Protecting resources for primary health care under fiscal federalism: options for resource allocation
1Health Economics Unit, University of Cape Town, South Africa.
2Department of Health Policy and Management, Trinity College, Dublin, Ireland.
*Corresponding author. Health Economics Unit, School of Public Health and Family Medicine, University of Cape Town, Anzio Road, Observatory 7925, Cape Town, South Africa. Tel: +27 (0)21 406 6576. Fax: +27 (0)21 448 8152. E-mail: Okore.Okorafor{at}uct.ac.za
 |
Abstract |
|---|
 Top Abstract IntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
The introduction of fiscal federalism or decentralization of functions to lower levels of government is a reform not done primarily with health sector concerns. A major concern for the health sector is that devolution of expenditure responsibilities to sub-national levels of government can adversely affect the equitable distribution of financial resources across local jurisdictions. Since the adoption of fiscal federalism in South Africa, progress towards achieving a more equitable distribution of public sector health resources (financial) has slowed down considerably. This study attempts to identify appropriate resource allocation mechanisms under the current South African fiscal federal system that could be employed to promote equity in primary health care (PHC) allocations across provinces and districts. The study uses data from interviews with government officials involved in the budgeting and resource allocation process for PHC, literature on fiscal federalism and literature on international experience to inform analysis and recommendations. The results from the study identify historical incremental budgeting, weak managerial capacity at lower levels of government, poor accounting of PHC expenditure, and lack of protection for PHC funds as constraints to the realization of a more equitable distribution of PHC allocations. Based on interview data, no one resource allocation mechanism received unanimous support from stakeholders. However, the study highlights the particularly high level of autonomy enjoyed by provincial governments with regards to decision making for allocations to health and PHC services as the major constraint to achieving a more equitable distribution of PHC resources. The national government needs to have more involvement in decision making for resource allocation to PHC services if significant progress towards equity is to be achieved.
Key Words: decentralization, financing, Fiscal federalism, primary health care, resource allocation
KEY MESSAGES
|
|
Introduction |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
A well-functioning and organized primary health care (PHC) system is critical to the promotion of good health in any country, more especially in the developing world. As observed by Bengoa et al. (2003
), it is an important component in the move towards achieving the Millennium Development Goals (MDGs). The global trend towards increased decentralization of expenditure responsibilities to lower levels of government (Ter-Minassian 1997; de Mello 2000) creates an added dimension to the challenge of developing a well-functioning and equitable PHC system. A key concern for the health sector is that decentralization of spending and decision making can increase inequities in the financing and hence provision of health services (like PHC) across geographic areas (Gilson and Mills 1995). In certain cases, the decentralization of spending responsibilities has been reinforced by the introduction of fiscal federalism, giving considerable decision-making autonomy to sub-national governments (SNGs) for resource use. The provision of PHC has traditionally been the responsibility of lower levels of government. In this regard, a key question for PHC under a fiscal federal system is ‘how does the central government ensure an adequate level of PHC service provision and financing across geographic areas within the country?’ In this study we use South Africa as a case study to explore the implications of this question. One interesting feature of the South African case is the overriding government policy objective of addressing the appalling inequities in the health sector that were entrenched by the apartheid health system. Improving equity was the driving concern of health policy in South Africa following the democratic change. Nevertheless, the key issues raised are of relevance to all decentralized health systems, and especially to those operating under a fiscal federal system.
The appropriate financing and allocation of resources for PHC in South Africa is an important issue considering the priority given to PHC by the National Department of Health (NDoH) over the last decade (Thomas et al. 2005). However, recent work has revealed that there exists an inverse relationship between the need for and funding of PHC services across provinces and districts; and a large funding gap between that required to deliver the nationally endorsed PHC package1 and current levels of funding (Thomas et al. 2003). Further research revealed that before fiscal federalism was introduced with a new constitution (in 1996), early gains were made in re-orienting service provision towards the needs of the population and more equitable resource allocation policies. However, following the devolution of substantial health expenditure responsibilities to provinces, further attempts at redistributing resources for PHC, particularly in the most deprived districts, have been undermined (Thomas and Muirhead 2000; Thomas et al. 2003). For example, in 2002/03 the province with the highest PHC per capita expenditure2 (Gauteng province, ZAR 225) spent more than three and a half times per person than the province that spent the least (Mpumalanga, ZAR 62).
Such research has led to substantial concerns about the gap between the policy commitment of the NDoH and actual funding levels for PHC, especially in deprived districts. With the adoption of a fiscal federal system, much of the decision on the funding for the health sector has been decentralized to provinces, with the NDoH retaining responsibility for policy development (Doherty et al. 2002), resulting in a divide between policy development and policy implementation. While such a division may be common across countries, in the South African case there has been no funding mechanism available to allow national policy to be appropriately resourced. NDoH would have liked to have imposed expenditure responsibilities on provincial governments, but it does not have the power to do so. Hence there has been no mechanism to ensure that provinces will in fact honour national policy. Better-resourced provinces have had no incentive or mechanism to subsidize other provinces and little inclination to adequately fund their own rural and deprived areas. Research also reveals that in South Africa, approximately 60% of people in the lowest socio-economic quintile depend on publicly provided health care services, with only 15% utilizing private facilities (Wadee et al. 2003). This further emphasizes the importance of promoting equity in allocations to health and PHC within the public sector.
Recent research estimated a funding gap of ZAR 3.3 billion for the financial year 2005/06 (approximately equivalent to US$500 million3 or an extra 36% on top of spending on PHC in 2005/06) for appropriate funding of PHC according to national guidelines (Chitha et al. 2004). Nevertheless, even with identified resource requirements, questions arise as to how to guarantee that extra resources allocated for PHC services will not be redirected to other needs, within or outside the health sector.
|
Aim |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
The aim of this paper is to review what kind of resource allocation mechanisms could be employed under the current fiscal federal system to ensure a more equitable distribution of PHC resources, and which also ensures that the PHC package is sufficiently well funded across all districts. Given the effect of the introduction of fiscal federalism on progress in the achievement of an equitable distribution of PHC resources, this paper also addresses some very important questions:
- How does fiscal federalism affect resource allocation mechanisms and, from international experience, what options are there to protect funding for policy priorities?
