The evolving role of the state, donors and NGOs providing health services in a liberal environment, Some insights from Uganda

Nicolas de Torrente is questioning how far have the liberal economic reforms in Uganda seeking to rebuild and reshape the health care system, how are reforms affecting the actual delivery of health services and what are the implications of these reforms in terms of the sustainability and equity.

INTRODUCTION

Since the early 90s, Uganda has been undergoing important reforms which aim to both revive and reshape the delivery of health services in the country. Donors have been instrumental in this process which is underpinned by a neo-liberal philosophy. On the one hand, rebuilding the health sector essentially involves increasing the level of resources available for health. While overall spending remains stagnant, donors funding of the health sector has increased in recent years, especially as poverty reduction has taken the centre stage of official development strategies in Uganda. On the other hand, transforming the health sector forms part of a wider liberal agenda of change, which is essentially based on enlarging the role of the market vis a vis the state. While it was initially restricted to the economy (narrow “structural adjustment”), this reform agenda has expanded in recent years to emphasise the importance of social services and to push for changes in the way they are delivered. It assigns a limited role to the central state, decentralises decision- making powers to local governments, recognises NGOs as key players, and encourages the participation of the local population in the funding and management of health services. Moving away from the centralised welfare state model of the 60s, the intended cumulative impact of these reforms is to improve the effectiveness and efficiency of health service delivery, thereby improving the health status of the Ugandan population.

Numerous concerns have been expressed about the provision of health services in poor countries implementing liberal economic reforms under donor pressure. In particular, the disengagement of the state from its central responsibility of providing health services to all its citizens, cutbacks in social service budgets and the introduction of “user-fees”, as well as the proliferation of NGOs “filling the gap” have raised concerns about the effectiveness, equity and sustainability of reformed health services. Critics have charged that implementing the neo-liberal agenda actually worsens the health status of poor people, in particular vulnerable groups. This critique forms the backdrop of this study, which seeks to examine the evolving roles of the state, donors and NGOs in Uganda’s health sector, and to assess the impact of their changing roles on way health services are delivered.

Uganda’s experience is very useful in examining these issues. “Star pupil” of the IMF and the World Bank, Uganda has undergone progressive, yet profound, structural adjustment for the past 12 years encompassing both stabilisation and liberal reform of the economy. This process has been heavily assisted by donors who provide considerable funding : aid covers approximately 50% of the government budget, and represents around 12% of GDP. The economy has been extensively liberalised, and the state has been considerably reformed, notably by cutting the number of civil servants and by redefining the relations between the central and local governments (decentralisation). NGOs have multiplied in recent years and are very active in the field of health and education. The health sector, which was in shambles in the mid-eighties, has been affected by these general developments and has been the object of specific reforms seeking to improve its performance.

This study is organised in 4 parts. Firstly, the research questions and the methodology of the study are described. Secondly, the general rationale and history of liberal economic reforms, and the history of their implementation in Uganda, with a particular focus on the health sector, are presented. Thirdly, the main part of the research is devoted to findings and analysis from field studies conducted in 3 districts, Rakai, Arua and Soroti. The main purpose of these specific studies is to illustrate general issues concerning health care delivery in Uganda, thus shedding light on the research questions described above. Finally, a conclusion summarises the main findings, and recommends some directions to address problems and constraints the study has identified.

Research Questions and Methodology

Under the general framework of liberal economic reforms, a “mixed” system of health care provision, combining the state, donors, NGOs and beneficiaries, is emerging in Uganda. The various reforms aiming to rebuild and restructure the health sector have a common stated intention. They aim for enhanced resources to go to a decentralised, more effective and limited state, integrating NGOs and

involving the population in health care delivery. Their ultimate aim is to improve the effectiveness and efficiency of the delivery system, thus improving the health standard of the Ugandan population.

The main questions this research seeks to address are thus the following:

How far have these reforms seeking to rebuild and reshape the health care system actually gone? What are the roles of the central state, local governments, NGOs and donors in the current mode of health care delivery? Are the various actors living up to the expectations which the liberal agenda assigns to them?

How are reforms affecting the actual delivery of health services, in particular as it concerns vulnerable population groups (rural poor, women and children, etc) ? Are essential health services more available? Is service delivery more effective?

What are the implications of these reforms in terms of the sustainability and equity of the health care delivery system? Are reforms going in the right direction?

This research focuses on institutional, managerial, financial and political issues arising from the evolving roles of the state and NGOs in Uganda’s health care system. As the researchers are not epidemiologists or public health specialists, and as data is scanty and unreliable, the actual result of these evolving state and NGO roles, i.e. changes in the population’s health status cannot be systematically or “scientifically” assessed. Rather, the research will broadly examine the general availability of general and specialised health services, focusing on “conventional” health care facilities (number, staffing, services, cost, etc). The quality of services will necessarily be a subjective assessment (staffing, drug availability, satisfaction of authorities, users etc), as epidemiological data on service effectiveness (i.e. the evolution of the population’s health status) cannot be considered.

The methodology of the study is essentially qualitative. A review of policy statements, project documents and funding trends provides background facts and figures. Interviews with key informants, that is state, NGO and donor officials, both at central and district level, furnishes more specific and contextual information. Finally, field observation, namely the visit of health units, provides subjective impressions on the status of health care delivery in the districts.

Examining the above research questions, this study takes the form of a general discussion on the evolving roles of the state and NGOs in Uganda’s health care system, buttressed by illustrations from three districts visited in early 1999. The district studies are an invaluable component of this research. They provide key insights and supporting evidence, thus shedding light on the research questions. However, given the scope of the questions themselves, the amount of information required to answer them fully, and the available time for the field research, the district studies cannot provide a comprehensive and systematic examination of all research questions listed above.

Three among Uganda’s 45 districts were chosen to be examined. They were selected in an effort to reflect Uganda’s geographical, political and economic diversity, and to portray the varying presence and input of donors and NGOs in district health services.

The first district is Rakai, located in the south of the country, along the shores of Lake Victoria and the Tanzanian border.Facts and figures about the three districts are taken from M.O. Rwabwoogo, Uganda Districts Information Handbook, Kampala, Fountain Publishers, 1998. This district is a peripheral part of the kingdom of Buganda, Uganda’s most

developed economic region. Populated mainly by Baganda and other Bantu ethnic groups (Banyankole, Banyarwanda), Rakai numbers 383 501 inhabitants spread out over 4 973 km2. As would be expected with its geographical location and ethnic composition, Rakai district is firmly pro-NRM: Museveni won 81.9% of the vote here in 96. Largely poor and rural, Rakai is a calm and stable area. Also noteworthy is the fact that the first cases of AIDS in Uganda were detected in Rakai in the early 80s, and that the district was devastated by the pandemic. Rakai was also among the first batch of districts to experience decentralisation in 93. AIDS and early decentralisation are the two factors which have attracted heavy donor support and numerous NGO activities to the district. The team visited Rakai in early March 1999.

Arua is the second district to be considered. Located in the West Nile region, Arua is a large district of 7 830 km2, populated by some 637 941inhabitants from a variety of Nilotic and Sudanic ethnic groups, mainly Alur, Lugbara, Kakwa and Madi. Situated in Uganda’s north-western corner which borders Sudan and Congo, Arua is in a sensitive geo-political situation. Originally home of ex-President Idi Amin, Arua has suffered from bouts of violence in the course of a somewhat tumultuous history. Since the NRM take-over, the district has witnessed the presence and fluctuating activity of numerous rebel groups (WBNF, UNRF II), largely composed of disgruntled ex-soldiers who feed on the area’s latent antagonism with the NRM government. Museveni only managed to win 17.3% of the vote here in 96. The proximity with the Sudan has also brought an influx of refugees into the area. Arua’s geographical location is however also an asset, as Arua town has developed into a major trading centre with access to deprived regions of Congo and South Sudan. Like Rakai, Arua was a pilot district for decentralisation as early as 93. Its proximity to the Sudan, the presence of refugees and the high poverty of a predominantly rural population have attracted large numbers of NGOs and donors to the area, not all of whom however work in Arua district. With Islam, Catholicism and Protestantism all boasting sizeable numbers of followers, there is also a strong tradition of missionary and religion-based social institutions affiliated to each faith present in the region. Arua was visited in late March 99.

The final district is Soroti, located in the north eastern part of Uganda. With Kumi district, Soroti is a component part of the Teso region, populated mainly by the Nilotic Itesot ethnic group. As the northern-most part of the district was granted separate district status in 97 under the name of Katakwi, Soroti now encompasses 450 390 people spread out over 5 630 km2. Soroti experienced a devastating and protracted insurgency war between 87 and 92 pitting the NRM government against various rebel groups, mainly composed of aggrieved ex-soldiers and security operatives from Obote’s regime. The war resulted in considerable loss of life and outmigration, widespread destruction of productive assets mainly cattle, and substantial damage to the district’s infrastructure. Since 92, the situation has progressively normalised, and relations between the district population and the central government have much improved. Museveni even managed an electoral victory in 96, gaining 60.6% of the vote. However, the set-back caused by the war is only slowly being overcome: the district was decentralised at a much later stage (95) and has attracted much less donor and NGO interest. Soroti was visited in April 99.

LIBERAL REFORMS AND THE HEALTH SECTOR IN UGANDA

This chapter introduces both the general agenda of liberal economic reforms in Africa and Uganda’s experience of these reforms, in particular as it affects the health sector. Whereas the main part of the study draws heavily on observations from field visits to 3 Ugandan districts, this part is mainly based on existing literature and documents.

First, the rationale and history of neo-liberal economic reforms in Africa is presented. In the 80s, a radical version of “structural adjustment” was introduced by the international financial institutions to address Africa’s deepening economic crisis. It focused on macro-economic stabilisation and on structural reforms aiming to “roll back the state” in favour of market forces. This liberal agenda was much criticised and generally, its results were not very promising. In the 90s, a re-assessment of the paradigm was spearheaded by the same institutions which had initially promoted it, in particular the World Bank. Without modifying its fundamental tenets, the “reformed liberal model” attributes a larger role to the state, puts more emphasis on the alleviation of poverty and on the importance of the social sectors. It also establishes a clearer link between economic liberalisation and political change (“democratisation”); emphasising the importance of “civil society” and NGOs as “agents of restraint” and alternatives to the state.

Second, Uganda’s experience of neo-liberal economic reforms in the past 12 years, and specifically its impact on the health sector is presented. 2 main phases are distinguished, which broadly correspond to the two phases of the neo-liberal agenda described above. In the 87-92 period, the health sector was neglected due to an overriding focus on macro-economic stabilisation. Since 92, the health sector is undergoing a twin-headed process of reconstruction and of restructuring in line with the modified donor-driven development strategy. The main elements of the health sector’s reconstruction, namely increased financial resources mainly through donor programs, and of its reshaping, namely the civil service reform, decentralisation, the integration of NGOs and the participation of the local population, are described.

I.) Liberal reforms in Africa and their critique

In the early 1980s, Africa was in the grips of a profound socio-economic crisis. The immediate cause was a collapse of the price of exports (primary commodities) and the rising cost of essential imports such as oil. Against a background of anaemic economic growth, spiralling inflation and huge budget deficits, the result of this external shock was that many countries were unable to meet their current expenditures and defaulted on the service of their debt. More fundamentally, this crisis prompted the recognition that the state-led development model introduced after independence in the 1960s was failing, as it stifled private entrepreneurship and burdened the economy with an expensive and unproductive state apparatus.

Structural adjustment programs or SAPs, devised by the international financial institutions (IFIs) namely the International Monetary Fund and the World Bank, were designed as a means to get African economies back on track, above all to restore economic growth on a sound footing. Their underlying philosophy is that of neo-classical liberalism, which generally believes that private economic forces competing in free markets lead to rational outcomes, maximising both individual benefits and public welfare. This applies both internally, and internationally, providing a rationale for opening domestic markets to free trade and international competition. In this strict version of SAP, the state’s role is to provide enabling conditions for markets to operate efficiently, namely by ensuring law and order, guaranteeing the sanctity of private property and contracts and correcting market failures.

SAPs come in a variety of forms, yet are generally considered to have two main components, namely the stabilisation of the economy focusing on macro-economic indicators (inflation, growth, balance of payment deficit) and structural reforms within the economy (liberalisation of prices, dismantling state monopolies, privatisation etc). Both are linked, yet stabilisation usually comes first as it primarily entails general policy measures such as maintaining a tight money supply, whereas structural reform involves more complex and drawn out processes of change with institutional, legal and human implications.

Social services such as health and education are doubly affected by the neo-liberal paradigm permeating SAPs. “Rolling back the state” is not restricted to economic affairs. In order to curb budget deficits and stabilise the economy , a reduction of the state’s general expenditure, which includes budget cuts affecting social services, is also required. Further, structural reforms are seen as necessary to improve efficiency in the delivery of services, notably by moving away from the state’s monopoly in the social sector.

Due to their precarious financial position, most African countries have been compelled to resort to the IMF for financial aid in the early 80s. They had very little choice in this matter, as their lack of creditworthiness impeded them from obtaining fresh money from other sources such as commercial banks. Bilateral donors also increasingly tied their assistance to the conclusion of an agreement with the IMF (“cross-conditionality”). Using its very strong position of “lender of last resort”, the IMF has made the release of financial support conditional on the adoption of a structural adjustment program. In the era of SAPs, “conditionality”, or the requirement that internal reforms must be adopted for funds to be released, has become a standard and accepted procedure among donors, in sharp contrast

with the donors formal respect of poor states’ sovereignty and the “laissez faire” attitude that prevailed up to then. The end of the Cold War, which has stripped aid of much of its role of supporting regimes for strategic or geopolitical reasons “no questions asked”, plays a key role in this introduction of economic conditionality. The argument for conditionality is that the provision of aid would not be successful unless the situation which caused economic collapse was adequately addressed, and that donors know better than African governments what the necessary reforms are. As a result, SAPs have been implemented in most African countries since that time, albeit to a varying degree, leading to a general liberalisation of economic conditions across the continent.

Critics of structural adjustment have sharply denounced this phenomenon of the state’s withdrawal and cutbacks in social services expenditures. They have accused donor countries of taking advantage of the financial and institutional weakness of African countries to promote a neo-liberal agenda which disregards the fundamental needs of the population. Their main concern was with the “social costs” of the liberal adjustment process. The state’s withdrawal is seen as translating into a reduction of essential services, affecting primarily vulnerable groups, especially the poor, isolated rural groups, women and children. This negative impact is not only directly linked to budgetary cuts for the social sector and the introduction of user fees (“Bamako Initiative” in health for instance), but also indirectly caused by the widening poverty of vulnerable groups in a liberalised economic environment. Critics charged that structural adjustment and liberal reforms do not necessarily lead to economic growth, and that when growth has occurred, it has been narrowly confined to the export-oriented and internationally-linked elite sector of the economy, neglecting or actually worsening the situation of the poor.

This critique was at first rebuked by the World Bank and other institutions. However, by the late 80s, as the economic performance of reforming countries didn’t conform to rosy expectations, and as much resistance was exerted against such drastic introduction of market forces, the World Bank began to modify its position. As early as 1989, the World Bank’s seminal report “Sub- Saharan Africa: From Crisis to Sustainable Growth” conceded the importance of political “ownership” and capacity in actually implementing liberal reforms, moving away from the position that the market alone had all the answers. Blaming autocratic regimes and oversized bureaucracies for failing to adopt positive economic change, it stressed that “good governance”, namely accountability, transparency and the rule of law were important pre-requisites for reform and development. Implicitly, this report went further than calling for a more efficient state. It also implied that democracy was the political system which would deliver “good government”, but was prevented actually writing it by the World Banks’s “apolitical” mandate.The World Bank, Sub-Saharan Africa : From Crisis to Sustainable Growth, Washington D.C., 1989. The winding down of the Cold War obviously played a major role in promoting this view.

In 1997, the World Bank further acknowledged that the state should continue to play a vital role in the socio-economic development, discarding the extreme position that “a smaller state is necessarily a better state” which imbued the initial SAPs of the 80s. The World Bank’s “State in a Changing World” Report stresses that “markets and governments are complementary”. It defines the state as “a partner in development” and makes the basic point that “for human welfare to be advanced, the state’s ability to undertake and promote collective actions efficiently must be enhanced”. The Bank’s emphasis is that a state should focus its actions based on its capacities, but at least fulfil 5 fundamental tasks, namely establish a foundation of law, maintain sound economic policies, invest in social services and infrastructure, protect the vulnerable and protect the environment, without which “sustainable, shared, poverty-development is impossible”.The World Bank, World Development Report 1997 : The State in a Changing World, Oxford, Oxford University Press, 1997, pp.1and 4. Exit the “minimal watchdog state”, enter the “focused, caring, effective state”.

As a parallel development of the 90s, the World Bank, followed by other donors, began focussing on the eradication of poverty as the key objective of liberal economic reform efforts. The priority on “broad-based” economic growth, seen as the most important way of lifting people out of poverty, is maintained. However, as seen in the list of the state’s core functions above, the importance of social services has been re-appraised in the context of poverty eradication. Instead of being seen essentially as a drag on state resources, social services such as health are now also viewed as an necessary investment to increase human productivity and therefore to combat poverty.See for instance The World Bank, Uganda : the Challenge of Growth and Poverty Reduction, Washington D.C., 1996.

