Economic Viewpoint - August 2006

Health Financing in Africa - Further Thoughts on Abuja

• Related Event
• Featured Reading
• Editorial

Author:  Chris Atim, Specialist on Health Economics and Financing at the HLSP Institute, London 

In this piece, we discuss some issues arising from the paper written for the African Union Special Summit on HIV/AIDS, TB and Malaria, held at the beginning of May 2006. The following summarises the main questions that we will address in this paper, based on the principal issues arising from the Abuja paper.

The first issue is the low levels of health spending in most African countries. However, the more interesting question perhaps is whether the Abuja target will make an impact, or is it mostly a red-herring? We will examine the strengths and limitations of the Commission on Macroeconomics and Health’s (CMH) recommended target for spending on health of $34 per capita. Next, we will discuss the important question of extra funding required for the priority health interventions. Given that whatever method used, estimates of the amount of resources required to achieve the Millennium Development Goals (MDGs) all show huge sums are needed, well beyond what the low income countries could possibly afford, the question is: Where should the extra resources come from? While no magic bullet can be offered, we will examine further some of the issues concerning external aid and the potential for domestic revenue mobilisation.

The discussion is based on secondary data from various sources.  It is recognised that these data may have limitations arising from the sources and methods of gathering that data.

Low health spending and the Abuja target

Evidence has already demonstrated, without much controversy, that by almost any measure, most African countries are spending far less on health than they need to in order to either ensure a reasonable package of priority health services for their populations, or to meet targets signed on to by African leaders.

In 2001 African Union (AU) countries set the Abuja target of attaining a 15% share of national budgets for the health sector.    Although no corresponding deadline was set for achieving the target, it is significant to note that by 2005 (according to the AU’s own data), only about a third of Sub-Saharan African (SSA) countries are allocating 10% or more of their national budgets to the health sector. Of the other two-thirds, close to half (or one-third of all AU members) are allocating below 5%.  The AU’s data also show that West and Central Africa contain most of the countries that allocate less than 5%; while Botswana is the only country that achieved the Abuja target of 15%.

Leaving aside for now questions about the appropriateness of this target as well as possible data comparability problems, what should be of more immediate concern, in terms of progress towards achieving the target by countries, is the apparent trend in the data. WHO data based on NHA analysis on the period from1998 – 2002, showed three countries exceeded or were close to exceeding the Abuja target of 15%. Mozambique, Congo DR, and Tanzania all achieved or surpassed the 15% target (see Figure 1 below);   while a further dozen or so countries achieved the regional average of about 9% or above. The AU’s data for 2005 (again barring issues such as data reliability and comparability between different sources alluded to above), would therefore seem to imply that, with respect to the earlier data, some AU countries have regressed.

But rather than dwell on what progress countries are making towards the Abuja target, perhaps a more useful, and at least equally interesting question is whether the Abuja target is in some way a red-herring?

If the reason behind the 15% target was to signal to the world the determination of African leaders to provide more resources to the health sector in the face of Africa’s serious health challenges, then that headline-grabbing figure was an effective message.    In terms of matching this bold pledge with appropriate action, however, the available evidence so far is at best not very encouraging. Also, if African leaders thought that setting the Abuja target would be the starting push to seriously address the health financing challenges of their countries, i.e. to try to ensure a reasonable package of priority health services for their populations, then this particular approach is questionable. The basic problem with this approach is well illustrated in Figure 1: Government Expenditures on Health in Africa (pdf 19KB).

The Abuja target line in the figure shows the three countries (Mozambique, Congo DR and Tanzania) mentioned above that met the target in 2002. However, looking at the per capita public health spending in those countries, 15% of the budget will purchase in dollars at average exchange rates,   $8, $1 and $7 of health from public funds in Mozambique, Congo DR and Tanzania, respectively. In terms of total health spending from all sources, the corresponding figures for those countries are $11, $13 and $4 respectively – still well below the CMH’s $34 target (from all sources) for ensuring the WHO-defined essential package of health services for their populations.  It is also an interesting coincidence that the three countries concerned are among those with low per capita incomes in Africa.