- How are decisions for allocation made in Provincial Departments of Health in South Africa? What role do districts and sub-districts have in resource allocation?
- What are the different options available for ensuring that PHC activities receive adequate funding and how acceptable are they?
To answer these questions the authors distil the international literature on fiscal federalism and health before using South Africa as a case study to explore in detail the key issues around resource allocation for PHC.4
|
Methods |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
To achieve the objectives set, the authors employed both secondary and primary data collection tools in a case study setting, i.e. South Africa. A case study approach was thought useful in that it typically allows an in-depth exploration of critical issues in one context (Yin 1994
). In terms of secondary data collection, the authors reviewed international literature concerning fiscal federalism and the health sector. This review focused not only on the theoretical literature on mechanisms of fiscal federal structures, but also on the experiences of other countries with fiscal federalism in protecting the deployment of resources for priority areas. Government publications and other literature on resource allocation, budgeting and the health sector in South Africa were also reviewed. This provided a better understanding of the governance structures and their linkages to resource allocation to PHC.
Semi-structured interviews were conducted with public sector officials and NGOs involved in PHC financing and delivery. Government interviewees were located at different levels of the South African system (National, Provincial and District) across three of the nine provinces: Limpopo, North West and Western Cape. The study sites were chosen based on PHC expenditure per capita at the provincial and district levels to ensure there was representation of well and poorly funded areas. In the 2002/03 financial year, Western Cape spent ZAR 202 per capita compared with ZAR 136 and ZAR 138 in Limpopo and North West provinces, respectively. The average per capita PHC expenditure for South Africa was ZAR 143.
A qualitative approach was used as it allowed for the perspectives of different actors in the health system to be captured in all their richness. Further, a semi-structured approach allowed flexibility to probe for details on issues, particularly where they were unexpected (Miles and Huberman 1994; Bowling 1997).
The authors also made use of interview material collected from an earlier study where provincial, district and local government managers were interviewed in three other provinces: Gauteng, Free State and Eastern Cape (Khumalo et al. 2003; Mubangizi et al. 2003; Okorafor et al. 2003a). Eastern Cape had one of the lowest per capita PHC expenditures of ZAR 89 in 2002/03, while Free State and Gauteng had much higher per capita expenditures of ZAR 173 and ZAR 225, respectively.5,6
Given that the authors were involved in the previous study and similar information was required from the same level of government officials, the interview tool was not piloted. In brief, interviewees were questioned on resource allocation practices, sufficiency of PHC funding, strategies for protecting PHC funds and promoting equity, and intergovernmental relations. Questions varied with the responsibility of the officer interviewed. In total, data from interviews with 39 government officials were used. Officials interviewed were senior provincial health officials, provincial district managers, provincial financial officers, district and sub-district health managers, ward councillors and facility managers. One official from the National Treasury was interviewed. Since these interviews were conducted in six out of the nine provinces, interview data are reasonably representative of views of policy makers and managers across the country. Although interview materials from an earlier study were used, the results of the data were similar.
Key informants from NGOs involved with PHC delivery and the strengthening of the District Health System were also interviewed. All interviews were transcribed with the consent of the interviewee and all interviewees were guaranteed anonymity to allow a better flow of data.
|
Literature review |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
Fiscal federalism can be defined as the devolution of expenditure responsibilities to sub-national levels of government. This has become a global trend in the past few years (Ter-Minassian 1997
; de Mello 2000). This is partly a reflection of the political evolution towards more democratic societies. Also, economic literature has presented the view that fiscal decentralization can entail substantial gains in terms of both efficiency and welfare. According to this view, such gains are best achieved by assigning responsibility for each type of public expenditure to the level of government that most closely represents the beneficiaries of these outlays (Musgrave and Musgrave 1989; Bird and Vailliancourt 1997; Ter-Minassian 1997; de Mello 2000). However, the literature also indicates that fiscal decentralization can entail significant costs in terms of distributional equity (McIntyre et al. 1998). Theoretical efficiency gains from decentralization can be significantly undermined by institutional constraints such as:- Weak administrative capacity in SNGs, overstaffing, poor technical skills at lower levels, and the existence of corruption;
- SNGs may not have developed modern and transparent public expenditure management systems; and
- The size of the local jurisdiction (which is often a result of historical developments or political factors) is not always consistent with the full realization of potential efficiency gains from decentralization (Ter-Minassian 1997).
Although many countries have substantially devolved expenditure responsibilities to lower levels of government, the form of decentralization and level of autonomy enjoyed by SNGs differ, reflecting their particular context (Bird and Vailliancourt 1997). In most countries operating a fiscal federal system, large expenditure responsibilities are decentralized to sub-national levels of government while the central government collects revenue from most of the major taxes. This usually results in large funding gaps at SNG level. There are generally four solutions to this problem. The first is to increase revenue at the sub-national level by transferring more revenue raising power to SNGs.7 A second is to reduce local expenditure. A third option is to transfer expenditure responsibility up to the government level with more revenue. The fourth option is to transfer some centrally collected revenues to lower levels of government; and this last option usually prevails (Bird and Vailliancourt 1997). Fiscal decentralization can also lead to inequities in the distribution of financial resources across SNGs. SNGs usually have differential capacity to generate their own revenues (Ter-Minassian 1997), leading to unequal capacities to carry out similar functions.
Intergovernmental transfers
In most countries, these funding gaps are addressed through intergovernmental transfers, which refer to the transfers of funds from the central government to lower levels of government. The key issues in intergovernmental transfers are around deciding on the type of transfers and the criteria for the size of transfers made to SNGs. The type of transfers used to correct these funding gaps has varying impacts on the results which they aim to achieve. The results of such transfers, whether good or bad, will depend on the incentives (built into the transfer system) they create for central and local governments and, indirectly, for residents of the different regions of the country (Bird and Smart 2002). Intergovernment transfer mechanisms can be grouped into two broad categories: revenue sharing and grants. Whether transfers are in the nature of revenue sharing or grants, there are basically three ways to determine how much is to be distributed:
- As a fixed proportion of government revenues;
- On an ad hoc basis, in response to specific claims; and
- On a formula-driven basis (Bird and Vailliancourt 1997).