These modifications have not fundamentally altered the philosophical basis of neo-liberal reforms. Economic growth based on free-markets and prudent economic management remains the centrepiece of the strategy promoted by the international financial institutions. However, the emphasis of the reforms has shifted, and their scope has been broadened. In particular, politics have come onto the scene, with the argument that economic and political liberalisation (should) go hand in hand. The end of the Cold War bolstered the argument that unresponsive and repressive states are impediments to economic progress. The so-called “Washington consensus” that “free markets and democracy” are the appropriate recipe for sustained stability and development have become part of a trend espoused not only by the US, but by a number of other international institutions and donors as well.

In this perspective, “civil society” is seen as the “missing link” of earlier attempts at economic reform. As it has been said, in the 1980s, the motto was “free markets”, whereas in the 90s, it is “free markets and civil society”. “Civil society” is considered by liberal theorists to be a necessary ingredient of democracy. Although definitions vary, a mainstream definition of civil society sees it as “the realm of organised social life that is voluntary, self-generating (largely), self-supporting, autonomous from the state, and bound by a legal order or a set of shared rules. It is distinct from society in general, in that it involves citizens acting collectively to express their interests, passions, ideas, exchange information, achieve mutual goals, make demands on the state and hold state officials accountable”.L. Diamond, « Rethinking Civil Society : Towards Democratic Consolidation », Journal of Democracy, vol. 5, no. 3, July 1994, p. 5. Civil society is generally seen as comprising a free media, civic and non- governmental organisations, trade unions and possibly political parties. In this perspective, NGOs,

as a part of civil society, are widely perceived as a positive political force, empowering society vis à vis the post-colonial state criticised for its excessive bureaucratisation, its inefficiency, its high cost and its repressive tendencies. NGOs are seen as emanating from society, representing the private initiative of citizens taking affairs into their own hands, contesting the state and holding it accountable. NGOs are seen as providing an alternative to the state, taking the self-help initiative to step in gaps left by its deficiencies. Here the case for NGOs which derives from liberal economic reforms implicitly dovetails into an argument that NGOs are conducive to political liberalisation and democratisation.

Since the 80s, the African continent has witnessed a proliferation of community and non- governmental organisations, both international and local, performing numerous roles. They have provided a variety of services, mainly in the social sector. As explained above, the main factors behind this growth of the “private non-for-profit” organisations are linked to the emergence of the neo-liberal agenda. Moreover, given the domestic weakness of African economies and “civil societies”, donors have been a driving force in promoting and supporting NGO activities, both for reasons of practical efficiency and economic/political philosophy.

The modifications that the neo-liberal agenda underwent in the 90s have largely taken the edge out of the sharp critique addressed at initial SAPs. The “new” development agenda seems to be less extremist and better intentioned than initial SAPs, with its focus on poverty eradication, its emphasis on progressive and democratic political change and its recognition of the various and complementary roles of state and “civil society”. Social services are again recognised as part of the core responsibilities of states, to which increased attention and funding must be devoted.

However, many questions and concerns still remain. The extent to which the “reformed liberal agenda” really represents a significant shift from initial SAPs has to be posed The fundamental thrust of reducing and redefining the state’s role and of increasing the participation of the “non- state sector”, i.e. NGOs and the local population, has been maintained. Is the state’s withdrawal compatible with effective, equitable and sustainable services? Can NGOs really compensate for the state’s retreat, and are they really part of “civil society”? Aren’t NGO activities fragmented, lacking in continuity and co-ordination? Isn’t the state’s legitimacy being eroded by privatisation of fundamental public services? It is such questions, as they relate to Uganda’s health sector, that the main body of the research will seek to address.

II.) Reviving and restructuring the health sector in Uganda:

As outlined above, Uganda has experienced substantial and extensive reforms inspired by the neo-liberal agenda since 1987. The cumulative effect of stabilisation and structural adjustment, NGO penetration and donor inputs have had a marked, yet complex effect on Uganda’s health sector. In order to better understand their impact, it is important to retrace the historic evolution of structural adjustment in the Ugandan context, and to situate the health sector’s place within the process of reform.

When Y. Museveni’s NRM seized power in January 86, Uganda’s formal economy had virtually collapsed, and state institutions were merely empty shells. The regime’s initial priority, beyond military and political consolidation, was to urgently revive economic activity. The first attempt in that direction, an autonomous state-led policy bypassing the international financial institutions, produced disastrous results. This failure gave Museveni no choice but to resort to the IMF. In May 87, the first of a series of “classical” structural adjustment programs was initiated, and was accompanied by increasingly generous financial support from the donor community.

Between 1987 and 1992, the SAPs’ emphasis was squarely on macro-economic stabilisation, especially curtailing inflation, while at the same time reviving growth. However, due to divergent views on the merits of adjustment within the Ugandan government, implementation was patchy and results were mixed. Although growth rapidly resumed, stabilisation remained an elusive goal, with persistently high inflation, periodic devaluations of the Uganda Shilling’s official exchange rate and large budget deficits. Fiscal and monetary discipline was lacking. In this context of uneven economic performance, the recurrent priority was to “get the basics right” and stabilise. Improving the performance of the social sectors was not high on the agenda.

In the late 80s, the health sector was in a shambles, representing a part, and a reflection, of the general breakdown of the Ugandan state. This presented a remarkable contrast to the initial post- independence period and resulted from the decay produced by dictatorship and war which dominated Uganda’s history in the 70s and early 80s. In the 60s, the health sector was considered one of the best in Africa, with the state investing heavily (with donor support) to provide free public medical services throughout the country. 22 rural hospitals were for instance built during this decade. Health indicators improved steadily. The 70s, under Idi Amin and the early 80s under Obote’s second presidency, combined to destroy those achievements. Many qualified personnel left the country, and public institutions suffered from neglect. The economy was dominated by the informal sector, as people were keen to “escape” from a repressive and unpredictable state. Social service provision likewise became largely privatised and informal. In sum, by the late 80s, the official health sector was only a shadow of its former self. It was governed by an over-centralised bureaucratic structure and manned by a plethora of underqualified and underpaid staff. Health personnel were working in a decrepit infrastructure, lacking the most basic resources such as drugs and medical equipment. Staff motivation and commitment was an all-time low, as was patient attendance. As a result, the public sector virtually ceased to exist as such. Although the staff and buildings were still nominally public, services were informally privatised, with everything on sale in exchange for variable “under the table” payments. Not surprisingly, health indicators plummeted.

As other social sectors, the health sector was neglected in the initial SAPs. However, this was not really as a result of expenditure cuts, as required in “classic” SAPs and denounced by critics. In Uganda, and this is one of the key differences with other reforming countries, “rolling back the state” was never a central component of structural adjustment. In fact all sides, including the donors, agreed that the state actually needed to be rebuilt, and its role in the economy (revenue and expenditure) needed to grow, even as structural adjustment was implemented. This deviation from the standard model is due to the twin historical facts of the collapse of the Ugandan state and the mushrooming of the informal economy: by 86, the state’s expenditures had shrunk to around 8% of GDP, and its revenue was even lower. This is very low compared to Sub-Saharan averages of around 25%. As a result, the economic size of the state actually increased nearly 5 times in real terms between 87 and 93, and the health sector was part of this overall trend.C. Harvey and M. Robinson, Economic Reformand Political Liberalisation in Uganda, Brighton, Institute of Development Studies, 1995, pp. iii and 1. Contrary to the “classical” model, expenditures towards health in Uganda did not decrease, even under the harshest phase of SAP between 87 and 92.

The problematic issue at that time was not really budget cuts, but rather that health was not given a priority status. There were no real efforts to promote or rebuild the health sector, the only endeavours in that direction being stop-gap measures by donors, either under the guise of “humanitarian aid” or “relief and rehabilitation”. These donor inputs were placed within a vacuum. Due to the collapse of state machinery and the absence of a functional Ministry of Health, donors initiated “vertical programs” dealing with specific issues. The donors’ priority was to provide a given service and to achieve results. They realised that if they worked through the existing weak and inefficient government institutions, their only achievement would be to lose funds. Donor therefore set up and funded parallel structures, often “advised” by expatriates, specifically to insulate and manage their programs from the rest of the administration. Some of these were autonomous bodies, such as the pioneer Uganda National EPI program (UNEPI) funded and managed by UNICEF, or the National Medical Stores “Essential Drugs Program” funded by DANIDA and managed by the Danish Red Cross. Others, such as the World Bank’s “First Health Project, 88-92”, an ambitious rehabilitation programme marred by mismanagement and delays, were run by “project implementation units” within the Ministry of Health. Despite their individual achievements, these programs could not achieve a beneficial overall impact as they were too specific and disjointed. With their mode of implementation, they also initiated a process of institutional fragmentation within the Ministry of Health.

Since 1992, the reformist camp within the Ugandan regime, committed to full-fledged liberal economic reforms, has, for a variety of reasons, gained the upperhand. Monetary and fiscal discipline have been enforced with a considerable degree of stringency. The economy has stabilised with inflation averaging less than 5% per year, while growth has continued apace, averaging over 7% per year. At the same time, the structural reforms which had somewhat half- heartedly begun between 87 and 92, were brought into full swing: the exchange rate and trade regime were liberalised, the Coffee Marketing Board was dismantled, and privatisation of the parastatals began in earnest. Donors have since the early 90s shown a keen interest in Uganda’s economic success, and have been prepared to provide sizeable contributions of aid to promote and help deepen the reform process. Total aid to Uganda increased from 500 million USD per year in the early 90s to approximately 800 million USD per year by 97/98. This comprises both general budget support, assistance for particular reforms (such as the creation of a Uganda Revenue Authority or decentralisation) and classical development assistance (infrastructure, technical capacity building etc).

During this second major phase of the SAP process, stabilisation has been largely achieved and the emphasis is more distinctively on structural reforms. Initially neglected, the social sectors, health included, are now more squarely in the picture. They have been affected by two major changes, which have the overall ambition of improving the delivery of health services to the Ugandan population. First, rebuilding the health system has emerged as a new priority for donors, and by implication, for the government. Second, a return to the status quo ante is ruled out. Reshaping the system is seen as a necessary measure to improve the effectiveness of health service delivery.

Rebuilding the health system has risen on the agenda, mainly due to increasing donor interest in the social sectors. A review of aid allocations to Uganda shows that health has consistently featured among the most favoured recipients of assistance, and increasingly so as of the early 90s. Donor commitments currently total approximately 60 million USD/year (cf. table). Both as a response to severe public health problems, in particular the AIDS pandemic, and in an effort to generally rehabilitate overall health services, donors have launched a number of major new programs. The World Bank has initiated the “District Health Support Program” (95 to 2002, 45 million USD) and the “Sexually Transmitted Infection Program”. (94 to 2000, 50 million USD). Other donors such as the EU and bilaterals like the UK, Denmark and the US have also started substantial programs in the health sector, most often aimed at reviving and improving health service delivery in particular regions of the country and frequently involving partnerships with local governments and NGOs.

	1993/94	1994/95	1995/96	1996/97
Total Disbursement	472.9 $	583.2 $	605.6 $	456 $
Health	39.3 $	44.6 $	52.7 $	64.5 $
Health %	8.3%	7.6%	8.7%	14.1%

Source: Aid Data Unit, Ministry of Finance, Uganda
NB: Not all this aid goes through the Government, nor is all of it reflected in the Government’s budget.

A large part of the Ugandan government’s budget is donor funded. Although the precise percentage of aid in the health budget cannot be derived from available national statistics, health is no exception to the general rule.

	89/90	90/91	91/92	92/93	93/94	94/95	95/96	96/97	97/98	98/99
Aid as % of:
Total Expenditure	58.6	59.5	60.6	63.1	56.7	45.4	46.3	40.1	45.1	42.9
Development Expenditure	72.8	75.7	81.5	90.9	89.2	81.5	87.3	76.8	83.1	76.8

The Ugandan government’s expenditure on health, inclusive of donor contributions, is globally stable in proportion to general budgetary expenditures. As stated above, the share allocated to health did not decrease during the initial phase of SAP (up to 92), although health is clearly wasn’t a priority. Since then, real allocations have increased, as the size of the government’s budget has expanded, the economy has grown and donor interest has picked up. The government’s overall expenditures more than doubled from 8.2% of GDP in 89/90 to 17.3% in 98/99, while GDP itself almost doubled in real terms over the same period. Figures for health are more difficult to interpret, but it is safe to assume that real expenditures on health, while remaining stagnant in percentage terms, grew in line with general government spending over this period.

94/95	95/96	96/97	97/98	98/99
Total Budget	605.59	643.47	798.44	882.00	1127.13
Health Budget	47.82	62.95	59.34	60.24	71.29
% Health	7.8%	9.7%	7.9%	6.8%	6.3%

Figures in millions of UgSh (current). The exchange rate between the UgSh and the USD has slid from around 1000 UgSh for 1 USD in 91/92 to around 1450 UgSh for 1 USD in mid 99.

Source : Background to the Budget 1999-2000, Ministry of Finance, Uganda. These figures include direct foreign aid to the governement.

However, despite an increase in donor interest and a general growth in Ugandan government expenditure, health spending does not amount to much. In 98/99, the Ugandan Government spent 71.3 Billion UgSh on health, corresponding to approx. 50 million USD. This is approx. 2.5 USD per inhabitant, and corresponds to 0.9% of GDP. In terms of % of GDP, the level of health spending remains very low, and is not increasing. It has been 0.97% in 94/95, 1.13% in 95/96, 0.98% in 96/97, 0.84% in 97/98. As the Background to the Budget document for 98/99 states, « the most critical budgetary issue in the health sector is the overall level of resources. The Government of

Uganda currently spends less than 1% of GDP on health (compared with 2.8% of GDP on education) ».Republic of Uganda, Background to the Budget 1998-99, p. 82. In general, it is estimated that expenditure for providing a « cost-effective » package of essential health services amounts to 12 USD per person per year.T. Barton and D. Bizimana, Health Sector Review, Oxfam Uganda, 1995, p. 7 In Uganda, the combination of public (2.5 USD/year) and private expenditure (which was estimated at 4.91 USD in 95/96 and is generally considered to be approximately twice as high as public expenditure) is estimated to be included between 7 and 12 USD per year.Republic of Uganda, Background to the Budget 1998-99, p. 82. This is lower than the desired amount, and moreover, certainly not all of it is spent on « essential health services », as the majority of private spending is carried out by the better off, while the poor rely more heavily on public services. Objectively, there is still a funding gap.

New developments are increasing donors’ interest in the social sectors, health included. In particular, the emergence of poverty alleviation as the guiding principle for WB and other donors’ policies in Uganda has placed social services, including health, under the spotlight. Since the early 90s, poverty has become a key concern for the World Bank and other donors. Once considered a marginal side-effect of liberalisation to be dealt with by piece-meal palliative interventions (such as « PAPSCA » (Program to Alleviate Poverty and the Social Costs of Adjustment), poverty has now become the centrepiece of the World Bank’s intervention strategies. This is due to the fact that, despite inconsistencies and controversies in the measurement of poverty, it is generally acknowledged that poverty remains very high, and that, even in the most optimistic assessment, it has not been significantly decreasing in the past years. The Government, based on a study by the Oxford-based “Centre for the Study of African Economies”, maintains that general poverty has decreased from 56% to 46% of the population between 92 and 96.Republic of Uganda, Background to the Budget 1998-99, p. 51. However, the Government also concedes that the situation of the poorest 20% has actually worsened over the same period. The latest UNDP « Human Development Report » embodies the official consensus : “despite the modest decline in the extent of poverty, the country’s remarkable economic performance has not generated wide enough benefits”.UNDP, Uganda Human Development Report 1998, p. 39. In the 97 «Poverty Eradication Action Plan », a Government of Uganda document which resulted from a lengthy WB initiated consultative process, a two pronged approach to « eradicate poverty » was adopted, with prong one being improving people’s incomes, and prong two being improving theRepublic of Uganda, Uganda National Integrated Household Survey 92-93, Volume I, Table 1.51.12. quality of poor people’s lives. In this context, the low health standards of the Ugandan population are seen both as a cause, and a result of poverty. To wit : Ugandans lose 4.3 days per month on average of work due to bad health. Hence, improving health makes sense as a means of fighting poverty. It is an investment in human capital, improving the productivity of poor people. The PEAP, backed by the World Bank, therefore advocates a radical increase in government expenditure on social services, including health in the coming years. Debt relief under the “Highly Indebted Poor Countries” initiative, which is reducing Uganda’s debt to multilateral bodies such as the World Bank and the IMF, is also made conditional upon the channeling of the saved money (approximately 30 million USD per year) to the social sectors. The “poverty action fund”, which amounts to approximately 50 USD per year is a fund composed of HIPC money and additional donor contributions which focuses on education, rural roads and health in support of the PEAP.Republic of Uganda, Background to the Budget 1999-2000, p.70. Although these initiatives have yet to be translated into significant funding increases for the health sector, they are significant in two respects. First, poverty reduction provides the first major cogent justification for increasing social expenditures other than « humanitarian » reasons. Second, it denotes a change in development strategy on the part of the World Bank, once one of the main exponents of « classical » structural adjustment with its “anti- social sector” bias.