In contrast, countries such as Seychelles, Botswana, Mauritius, South Africa, Tunisia and Gabon, which enjoy considerably higher levels of per capita income, all spent less than 12% of their national budgets on health and yet they all exceeded the CMH target from their public spending alone! There are in fact 12 countries from Figure 1 that have exceeded the CMH target while allocating less than 15% of their national budgets to health, and nearly all are middle income or low-middle income countries.  Whereas, low income countries such as Mozambique, the DRC (which exceeded the 15% allocation with just $1 per head) and Tanzania need to devote a larger share of their smaller budgets to approach anything close to what is required under the CMH target.

The obvious point here is that what 15% of the budget will buy a country depends on the size of its economy (GDP) and its population (as well as the efficiency with which resources are used, which we are not examining here for now). More specifically, better off countries such as middle income ones in Africa may be able to afford reasonable health care for their populations without devoting 15% of their budgets to health.  Conversely, for the poorer countries, allocating 15% of their budgets may do little to address the chronic under-funding of health services in their countries. Analysis of what difference achieving the Abuja target would make confirms this to be the case for many low income countries.

It is necessary to stress this otherwise trivial observation because of the risk that the Abuja target, and league tables such as the AU’s monitoring data which are based on it, might lead national policy makers to believe, consciously or unconsciously, that reaching that mark is all that it would take to adequately fund their priority health interventions (say those defined in the MDGs, which African leaders have also signed onto) and the required health system improvements. A growing, dynamic economy would clearly appear to be more important in terms of how much per capita can be made available for funding health service improvements; although a high per capita allocation alone would not ensure good quality and affordable health care at the point of delivery, or the achievement of the MDGs. For these to be attained, additional issues such as good sector leadership, the efficiency of resource use, removal of bottlenecks in the organisation of health service delivery, and lifting of barriers to access especially for the poor must also be addressed.

The CMH $34 target

A logical conclusion from the discussion so far is that the CMH approach (of a per capita spending target) is a better one. This makes intuitive sense in general, in so far as that approach stresses that a certain minimum level of per capita spending is required to achieve priority health goals. But there are questions about whether the CMH figure truly represents the level of resources required in all low-income countries?

The issue deals with the generic character of the CMH figure, especially if we understand the CMH to say that US$34 would buy the same essential package at average exchange rates in each of these countries. Given the differential purchasing power of the dollar across countries, this clearly cannot be the case. Moreover, the disease burden differs significantly across countries.  Countries with a high burden of AIDS cases would be expected to require greater resources than those with a lower burden, all other things being equal. A similar case can be made with regard to malaria endemic countries which have switched to ACTs due to high resistance to the older and cheaper anti-malarial drugs.

Therefore, although it is an improvement over the Abuja approach, the CMH method shares the same limitations with the Abuja target.  In trying to find an attention-grabbing figure to focus international attention on what is required to resolve critical health problems in low income countries, these targets ignore specific country realities and important differences.   It runs the risk that the figure arrived at may not be relevant in a country setting.

In view of the potential problems with a universal, generic target, we would argue for setting country-specific spending targets and focusing on the achievement of priority interventions such as the MDGs. Such an approach arguably would not be as headline-friendly or attention-grabbing on the international scene as a simple figure like $34, but it would be operationally far more useful to countries. Stakeholders including governments, civil society and development partners in each country would know how much they need to aim at mobilising (from both domestic and external sources) to achieve the health targets to which they have subscribed.

Estimates of required per capita spending needed to fund priority services are of course easily translated into total amount of resources required. And for policy makers, it is arguably such estimates of overall resources (in a country-specific way) that would be operationally most useful. However, the best estimates of required resources available at this stage are based on global level calculations of total or incremental resources required to achieve the MDGs. In the absence of an aggregate of country-specific estimates, these global level calculations are what we have to work with. And whatever the method used, there is near unanimity on two issues. The first is that the scale of the overall resources required to meet those internationally-agreed health targets is huge, ranging from $20 billion - $70 billion per annum until 2015.  The second issue is that such a scale of resources far exceeds anything that low income countries will be able to mobilise from domestic sources alone in the foreseeable future, on the basis of current revenue-mobilisation efforts.