Revenue sharing arrangements are usually geared towards correcting vertical funding gaps. Sharing of tax revenues can be on a tax-by-tax basis, with different coefficients of distribution among levels of government for each tax, or on the entire pool of central government tax revenues. Tax-by-tax sharing is practiced in countries such as Argentina, Brazil, Hungary and Russia. However, a major disadvantage of such sharing is that it provides an incentive for tax administration at central government to concentrate its collection and enforcement on the taxes that are not shared or are shared to a lesser degree (Ter-Minassian 1997). Furthermore, tax-by-tax sharing provides the central government with incentives to concentrate increases in rates (for instance for stabilization purposes) on the shared taxes. Therefore, revenue sharing based on the entire pool of government revenues may be preferable (Fjeldstad 2001).
In general, grants can be grouped into two:
- General purpose grants: unconditional transfers aimed at addressing vertical and horizontal imbalances;
- Conditional (specific purpose) grants: transfers carrying conditions regarding the use of the funds and/or the performance achieved in the programme(s) financed through them. Some conditional grants may require matching elements by recipient authorities.
Most countries use a combination of revenue sharing and grants. In general, the former forms the basic revenue for SNGs. Grants are additional transfers made to certain (or all) sectors of SNGs either to increase the overall expenditure capacity of certain jurisdictions (usually general purpose grants) or to influence the level and distribution of particular services across all jurisdictions (usually conditional grants).
The choice between conditional and unconditional transfers should be based on a number of considerations. On the one hand, the imposition of conditions clearly reduces the level of autonomy at lower levels with respect to decisions around ‘how much’ to spend and on ‘what’. This is contrary to the welfare and efficiency arguments in support of decentralization. On the other hand, the imposition of conditions may be justified by other considerations. For example, it may be necessary to attach conditions to funds to realise uniform or minimum expenditure on issues of national concern, such as PHC (ibid).8 The design of grants to SNGs has significant implications for effective and efficient use of the resources and for the achievement of the stated goals for which the funds are disbursed (Bird and Smart 2002).
If any grants are used, some choices can be identified:
- Whether the transfers should be made on a conditional or an unconditional basis. It is to be noted that an unconditional grant simply increases the SNG’s income without altering their spending priorities, which are dictated by local preferences. The main justification for conditional grants over unconditional grants therefore must be that local decision-making fails to produce the socially optimal outcome. However, many developing countries have relatively weak capabilities in expenditure management at the SNG level. The use of conditionality and performance criteria for a special purpose grant may then generate confusion and pro forma fulfilment of the needed criteria (Ahmad and Craig 1997). Conditional grants are more appropriate where SNGs lack the capacity or desire to allocate resources to stated priorities, as the conditions attached to the funds dictate the terms of how the money is to be spent. However, where the conditions for use (and performance) are such that they require a high level of managerial capacity to fulfil the stated criteria, managing conditional grants at lower levels could become very difficult. Therefore, unless SNGs possess the capacity to monitor and manage the conditionality for grants, it may be better if central governments simplify the design and conditionality of special purpose grants, and/or supplement these with lump-sum transfers, which could then be seen as ‘own’ resources by recipient governments (ibid).
- Whether there is to be some redistribution in the transfer mechanism or whether the transfers will be made based on efficiency (or other) criteria to each member of the defined population in each region.
- Within the category of conditional transfers, whether the central government should require SNGs to undertake some matching of funding of programmes by lower level governments. This might be done to ensure that SNGs spend resources on this priority activity, and not on other activities. It may also be done to pave the way for the transfer of the activity to SNGs, by gradually decreasing the proportion of funding paid by central government (Ahmad and Craig 1997).
Intergovernmental transfers and autonomy
The nature of intergovernmental transfers to SNGs may depend on the public good/service that they finance. For certain public services, their outcomes are of national interest and therefore the central government may see a need to interfere in fiscal operations at lower levels to realise a more ‘desired’ outcome. For example, in Australia, in pursuit of national policy objectives, sectors such as health, education, social welfare and housing are largely funded through specific purpose grants. In Canada, the major general purpose grants are transferred to provinces with below average tax capacity, while specific purpose grants are employed to fund health and, more broadly, the social sector. In Italy, conditional grants have been used to influence the level and distribution of sub-national expenditure on health and public transport, which are deemed to be of national concern. In Bulgaria, specific purpose grants are given to municipalities for capital expenditure purposes only; while general purpose grants are the dominant form of transfers to municipalities (Bogeti
1997). Whether the central government interferes in fiscal operations directly (attaching conditions to transfers) or indirectly (laying down norms, standards or such regulations), such interferences reduce the level of autonomy of SNGs.
The health sector
A major concern for the health sector is that the introduction of fiscal federalism can lead to inequities in the distribution of health care resources between geographic areas (see Thomas et al. 2003). With increased autonomy at lower levels, SNGs have greater influence on the budgetary allocations to the health sector. Differential capacity to generate and to utilize resources coupled with different local preferences will most likely yield different levels of funding and provision of health care services across SNGs. However, the type of transfers to lower levels for the health sector can significantly influence the distribution of resources for the health sector across areas covered by different SNGs.
