27) Kent, R. and Von Hippel, K. Social facilitation, development and the diaspora: support for sustainable health services in Somalia. 2004 Nov. (28) OECD. Paris Declaration on Aid Effectiveness, Ownership, Harmonisaton, Alignment, Results and Mutual Accountability. Paris: OECD; 2005. 46 (29) OECD. Principles for good international engagement in fragile states and situations. USA: OECD; 2007 Apr. (30) Oswald, S., Base, K., and Meyer, F. Donor harmonization to increase efficiency. Germany: D+C Magazine for Development and Cooperation; 2004. (31) OECD, OGDE, The World Bank, ADB, European Bank. High Level Forum - Harmonization, alignment, results: report on progress, challenges and opportunities. 2005. (32) Anti-Corruption Resource Centre. Making use of the DAC principles. www.u4.no .: U4 Anti-Corruption Resource Centre; 2004. (33) Medecins Sans Frontieres. The need for independent humanitarian aid. Washington DC: MSF; 2007. (34) MSF. Head of Mission David Michalski. Personal Communication. 2007 Apr 14. (35) UNAIDS and WHO. AIDS Epidemic Update. Geneva: UNAIDS; WHO; 2006 Dec. (36) World Health Organization. TB country profile Somalia, 2000-2005.Personal Communication. www.who.int/globaltlas.; 2006. (37) World Vision. 2006 data on TB country profile Somalia. Personal Communication. 2007. (38) UNICEF. 1999 and 2006 EPI data: MICS. Personal Communication. 2007. (39) WHO and UNICEF. 2000-2005 EPI Data - Review of National Immunization Coverage 1980-2005. (40) WHO. The Health of the People - The African Regional Health Report. Switzerland: WHO; 2006. 47