Although rebuilding the health sector has risen in significance, it does not entail simply reviving the health delivery system of the past. Rebuilding cannot be dissociated from reshaping the health system in line with the prevailing liberal ideology. This essentially means moving away from the centralised, welfare state model of the 1960s. In that perspective, several cumulative reforms have taken place or are underway, affecting the state in general and the health sector in particular. Taken together, four reforms are having wide ranging effects on the way health services are being delivered.

First, a vast civil service reform program has been implemented with substantial donor backing. Initial studies undertaken in the late 80s confirmed the general perception that the Ugandan state apparatus was bloated with underpaid, largely corrupt and demotivated staff and was, as a result, failing to carry out its key functions. The pay of civil servants was so low that they were forced to resort to a number of survival strategies, including taking on extra jobs, selling their services (“corruption”), thieving and diverting resources. Backed by donors, the civil service reform program initially focused on reducing the size of the civil service by /peliminating ghost workers and retrenching the least qualified staff. By freeing up resources, it was expected that the pay of the remaining state employees could be increased, thus boosting efficiency. These reforms have succeeded in bringing the number of public employees from 320 000 in 1990 to 164 632 in 1998, and in sizeably increasing wage packages.Republic of Uganda, Background to the Budget 1998-99, pp.19-20. As was the case in other civil service departements, significant numbers of health personnel were retrenched, and salaries of health personnel paid by the central government have considerably increased over the years. However, they still remain lower than their private-sector equivalents, and for a large number of staff, they are still inferior to the so-called “minimum living wage”The total package includes the basic salary, which is variable according to function and me.comexperience and currently ranges from 49 548 UgSh/month for all support staff, 103 096 UgSh for enrolled nurses/midwifes to 249 544 UgSh for senior nursing officers/doctors-medical officers. The top salary is 430 698 UgSh/month for senior doctors-medical officers. Moreover, « pay-as-you-earn » income tax is deducted from these salaries, as is graduated tax of 80 000 UgSh per year. A uniform « lunch allowance » of 66 000 Ug Sh per month is added on top of all salaries, including for support staff. All staff are also supposed to have housing provided, but in practice this is not always the case. The reforms also intended to improve institutional efficiency by restructuring the state’s organisation and management systems. However, institutional change, such as the introduction of “results oriented management”, has been a lot more difficult to achieve.

Second, a significant and ambitious process of decentralisation was launched, again with considerable donor funding and technical assistance. Decentralisation was aimed at radically transforming the power relations between the central government and the initially 39 (45 since 1997) local governments at the district level, by gradually devolving substantial administrative, political and financial authority from the former to the latter. The process started by transferring limited responsibilities to 13 pilot districts in 93, and gradually expanded, including the whole country and increasing the amount of devolved authority from the centre to the local units. In 95,decentralisation was enshrined in the Constitution, and since 97, all districts were granted wide ranging powers under the Local Government Act. District Councils, headed by a directly elected Chairman (“LCV chairman”), are entitled to make their own policies on a wide range of issues, and are responsible for managing activities in a number of sectors, including road maintenance, health and education. In these areas, the central government’s role is restricted to policy making, monitoring and supervision. Civil servants working in the districts are no longer appointed and deployed by the central government, but are squarely under the responsibility of the district administration. Local governments are entitled to collect their own taxes, in particular general graduated tax (assessed tax payable by all heads of household) as well as market dues, trade levies etc. However, the bulk of district funding derives from unconditional (“block grant”) transfers from the central government for general duties, and from the central government’s earmarked conditional transfers for specific activities. Decentralisation, which will be examined in greater detail in the following sections, involves a considerable transformation in the way the public services such as health are planned, managed and funded.

Third, in line with donor perspectives on the appropriate role of the state in a liberal environment, the private sector, in particular NGOs, are increasingly being brought into the mainstream of public service delivery. The underlying rationale is that the state, while retaining an overall responsibility for the supply of “public goods” such as health, should not necessarily be the agency actually providing the health services. If other agencies possess a “comparative advantage”, and can provide services more efficiently, the state should willingly recognise and encourage their efforts. In this perspective, the state’s role is to set out a general framework (“policy and standard setting”), while actual service delivery is left to the most efficient provider.

A number of studies undertaken in Uganda do show that the NGO health units are more efficient than the state-run facilities. The Ministry of Health’s “White Paper on Health Policy 93” recognises that “Government has 60% of the beds, but NGOs [mainly church-run institutions] serve more inpatients. Capacity utilisation is 50% higher in NGO hospitals”.Republic of Uganda, Ministry of Health, White Paper on Health Policy, Update and Review, 1993, p. 10.Based on the “National Health Manpower Study” of 1990, the World Bank also calculates that staff productivity, expressed as the number of patients handled per staff per year, is much higher (and therefore better) in NGO institutions than in the public sector. NGO doctors treat as much as 5 times more patients than government doctors.The World Bank, Uganda, The Social Sectors, Washington D.C., 1993, p.58. The World Bank also quotes a study of hospital costs in three government and two NGO hospitals, which shows that spending per inpatient in 89/90 in government hospitals was approximately double that in the NGO hospitals.The World Bank, Uganda, The Social Sectors, Washington D.C., 1993, p. 59. Study by Dr. P.Y. Kadama, the Health Sector Costs and Efficiency Study in Uganda, IHPP, 1991.

There is a need to distinguish between 2 main categories of institutions, which are both lumped together as “NGOs” (or “Public Non for Profit” PNFP) in Uganda. On the one hand, there are the church-based and run institutions, generally endowed with permanent infrastructure and a high degree of independent funding and who often engage in basic (curative) service provision with a long-term perspective. Hereafter, they are referred to a “church-run services”. On the other hand, there is a great variety of international and local NGOs and CBOs. In spite of their diversity, most of them run programs in specific geographical areas and/or dealing with specific health issues or problems (such as AIDS). Their focus is generally more preventive than curative, with programs such as health education, community based health care (CHW, TBA training), AIDS awareness etc. When NGOs are mentioned in this text, it is this type of NGOs that is being referred to.

It must be underlined here that church-based health facilities, have a long and well-established presence in the country. As early as 1987, a Government appointed “Health Policy Review Commission” noted that in the 70s and early 80s, these institutions had prevented the health sector from collapsing all together, as they maintained essential services while the government-run facilities had all but ceased to exist. At that time, the government and church-based facilities entertained informal, and often collaborative, relations. In the 90s, the policy of integrating “NGO sector” represented, in the case of the church-based facilities, a formal acknowledgement of their existence and performance by the state. The situation is quite different with the other types of NGOs, namely international and local non-governmental and community-based organisations. Their history in Uganda is much more recent, and, as will be examined later, their activities often much more of a limited, temporary and issue-oriented nature. Their integration therefore poses particular questions.

The policy of integrating NGOs, pushed by the donors, is both a pragmatic acknowledgement of the state’s limited capacity, particularly acute in the Ugandan case, and a ideological shift in mainstream conceptions about the desired role of the state. As will be examined in subsequent chapters, this has led to an integration and institutionalisation of NGO activities within the public health sector, which represents a major change in the way health services are being delivered in Uganda.

Fourthly, again reflecting the new consensus particularly among donors, the local population has been called upon to participate and contribute more actively in the delivery of social services. Here again, this reform is a product of pragmatism and ideology. On the one hand it follows from the recognition that the state’s organisational and financial ability to provide services is limited, and that the local population’s efforts are needed to supplement it. On the other, it reflects the view that local ownership and control of public services is necessary to “tame” the state, thus improving the efficiency of services by limiting abuses and making them more responsive to local needs and concerns. The intention of moving away from the centralised and distant welfare state managed by bureaucrats is very apparent here. Two main forms of local participation have been introduced. First, there is the “user-fee” system, loosely based on the internationally accepted “Bamako Initiative” principles. In Uganda, user fees have been implemented “under the carpet”, as Parliament in 1990 refused to abolish the ideal of universal free service and to officially sanction their introduction, as proposed by the Ministry of Health and the Cabinet. In practice however, public health institutions have been allowed to set up their own fee collection system, with district health authorities issuing general guidelines on fee levels and utilisation. Although not specifically legal, the practice has been de facto institutionalised throughout the country. As Districts have the right to raise local revenue, the collection of user fees has been considered to be approved under that authority.Republic of Uganda, Local Governments Act, 1997, Article 81 (1). User fees have been heavily promoted by donors, in particular the World Bank, as the major way of easing the revenue constraint in the funding of health services. In the “White Paper of 93”, which is described as an “excellent basis” by the World Bank, and subsequently in the “Three Year Plan 93-95”, the Ministry of Health estimates that 15% of Uganda’s total expenditure on health will be derived from user fees by the end of 95.Repubic of Uganda, Ministry of Health, The Three Year Health Plan Frame, 93/94-95/96, 1993, p. 48.

Secondly, health management committees (HUMCs) have been set up to guide and supervise the operations of public health units. Part of the general philosophy of popular participation introduced by the NRM government, these committees include members of the elected local councils. The HUMC’s chairman is an elected member, whereas the incharge of the health unit acts as a secretary to the committee. Linking health units to the local population, HUMCs tasks include overseeing management, planning activities and supervising the utilisation of the collected user fees. At the district level, elected district councillors, in particular the “secretary for health”, also play a similar role of providing public scrutiny and supervising the operations of the health services. By encouraging the participation of beneficiary populations in the management of services that directly affect their lives, these reforms are potentially reshaping the mode of health service delivery in Uganda.

To sum up, this overview suggests that, contrary to the strict neo-liberal model where the state is compelled to withdraw from social service provision leaving NGOs to pick up the pieces in its wake, the situation in Uganda is far more complex. In line with the “new” liberal agenda of the 90s, liberal reforms in Uganda have been associated with a reconstruction and redefinition of the state’s role, rather than a simple reduction of the state’s capacity to provide essential services. As opposed to the “welfare state” model of the post-independence period (notwithstanding the important role played by church-based social services), the key difference is that state no longer claims a monopoly in the planning, funding (through general taxation) and delivery of social services. While the state’s capacity in health has increased since 86 supported by several donor programs, other actors are also included in an increasingly systematic fashion. NGOs provide both general and specialised health care, donors participate in funding services and in elaborating health policies, and users are required to directly participate in the management and financing of health services.

We now turn to the district studies, in order to assess, following the research questions, how far these liberally inspired reforms have gone, and whether they are achieving their stated objective, improving the delivery of health services and, by implication, the health status of the Ugandan population.

DISTRICT SPECIFIC FINDINGS AND DISCUSSION

I.) Increased attention to health, yet poor effectiveness of services:

The first major and general finding from the district studies supports accounts made by other analysts, and is even echoed in Ministry of Health and donor publications. It is that, although health has been receiving increasing attention, in particular from donors, this heightened effort to rebuild the sector does not seem to be translating into visibly more effective services, nor a marked improvement in the population’s health status. The following elements from the districts can be taken into account.

The paradox of low, yet abundant, financial resources in the health sector:

The district budgets and interviews of district officials in Rakai, Arua and Soroti reveal an apparent paradox concerning the funding of the health sector. The overall level of expenditure, expressed for instance in per capita terms, is low. Donors and NGOs have a substantial presence, except in the curative sector, which is well funded from the government’s recurrent budget. However despite this overall low funding base and heavy donor presence, health is well funded compared to other sectors. Also, both Ministry of Health and district officials state that the health sector is sufficiently funded.Dr. Prosper Tumusiime, Head of the Planning Unit, Ministry of Health, interview, Kampala, May 20, 1999. The problem therefore seems to be less with overall levels of funding, than with the structure of the resources, and the ability/capacity of the district to spend the available resources efficiently and effectively.

The overall level of expenditures on health are low : Soroti was due to spend 3599 UgSh (or 2.5 USD) per inhabitant on health in 98/99, including donor funds. The respective levels for Rakai and Arua are 3721 UgSh and 6085 UgSh. Not surprisingly, these figures are lower than the national averages (except for Arua), as a considerable percentage of the health budget is spent centrally, in particular for the Ministry of Health and Mulago Hospital, the national referral hospital based in Kampala. Mulago Hospital alone absorbed on average close to 20% of total health expenditures in

Uganda in the period between 94/95 and 98/99.Calculated from : Republic of Uganda, Background to the Budget 1999-2000, pp. A 18-A23. 23 Mr. Okoropot, Chief Administrative Officer, Soroti District, interview, Soroti, March 24, 1999. 24 Republic of Uganda, Population Secretariat, 1998, quoted on www.uganda.co.ug/nadic

Again reflecting national trends, donors comprise a significant portion of health expenditures at the district level. Visible donor funds (including the donor-backed Primary Health Care program) account for between 25% and 40% of total expenditure on health in Soroti and Arua. This is roughly equivalent (somewhat less in fact) to the donor share in Uganda’s national budget. Rakai District, which doesn’t have a regional hospital absorbing sizeable central government funding, is in the same bracket (34% in 98/99) due to the relative scantiness of donor funding in health in that District.

This does not include other « invisible » donor funds, such as direct donor funding of NGOs (which are often not reflected in the district budget) or the support donors provide towards funding the central governement’s budget deficit. This means that even recurrent expenditure which appears to be wholly government funded, such as salaries of district hospital staff, in fact contains a percentage of donor funds.

Compared to other sectors (excluding education with the heavily donor-supported “Universal Primary Education Program”) such as agriculture or infrastructure, health is relatively well funded. In Arua, health spending amounted to 3.8 Bn Sh in 98/99, which represents 23% of the district’s total budget. This is significantly higher than agriculture (less than 2%) or works (less than 4%). Most respondents, in particular members of district health teams, as well as district officials and politicians supervising health, acknowledged that health was receiving a lot of money. As stated by the Chief Administrative Officer of Soroti District, the level of money is not the problem, rather that “even though it is well funded, health is lagging behind”.Mr. Okoropot, Chief Administrative Officer, Soroti District, interview, Soroti, March 24, 1999.

An improved, but still insufficient, health infrastructure:

The infrastructure for health delivery has increased markedly over the past years. A large number of health units have been either rehabilitated or constructed in the districts examined. In Rakai District, an entire hospital was newly built using World Bank funds (Rakai Hospital, First Health Project) and another was rehabilitated (Kalisizo). In Soroti, which is lagging behind essentially due to the long period of warfare and insecurity, the regional hospital is currently being substantially rehabilitated, again by the World Bank (DHSP). All districts stated that numerous peripheral health units had, or were, being rehabilitated, with Soroti again clearly the worse off.

Despite this effort, the existing infrastructure, when assessed according to objective indicators, is still insufficient to cover the essential health needs of a fast growing population (2.7%/year) Republic of Uganda, Population Secretariat, 1998, quoted on www.uganda.co.ug/nadic.

There were for instance 2.18 beds per 1000 people in Soroti, 1.6 beds/1000 in Arua and only 0.8 beds/1000 in Rakai, whereas the national average was 1.26 beds/1000. Whereas 49% of the population nation-wide was living within 5km of a health unit, in Arua the percentage was 48%, in Rakai 39% and in Soroti only 19%.

The Ministry of Health has recognised the problem of insufficient health infrastructure, and is implementing a controversial policy of creating “health sub-districts” in an attempt to improve the situation. “Health sub-districts” are in fact a further stage in decentralisation, and represent a geographical division of the district around a number of hospitals or high-level health centres. These centres of the “health sub-districts” are expected to carry out much of the planning, management of supervision of health activities in their area, leaving the district health team to focus on broader policy issues. In particular, peripheral health units are supposed to report to the centre of the health sub- district, not to the district health team for day-to-day matters. As hospitals and high-level health centres are currently insufficient in number, the policy entails an “upgrading” of several health centres per district to include basic surgery, blood transfusion, maternity services, as well as the posting of qualified staff (at least one medical officer and two medical assistants per centre). This ambitious policy, which is currently being implemented with much difficulty using “Primary Health Care” programme funds for the infrastructural development, would lead to an expansion of the country’s existing health infrastructure, as well as a further redefinition of the management of health services.

A lack of qualified human resources:

The number of qualified staff is still very limited, especially in remote rural areas where the bulk of the population resides. In Arua, there were 427 trained personnel in 1993, 291 in Soroti, and 230 in Rakai. This represents 0.66 trained staff for 1000 people in Arua, 0.64/1000 in Soroti and 0.59/1000 in Rakai, all lower than the national average of approximately 0.8/1000.Republic of Uganda, Statistical Abstract 1998, p.11According to the Ministry of Health, only one third of the positions that should be filled by qualified staff actually are.Dr. Prosper Tumusiime, Head of the Planning Unit, Ministry of Health, interview, Kampala, May 20, 1999. Related to the issue of an overall lack of qualified staff is the familiar problem that trained personnel, especially doctors, are concentrated in hospitals, mainly located in urban centres. Arua has 12 doctors attached to its main regional hospital, and 5 others for the remaining government health facilities in the district. Soroti’s regional hospital is staffed with 11 doctors, but except for the DMO and 3 doctors at the church-run Luwala Hospital, there are no other doctors working in public facilities in the district. Compounding this problem is the fact that many, if not most, peripheral health units up to health centre level are staffed only with enrolled nurses or unqualified nursing aides. There are no doctors, hardly any medical assistants, and only few registered nurses at this most important level of health care delivery. This lack of qualified staff represents a major constraint on the availability and quality of the health services delivered by the public and “non-for- profit private” sectors.