The outcome of the health sector is of central interest for most countries. It is recognized that everyone has the right to enjoy the highest attainable level of physical and mental health (OHCHR-UNOG 1966). In this regard, central governments in most countries operating a fiscal federal system influence fiscal operations to achieve a desired distribution of resources and provision of health services within the country. There are different ways in which central governments have influenced these within health systems. In some cases higher levels of government retain expenditure responsibilities for health services, with the central government maintaining overall control of activities in health sector financing and provision, as in Australia and Canada (Ahmad and Craig 1997; Krelove et al. 1997). In other cases, all tiers of government share the responsibilities for financing and delivery of health care, as in Argentina (Schwartz and Liuksila 1997) and Nigeria (Ayodele 2003). Where financing and provision of health services are decentralized to lower levels, the central government transfers funds for health as a specific purpose grant with conditions on how the funds are to be used. In other cases, specific purpose grants are used to finance only specific programmes within the health sector. In Chile and Colombia, fiscal decentralization of the health systems achieved a more equitable distribution of public health resources. Although decentralized units enjoyed ‘moderate’ choices over financing of health, equity was achieved through centrally enforced resource allocation criteria for services provided by the decentralized units (Bossert et al. 2003).
For all options outlined, the central/national government still retains some control over expenditure responsibilities, or at least participates in the allocation criteria to health or the spending on and providing of health services. The case of India however, is similar to that of South Africa. States are fully responsible for the provision of PHC from own-revenue and unconditional transfers from the central government. This has resulted in widely varying levels of quality and quantity of health care provision across states, reflecting their varying levels of economic development, their health sector priorities and their current and past investments in health (Bajpai and Goyal 2005).
PHC provision under a fiscal federal context
The PHC approach advocates community participation and greater responsiveness to the needs of the community (World Health Organization 1978). This approach thus prescribes expenditure responsibility for PHC services at lower levels of government, following the assumption that local governments will have a better understanding of local preferences for local public goods. Given that local preferences may vary across jurisdictions, local outputs have to vary accordingly to maximize overall welfare (Oates 1999). However, PHC services have strong merit good characteristics, requiring uniform access across local jurisdictions, and therefore some regulation on their funding and provision across local jurisdictions.
In South Africa, equity is a primary objective in the financing of public sector health services and PHC (Gilson et al. 1999; Okorafor et al. 2003b; Thomas et al. 2003). There is broad consensus in the literature that it is more appropriate for the responsibility for redistribution and equity to lie with the central government (Shah 2005), due to the nature of resource mobility and the openness of local autonomy (Buchanan and Wagner 1971; Inman and Rubenfeld 1997). Another rationale for assigning such responsibility to the central government is that this kind of policy (on equity) should be set and implemented by one level of government (Smith 1985). Differences in local preferences arising from perpetuated inequities in South Africa may support the argument for local financing and provision of PHC, but the extent of inequities in resource allocation across local jurisdictions also make a case for the central government to intervene in order to achieve a more equitable distribution of allocated resources for PHC. Oates (1999) argues that intergovernmental fiscal arrangements may not necessarily conform to the traditional theoretical framework for the assignment of functions to different levels of government. The nature of the public service (or good), the context and the time within which the service is provided may result in differences in the pattern of goods and services provided by different levels of government. Following this line of thought, significant weight is therefore placed on the South African context in discussing options for providing and financing PHC.
Discussion
The theoretical and international literature reviewed allows us to draw some preliminary findings. There are various transfer options for financing public services, and the type of transfer for funding PHC should depend largely on the extent to which the level of distribution of PHC expenditure and provision are of concern to the central government relative to other public sector programmes. If the level of distribution of PHC expenditure and provision is of national concern, then fiscal intervention may be necessary to achieve a more desirable outcome. Indeed, as we shall see in the next section, South Africa may well be out of line with the countries reviewed above, all of which exercise some central control over funding of priorities, such as the health sector. In no countries reviewed (with the exception of India) is the funding of health priorities completely left to the devices of SNGs, as in South Africa.
From our review, there are three choices available to the central government to exercise some control over geographic spending on PHC. Firstly, it could finance all public health sector activities through conditional grants, or just key health programmes such as PHC. A second option is to allow SNGs to negotiate allocations to PHC, based on guidelines/norms set by the central government (e.g. minimum expenditure per capita). The third option is for the central government to relieve the SNGs of the responsibility of funding and providing PHC, i.e. recentralize spending. In general, the performance of these options in achieving a more equitable distribution of PHC funds depends largely on the context of the country; and the existing intergovernmental fiscal arrangements. Detailed information on the costs and benefits of these options are not available from the international literature. However, based on reviewed literature, a summary of the advantages and disadvantages of these options is given in Table 1.
|
|
The South African case study |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
Context
Following the democratic transition, McIntyre et al. (1995) summarized the key challenges for the health sector in the post-apartheid era. The first issues related to developing effective and affordable primary care services and ensuring a more equitable distribution of public sector health finance between provinces and between localities within each province. More than 10 years since these issues were raised, significant geographic inequities in PHC funding still exist. This raises the question of what has been done to move towards more equitable resource allocation for PHC. The policy response was quite explicit in the prioritization of the PHC approach in the ANC National Health Plan, the White Paper and the recently introduced National Health Act. All three emphasize that health care is to be provided in accordance with the PHC Approach and the District Health System is to be the main service delivery vehicle for primary care (African National Congress 1994; Republic of South Africa 1997; Republic of South Africa 2004).
Nevertheless, the development and implementation of changes to resource allocation patterns has had to occur within and through the normal process of government budgeting. It has thus had to adapt to the broader evolution of these budgeting processes (Thomas 2003). Indeed, there were two very different phases of resource allocation policy between 1994 and 2004. The first was characterized by a health sector formula that supported re-allocations of budgets between provinces on the basis of population and weighted by need (Gilson et al. 1999). Under the direction of the Director General of the NDoH, the Function Committee devised and oversaw the formula which was to achieve its goal of equitable resource allocation within 5 years. Substantial redistribution was planned for the first year (Gilson et al. 1999).
The second phase, the era of ‘fiscal federalism’, resulted from the introduction in 1996 of the new constitution, which finalized the governance structure of the country after transition. From 1997/98 the National Department of Finance (now the National Treasury) allocated block grants to provinces on the basis of a formula intended to reflect differential levels of overall provincial ‘need’, though there has been disquiet about its weightings and limited redistribution effects. The divisions have to be reconciled with provincial and sectoral budgets. This occurs with the development of the provincial and sectoral Medium Term Expenditure Frameworks (MTEFs), which are 3-year rolling budgets.
Figure 1 sets out the annual budget cycle. Initial budget plans are prepared by all spending agencies at provincial and national levels (Step 1). These are reviewed by the Minister's committee on the budget (MinComBud) and intergovernmental technical forums to determine an indicative division of revenue (Steps 2 and 3). Departments then submit their revised budget plans. These are considered to ensure that the figures fall within the guideline allocations determined through the vertical and horizontal divisions (step 4). The estimates developed then go on to form part of the final national budget presented to parliament (Step 8) (National Treasury 2006).
|
This process has effectively undermined the NDoH, leaving it unable to determine health resource allocations across provinces, with health allocations subject to competition from other sectors at the provincial level (Gilson et al. 1999
; Thomas and Muirhead 2000). While the NDoH had an active voice on the Function Committee, the move to a system of fiscal federalism under the new constitution also reduced its power substantially in the resource allocation processes. Indeed there have been concerns that, given the decentralization of the sectoral allocation at the provincial level, allocations to the health sector would be left hostage to politics (McIntyre et al. 1998; Gilson et al. 1999). Nevertheless, once they have received their budget, Provincial Departments of Health do have some power to decide how to allocate resources within the health sector. Precisely how such allocation decisions are done is not well documented. This is therefore explored further later in this report.
One way of preserving some national control over health sector allocations has been to develop ‘conditional grants’ under the management of the NDoH. These earmark funding for specific designated purposes. Several health conditional grants were introduced in 1997/98 but these relate primarily to higher-level services (see Box 1). Interestingly, and prior to the formulation of the conditional grants, the Financial and Fiscal Commission proposed a ring fencing of funds, through a ‘Minimum National Standards Grant’ to support PHC and the District Health System (Financial and Fiscal Commission 1996). Further, there was an attempt by the NDoH to obtain a conditional grant for PHC activities in 1997. Neither proposal succeeded however. Instead, the National Treasury sanctioned the protection of higher levels of hospitals. Currently, then, despite its importance to national health policy, there is no resource allocation mechanism that protects the funding of PHC.
| Box 1 Health conditional grants introduced since 1997/98 The health sector conditional grants cover: 1) Research and training of health professionals (all provinces); 2) Central hospital specialist services situated in Gauteng, Western Cape, Free State and Kwa-Zulu Natal to cover the costs of the use of these services by residents of other provinces; 3) Hospital rehabilitation and construction (the ‘redistributive’ grant); 4) The primary school nutrition programme. The NDoH releases these funds to provincial health departments only when they meet certain conditions: the submission of business plans detailing how the resources will be used and demonstration of appropriate use of the previous year's conditional grant funding. Source: Gilson et al. (1999).
|
Interview results
This section presents information collected through the interviews of government officials and NGO informants on resource allocation practices. Thereafter, the section will highlight key issues for choosing a preferred allocation mechanism.
Existing practice
As previously mentioned, PHC budgets are largely decided by the Provincial Department of Health, with little or no interference from the NDoH. Each province conducts its negotiations independent of others. The interviews reveal that allocations to PHC activities are generally done incrementally, on a historical basis. Given the current inequities in resource allocation to PHC, this simply maintains the status quo regarding the inequity of resource allocation. Although officials are aware of this, there is some inertia with regard to moving away from incremental budgeting towards a needs-based approach. Some officials have attributed the prevalence of incrementalism to lack of technical expertise at lower levels to manage a needs-based approach. Others attribute the proliferation of incremental budgeting to system constraints.
Also, in one province managers involved in trying to develop a needs-based formula had problems with the quality of data used. According to them, available data for developing a needs-based formula were not very reliable. In contrast, in another province the resistance to formula-based budgeting came about because of vested interests rather than a lack of technical expertise. District and provincial officials had attempted to use a needs-based formula for budgeting, but this met stiff opposition. Some parties were opposed to having cuts in the PHC budgets to certain districts, irrespective of relative need. Subsequently, the initiative was dropped. Based on both technical and political constraints, provincial authorities have frequently found it difficult to budget effectively for PHC services and to recognize the differential needs of the population they serve.
Further problems related to a lack of communication between different levels of government and the availability of appropriate financial information. Most officials at the district level commented that PHC activities were under-funded, and that this was in part due to the limited involvement of district-level officials in the budgeting process.
A compounding problem for proper budgeting of PHC is that it has been difficult to identify and monitor PHC expenditure for budgeting purposes. In some districts, there is no proper accounting system in place to log PHC expenditure.
Further, PHC activities are not carried out in only one type of health facility. They are rendered in clinics, district hospitals and community health centres. It is difficult then to estimate how much of the services provided in, say, district hospitals are PHC services. There are no guidelines on how to calculate PHC expenditure.
Protection of PHC funds
Officials were asked if PHC funds should be protected in some way. Some managers felt that PHC funds should be protected, either as a conditional grant or in a ring-fenced budget. In contrast, other managers opposed such protection. They felt that such protection was not ideal for funding PHC under the current dispensation. An official commented that the use of ring-fenced budgets could back-fire on PHC managers if budgets allocated to PHC were not fully utilized. This could result in budget cuts for the next year.
From our interviews, it appears that the use of conditional grants to finance PHC may not receive sufficient support. This is partly because national financing authorities are generally moving away from the use of conditional grants, placing more value on the autonomy of decentralized government. Some provincial health managers also had concerns around the use of conditional grants. They felt their use could substantially reduce the budget for other health programmes.
Discussions around strategies to protect and increase the size of PHC budgets produced no one solution that was endorsed unanimously by all interviewees. The failure of the NDoH to get a conditional grant for PHC in the mid to late 1990s and the current opposition to this approach may indicate that this option is difficult to pursue. Other actors suggested developing consensus for PHC funding with respect to desired levels of service provision, nature of the service and amount of services included within the PHC package. Nevertheless, given that the policy is already in place, the issue would appear to be one of policy implementation and effective enforcement, not design.