A mix of health service providers:

In all three districts visited, the Government remains the largest provider of health services, in line with the national situation. The Government runs the main hospitals and the vast majority of the peripheral health units, and employs the majority of the health staff. National evidence from health users’ survey shows that the public sector is both the most used (40% of outpatients resort to the public sector, as compared to 35% private and 25% NGO)Republic of Uganda, Ministry of Health, White Paper on Health Policy, Update and Review, 1993 p. 10., and the preferred, service provider (83% of health users say government health services are their preferred service provider).T.Barton and D.Bizimana, “Family and Household Spending Patterns for Health Care”, in: R. Claesson (ed.), Health Care Financing: Proceedings of the Dissemination Seminar, Health and Economics in Uganda, Kampala, March 95, pp. 10-11.

However, in all three districts, again reflecting national trends, the non-governmental sector was a major player in health delivery. Private “for profit” and traditional medicine is clearly a large provider, but was not closely examined in this study. More attention was given to the “private non- for-profit” (or PNFP) sector, which actually comprises two main categories often misleadingly lumped together as “NGOs” in Uganda.On the one hand, there are the church-run services, which have permanent health facilities independently providing basic services. These are well established institutions with a long term perspective: “we have been here 50 years, and God willing, will be here another 50” as the doctor in charge of the Protestant “Church of Uganda”-run Kuluva Hospital in Arua District put it.Dr. K. Neudeck, Acting Medical Supervisor, Kuluva Hospital, Arua District, interview, March 23, 1999. They are significant service providers, especially in hospital based care. There are 2 church run hospitals in Arua district out of a total of 4, one in Soroti out of 2, and one large health centre in Rakai (2 public hospitals).

On the other hand, there is the very diverse group of international and local NGOs and CBOs (community-based organisations). They generally have a more recent presence in the districts and are engaged in more limited programs. The limitation can either be geographical (part of the district), thematic (one specific illness, problem) and/or temporal (a specific time period). Unlike the church-run services, they are in general not involved in basic service delivery, emphasising prevention rather than curative or integrated services. As an example of the issue-oriented nature of NGO programs, almost all health NGOs in Rakai are involved in AIDS related activities (Médecins du Monde, World Vision, RAIN). Their programs are comprehensive, including prevention with AIDS awareness and treatment with home-based care and counselling. However, they do not geographically cover the whole district, even when they are all put together. The situation in other districts is more diverse, with the considerable number of NGOs present and engaging in a variety of activities, ranging from health education, training of community health workers, technical support to the districts, sleeping sickness control etc.

An issue common to all districts is the variable, but generally low rates of patient attendance rates in public service institutions. In Arua, according to the district health team, only approximately 20% of sick people attend public (or PNFP) health facilities. Patients continue to heavily rely on private medicine (drug shops) and traditional practitioners for several reasons among which are an preference for auto-prescription and medication, belief patterns and distrust linked to the poor quality of the public service. Financial considerations associated with user fees also play a role. This seems to reflect an even poorer rate of government service use than national averages: the Ministry of Health states that the government covers 40% of curative services delivered, and that approximately half of the population uses some form of care outside the organised modern health sector.Republic of Uganda, Ministry of Health, White Paper on Health Policy, Update and Review, 1993, p. 10.

More generally than the respective merits of different health care providers, poverty is a powerful explanatory factor for the low level of health service use. According to the National Household Survey on utilisation rates, 51% of Ugandans do not seek medical attention when they are sick! The same study points out the economic constraint, stating that only 56% of Ugandan households say they could afford the health care they wanted, the last time someone in the household was ill.Republic of Uganda, Background to the Budget 1999-2000, p. 60 and 65.

The health status of Uganda’s population: mixed trends, but a still “unacceptable”situation

The effectiveness of the many reforms seeking to rebuild and restructure the health delivery system is open to debate. Evidence concerning their intended aim, the betterment of the Ugandan population’s health status, is uncertain. Global indicators concerning the evolution of the health status of the Ugandan population are both very tentative due to poor statistics, and mixed in their actual results. On the one hand, infant, under five and maternal mortality rates seem to be decreasing, with UNDP estimating that infant mortality rates dropped from 121/1000 in 94 to 88/1000 in 96.UNDP, Uganda Human Development Report 1998, Kampala, p. 18. Life expectancy is also increasing according to UNDP, improving from 41.8 years in 1991 to 50.4 in 1996, but others, such as the official Uganda Population Reference Bureau, state that life expectancy is actually dropping, and currently stands at 40 years for men and  years for women in 1998UNDP, Uganda Human Development Report 1998, Kampala, p.25 and Uganda Population Reference Bureau, 1998, quoted at : www.uganda.co.ug/nadicOn the other, vaccinal coverage seems to be decreasing: immunisation rates against measles dropped from 70% in 96/97 to 49% in 98/99, with coverage rates for other vaccines showing a similar trend.Republic of Uganda, Background to the Budget 1999-2000, p. 22. Total fertility rates are very high (7.3), and not declining.The World Bank, Uganda, Social Sectors, Washington D.C., 1993, p. xiii. In any case, Ugandan health indicators among the world’s worst, and are below regional and Sub-Saharan African averages. The Ministry of Health acknowledges that the situation is “unacceptable”, and presses for a deepening of the reforms to actually improve health indicators.Republic of Uganda, Ministry of Health, Ten Year National Health Policy (draft), November 1998, p. 2. A further indication that Uganda’s health system perfoms poorly can be found in the high level of distrust of those who can afford to avoid it. In 1996, the Government spent 421 920 USD, and 1 503 451 USD in 97 for the overseas medical treatment of the country’s political elite… Figures quoted in M.K. Mulera, « Adyebo"s unwanted medical bills, tears District level indicators show that Rakai, Soroti, and Arua are no exceptions to the generally poor health status of the Ugandan population. In a “quality of life index” computed for an OXFAM 1995 st  Health Sector Review based largely on health-related indicators, T. Barton ranked Arua 21, Rakaird th 23 , and Soroti 27out of the 32 districts examined. Arua’s infant mortality rate was 137, Rakai’s 119 and Soroti’s 116, while the national average at that time was 122/1000. Other indicators all show the 3 districts either within, or under, national averages. Beyond statistical figures, there is a wide unity of view among the district officials who were consulted and who all subjectively considered the health situation in their districts to be poor, inspite of the efforts undertaken to date. To sum up, in general, it appears that the albeit still limited, mainly donor-funded investment in health services has failed to be translated into visible improvements in terms of health service delivery, and in the health status of the Ugandan population. Why? There are certainly many factors at play, some of which have been alluded to above, such as the mediocre starting point, the low level of health facility use, the lack of qualified and motivated staff, the objective funding gap and the poor quality of health services provided. Others, such as the effects of the overall socio-economic evolution on the population’s health status, are clearly beyond the scope of this study. However, additional analysis from the district studies presented in the remainder of this study can help shed light on this disjuncture. Instead of looking at general issues, the ensuing discussion focuses on the effects that the reforms, which intend to improve the situation, have actually had, in order to show that some of the problems are linked to the reforms themselves.

II) Unsatisfactory Division of Institutional Responsibilities between Districts and the Ministry of Health

One of the main factors impeding the effective delivery of health services in Uganda is the shifting and as yet unclear definition of roles and responsibilities between the central and local governments. Clearly, decentralisation is a major and potentially meaningful change in the way services are funded, planned and delivered in Uganda. However, numerous constraints are preventing it from realising its potential.

The major thrust of the initial 1993 decentralisation reform was to empower the local administration vis a vis the central government. The Chief Accounting Officer (CAO) in effect became the most powerful figure in the district, since he/she controlled the purse for an increasing number of delegated activities and responsibilities. Since 1997, decentralisation has made an important step forward by injecting meaning and competition into local politics. This is because the 97 Local Governments Act gives increased authority to the elected politicians, especially the directly elected District Chairman (“LC V”), over the local administration. The District Chairman appoints a “cabinet” of permanent councillors (“secretaries”) with particular responsibilities over “directorates”, such as works or health. In certain districts such as Rakai, where the relations between the district administration and the newly elected District Council is tense, the political leadership conceives of the secretaries as instruments of surveillance over the civil servants and as “tools to cut them to size”.Captain D. Matovu, District Secretary for Health, Rakai District, interview, March 2, 1999. Significantly, the secretary for health’s office is next door to the District Director of Health

Services (DDHS, formerly the District Medical Officer or DMO), and he shadows the DDHS in many of his official functions. Several other changes have likewise reinforced the district’s political authority. All district-based personnel, the CAO and the DDHS included, are to be hired and fired at the district level. The link with the central government has thus been severed. A “district service commission”, with an independent status, is to handle all personnel issues in order to minimise political interference. It has however become increasingly clear that decentralisation has gone hand in hand with a “localisation” (some say a “tribalisation”) of the district administration, and an increasing leverage of the political leadership over the civil servants. The allocation of budgetary and aid-related resources is also increasingly being oriented at the local level, as all budgets, aid inputs included, have to be voted by the District Council.

The district administration, although it has lost some of its powers to the district politicians since 97, has also globally gained an enhanced authority as a result of decentralisation. Most importantly, the district administration is directly responsible for implementing budgeted activities. This entails the central role of spending money. Not surprisingly, it is over this issue that the tension between district administration and politicians has peaked. In Rakai, the newly elected political leaders wanted to be the ones signing the cheques… an audacious move which was successfully defeated by the civil servants. The enhanced budgets which have come as result of decentralisation, including the influx of aid-related funds and activities, put the district administration at the centre of resource allocation and use. As a result of the heightened responsibilities and resources vested in the district administration, qualified staff have been returning to the districts. Whereas several years ago, Kampala was the only place to be for university graduates and trained staff, there is now an incentive for such personnel to go back to the districts, often in remote rural areas. In the health sector, district health teams have been strengthened, and have a bigger say over the management of the health sector.

Despite increasing the authority of the district political leadership and administration, thereby making the district an important hub of power and resource allocation, decentralisation can still be described as incomplete and half-hearted. There is a very long way to go before reaching full decentralisation, which partly explains why the expected benefits from decentralisation, in particular effective service delivery, hasn’t (yet) materialised. In particular, the planning, determination of priorities and actual control over resources for activities whose responsibility now squarely lies with the district is not effective, despite the rhetoric. The districts are still very much at the receiving end of policies and budgets which are determined at the centre between the government and the donors.

There are many impediments to full decentralisation, the most important of which is the structure of resources, which does not reflect the theoretical division of roles and responsibilities between the central government and the districts. District resources are supposed to come from 2 main sources, local revenue and government transfers. Concerning the latter, there are 2 main types of transfers. The first is the so-called unconditional grant (or “block grant”) which is meant to provide core funding for districts to spend according to their priorities. The second are conditional grants, which are funds earmarked by central government for specific activities which districts should implement, such as the maintenance of feeder roads or the delivery of primary health care, further examined below. Delegated funds, such as the funding for regional hospitals, is essentially a particular kind of conditional grant.

The main problems with government transfers are the hidden earmarks and the low level of funding which affects the unconditional or block grant. A large part of the unconditional grant is in fact earmarked for the payment of salaries to ex-central government staff who are now under the responsibility of the districts. In Arua, out of 1.9 Bn UgSh received per year as a block grant, almost 1.2 Bn UgSh is earmarked for such salaries, which include the wages of trained health personnel. When all salaries have been paid and the operations of the local administration provided for, almost nothing remains for non-wage expenditure in the different sectors. In Arua again, the block grant amounts to approximately 160 Million Sh per month, of which salaries and operating expenses consume 130 Million, leaving about 30 Million to be split among the activities of 8 directorates… Essentially, this means that directorates without aid-funded or conditional grant funded activities can do next to nothing.T. Opiti, District Financial Officer, interview, March 24, 1999.

This sorry state of affairs is also due to the fact that local revenue is extremely poor, and contrary to expectations, is not increasing. In Soroti, local revenue was 685 Mn UgSh last year, which represents approximately 1520 UgSh per person, in Arua, it was 938 Mn UgSh, which amounts to 1540 UgSh per person, and in Rakai, it was 577 Mn UgSh, representing 1504 UgSh per person annually. In each district, local revenue represents just over 1 USD per capita per year. This is obviously very low, and largely below budgetary expectations. The district administrations attribute the low tax collection rate to a number of factors, including widespread poverty, the narrowness of the tax base, climatic problems (El Nino), political campaigns, and inefficiency and incompetence of tax collectors – chiefs. Certainly, these all play a role to a certain extent. Concerning the narrowness of the tax-base for instance, there are only 39 000 tax payers in Soroti, where the minimum graduated tax is 10 000 UgSh and the maximum is 80 000 UgSh (employees, civil servants). In Arua, it is estimated that, out of a total of 75 000 taxpayers, the 7000 employees and civil servants in formal employment, in particular the 4000 teachers, contribute to 2/3 of the total tax revenue collected.

One factor however stands out to explain the dismal level of tax returns, namely the « ever negative attitude of the population towards payment of taxes »Soroti District Council, Preamble to the 1998-99 Budget, 1998, p.a. or the « unwillingness of tax payers to realise the importance of paying taxes »Arua District Council, Background to the Budget 98-99, 1998, p. xiii. Only a « small proportion [..] willingly walk to the sub-county offices to settle their tax obligations. The majority have to be coerced to pay taxes »Soroti District Council, Preamble to the 1998-99 Budget, 1998, p.b.

The expectations of decentralisation theory were that local revenue would increase. It assumed that people would pay more willingly to the local government than to a distant central state, as they would see a more direct link between their payment on the one hand, and decisions and service delivery at the local level on the other. Local democracy would thus be beneficial to revenue collection. In fact, the contrary may be true, as the Rakai budget statement for 98/99 observes : in the past financial year, « most of the time was spent electing local leaders », so that « the district administration feared to harass the electorate during campaign days to force them to pay graduated tax and other dues ».Rakai District Council, Budget 1998-99, 1998,This persistent lack of local revenue raises a big question mark and imposes a great burden on the effectiveness of decentralisation.

Essentially, the current structure of local government revenue, with limited unconditional funds and local revenue, means that district can’t plan and decide on their own priorities. They are constrained by the money that is available, and the conditions that are attached to it. As shall be seen in the following section, donor funds and activities reinforce this trend.

A clear and significant illustration that districts fail to control resources and allocate them according to their priorities relates to the payment of staff. As stated above, the salaries of qualified staff, who were formerly employees of the Ministry of Health, are included in the “block grant” paid by central government to the districts. In general, this means that qualified staff are paid regularly, as health is a “priority program area”, protected from budget cuts. However, unqualified staff, or staff hired by districts after decentralisation, are not included in this central government allocation. Unqualified staff, especially those at the health unit level, were previously paid by the Ministry of Local Government, and now fall on the shoulders of the districts. They are to be paid with the remainder of the block grant and/or with local revenue. In fact, due to the poor level of the block grant and the low revenue collection, this category of personnel is not paid, or only paid after a long delay. Arrears amount to approximately one year in Arua, and 6 months in Soroti.

The rules of local revenue use make it even more difficult to pay this category of staff: as per decentralisation guidelines, the district only retains 35% of collected revenue, 65% staying at the sub-county level for administration and activities there.Republic of Uganda, Local Governments Act 1997, Art.86 (2). It is out of this 35% that these salaries are supposed to be paid… In Arua, 35% of the total expected local revenue, if collection is 100% effective, is 489 Mn UgSh per year, but the total local wage bill is 428 Mn UgSh. As collection averages generally only 70% of its target, there is a constant deficit, which means that staff cannot be fully paid and arrears accumulate.

This is really a major problem, since unqualified staff such as nursing aides or enrolled nurses form the majority of the health personnel at the peripheral level, often being in charge of the health units. Moreover, this shortage of funds prevents the district from hiring staff, who are available and necessary to improve the health services. This applies to qualified as well as non-qualified staff, as the Ministry of Health is not longer recruiting (except for delegated services such as regional hospitals). Human resources are now solely the district’s responsibility. As a result, nursing schools produce new graduates, but districts cannot hire them. Health units are therefore still headed by (unpaid) nursing aides. Bureaucratic rules and district weakness combine to achieve unfavourable human resource management, thus undermining the effectiveness of the health delivery system.

Besides financial issues, other constraints militate against the effectiveness of decentralisation. The main ones include the lack of management skills and organisation (“capacity”) at the district level, the mentality and habits of centrally-oriented civil servants and the resistance from central ministries like the Ministry of Health to hand over effective authority to the districts. These constraints are more qualitative than material: in all three districts visited, the district health teams had decent offices, computers, photocopiers and vehicles, all courtesy of various aid programmes. The problem lies rather in the requisite skills to plan, organise and manage health services. For many civil servants accustomed to receiving directives from a far-away centre, decentralisation represents a considerable change in work attitude and requirements. Instead of being reactive and executing instructions from above, district level civil servants are now required to be proactive, initiating and directing activities according to their local situation. Decentralisation being a recent reform, it is not surprising that behavioural traits and competencies are still more adapted to the old system, reinforced by a tendency from the central ministry to retain its old powers and privileges. In Arua for instance, district civil servants often state that they have a “double allegiance”, to the district and to their old ministry. The Ministry of Health, as will be seen with the Primary Health Care program, continues to rely on a “top down” approach, giving little space to the district initiatives and priorities.