Another approach would be to develop costed norms to advocate for an increase in PHC resources. This would detail across provinces the resources required for the NDoH to meet its policy objectives and provide comprehensive PHC. The advantage with such an approach is that it demonstrates quite clearly how funds will be used in each province and acts to protect funding of PHC from health sector claims. Nevertheless, there are concerns about such an approach. First, key stakeholders are worried it would be very expensive, noting that a similar initiative in the education sector resulted in a wish-list approach to norms with associated high expenditure. Relatedly, there are concerns about the problems of producing a consensus on data for costing a defined PHC package. Further, the approach would buck the general national trend which is away from conditionalization towards increased autonomy.
The other option raised is for the NDoH to bid for increased health funding across the board, on the understanding that it is not just PHC which is under-funded but the entire health sector (Segall and Brijlal 2003). Such an approach has the advantage that PHC will not be in competition with other services and activities for scarce resources, as all will receive more. Nevertheless, the likelihood of the entire health sector getting a substantial jump in funding is probably small, especially given competition from other sectors and the current prioritization of job creation in the economy. Even if plentiful resources were available, there is also no guarantee that provinces would change their relative funding distribution across districts.
It appears then that a stalemate has been reached.9 Figure 2 attempts to summarize the factors that affect the choice of the appropriate transfer mechanism to protect PHC funding. The importance of different sets of values and the capacity of SNGs will have a significant bearing on the choice of transfer mechanisms. For instance, there are tensions between values that relate to the actual distribution of resources and those that concern the process of deciding the distribution of resources. At the same time such values have to be balanced against the reality of local capacity, be it in relation to resource management or generation of own revenue.
|
Despite the commitment to PHC and to equity in resource allocation in South Africa, it appears that the desire to grant SNGs autonomy over provincial decision-making and the desire to build consensus over the issue of appropriate PHC funding currently hold sway. Hence procedural values win out over distributive goals in this context. Further, while inequities are acknowledged, it is seen that the development of local capacity is a precursor to the effective use of resources. Therefore a resource mechanism which relies on enforcing some form of conditionality is not pursued. At face value, weak capacity and the importance of procedural values limit the acceptability of conditional transfer mechanisms (whether through a specific purpose grant or norms and standards). The status quo thus seems difficult to shift, with the consequence that funding remains grossly inequitable.
In general, the opinions of officials on constraints to achieving equity and strategies for achieving equity were similar across provinces visited, whether well or poorly resourced. All provinces cited the use of incremental budgeting and the lack of support from the province to districts as key constraints. The main difference between the two groups of provinces (well resourced and poorly resourced) is on the subject of protecting resources for PHC. Officials from the well-resourced provinces were concerned that if PHC funds are transferred as conditional grants or allocated based on a defined norm/standard, the lack of managerial capacity in the more rural areas within the provinces would result in under-spending of their (rural areas) budget. This, they feared would attract budget cuts in the subsequent years. Based on the responses from officials in poorly resourced provinces, this problem is even more severe. They face an even more acute shortage of management skills. A higher proportion of the poorly resourced provinces consists of rural areas, and attracting management skills to these areas has been extremely difficult. The lack of managerial skills in the less funded areas may also be limiting their ability to articulately justify increases in their budgets.
Literature on fiscal federalism highlights the uniqueness of fiscal federal relations and structures for any country. Since no single ‘general’ solution is ‘suitable’ for the South African case, with stakeholders holding differing views, it is important to review whether there is any localized solution which might work within context. This is discussed in the next section.
|
Conclusions and recommendations |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
Resource allocation to PHC in South Africa is dependent on budgetary negotiations within the Provincial Departments of Health. Evidence shows that huge inequities in resource allocation across districts within provinces still exist (Thomas et al. 2003
). Current practices for budgeting for PHC funds mean that the existing inequities are perpetuated, until such a time as a redistributive element is factored into the allocation process. Even where redistribution according to need is considered necessary, there are political, technical and system constraints that prevent the adoption of the required budgeting process/outcome. What is clear at this point is that the current resource allocation process under the management of the provinces cannot deliver an equitable distribution of PHC resources. While the National government is concerned about the inequity in resource allocation, it has little influence over how financial resources for PHC are allocated across districts within provinces. Fiscal federalism has, in the South African case, undermined the leadership of the National Department of Health. While it can still choose amongst policy priorities, the ability of the NDoH to develop appropriate resource allocation mechanisms is much diluted. The constitution developed in 1996 endorsed the fiscal federal system to preserve the autonomy of the provinces and help create political stability in the country following the democratic change. Funding is now under the control of the National Treasury and then Provincial departments. This has dislocated the realization of health policy. At best the NDoH is reduced to cajoling and lobbying provinces to address the inequalities in the funding of PHC across districts. While in the mid-1990s there were intense concerns about political stability and the representation of minority regional groups, times have moved on and it is unclear that the arguments for provincial autonomy have the same force. Documented international experience suggests that, in a fiscal federal context, the actual responsibility for resource allocation to health (or programmes and sectors whose outcomes are of national concern) lies primarily with the national government. In some cases SNGs share the responsibility for financing and providing health care, while national governments retain major control over the financing and provision of the services (through conditional/specific purpose grants). In no case reviewed (with the exception of India) do countries leave the funding of health priorities under the control of SNGs, as in South Africa. South Africa appears to be out of line for reasons noted in the preceding paragraph.
While interference is needed from national government in fiscal operations at lower levels of government, it will reduce the level of autonomy of SNGs to allocate resources. Although this goes against the basic principle of fiscal federalism, central government interference may be ‘necessary’ in certain cases and this is validated by international experience. PHC, as a national priority for South Africa, would seem to be one such special case. It is important that intergovernmental relations and transfers should be designed to achieve a good balance between central objectives and local preferences. Currently, central objectives are still being ignored in resource allocation practices for PHC. The result is persistent inequity in the funding and delivery of PHC across districts.
Yet, there may still be a way forward for South Africa, despite the system constraints, though it will require stakeholders to alter their positions. Subsequent to the main research for this article, discussions (by the authors) with officials of the National Treasury revealed that they are increasingly concerned about inequities in health sector spending across provinces and are more receptive to initiatives from the NDoH to deal with this problem. The NDoH could engage with the National/Provincial Treasury offices and the different Provincial Departments of Health in drawing up overall national health priorities. This would involve the NDoH convincing its provincial counterparts to include equity in PHC allocations as a national priority. If this is achieved, the National Treasury and the Provincial Treasury offices could then ensure that allocations to PHC across provinces are according to the overall national health policy on equity in PHC financing.10 However, the condition of this would be that the NDoH must provide a detailed plan (in consultation with Provincial Departments of Health and the National Treasury) for strengthening managerial capacity in the utilization of extra resources in under-funded areas, including equity targets and benchmarks for assessing equity. Ultimately, a coordinated approach, involving all stakeholders (NDoH, Provincial Departments of Health, National Treasury and Provincial Treasury Departments) may be the only foundation for effecting sustainable change in the distribution of resources for PHC in South Africa.
|
Acknowledgements |
|---|
The authors would like to thank the Danish International Development Agency (DANIDA); the National Department of Health, South Africa; and Health Systems Trust, South Africa, for funding the projects that have contributed significantly to materials used in this article.
|
Endnotes |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
1 The nationally endorsed package is a comprehensive PHC package.
2 Note that per capita expenditure refers to the total expenditure divided by the population without access to medical aid. This is for better reflection of overall dependence on public health services within the provinces. The proportion of the population with access to medical aid mostly uses private health services (Wadee et al. 2003). Also, the mean exchange rate for the period April 2002 to March 2003 is US$ 1 = ZAR 9.742.
3 Rate of conversion used corresponds to 2005/06 rates.
4 It is the authors’ intention to focus on issues of resource allocation and government organization. Boosting implementation capacity of SNGs is not addressed though it may affect, and be affected by, resource allocation. This topic is covered in more detail in Thomas et al. (2007).
5 The choice of Limpopo, North West and Western Cape provinces in the more recent survey was in order to include data from more provinces. This was at the loss of more in-depth analysis had the same officials interviewed in the earlier study been re-interviewed.
6 Thus, three of the provinces from the combined studies (Gauteng, Western Cape and Free State) had per capita PHC expenditure that was higher than the national average of ZAR 143 (2002/03) and are therefore considered to be well-resourced provinces. The rest (North West, Limpopo and Eastern Cape) have lower PHC expenditure per capita than the national average and thus are considered to be poorly resourced.
7 Revenue raising power/responsibility refers to the authority of a government level to collect tax returns. Expenditure responsibilities refer to the decision-making authority of a government level over the use of financial resources within its budget.
8 Possibly also to support specific standards or levels of service provision.
9 As of end 2006 the authors are not aware of any coordinated efforts involving all relevant stakeholders to resolve this impasse.
10 The National Treasury and Provincial Treasury offices have the authority to enforce spending patterns in line with overall national priorities as agreed by the NDoH and Provincial Departments of Health.
|
References |
|---|
|
TopAbstractIntroductionAimMethodsLiterature reviewThe South African case...Conclusions and recommendationsEndnotesReferences |
|---|
African National Congress. A National Health Plan for South Africa. (1994) Prepared by the ANC Health Department, with support and technical assistance of WHO and UNICEF. Maseru, Lesotho: Baar Mapping and Printing.
Ahmad E, Craig J. Intergovernmental transfers. In: Fiscal federalism in theory and practice.—Ter-Minassian T, ed. (1997) Washington, DC: International Monetary Fund.
Ayodele J. Fiscal federalism: the Nigerian experience. Economic Commission for Africa Ad-hoc Expert Group Meeting on Fiscal Policy and Growth in Africa: Fiscal Federalism, Decentralisation and the Incidence of Taxation. (2003) October 7–9. Addis Ababa: United Nations Conference Centre.
Bajpai N, Goyal S. Financing Health for All in India. In: DGSD Working Paper No. 25. (2005) New York: Centre on Globalization and Sustainable Development, The Earth Institute at Columbia University.
Bengoa R, Adams O, Kawar R. Primary health care: a framework for future strategic directions. (2003) Geneva: World Health Organization.
Bossert TJ, Larranaga O, Giedion U, et al. Decentralisation and equity of resource allocation: evidence from Colombia and Chile. Bulletin of the World Health Organization (2003) 81:95–100.[Web of Science][Medline]
Bowling A. Research methods in health: investigating health and health services. (1997) Buckingham and Philadelphia: Open University Press.
Bird RM, Smart M. Intergovernmental fiscal transfers: international lessons for developing countries. World Development (2002) 30:899–912.[CrossRef][Web of Science]
Bird RM, Vailliancourt F. Fiscal decentralisation in developing countries: an overview and perspective. Prepared for International Seminar in Public Economics. (1997) Tokyo. August 1997.
Bogeti 
. Bulgaria. In: Fiscal federalism in theory and practice.—Ter-Minassian T, ed. (1997) Washington, DC: International Monetary Fund.
Buchanan J, Wagner R. An efficiency basis for federal fiscal equalisation. In: Analysis of public output.—Margolis J, ed. (1971) Princeton, NJ: Princeton University Press for National Bureau for Economic Research.
Chitha W, Cleary S, Davauid E, et al. Financial resource requirements for a package of primary health care services in the South African public health sector. In: Results of a workshop held on 19 May 2004 and subsequent iterative modelling activities: final report. (2004) Cape Town: Health Economics Unit, University of Cape Town.
de Mello LR Jr. Fiscal decentralisation and intergovernmental fiscal relations: a cross-country analysis. World Development (2000) 28:365–80.[CrossRef][Web of Science]
Doherty J, Thomas S, Muirhead D. Health financing and expenditure in post-apartheid South Africa, 1996/97–1998/99. (2002) Johannesburg: National Health Accounts Project Team.