III) Donor-induced distortions of priorities and modes of service delivery:

The second important factor which sheds light on the inadequate manner in which health service are delivered in Uganda is the important and ambiguous role played by donors. On the one hand, donors provide indispensable funding and support to the health sector, without which much of the sector’s health activities simply wouldn’t exist. On the other hand, donor-driven policies and donor- funded programmes significantly affect the way in which health care is provided, leading to a distortion of priorities and modes of service delivery.

As donors compose a large part of government revenue, both generally and in the health sector, they have a large say in the determination not only of government spending patterns, but also of priorities and policies. In that sense, their leverage is both quantitative and qualitative. Policies such as the introduction of user fees, which were “smuggled in through the back door” when Parliament refused to authorise them, were in fact heavily pushed by donors, in particular the World Bank, as Okunozi and Macrae’s work clearly shows.S. A. Okuonzi and J. Macrae, « Whose policy is it anyway ? International and national influences on health policy development in Uganda », Journal of Health Policy, 1994. The policy on the recognition and integration of NGOs was also clearly impressed upon the Ugandan Government by donors, again with the World Bank in the lead. It derived from the neo-liberal development philosophy on the appropriate role of the state, endorsed by the Ugandan Government in a strategy document entitled “The Way Forward II” published in 1992. This is not to say that the Ugandan Government is passive or powerless, or that all health policies are a result of donor imposition. However, these examples testify both to the intellectually persuasive capacity and the sheer financial power wielded by donors in their relation with the Ugandan Government.

Decentralisation in itself is a reform which has received considerable donor support. Donors are decentralising funding and are implementing “development” and “capacity building programs” at the district level. Increasingly, districts develop direct relations with donors, analogous to the relations between donors and the central government. Rakai for instance has a annual “donor conference”, where representatives from Kampala-based embassies and international institutions are invited to support the district development plan.

In each of the three districts visited, a major donor had set up a direct programme of support to the district, albeit with varying modalities. In Rakai, Danida have embarked on an ambitious 15 year “Rakai Development Program”, worth approximately 4 million USD per year, which combines capacity building of the district administration and political leadership, with large-scale development activities. The objective is to enable the district to “take off” with a massive injection of technical assistance and financial resources. Danida provided a considerable part of its funding directly through the district administration, including an unearmarked “block grant” of approximately 200 000 USD per year (within Danida’s 750 000 USD yearly recurrent contribution to the district budget). In Soroti, the Dutch Government has initiated a “Soroti District Development Program”, which has the particularity of being based on “participatory rural appraisals” undertaken at the sub-county level to define its priorities and activities. Also a long-term undertaking, implementation of the SDDP is integrated into the district administration. Initiated in 96, SDDP has provided an average of 2.6 million USD to the district in the past 2 years, mainly spent on a community-based cattle restocking and rural credit scheme. In Arua, the Dutch government has taken a different approach, choosing to support development activities through purposefully set up programs (Community Action Program, Women’s Empowerment Programme) run by SNV (Dutch government “NGO”) outside the district administration. One of the Programmes, the “Capacity Building Programme” is integrated in the CAO’s office, and aims to strengthen district management by providing equipment and training (over 0.5 million USD per year). The whole SNV programme is in the process of being integrated into the district administration over the next few years.

This increased direct donor presence at the district level in Arua, Rakai and Soroti only affects the health sector in those districts in a marginal fashion. None of these programs has a specific health focus. Danida’s large programme does have some health projects, especially in the area of nutrition and reproductive health, but these are very minor compared to the bulk of the programme activities. In Soroti, health came in 5 th or 6 th position of community-expressed needs in the first subcounty “participatory rural appraisals”. In sub-counties lacking health facilities, health had a higher priority ranking, often placed in 2 nd position behind agriculture. SDDP has however been reluctant to build health units because of the recurrent cost implications linked to the district’s problems in hiring and paying qualified staff. In 98, an SDDP policy review actually recommended to stop all health activities, but this was not fully taken on board. SDDP has tried to strengthen and reorient the planning of health services in the district by hiring a “technical advisor” to the district health team. Attached to AMREF, this expert, who is originally from the district and who came back from exile to take up this post, is contracted by SDDP. SDDP are also funding the rehabilitation of some health units, especially to install maternity wards.Peter de Lange, SDDP manager, interview, Soroti, April 20, 1999. In Arua, the “Community Based Programme”, which also relies on “participatory rural appraisals” to define its activities, has undertaken some community-based health activities, namely health education and training of TBAs. The few infrastructure projects that were carried out (2 dispensaries, and 2 maternity wards) proved to be problematic, as the inputs and the staff promised by the district medical office were very slow to materialise. E. van der Werf, SNV West Nile Program manager, interview, Arua, March 23, 1999. All in all, these sizeable development programmes set up by donors at the district level have not resulted in significant activities in the health sector.

The major part of donor-funded health activities at the district level take the form of vertical programs. The most important are the World Bank funded DHSP and STIP, but the EU’s “District Health Secotr Support Program” and “European Development Fund”, the Ministry of Health’s PHC and UNICEF programs all conform to the same basic model. Overall funding decisions, definition of objectives, priorities and an outline of recommended activities are all decided at the central level. The district’s role is to establish a workplan reflecting these instructions, essentially describing activities and explaining how the money will be spent within the guidelines. The PHC guidelines for instance earmark certain funds for the upgrading of specified health facilities, then orient the use of unearmarked funds, suggesting “probable allocations” for specified program areas such as 5% for “school health programme”, 15% for “safe motherhood and infancy”. They suggest authorised activities, such as “condom supply through outreach centres” and prohibit others, such as “purchase of medical/transport equipment” or “sitting or committee allowances”.Republic of Uganda, Ministry of Health, Guidelines on the Use of Conditional Grant for Primary Health Care (PHC) 1998/99, 1998, p. 9.

Although the range of authorised activities is broad, the whole process is directive and top-down. The district is essentially a recipient of centrally-defined funding and programs, which still basically conform to the vertical model initiated in the 80s, when the state had collapsed.

Districts are required by the Ministry of Health to establish an annual overall workplan for the health sector. Examining one of these workplans, for instance Arua’s, shows that the health sector is essentially an accumulation of donor-funded vertical programs, aside from core funding for staff salaries and operating costs for hospitals which come from the government. The following budget presentation also reflects this division between governement funded expenditure and the collection of donor funded programs.

ARUA Health Budget (in millions of UgSh. Shillings)	1998-99
Government
Arua Regional Hospital, Operating Costs	410.0
Arua Regional Hospital, Salaries.	350.0
District Hospital (Yumbe), Operating Costs.	150.0
District Hospital (Yumbe), Salaries	/ (incl. HU)
Lunch Allowances, Regional and District Hospitals	284.0
Salaries, Health Units.	429.5
Lunch Allowances, Health Units	318.2
NGO Hospitals, Operating Costs	143.9
Arua Nursing School	114.6
Primary Health Care Grant (donor-funded, from debt relief to the Government)	332.7
Total	2 589.6
Donors
World Bank (DHSP)	262.5
World Bank (STIP)	193.8
European Union (DHSSP)	314.7
UNEPI (immunisation)	216.7
Danida (Essential Drugs)	99.0
WHO	20.0
Others	129.0
Total	1 235.7
Grand Total	3 882.0

Examining the description of activities in the health plan is even more explicit. While government funded activities are general, such as “payment of salaries”, donor-funded activities are specific and detailed, such as “training workshops for 18 health unit management committees: 11 participants, 2 facilitators x 5 days, 16 380 million UgSh” (EU funded), “continue quarterly meeting of STIP implementation team, 1 480 million UgSh” (STIP) or “monthly supplies of 10 litres o kerosene to 14 primary health units for sterilisation equipment, 1 680 million UgSh” (PHC).Arua District, Arua District Health Workplan 1998/99 by Donor excluding Hospitals, 1998.Each activity, even holding a meeting or conducting a supervision visit in a health centre, has to be planned, with an appropriate price tag attached.

This research was unable to evaluate the effectiveness of these donor-funded programs, and did not have access to any specific evaluation studies. It is therefore difficult to say whether these programmes have met the objectives they have set out for themselves. However, beyond the narrowly defined programmatic or “instrumental” effectiveness, there is a wider definition of effectiveness of aid, which Carol Lancaster names “contextual” effectiveness. This has to do with the wider impact of aid programmes in social, political and economic terms.C. Lancaster, Aid to Africa : So much to do, so little done, University of Chicago Press, Chicago, 1999, p. 5. It is this type of effectiveness that we now examine.

A significant institutional consequence of these donor-funded vertical programs is that the process of fragmentation, which was experienced by the Ministry of Health at the central level, is replicated at the district level. There are no “program implementation units” as such at the district level, but each key member of the district health team has an interest in managing, or being the “focal point” of, a (lucrative) donor funded program. The incentive is clear: all activities conducted under these programmes, from writing a report to holding a meeting, are specifically funded. Moreover, there is usually a “capacity building” component included, which ranges from a vehicle and/or a computer, to workshops, trips to Kampala and training seminars. As there are usually several donor-funded programmes, each of the key members of the district health team usually gets a fair share. However, one of the important consequences is that donor-funded program activities have a priority ranking, and that district health team members model their work around (funded) program requirements. This can certainly have perverse effects, as donor-funded program activities may not match the district’s actual health priorities.

Put together, the effects of these donor-funded programmes have considerable implications for the working patterns of district health staff. Workplans are reactive, and are made to conform to funding requirements and programme guidelines, not necessarily to district priorities. Essential health service delivery is neglected to favour funded “activities” such as health education, training, sensitisation etc. These are potentially useful, yet they should not be undertaken at the expense of the backbone of the health sector. The mentality of health personnel is affected by an “allowance culture”, as there are monetary inducements attached to all donor-funded activities such as seminars and workshops. For the unpaid district staff, it is part of their survival strategy. To provide essential services in the health unit and not be paid, or to attend an “orientation workshop on mother-baby packages (WHO, 3.5 million UgSh)”, a training on the “integrated management of childhood illness (DHSP, 7.2 million UgSh)” or an “awareness meeting among women’s groups and cyclists about trauma at the parish level (UNICEF, 14 million UgSh)”, all of which pay allowances and per diems; the choice is easy to make.Examples from the Soroti District Health Workplan, 1998-99. For the district management team, the “allowance culture” also operates, despite the fact that basic salaries are paid by the Government. Managing donor programs, rather than managing the overall health sector, is the priority, and the two do not necessarily overlap.

On the financial level, the combination of government and donor priorities leads to a distorted allocation of resources, which doesn’t address the priority health needs of the majority of the Ugandan population. The “essential health package”, which the Ministry of Health recognises to be the key objective it is aiming at, is not available.Republic of Uganda, Ministry of Health, Ten Year National Health Policy (draft), November 1998, p. 4. The bias towards urban-based curative services, especially Mulago and other regional hospitals, is essentially due to government’s political priorities. As cited above, Mulago alone absorbs on average 20% of the health budget, with district hospitals taking a further 18.6% (in 98-99). While donors attempt to focus on primary health care in rural areas, their funding patterns, although they justifiably emphasise preventive actions, are biased towards activities such as training, workshops, seminars, mobilisation and sensitisation. One of the main reasons for this bias, as will be described in more detail below, is that donors want to focus on “development activities” and refuse to fund recurrent expenditure, such as salaries.

The provision of essential health care at the peripheral health unit is arguably the most important part of the entire health service delivery chain. It encompasses curative and preventive actions, and is the health system’s first and most important point of contact with the community. It is also probably the most neglected component of the health system. As many of the health workers are not paid, the health units attempt to maintain their activities by drawing on a combination of resources. First, the regular (if often delayed) supply of essential drug kits from National Medical Stores (Danida funded) is vital. Second, donor programmes assist indirectly, periodically providing allowances to staff (training, workshops). Allowances for immunisation activities from UNEPI are important supplements. Thirdly, as will be seen later, user-fees contribute a small amount of cash, insufficient to be put to productive use but useful to top-up the payment of staff at the health unit level. “Informal” or under the table payments also still occur. What must be stressed here however, is that despite their primary health care focus, donor programmes don’t squarely help alleviate the basic problems of essential health care delivery, both curative and preventive, which is to be routinely delivered at the health unit level.

There are two additional important issues related to donor-funded programmes which need to be addressed, namely how they affect the equity and the sustainability of health care delivery.

On the one hand, it can be stated that donor-funded programmes are generally associated with inequity. There are two aspects to this general point. Firstly, by definition aid chooses and cannot be applied equitably. Those who don’t receive aid, for instance certain districts or parts of districts, local-based staff and non-donor funded activities are discriminated against. Under the current system, since local revenue is weak and equalisation grants between districts, although they are provided for in the Constitution, aren’t implemented, “poor people” leads to “poor services” in the absence of donor funding. Therefore, competition among districts to attract donor funding is great, and the necessarily uneven spread of aid ensures that these inequalities are not alleviated, but reinforced. Secondly, the logic of donor-funded programmes, which allows certain expenditures but prohibits others, in particular recurrent expenses, leads to inconsistencies. For instance, Soroti District cannot afford drivers, because their salaries have to be paid out of local revenue. Therefore, health inspection activities are not carried out regularly, despite the availability of donor- funded allowances, but not salaries, for drivers.

On the other hand, the sustainability of aid programmes is doubtful at best. One of the main reasons is the abusive terminology of “development” attached to most donor funded programmes. As donors want to their programmes to have a development focus, they refuse to fund recurrent expenditure, putting their money solely into “development activities”. While this may be

Apparently this condition of not funding recurrent expenditure permeates even funding nominally under the Government, such as the PHC programme funded by the “Poverty Action Fund” underwritten by donors. While the PHC programme represents a new injection of money, it does not fund needed recurrent costs (for instance the salaries of health staff which will be deployed in the newly upgraded « health sub-district » centres). Rather, as

understandable in terms of preferred role sharing between donors and the Government, in practice the distinction is both theoretically difficult to make, and practically impossible to carry through. In fact, many donor-funded programmes have a long-term orientation, and are “self-contained”, meaning that they fund almost all activities which are undertaken under the programme. Although the Government should theoretically contribute sufficient “counterpart funds” to at least pay for all recurrent expenses associated with development programmes, in practice this is not the case. As a result, what should really be considered recurrent expenditure (such as “support – supervision” of peripheral health units under DHSP) is disguised as “development” or “capacity building” expenses.

Clearly, many such activities, which are recurrent inasmuch as they are conducted year after year and form part of the normal expectations and revenues of health workers (training, workshops, etc) are not sustainable. They would cease due to lack of donor funds. Beyond the simple financial burden that taking over donor programs poses on districts, there is a perverse effect of “costing” every single activity within the district health workplan. This practice is undertaken to justify donor funding allocations. As a result, every activity is now considered to have a price (writing a report, inspecting a health unit, giving a training). Since much of these activities are donor-funded, it is feared that they will cease when funding stops, even when they are within the normal attributions of posted (and paid) staff.

Moreover, the recurrent cost implications of donor funded programmes are frightening for districts. They are linked on the one hand to the fact that some donor programmes substitute for important recurrent expenditure that the government should actually be paying, and on the other that development-oriented donor programmes increase the health infrastructure and services offered. Some activities funded by donors are really meant to continue, and it is far from certain that districts will be able to manage. Several examples stand out, such as integrating of services previously run by NGOs (for instance MSF run sleeping sickness programme in Arua), managing expanded essential services following the creation of health sub-districts and continuing drug purchase and delivery if/when the Danida-funded “esential drugs program” phases out. For instance, the PHC grant provides for recurrent costs of health sub-districts on paper, yet staff costs are not included. As a result, when Arua District recruited a medical doctor to be posted in the upgraded Koboko health centre, the Ministry of Health told the District that it was the one who had to pay. As explained above, this is beyond the financial capacity of the district, and the appointment had to be cancelled.

Likewise, it is difficult to see how Arua District will be able to fund and manage the sleeping sickness treatment and control programme initiated by MSF. MSF identified and “filled a gap” in the health delivery system by taking care of a serious public health problem caused by a high incidence of sleeping sickness in the district. The district didn’t have the capacity or the expertise to tackle this problem. In doing so, MSF implemented a vertical and autonomous program, which almost totally substituted for district health services. Currently, as prevalence rates are appearing to decline, and as a number of staff have been trained, MSF feels that its active presence is less justified, and wants to hand over responsibility to the district. On the one hand, there is the large task of bringing the district health team into the planning and management of the programme, and of actually reintegrating formerly separate health facilities and staff into the district health system. This task is complicated by the autonomous and vertical manner in which MSF implemented its activities, but these difficulties can be overcome. On the other hand, there is the more intractable question of the additional burden that taking over this program places on the district’s shoulders. Sleeping sickness is a serious public health problem, but it is not ranked among the 10 priority other donor-funded programmes, PHC money is restricted to “development activities”, such as the upgrading of health facilities and the establishment of health sub-districts. diseases in Arua. It is therefore not high on the district health team’s agenda. Moreover, additional staff and facilities to maintain, as well as drugs to buy, imply recurrent financial costs. Given the district’s tight financial situation, it is unclear how they will be able to cope if central government doesn’t step in. Finally, the drugs themselves are rare and need to be sourced on the international market. Treatment protocols are complicated, and evolve according to the availability of drugs and the evolution of scientific knowledge. It is far from certain that the district will have the required expertise and access to maintain supplies and to manage such a complex and changing treatment programme.