Financial and Fiscal Commission. The Financial and Fiscal Commissions Recommendations for the Allocation of Financial resources to the National and Provincial Governments for the 1997/98 Financial Year. (1996) Midrand, South Africa: Financial and Fiscal Commission.
Fjeldstad O-H. Intergovernmental fiscal relations in developing countries: a review of issues. In: CMI working paper WP 2001. (2001) 11. Bergen, Norway: Chr. Michelsen Institute.
Gilson L, Mills A. Health sector reforms in Sub-Saharan Africa: lessons of the last ten years. Health Policy (1995) 32:215–43.[CrossRef][Web of Science][Medline]
Gilson L, Doherty J, McIntyre D, et al. The dynamics of policy change: health care financing in South Africa, 1994–1999. Monograph 66. (1999) Johannesburg: Centre for Health Policy, for the Centre for Health Policy (University of the Witwatersrand) and the Health Economics Unit (University of Cape Town).
Inman RP, Rubenfeld DL. The political economy of federalism. In: Perspectives on public choice.—Mueller DC, ed. (1997) Cambridge: Cambridge University Press.
Khumalo N, Funani I, Muirhead D. The Local Government Financing Project: case study of Tshwane Metro. (2003) Johannesburg: Centre of Health Policy, University of the Witwatersrand.
Krelove R, Stotsky JG, Vehorn CL. Canada. In: Fiscal federalism in theory and practice.—Ter-Minassian T, ed. (1997) Washington, DC: International Monetary Fund.
McIntyre D, Bloom G, Doherty J, Brijlal P. Health expenditure and finance in South Africa. (1995) Durban: Health Systems Trust and The World Bank.
McIntyre D, Baba L, Makan B. Equity in public sector health care financing and expenditure in South Africa: an analysis of trends between 1995/96 to 2000/01. In: Technical Report to Chapter 4 of the 1998 South African Health Review. (1998) Durban: Health Economics Unit, University of Cape Town, and Health Systems Trust.
Miles MB, Huberman AM. Qualitative data analysis: an expanded sourcebook. (1994) London: Sage.
Mubangizi D, Okorafor O, Thomas S. The Local Government Financing Project: sub-district equity analysis – a case study of Thabo Mofutsanyane District, Free State Province. (2003) Cape Town: Health Economics Unit, University of Cape Town.
Musgrave RA, Musgrave PB. Public finance in theory and practice. (1989) New York: International Students Edition McGraw-Hill.
National Treasury. Medium Term Expenditure Framework; Treasury Guidelines: Preparing budget proposals for the 2007 MTEF. (2006) Pretoria: National Treasury, Republic of South Africa.
Oates WE. An essay on fiscal federalism. Journal of Economic Literature (1999) 37:1120–49.[Web of Science]
OHCHR-UNOG. International Covenant on Economic, Social and Human Rights. (1966) Geneva: Office of the High Commissioner for Human Rights.
Okorafor O, Mubangizi D, Thomas S. Local Government Financing Project: case study of Alfred Nzo District. (2003a) Cape Town: Health Economics Unit, University of Cape Town.
Okorafor OA, Thomas S, McIntyre D. An evaluation of public primary health care financing in South Africa: 1992/93–2002/03. In: Report prepared for Valley Trust. (2003b) Cape Town: Health Economics Unit, University of Cape Town.
Republic of South Africa. White Paper for the Transformation of the Health System in South Africa: Towards a National Health System. In: Government Gazette. (1997) 382, no. 17910, Notice 667 of 1997.
Republic of South Africa. National Health Act. Act No. 61 of 2003. In: Government Gazette 469. (2004) 23 July 2004, no 26595.
Schwartz G, Liuksila C. Argentina. In: Fiscal federalism in theory and practice.—Ter-Minassian T, ed. (1997) Washington, DC: International Monetary Fund.
Segall M, Brijlal V. Essential health care for all South Africans – an investigation into public health financing and provincial health resource distribution. (2003) Pretoria: Department of Health.
Shah A. Fiscal federalism and macroeconomic governance: for better or for worse? In: Policy Research Working Paper Series 2005. (2005) Washington, DC: The World Bank.
Smith B. Decentralisation. (1985) London: George Allen and Unwin.
Ter-Minassian T, ed. Fiscal federalism in theory and practice. (1997) Washington, DC: International Monetary Fund.
Thomas S. Managing actors in South African health financing reform: testing a conceptual framework. (2003) Thesis presented for the Degree of Doctor of Philosophy in the Department of Public Health, University of Cape Town.
Thomas S, Muirhead D. National Health Accounts: The Public Sector Report. (2000) Cape Town: NHA Consortium.
Thomas S, Mbatsha S, Muirhead D, Okorafor O. Primary health care financing and need across health districts in South Africa. (2003) Cape Town: Health Economics Unit, University of Cape Town and Centre for Health Policy, University of Witwatersrand.
Thomas S, Okorafor O, Mbatsha S. Barriers to the equitable funding of primary health care in South Africa. Applied Health Economics and Health Policy (2005) 4:183–90.[CrossRef][Medline]
Thomas S, Mooney G, Mbatsha S. The MESH approach: strengthening public health systems for the MDGs. Health Policy (2007) in press, Feb 6 [Epub ahead of print].
Wadee H, Gilson L, Thiede M, et al. Health care inequity in South Africa and the public/private mix. In: Paper prepared for the RUIG/UNRISD project on Globalisation, Inequality and Health. (2003).
World Health Organization. Declaration of Alma-Ata. International Conference on Primary Health Care, Alma-Ata, USSR, 6–12 September 1978. (1978) Online at: http://www.euro.who.int/AboutWHO/Policy/20010827_1.
Yin RK. Case study research: design and methods. (1994) second edition. Applied Social Research Methods Series, Volume 5. Thousand Oaks, CA: Sage Publications.
Accepted for publication 9 July 2007.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||