Clearly, MSF’s decision to phase out is logical, as international NGOs have little justification in becoming permanent health service providers in Uganda. The state should take responsibility in providing public health services. Also, the very success of MSF’s work in helping to bring down the prevalence rate will ease the district’s task, as there should be less patients and the treatment costs should be lower. However, if the district doesn’t perform well in surveillance and control of the disease, thereby maintaining a low prevalence rate, the situation could rapidly deteriorate especially due to the vicinity of Sudan and Congo, both “no-state” areas. This would again require the costly intervention of an NGO…

The difficulty of MSF’s integration points to a real dilemma concerning the role of NGOs in Uganda’s health care system. On the one hand, as the next section will describe, for NGOs to become institutionalised auxiliaries of the state’s health care system is unjustifiable. This is clearly the path that most NGOs have taken, at the donors’ insistence and with the state’s acquiescence, yet its logic based on “cost-effectiveness” is faulty and its benefits in terms of sustainable and equitable health care delivery are dubious. On the other hand, NGOs which show a degree of leadership, taking on tasks which fall beyond the capacity or means of the state, have a very difficult time actually phasing out and handing over to the state when their presence is no longer justified.Another example concerning the shaky sustainability of NGO programs trying to lead the way and then hand-over to the state was International Care and Rehabilitation (ICR)’s activities in Rakai. This NGO, with DfID funding, revived a health clinic, in particular by paying for the deployment of a medical assistant. When the funding was over, the NGO expected the district to take over, but this failed to materialise due to a lack of funds. Services have collapsed. Surely, the way these NGOs implement their activities, especially the degree of substitution they embody, plays a role in determining the process of (re)integration. MSF’s vertical and autonomous program is hence difficult to integrate. However, there are important human, technical and financial constraints to taking over donor-funded NGO activities, which, in the current state of the decentralisation process, districts are not able to overcome on their own.

A final point relates to the increasing use of expatriate “technical advisors” posted at the district level. This is somewhat reminiscent of the colonial days, as Uganda had not seen “mzungus” (white people) posted in District administrations since the British! It reflects the pattern set at the central level, where the number of technical advisors and experts has expanded along with the rise in aid funds. With decentralisation, donors want to control/supervise the activities and use of the funds they are now allocating to the districts. In a repetition of what has occurred at the central level, they are posting “technical advisors” (often belonging to NGOs) at the district health team level. In the most extreme case examined, the EU posted 2 expatriate advisors to the DMO in Arua, one from CUAMM and one from SCF, even though the EU funded DHSSP programme represents only about 10% of Arua’s overall health budget! Given the record of technical assistance, which is often no more than substitution for locally available expertise and experience, and which is known to deresponsabilise the office holders, can this really be considered “capacity building”?

Another example concerning the shaky sustainability of NGO programs trying to lead the way and then hand-over to the state was International Care and Rehabilitation (ICR)’s activities in Rakai. This NGO, with DfID funding, revived a health clinic, in particular by paying for the deployment of a medical assistant. When the funding was over, the NGO expected the district to take over, but this failed to materialise due to a lack of funds. Services have collapsed.

IV) The uncertain benefits of NGO integration:

One of the central tenets of the neo-liberal agenda is that the non-state sector, in particular NGOs, should have an important role in service delivery. In a new division of labour, the state should give up its monopoly on service provision and focus on policy-making and supervision, while NGOs, where they are more efficient, should deliver services. As stated above, there is a need to distinguish church-based and run institutions from the great variety of international and local NGOs and CBOs. While this policy is sensible in the case of long-established church-run “private non for profit” institutions, its logic is shaky and its benefits are controversial in the case of international and local NGOs. By providing NGOs with state money, often derived from donor sources, NGOs become institutionalised players in health service delivery, a trend welcomed by most NGOs. However, NGO activities, while they may be effective in narrow terms, by definition lack equity in duration and coverage.

It is generally assumed that NGOs “fill the gap” left by the collapse or withdrawal of the state, thus providing services and fulfilling functions that would otherwise not be fulfilled. One of the key findings of this study is that this is only partly the case, and that when it happens, it is due to donor requirements. The numbers, funding and activities of NGOs in Uganda generally, and in the health services of the studied districts in particular, in fact greatly increased after the initial phase of SAP, as the general economic environment improved and reconstruction and reform got seriously underway. NGOs accompanied the state’s efforts and followed the donors’ increased engagement and funding. In that sense, they did not really lead, or work to fill a vacuum when it was arguably the most needed, when, in the initial phase of SAPs, state funding for health was very low and state services had all but collapsed. In Rakai for instance, there were very few NGOs tackling the AIDS problem in the late 80s and early 90s, when the epidemic was arguably at its peak. The first AIDS-related organisation, Rakai Project, started as a research venture in 88. In fact, it is not strictly speaking an NGO, as it is jointly run by the government’s National Institute of Health and Columbia University. In 92, Rakai Project’s counselling and treatment services began in a limited area, and an off-shoot, RAIN, was created to implement the same activities in other areas not covered by Rakai Project. In fact, RAIN was specifically created on the request of a donor (Danida) who didn’t want to fund a state institution for such treatment related activities. World Vision, began activities in 91 with World Bank “PAPSCA” funding, while MdM started its activities in 93. None of these NGOs, except MdM at Kakuuto health centre, provided curative services for AIDS patients, which were almost non-existent in the district before Rakai Hospital was built (with World Bank funds) in 95-96, and Kalisizo Hospital renovated (again with World Bank funds). This example shows that NGOs are not necessarily ahead of the state and the donors, either time-wise or in terms of the type of services provided.Soroti District could provide a similar illustration of this phenomenon, as there was hardly any assistance apart from MSF-Holland during the difficult period of insurgency. It was only after the Government and donors had started reconstruction that NGOs began to move in. In fact, they usually follow the Government and the donors’ agenda in terms of timing, priority activities and areas of intervention.

The current emphasis on the recognition of church-based and NGO activities, and their integration into the mainstream of health sector delivery takes several forms. In particular, NGO can access government funds, most often provided by donors, for specific “development” programs undertaken in the district. Donor programs such as the World Bank’s STIP and DHSP release funds to the district, conditional upon the subcontracting of NGOs to carry out the actual program activities. The district health team plays a policy, monitoring and supervisory role, while NGOs actually implement the programme. NGOs are presented as more competent and efficient than the state, as they are already operational with a number of human and logistical assets. Moreover, program money only funds specific activities (workshops, training, allowances etc), making it necessary for the NGO to have core funding from other sources. In that sense, these programs are presented as a financial saving for the state. Their activities can be implemented relatively cheaply by the NGOs since the program only covers part of the costs. NGOs apply for donor funds based on specific program guidelines. Most of the activities deal with prevention and capacity building, typically ranging from community mobilisation and to health worker training. Essential (curative) service provision, or routine activities integrated into service delivery at the health unit level, are generally excluded. NGO applications are passed on to the Ministry of Health (project implementation units) through the district health team, who issues recommendations and advice on the NGOs’ project and on its performance record. The Ministry of Health then selects the NGOs and provides it with funds, which are disbursed at the district level through the district administration.

Church-based hospitals also receive operational/recurrent funding from the Ministry of Health (central government), in support of their public service delivery role. However, the level of funding is much lower than for Ministry of Health hospitals: for instance, the 250 bed regional hospital in Soroti receives 345 Mn UgSh per year (98/99) for operating costs, while the 135 bed Luwala hospital, also in Soroti District, gets only 26.3 Mn UgSh per year (98/99). Moreover, the staff in Soroti receive salaries (383 Mn UgSh per year) and lunch allowances (184 Mn UgSh per year) from the central government, which is not the case for Luwala. Church-based hospitals are also recognised as centres of “health sub-districts”, which represents a further step in decentralisation. Accordingly, instead of being supervised by the district health team, the health units within a particular sub-district are placed under the responsibility of the centre of the health sub-district, which can be a church-based hospital. As such, these institutions receive PHC funds (just like the Ministry of Health hospitals) to provide support and supervision to the peripheral health units, Luwala Hospital for instance receives 2 Mn UgSh per year for this purpose.

This trend of integration and provision of state funds is generally seen positively by church-based institutions. The main reason is that church-based hospitals are in a very difficult financial situation, as overseas funding for recurrent activities has been decreasing in recent years. In essence, they are largely surviving on user-fees, as will be seen in the next section. De facto, church-run hospitals are providing an important public service, and the allocation of government funding represents a recognition of that role. The main problems which church-based hospitals point out are on the one hand technical and administrative difficulties in fund disbursement, leading to delays in activity implementation and on the other the insufficiency of government funding, even for PHC activities. There is a continued need for core funding from other sources, even to fulfil activities which government has newly and wholly delegated to the church run hospitals, such as support and supervision of health units. For instance, Luwala Hospital does not have a vehicle of its own, making PHC outreach activities very difficult, if not impossible. Theoretically, they would have to use a vehicle from the district health team, but this is illogical and doesn’t happen in practice. Another (potential) problem relates to institutional responsibility, especially for supervision of health units. Most of these are government-run, with government- employed staff, yet they are to be supervised by church-run hospitals. This makes sense in medical terms, but doesn’t conform to administrative hierarchy. The church-based supervisor therefore needs to refer to the district health team in case any disciplinary or administrative action vis a vis the staff is required. The converse is also true: if peripheral health workers have a grievance against the church-based supervisor, they will have difficulty seeking redress. As the church-based supervisor is not employed by the district, it is unclear what authority the district health team has over his activities.

Most NGOs, especially the local ones, also welcome this integration and look forward to an institutionalised and permanent role in the health care delivery system. There are a number of indications to that effect. The clearest is that NGOs adapt their activities to donor-funded program requirements and guidelines. As the manager of HealthNeed in Soroti put it, his NGO built on its competencies and existing activities to expand and devise a program based on DHSP guidelines.R. Ochen, Manager HealthNeed Uganda, interview, Soroti, April 21, 1999. Other NGOs such as World Vision also clearly modify and expand their activities according to the availability and the requirements of donor funding.S. Tumwikirize, Program Manager World Vision, interview, Kampala, March 11, 1999. Some local NGOs go further by coopting district political and administration officials on their board of directors : RAIN for instance has put the DMO and the LCV Chairman as ex-officio board members.Ex officio means they are members by virtue of their position, not in their individual capacity. They do not have voting powers.RAIN says that this integration of officials within the NGO structure been helpful to its activities by enhancing the

transparency of its operations.D. Kitusibwa, Program Manager RAIN, interview, Kalisizo, March 4, 1999.

Many international NGOs, if and when they pull out, also seek to put a local NGO in their place to continue the activities they initiated. HealthNeed in Soroti was created after HealthNet International was told by a major Dutch donor (InterChurch Coordination for Development Cooperation) that funding could only continue if a local NGO took responsibility. Consequently, HNI transferred competencies, staff and materials to the newly created HealthNeed, maintaining a degree of financial and technical support during a handover period. MdM, looking to phase out of Rakai District, is also aiming to set up a local NGO based on its trained national staff and logistical assets to continue the activities it initiated. Such “ex-international” local NGOs are in fact prime candidates to access donor-funds via districts, as they generally have a sound logistical base (vehicles, office) as well as trained staff and established working procedures to build on. HealthNeed in Soroti has a good record of obtaining such funding, thereby moving to become a permanent health service institution in the district. The main complaints of NGOs concerning their new integrated role in the district’s health service delivery system mirror those of church-based institutions. Funding levels from DHSP or STIP are said to be too low and administrative difficulties are blamed for delayed fund disbursement and activity implementation.

Remarkably, district officials all adopt the official rhetoric concerning the integration of NGO and church-based institutions. They explain that NGOs offer “comparative advantages” over the state, and that a small district health team and limited government services cannot cover all the needs. It is however clear that, due to central government directives and the conditionality embodied in donor programmes, they have little choice concerning the general policy. On the positive side, some district officials view NGO activities, in particular the services of church-based institutions, as a “saving” of district funds. What the NGOs do, the district doesn’t have to pay for, or so the argument goes. It is a burden off the district’s shoulders in an era of tight resources. Also, district officials see the provision of funding to NGOs as a means to gain a measure of control over their activities. Finally, district officials, in particular staff of the district health team also get allowances from NGOs (from donor funds) for certain activities such as inspection or training. These monetary benefits help sweeten the pill of conceding a growing role to NGOs. On the negative side, district officials often have a mixed view of the actual performances of NGOs in the field. They also resent the fact that the policy decisions come from above. For instance, the Ministry of Health and the World Bank continue to select implementing NGOs for STIP and DHSP, while districts only play a consultative role. It is significant that district officials’ criticism of NGOs is generally targeted at local and international NGOs, while church-based institutions enjoy a much higher level of recognition and support.

Despite the overall policy, the integration of NGOs into the district’s health service delivery system is only partial. On the one hand, those NGO activities benefiting from donor funds channelled through the district are better co-ordinated. Based on donor guidelines, the district health team periodically calls meetings of participating NGOs to plan activities and review progress. However, for activities which are not funded through the district, there is no real integration and co-ordination is weak. In general, NGOs seek to operate with the least interference from the district officials. NGOs inform the district health team of their activities by sending periodic reports. However, there is no integrated planning and no operational co-ordination. Often, officials do not know the budgets of NGOs working in their district, and only have a general understanding of the programs being implemented. Despite a number of efforts, actual co-ordination is deficient, essentially because of the district’s lack of capacity and know-how to actually take a leadership role.

Critics have often argued that NGO activities weaken the state’s legitimacy by taking away its key function of service delivery. Evidence from the district studies suggest that this fear is not founded. In fact, politicians, both at the national and local level have factored NGOs into the political equation in a number of ways. In particular, they attempt to take credit for NGO achievements and try to influence NGO operations. The argument to convince potential voters is simple: “You see, I am a powerful man, I have brought you this NGO, so vote for me”. This type of politicisation of aid however varies in the different districts considered. In Arua, it is not visible. For instance, Members of Parliament are considered to be very detached and only rarely visit their constituencies. The MP for Terego County only once came to the MSF project in Omugo, during the exceptional period of cholera. In Soroti, politicisation is low-key, with only a few instances of politicians trying to influence NGO activities and aid allocations. As SODAN (Soroti Association of NGO Network) was instituted to co-ordinate NGO activities in the district and interface with the district administration, the wife of a prominent Soroti District Minister (later convicted of high level corruption) lobbied hard to facilitate and lead the new body. SODAN however failed to keep the NGOs support, and it has become largely ineffective.

In Rakai however, the politicisation of aid is very pronounced. Aid in fact became the single most important issue in last years’ local elections. As a corruption case involving the CAO had been discovered and DANIDA had suspended funding, the campaign revolved around the different candidates fitness and ability to re-establish Danida’s trust and to restart Danida funding. The incumbent district administration, probably using money gained during their period in office as campaign funds, backed one particular candidate. However, he was eventually defeated by another candidate campaigning on an “anti-corruption” ticket. The eventual victor was heavily supported by a powerful Member of Parliament, himself a master at the art of coopting and politicising the role of NGOs. He had notably facilitated MdM’s entry into the district, and gained much influence over MdM’s activities in his constituency. He had also infiltrated the NGO co- ordination meeting in the district (called RAJAC, or Rakai Joint Advisory Conference), prompting “community-based organisations”, often headed by politicians, to attend en masse and to use the meeting as a platform to present activities and make funding demands. Incidentally, he is also chairman of the “Uganda Community Based Organisations Association”… It is rumoured that this MP was actually behind the revelations of corrupt use of Danida funds, as the accused CAO was said to have wanted to launch a political career and to oppose the MP in the upcoming parliamentary elections in his constituency. In any case, the whole local government electoral campaign was fought over the issue of aid, clearly showing that aid money and NGO activities don’t undermine the legitimacy of the state. On the contrary, they are a source of legitimacy for those public office-holders who can claim credit for their achievements and, more controversially, who can successfully divert their resources.

To sum up, while integrating church-run institutions who have a permanent base, a long-term focus and who provide efficient essential services, is a logical step which should be supported, there seems to be no similar rationale for integrating NGO activities. NGOs, due to their fragmented and limited activities, both geographically and temporally, which are moreover not generally of an essential nature, do not represent a credible alternative to the state and its essential services. Moreover, by only covering a specific issue or area, NGO services are intrinsically inequitable. For example, World Vision in Soroti only covers one of the districts’ 25 sub-counties. Even by putting together all the NGOs in Soroti such as World Vision, HealthNeed and others, programs like STIP or DHSP don’t cover the whole district. Except for administrative constraints on flexibility of employment, it is not clear why the district medical office cannot carry out tasks such as health education or community mobilisation, and why NGOs necessarily have a “comparative advantage” in those fields. Integrating NGOs into the mainstream of service delivery makes only limited sense, as it follows the short sighted argument that NGOs are already on the ground and that the donor-provided program funds only cover part of their overall expenses. However, such type of donor funding is an incentive for NGOs to institutionalise their presence and become permanent features in the district’s health delivery system, in spite of their intrinsic shortcomings. Only few NGOs see their role as temporarily leading the way where the state isn’t able to go and letting the state take over as soon as it can. Most NGOs actually welcome the trend towards permanency and institutionalisation. This process leads to a de-responsabilisation of the state, without clear efficiency gains (if all NGO costs are taken into account). In any case, this is negative and unsustainable in the long-run, as NGOs are largely dependent on donor-support.

Politically, it is very doubtful that NGOs, in particular the local ones, actually play the “civil society” role expected of them by democratisation theory. Instead of leading and filling real gaps in the state’s capacity, NGOs generally follow donors and support the general process of state reconstruction. In fact, the process of integration itself, the narrowly defined service provision role that NGOs play, the economic motive evident in much of the NGOs behaviour, the co-optation of state representatives onto (local) NGO boards of directors, and the politicisation of NGO activities as evidenced in Rakai suggest quite the opposite. NGOs are a source of legitimacy and support for the state, they are not an instrument of restraint and accountability on its actions. Parts of the state, together with donors and NGOs, combine to form a mutually supportive structure, whose benefits in terms of essential service delivery is however far from evident.This conclusion is very similar to Susan Dicklitch’s views based upon a study of NGOs in Uganda. « NGOs are increasingly relegated to service provision and gap filling activities by the retreating state, but those supportive functions are not matched by increased political efficacy. » NGOs are « aid dependent unofficial parastatals, rather than development organisations coexisting alongside governments, but distinct from and not simply substituting for them ». S. Dicklitch, The Elusive Promise of NGOs in Africa, Lessons from Uganda, Macmillan, London, 1998, p. 3.

V) The mirage of popular participation:

An essential component of the reforms is to involve the “end-users” in the financing and management of the health services. Grass-root participation, both in financial and managerial terms, is introduced to enhance a sense of popular ownership of the health services, and to improve their accountability and efficiency. However laudable in its aims, this reform is not fulfilling its promise. User fee collection is generally low, except in church-based institutions, and generally hasn’t stopped the practice of “under the table” payments. Moreover, there are indications that patient attendance may be negatively affected by the charges. Summary evidence indicates that “health unit management committees” are generally not functional, and when they are, they are generally not fulfilling their assigned role of effectively managing the health units. The reasons for this are not well known, but they could be linked to the general decline of popular participation within the structures introduced by the NRM. “User fees” are payments charged for the delivery of health services. As Parliament refused to specifically approve user fees in 1990, the authority to raise these charges derives from the Local Government Act. The Ministry of Health, as well as District Health Team issue general guidelines on user fee collection and utilisation. Generally speaking, user fees are supposed to be “reasonable”, in the sense that they are not meant to fully recover the costs of treatment and drugs but only to provide a supplement to other sources of revenue (Ministry of Health, donor programs) for the health units’ income. User fees are supposed to be “all inclusive” and not “illness specific”, i.e. they should be the same whether a patient has malaria or acute respiratory illness, and should cover both the treatment and the necessary drugs. Some specific services, such as laboratory examinations or surgical operations, can be charged individually. User fees should not be a barrier to access, therefore destitute patients should be exempted, and payment should not be made a condition to life saving and essential care, especially in the case of emergency. Moreover, the treatment and prevention, of individual diseases with significant externalities (i.e. when individual care has widespread social benefits), such as immunisations, cholera or tuberculosis, should be exempted from payment. Finally, the utilisation of user fees should be productive, going to improve the availability and quality of services. According to Ministry of Health Guidelines, 30 % should go to staff welfare (“top up payments”), 10% to the maintenance of the health units, 3% for the operations of the HUMC and the rest, 57%, to buy supplementary drugs.Republic of Uganda, Ministry of Health, Fee-for-Service Implementation Guidelines, Vol. 1 Management Guidelines, Entebbe, 1997, pp. 14-15. All in all, user fees are intended as a reasonable financial participation from the users towards a betterment of the health services they enjoy.

Theory and practice however diverge widely in the case of user fee collection and utilisation. Firstly, the absence of national legislation leads to a lack of uniformity and consistency in the way user fees are collected and used. As guidelines are only indicative and thus not binding on the individual health units, significant discrepancies exist between Districts and even health units within particular districts. This is detrimental to transparency, as users have no clear point of reference and thus cannot contest charges. Moreover, the fee structure is often complex. In the three districts visited, first visit registration and consultation fees varied from 200 to 500 UgSh at different peripheral health units, with children in some instances charged less than adults. Revisits were charged differently, as were as inpatient services. At the hospital level, things get even more complicated. In Arua Hospital, in-patients are charged 2000 UgSh upon admission, and then 2000 UgSh per day, whereas in Kalisizo Hospital (Rakai), in-patients are charged 500 UgSh per day. They however need to pay 1500 UgSh to see the doctor, while seeing a clinical officer only costs 500 UgSh. There is no such additional charge for doctor care at Arua Hospital. Laboratory exams for tuberculosis (sputum) are charged 300 UgSh in Arua, as any other lab exam. Cursory evidence concerning the utilisation of the fees indicates that the bulk goes towards “staff welfare”, leaving very little for maintenance of the health unit and purchase of supplementary drugsF. Mwesigye found that, for 10 health units in Rakai District, 50% went to staff welfare, 22% for charcoal/ parafin/soap/stationery, 18% for drugs, 5% for repairs and maintenance, 1% for the HUMC and 4% for miscellaneous uses. Mwesigye, F. Priority service provision under decentralisation : a case study of maternal and child health care in Uganda, Makerere Institute of Social Research, Kampala, May 1999, p. 12.. Some units, such as Kalisizo Hospital, actually indicate that guidelines are not followed, and that 90% of the collected funds are used towards staff incentives since the Hospital is understaffed and needs to hire supplementary help.

Secondly, the introduction of user fees has not eliminated “under the table” or informal payments. The existence of such informal payments, de facto linked to the “privatisation” of public health services in the era of state collapse, is a powerful argument in favour of user fees. If free services are a myth it is argued, then why not formalise and bring above board a practice that is both already in existence and widely accepted by the local population? The problem is that, according to anecdotal evidence, “under the table” payments have not disappeared. In fact, they take a variety of forms. Perhaps the most common is for medicine to be charged extra, and not to be included in the treatment costs. Either the staff charge for the medicine at the unit itself, or, if the drug is not available, they ask the patient to buy it at a private drug shop. These shops are often linked to the health unit staff. Other forms of “informal” payments additional to the formal user fees include additional payment to see the doctor, for a specific operation etc. This study was not able to go beyond anecdotal evidence, but others have looked at this problem in more depth. Asiimwe and others subjected 12 health units in two districts to intense examination using a variety of methods, and concluded that health workers in all but two facilities routinely charged users beyond the formally agreed levels, and that the centrally provided drug supply was used as a source of additional income. Asiimwe’s conclusion is harsh: “the situation (…) can be summarised as the absence of a public health system. Almost all the elements of the system, which were once public, have been incorporated into the private business activity of the health workers. (…)Public health facility premises have become the sites on which private transactions are conducted. The result is that very few free services are delivered in the public health facilities, and almost none at all are delivered to the poor.”Asiimwe, D., Mwesigye, F., McPake B. and Streefland, P. « Informal Health Markets and Formal Health Financing Policy in Uganda », in : Bennett, S., McPake, B., and Mills, A. (eds.), Private Health Providers in Developing Countries : Serving the Public Interest ?, London, Zed Books, 1997, quoted in Ablo, E. and Reinikka, R., Do Budgets Really Matter ? : Evidence from Public Spending on Education and Health in Uganda, The World Bank, Washington, D. C., 1998, pp. 27-28. Obviously, the fact that salaries of many of the peripheral “unqualified” health workers are not being paid increases the likelihood of staff survival strategies, of which informal charging is a significant component.

Thirdly, formal and informal user fees can be a barrier to the access of health services, particularly for poor users. Here again, the evidence encountered by this study is only anecdotal. Moreover, there is no comprehensive study that seeks to link attendance rates with the level of user fees. However, it seems quite clear that high formal and informal charges can be significant obstacles to access. The “Uganda Participatory Poverty Assessment” echoed strong views from poor communities concerning user fees: “Cost sharing is not for the poor, (…) in the 90s, you have to pay for medical care and drugs. Many die in the villages because they cannot afford to pay the user charges. Those who have some money pay, but get insufficient treatment”The Republic of Uganda, Ministry of Finance, Planning and Economic Development, Uganda Participatory Poverty Assessment, Key Findings, Kampala, 1999, p.8.. Attendance in church-run institutions is generally low for “non-emergency” services, reflecting the high rate of user fees charged. Luwala Hospital (Soroti District) for instance charges 500 UgSh for registration, then 2000-3000 UgSh for a full course of antibiotic treatment, and only attracts 10 to 15 outpatients per day.

Fourthly, user fees have not lived up to their expectations in terms of providing a sizeable and consistent source of income for the health units. Research by Fred Mwesigye surveying the operations of 10 health units in Rakai District shows that actual user fees collected amounted to an average of approximately 200 000 UgSh per month. If Ministry of Health estimates of operating costs of 5 Mn UgSh per month (excluding salaries) is to be taken as a benchmark, this represents approximately 1% of health units’ total operating expenses.Mwesigye, F. Priority service provision under decentralisation : a case study of maternal and child health care in Uganda, Makerere Institute of Social Research, Kampala, May 1999, p. 8. This low level appears to be somewhat of a paradox, as the cumulative level of formal and informal fees is high, even amounting to a barrier in access. The reason for this disjuncture is not well established, but is most probably linked to the management of the user fees. It is plausible that significant amounts of the user fees are not reflected in the health units’ accounts, or are not used productively. This low level of revenue is obviously a disappointment for the proponents of the user fee system. It is to be recalled that the Ministry of Health, echoing World Bank predictions, expected 15% of the country’s total health budget to be raised by user fees. There is however a significant difference in user fee collections between government-run and church-run facilities. Church based institutions are more reliant on user fees for economic survival, as they have fewer sources of support than government institutions. Their rate of collection is generally higher than their government counterparts, reflecting higher charges and perhaps better quality of care. In fact, church-base institutions generally charge for drugs additionally to treatment fees, which makes their rates closer to full “fee for service” as in private practice rather than token “user fees”. Luwala Hospital for instance collects 2.5 to 3 Mn UgSh per month, which is enough to cover staff salaries, while Soroti Hospital, which is twice as large, collects the same amount. For Soroti Hospital, this amounts to around 5% of total costs, while user fees cover about 40% of total expenditure in Luwala and 30% in Kuluva Hospital. However, as seen above, the high fee collection also detrimentally affects patient attendance, although this problem appears to be less acute for emergency/in-patient services, where families are willing to sacrifice economic assets (sale of a cow, of food) to obtain services and where less treatment alternatives are available. Obviously, these fee collection figures do not include “under the table payments”, which can be quite sizeable. Finally, Health User Management Committees are somewhat of a fiction. In many districts, they are not operating, and when they are functional, it is unclear whether they actually manage the health units. In Soroti District, setting up the HUMCs was an objective of the district health team, indicating that very few HUMCs were actually operational. This seems to be the case for most of the health units visited. In Rakai District, the HUMC for Kakuuto health centre (supported by MdM) was indeed functional, yet its effectiveness was hampered by in-fighting over the management of user fees. This was the core issue the HUMC was concerned with, to the detriment of other issues such as staffing, patient attendance and mobilisation, disease patterns and prevention etc. Although they are strongly supported by donors, HUMCs are also very much in line with the ideology of “popular participation and management” introduced by the NRM since it captured state power in 1986. The fate of HUMCs may well be linked to wider trends that are affecting the NRM structures of popular participation, in particular the monetisation of public affairs. Whereas serving on local committees (or resistance committees, as they were known until 1995) was a matter on honour performed on a volunteer basis, it has now become a lucrative opportunity and a source of livelihood. This general shift in political culture may well be also taking the shine and the glory from participation in the HUMCs, thereby limiting the extent to which they actually represent the local population and blunting their effectiveness in holding health unit staff accountable. Members of HUMCs are appointed by their respective sub-county councils, and part of the user fees collected by the health unit form their monetary benefits. When these benefits are not met, HUMCs either play a very passive role or may simply not exist.

To sum up, “participation” of health users in the productive funding and the management of the health units is still minimal and marginal. The only significant ways in which users participate is through paying a combination of formal and informal user fees. This payment however does not generally give users increased access to quality and effective services. In fact, the opposite may well be the case, as payment for service can be a barrier to access. HUMCs, which are supposed to link the local population to the health units, are either non-existent, or mainly concerned about monetary benefits deriving from user fees. They are detached from the population, and do not form effective restraints or guides on the health unit staff. The reality of popular participation, which is not much more than a mirage, therefore contributes to perpetuate a “top-down” instead of the intended “bottom-up” approach to health service delivery.

CONCLUSIONS AND RECOMMENDATIONS

Conclusions : The perverse effects of well-intentioned liberal reforms

The cumulative effect of the reforms the health sector is undergoing is potentially positive for Uganda’s health delivery system. There is no doubt that the country’s health sector needs both to be revived and to be restructured. Decentralisation in particular is potentially a significant move in the right direction. However, the benefits of those liberal-minded reforms that seek to limit the role

of the state are ambiguous. The integration of NGOs, other than church-based institutions, appears to be misguided. Moreover, donor-funded programs, which are a “necessary evil” due to Uganda’s poor resource base, are implemented in a manner which leads to distortions. Restricting the role of the state appears to be detrimental to the sustainability and equity of health service delivery. Clearly, large obstacles remain before effective improvements in the availability and quality of essential services are realised. Some of the main difficulties are summarised below. As a result, the general feeling, which is reflected by the existing health indicators, is that the health status of Uganda’s population remains poor, and is not likely to soon improve significantly.

First, districts health teams do not have the resources at their disposal to plan and implement health activities according to the priorities they have identified. Significantly, the general feeling is that, despite generally low levels of funding in terms of USD per capita, the amount of money in the health sector is considered sufficient by the Ministry of Health and local health officials. The problem lies with the structure of resource allocation. For the districts, a key problem is the low “unconditional” or block transfers from the central government, and the dismal level of revenue collected locally. As a result, only activities benefiting from “earmarked” central government transfers, such as regional and district hospitals, and from donor programs, such as AIDS education, receive adequate funding. This resource structure leaves big gaps in the health delivery system, particularly affecting essential services at the peripheral health unit level. The fact that so- called unqualified health personnel, who in fact form the backbone of health care provision, are not paid by central government transfers and are left to the district, is major problem. In practice, it means that they are most often not paid at all, or with substantial delays reaching up to a year in the case of Arua District.

Second, there is a persistent lack of capacity in the health care system at the district level, particularly in terms of qualified staff and management expertise. Only one third of the positions that should be filled by qualified staff actually are so. Again, as the central government is neither hiring nor paying, the districts’ financial constraints hamper the engagement of qualified staff, even when they are available. Also, district health teams still lack the experience and expertise to actually take district health matters into their own hands. Although decentralisation has attracted a move of trained and qualified personnel to the districts, this is still insufficient and has yet to make an impact. Planning of health activities is still reactive, not proactive. The central ministries still resist effective decentralisation, and district level civil servants have yet to adopt the necessary attitude of self-help and responsibility. As a result, management is not effective, in spite of numerous computers, photocopiers and vehicles at the disposal of district health teams, all courtesy of various aid programs.

Third, donors are a mixed blessing. On the one hand, they provide much needed funds, and much of the reconstruction of the health sector can be attributed to their efforts. In particular, the rehabilitation of the health infrastructure visible in many of the districts visited can be credited to donor programs. Donor funded program comprise between 25% and 40% of the district-level health budgets. On the other hand, donor programs are not configured at the district level. They are designed at the centre and implemented vertically, leading to distortions in the definition of health priorities and in the mode of working of health personnel. They also have dubious consequences in terms of the sustainability and equity of the services provided. Essentially, the money included in donor programs, which covers activities presented as “developmental” and which generally exclude recurrent costs such as salaries, is the driving force behind the district health teams’ activity patterns. Workplans are made to conform to donor program requirements, and do not necessarily reflect actual priorities. As all activities, ranging from writing a report to attending a meeting, are funded, other duties are neglected in favour of “allowance seeking”. Workshops and seminars are favourites of health personnel, and form part of the generally unpaid “unqualified” staff’s survival strategy. District health teams themselves are fragmented as a result, with every member jockeying to be a focal point of lucrative donor-program. As a consequence of not wanting to fund “recurrent costs”, donor programs focus on preventive, educational and training activities presented as “capacity building” and “developmental” in nature. This does not squarely address the main problem of basic service delivery at the health unit level, which is arguably the highest priority, and possibly the most neglected area. Sustainability of activities initiated under donor programs is also a concern. Even if they are presented as “developmental”, donor programs become part of the normal expectation and revenue base of health personnel. Also some of their activities are meant to be carried on, such as the expanded infrastructure of the “health sub-districts” built under the PHC program, or the control and treatment of sleeping sickness, spearheaded by an NGO in Arua. In those cases, the limitations on the district’s capacity and revenue base, especially for the payment of staff and drugs, make successful integration and continuation very doubtful. Beyond the narrowly defined effectiveness of donor programs, which this study could not analyse, it is therefore clear that these programs generate a series of behavioural and financial consequences that are not favourable to the effective performance of the health delivery system.

Fourthly, the policy of limiting the state’s role and integrating the “private non for profit” sector as a permanent player in the health delivery system seems to be misguided, at least for the international and local NGOs. Church-based institutions, who have a long-term perspective and focus on essential services, are justified candidates for such integration. For the state to allocate limited funds for their recurrent operations is a recognition of their pubic service role, and should serve to tie them into an integrated system of health care delivery with common standards and practices. However, the rationale of including other NGOs, which have a shorter time-span, modify activities according to funding possibilities, and have a limited coverage and focus, is much less evident. This is clearly a donor-driven policy, whereby funding guidelines require districts to allocate funds for activity implementation to NGOs, while remaining with a general supervisory role. The so-called “comparative advantage” of NGOs seems to rest on short-sighted view that donor programmes fund only part of their costs, building on an existing logistical and personnel base funded by other sources. Except for bureaucratic rigidities in staff employment, it is unclear why districts couldn’t take on much of the activities undertaken by NGOs, such as health education or community mobilisation. In any case, NGOs, especially local ones, welcome this trend of integration, and are looking forward to become institutionalised permanent players in the health service field. They modify their activities to fit with the donor guidelines, co-opt district officials on their boards of directors to facilitate their operations, and are happy to be integrated into a triangular donor-state-NGO service delivery system. They rarely lead, or take the initiative, but rather follow the state and the mainstream donors’ concerns. In certain districts, especially Rakai, this integration of NGOs and donors has gone so far as to become the major issue in political affairs, implying that NGO and donor activities actually strengthen the credibility and legitimacy of political leaders and district officials. Therefore, contrary to liberal democratic theory’s optimistic view, NGOs can only very partially be considered part of “civil society”. They are much rather auxiliaries to the state. In sum, by “deresponsabilising” the state, the integration of NGOs is a short-sighted and intrinsically inequitable solution to improve service delivery.

Fifth, popular “participation” in the funding and management of health services hasn’t lived up to the expectations of its proponents. Lacking a strong and binding legal basis, user fees are collected and used in an inconsistent and widely opaque manner. Guidelines issued by the Ministry of Health and the districts are either not known, or not adhered to. The introduction of user fees in the 90s has not eliminated the practice of “under the table payments” established during the period of state collapse in the 70s and 80s. The financial burden on health service users is therefore heavy, and there are reasons to believe that this discourages patients, especially poor patients, from seeking medical treatment. Moreover, in what appears to be a paradox, the income derived from user fees apparently amounts to only a fraction of the required expenditure of the health unit. According to a study conducted in Rakai, only a fraction of the total costs was raised from user fees. Most of the fees collected are used on staff welfare, which proves to be a helpful supplement in the case of unpaid staff, but which leaves very little for additional drugs or maintenance. Church-based institutions charge higher fees, as they are more dependent on user charges for survival. They also boast a higher collection rate (up to 40% of their total income), but their patient attendance appears to be low, especially in the case of “non-emergency” services. Finally, HUMCs, which are supposed to provide an organic link between the community and the health services, are disappointing. Either they are non-functional, or when they do meet, they are mostly concerned with the monetary benefits from user fees collection and utilisation. Their poor performance may be linked to a general decline in the quality and commitment of popular participation in the structures introduced by the NRM, reflected in the widespread monetisation of public affairs.

In conclusion, it can be stated that poor health services in Uganda are not simply a result of “rolling back the state” as the standard structural adjustment critique assumes. The ineffectiveness of essential health care delivery in Uganda is much better explained as a combination of a weak state and distortions introduced by the aid system (donors and NGOs). This is taking place in a context of overall low resources and funding, even though donors are increasing their engagement. However, the structure of resources, where government continues to prioritise urban curative services and where the growing funds provided by donors do not go to improving the most needed essential services delivered at the health unit level, is the main reason the system is stagnating. Decentralisation of competencies has not been matched by effective decentralisation of resources. Aid money, with its emphasis on “development” and “capacity building” is failing to address the most pressing problems of primary health care. Moreover, the liberal philosophy of the limited state has generated a faulty policy of NGO integration and unrealistic expectations of community participation. As a result, the current patchwork of actors and funding mechanisms is not providing effective, equitable and sustainable health services in Uganda.

Recommendations : Towards a state-centred sectoral approach ?

Following the above analysis, four basic recommendations are presented here.

First, an integrated approach to health services is necessary. Instead of being an agglomeration of vertical programmes reflecting the interests and commitments of various actors, health services should be tailored to the actual burden of disease and priority needs of the Ugandan population. Ideally, resources should be matched to the most effective interventions in terms of reducing mortality and morbidity. Obviously, this is easier said than done, and political, socio-economic and institutional pressures will always skew resources away from the most effective use patterns. It is likely that tertiary level curative care for instance will always command a disproportionate level of attention and funding. However, it is encouraging that both the Ministry of Health and donor countries have recognised some of the pitfalls of the current system, and have made declarations of intention of moving towards an integrated, or “sectoral”, approach.

The Ministry of Health’s “Ten Year National Health Policy”, drafted in 98 but which has yet to be formally adopted, takes as point of departure that “75% of life years lost to premature death are accounted for by 10 well known and preventable diseases” with perinatal and maternal related conditions (20.4%), malaria (15.4%), pneumonia (10.5%), AIDS (9.4%) and diarrhoea (8.4%) together accounting for over 60% of the burden.Republic of Uganda, Ministry of Health, Ten Year National Health Policy (draft), 1998, p. 2. Accordingly, PHC remains the “basic philosophy and strategy in national health care development” and government will “focus on health services that are demonstrably cost-effective, have the largest impact on reducing morbidity and mortality with emphasis on protecting the poor and vulnerable population”. To this “Essential Health Care Package”, government will allocate “the larger proportion of its annual budget”.Republic of Uganda, Ministry of Health, Ten Year National Health Policy (draft), 1998, pp. 4 and 5. Following our analysis, achieving this would entail a significant reorientation in resource use. In particular, it would mean attaching priority importance to funding and support of peripheral health units, where the bulk of these health problems are addressed. Upgrading health unit personnel, and providing them with regular salaries should be top of the list67. It would also mean radically revising the mode of donor funding, so that donor money is effectively channelled to priority interventions. Instead of funding specific vertical programs, and thereby inducing various distortions, donors would be called upon to co-finance the health sector as a whole, based upon the “essential health care package” strategy. To a large extent, donors would have to give up earmarks, and back general expenditures. This so called “programme” or “sectoral”, as opposed to “project” approach, is being considered quite seriously by a number of donors.

There are however, a number of important obstacles before it can be put in practice. Vested interests and habitual patterns of behaviour, both on the donor and the Ministry of Health sides, are resisting change. Within the Ministry of Health, the donor-related “project implementation units”, where bureaucrats earn lucrative top-ups and allowances, are trying to preserve their privileged status. Within the donor community, abandoning the “flag-waving” attitude which specifically identifies certain activities to donor funding, is proving to be difficult. Moreover, and perhaps more fundamentally, there is the problem of control and accountability. For donors to give up vertical control over program content, activities and funds, and to put their money in a common pot for the Ministry of Health and districts to spend, they need to be confident that the resources will be well used. Corruption, in the form of diversion of funds and “leakages”, is an important concern in this context68. This is not to say that the current system of vertical donor-controlled projects prevents corruption. In fact, there is much evidence to suggest quite the opposite, the World Bank’s First Health Project being a notorious example. However, donors feel that the opportunities for corruption would be even greater in the “sectoral approach”. Also, for donors to back the Ministry of Health’s overall strategy, they would want to agree with that strategy, its activities and the use of the resources committed. Therefore, somewhat paradoxically, instead of being more “hands-off” than the project approach, a sectoral approach would probably mean more direct donor influence within the Ministry of Health in the form of “policy dialogue” and tight expenditure controls. The donors and the Ministry would formally become “co-deciders” in managing the health sector, an evolution that has all appearances of being a form of recolonisation. Despite all these difficulties, it is probable that a sectoral approach, if implemented in a coherent and integrated manner with donors sufficiently sensitive to sovereignty issues, would improve resource allocation and use.

Second, the state is a necessary evil. Despite all its deficiencies, the state needs to be strengthened and placed at the centre of health care planning, policy making and delivery. Except for church-related institutions, NGOs should not be systematically integrated into the health care delivery system. This is bad both for the state, as it leads to its deresponsabilisation and dereliction of fundamental duties, and for NGOs, as it robs them of their supposed “civil society” function. NGOs have no inherent comparative advantage over the state; there is therefore no reason why they should be encouraged to take over state duties. NGOs should be encouraged to play a role of leadership and innovation, going where the state does not, doing what the state can’t, instead of becoming auxiliary service providers integrated with the state and mainstream donors. In spite of its problems, in particular corruption, inefficiency and poor staff management, the state is the sole institution which has, by definition, a mandate to cover the entire population in an equitable manner. Unless the situation is considered to be so bad as to warrant giving up on the state, then the state system should be given the highest degree of attention and support. Modifying its institutional culture, improving standards and flexibility in human resource management (link incentives to performance, easier hiring/firing practices etc.) would be a priority concern. This could lead to an improvement in the quality of service provided by health care professionals in the public sector, which is generally considered to be very low. It would also enable the state to conduct activities currently undertaken by NGOs in a flexible manner, by hiring expertise on a temporary basis. Health is a public good, and the state is the best, if not the only, provider of public goods available.

Third, decentralisation is a potentially powerful and significant institutional transformation. However, for decentralisation to fulfil its potential, a revision of centre-district relationships is necessary. In particular, the pattern of resource allocation must be revised. Districts must be given the means to carry out the functions which the central government has devolved to them. It defeats the purpose of decentralisation if district-level staff performing key duties, such as health care providers in peripheral health units, are not paid. This means that either the block or unconditional grant should be considerably increased, or that local revenue should be enhanced. To achieve the latter, a reform of the taxation system would be necessary. Districts are currently only allowed to collect graduated tax as well as levies from markets and real estate. However, the bulk of the tax revenue, namely income tax, VAT and import duties are collected centrally by the Uganda Revenue Authority. This is detrimental to districts. For instance, Arua produces much of the tobacco used by Uganda’s largest cigarette producer, British American Tobacco (BAT), but receives no revenue from BAT’s income or from VAT on cigarettes. A reform of the tax collection or of the district budget funding system would ensure that Arua gets a better share of the revenue it contributes in generating. Decentralisation must also become more effective in the actual planning of health services. The “top-down” approach, where districts are essentially the recipients of strategies and funds decided upon centrally, has to be revised. A move towards the sectoral approach in terms of donor funding would be helpful, as it would eliminate the vertical nature of current donor programs. However, the Ministry of Health itself needs to allow districts to make their own choices, within the general framework of a nationally agreed-upon health care strategy and priorities. If the PHC guidelines are anything to go by, there is a real danger that a vertical Ministry of Health-district system might replace the current donor-district system, even under a sectoral approach. Districts should not be left out of national level strategy making, and should be given the means and support to actually plan at the district level.

Fourth, given the objective funding constraints a poor developing country like Uganda faces, donor support for the health sector is both inevitable and necessary. Uganda should not shy away from accepting grants, and from prudently taking on favourable loans, in order to meet the priority needs of its people. The donors’ move away from strict structural adjustment to a more humane version emphasising poverty alleviation and social service provision can only be welcome, notwithstanding the continued adherence to a liberal philosophy. However, the terms on which donor programs are being implemented should be seriously revised. First, the fuzzy and somewhat incoherent distinction between “recurrent” and “development” funding should be reconsidered. It is not justified to smuggle what should in fact be recurrent activities under development, and refuse to treat them as such, thus contributing to a perverse “allowance culture”. It would be more fruitful to recognise the recurrent nature of many of the donor program activities, and to give them adequate funding provided that they are clearly essential, and that a realistic phase-out and reintegration plan is devised. This would probably entail cutting many of the workshops and seminars which are implemented as way to bypass the current ban on funding recurrent activities, and to provide (part of) the salaries to the staff who actually conduct curative and preventive activities on a permanent basis. Second, the vertical organisation of most donor programs should be abandoned. The sectoral approach, described above, should be encouraged. Moreover, decentralisation should be made effective, and a “bottom-up” instead of a “top down” approach should be emphasised. Donors be open to funding (part of) the activities that districts have identified in a particular area, in line with the national “essential health care package” strategy, instead of determining the type and nature of funded activities from afar. Finally, funding for health, with additional donor money made conditional upon increased government allocations, should be increased. The conceptual justification of linking poverty and health care is convincing, and provides a cogent rationale for increasing funding allocations. It also justifies decreasing, or even eliminating user fees, as high payments for service are apparently a major barrier to access. The emphasis on taxation and service charges may be counterproductive in conditions of poverty, as people tend to avoid the use of the service in order not to pay. By negatively affecting the population’s health status, this “evasive” behaviour in turn reduces the country’s productivity, therefore the population’s ability to develop, pay taxes etc. A vicious circle sets in. Although abolishing user fees may be too radical a move, reducing and unifying the charges while improving their management, on the basis on binding national legislation, would be a major step in the right direction. If donors and the government increase funding towards health, the public service should be made the major recipient of these increased funds, in support of the “integrated” or sectoral approach outlined above.

Obviously, adopting these recommended actions would not solve all the problems Uganda’s health system is currently facing. It is an unavoidable fact that, for some time to come, Uganda will remain a poor country, with a significant financial constraint on health sector expenditures, scanty qualified human resources, entrenched behavioural patterns and pervasive difficulties in establishing sound management structures. These few recommendations do not have the ambition to resolve those daunting issues in any significant way. However, they would possibly address some of the problems that the current well-intentioned but flawed effort at reviving and restructuring the health system is actually creating. In that sense, they could possibly make a small impact in moving the health system towards most everyone’s intended objective, namely improving the health status of Uganda’s population.

 

Bibliography

Ablo, E. and Reinikka, R., Do Budgets Really Matter ? Evidence from Public Spending on Education and Health in Uganda, The World Bank, Washington D.C., 1998.

Arua District Council, Background to the Budget 98-99, Arua, 1998.

Arua District, Arua District Health Workplan 1998/99 by Donor excluding Hospitals, Arua, 1998.

Barton, T. and Bizimana, D. “Family and Household Spending Patterns for Health Care”, in: R. Claesson (ed.), Health Care Financing: Proceedings of the Dissemination Seminar, Health and Economics in Uganda, Kampala, March 95.

Barton, T. and Bizimana, D., Health Sector Review, Oxfam Uganda, Kampala, 1995.

Diamond, L. « Rethinking Civil Society : Towards Democratic Consolidation », Journal of Democracy, vol. 5, no. 3, July 1994.

Dicklitch, S., The Elusive Promise of NGOs in Africa, Lessons from Uganda, Macmillan, London, 1998.

Harvey, C. and Robinson, M., Economic Reformand Political Liberalisation in Uganda, Brighton, Institute of Development Studies, 1995.

Lancaster, C., Aid to Africa : So much to do, so little done, University of Chicago Press, Chicago, 1999.

Mwesigye, F. Priority service provision under decentralisation : a case study of maternal and child health care in Uganda, Makerere Institute of Social Research, Kampala, May 1999

Okuonzi, S.A. and Macrae, « Whose policy is it anyway ? International and national influences on health policy development in Uganda », Journal of Health Policy, 1994.

Rakai District Council, Budget 1998-99, Rakai, 1998.

Republic of Uganda, Background to the Budget 1998-99, Kampala, 1998. Republic of Uganda, Background to the Budget 1999-2000, Kampala, 1999.

Republic of Uganda, Local Governments Act, 1997, Kampala, 1997.

Republic of Uganda, Statistical Abstract 1998, Kampala, 1998.

Republic of Uganda, Uganda National Integrated Household Survey 92-93, Kampala, 1994.

The Republic of Uganda, Ministry of Finance, Planning and Economic Development, Uganda Participatory Poverty Assessment, Key Findings, Kampala, 1999

Republic of Uganda, Ministry of Health, Fee-for-Service Implementation Guidelines, Vol. 1

Management Guidelines, Entebbe, 1997

Republic of Uganda, Ministry of Health, Guidelines on the Use of Conditional Grant for Primary Health Care (PHC) 1998/99, Entebbe, 1998

Republic of Uganda, Ministry of Health, Ten Year National Health Policy (draft), Entebbe, November 1998

Repubic of Uganda, Ministry of Health, The Three Year Health Plan Frame, 93/94-95/96, Entebbe, 1993

Republic of Uganda, Ministry of Health, White Paper on Health Policy, Update and Review, Entebbe, 1993

Rwabwoogo, M.O., Uganda Districts Information Handbook, Kampala, Fountain Publishers, 1998.

Soroti District Council, Preamble to the 1998-99 Budget, Soroti, 1998.

The World Bank, Sub-Saharan Africa : From Crisis to Sustainable Growth, Washington D.C., 1989.

The World Bank, Uganda, The Social Sectors, Washington D.C., 1993

The World Bank, Uganda : the Challenge of Growth and Poverty Reduction, Washington D.C., 1996.

The World Bank, World Development Report 1997 : The State in a Changing World, Oxford University Press, Oxford,1997.

UNDP, Uganda Human Development Report 1998, Kampala, 1998.