External aid flows

So where are all those extra resources going to come from? In response to this question, a recent World Bank study concluded that “public expenditures on health must be increased and this additional spending must be financed largely by donor support, especially in the least developed countries”.  However, while the international context for responding to the challenges posed by the MDG targets and especially disease-specific interventions may have considerably improved in recent years, much of this due to the entry of Global Health Partnerships (GHPs) and new private foundations onto the international funding scene, both past experience and the economics of aid flows call for some caution regarding the high expectations from this source.

The rise of new GHPs and increased bi- and other multilateral assistance for specific health interventions have come under a number of criticisms, relating to both macroeconomic and fiscal problems associated with external aid flows. We will try to use mainly contemporary examples to help illustrate the following fiscal issues associated with the international aid flows:

A. Unpredictable and volatile funding;

B. Aid promises consistently exceeding actual disbursements;

C. Increased transaction costs for countries with large numbers of uncoordinated actors and interventions;

D. An emphasis on donor-driven priorities (often vertical, disease- and intervention-specific programs) which are often not integrated into countries’ ongoing priority programs.

With regards to point A above, a good example is shown in Table 1 below. This example illustrates how extremely difficult it is to plan against the uncertainties produced by such high levels of variation.

Table 1: Uganda's Donor Project Expenditure against Budget FY 2004/05                         

Source: Uganda Annual Health Sector Performance Report, Financial Year 2004/2005.                         

Notes: 1) Ug. Shs = Ugandan Shillings. 2) Donor expenditure of 254,847 billion Ug. Shs is equivalent to US$146.895 million; or per capita of US$5.74. See source for more details and notes on above data.                

A study of 72 countries found that volatility in aid flows far exceeds that in tax revenues, averaging about 40 times that in revenue.  Gottret and Schieber give more examples of wild fluctuations in donor aid commitments that occured in seven African countries between 1997 and 2001.

Yet such aid commitments are the basis on which countries often make planning and budgetary decisions, despite abundant evidence of the lack of direct correlation between commitments and disbursements. These variations may result from factors such as donor political and budgetary processes, donor administrative procedures, the recipient country’s absorptive capacity, exchange rate fluctuations and donor conditionalities.

While in the Ugandan case, the actual disbursements exceeded the planned commitments, the global picture shows that the opposite is more often true, especially with regards to the most vulnerable countries (point B above). The study on aid volatility cited above also found that aid predictability (measured as the difference between aid commitments and actual disbursements) is especially skewed against the poorest countries. The latter received in the order of 50 percent of what was promised, compared to countries on the upper income scale which tended to receive their promised resources.

The most recent illustration of disbursements falling behind commitments may be seen from the promises made by the G8 at the Gleneagles Summit in 2005. It has been estimated that in order to get from the 2004 baseline to the 2010 target commitments made by the G8, an increase in aid of US$3.5 billion in 2005 would have been needed. Figure 2 shows the annual ODA increases that would be needed to reach the 2010 commitments made at the Gleneagles Summit.

Figure 2

Source: DATA Report, July 2006

However, the first full-scale audit of the G8’s progress regarding their promises to the poor at Gleneagles shows that the donors came up with only U$1.6 billion in 2005. In fact, all G8 donors, with the sole exception of France, are off-track on their commitments with regards to their 2010 goals, and some – Germany and Canada – have even decreased aid to Africa since the G8 Summit. It has also been suggested that some donors used debt relief – the one area where the G8 promises were delivered – to mask decreases in their overall aid to African countries.

With regards to the large numbers of uncoordinated actors and interventions (point C), the diagram below of actors in the field of HIV and AIDS alone within Tanzania is highly indicative of this challenge. This picture is not untypical of the situation in many other countries.

Figure 3: Actors and donors in HIV and AIDS in Tanzania

With regards to point D, Figure 4 below from Uganda’s 2004-05 health sector report shows how for that period, donors allocated more than half of their spending to areas outside the country’s priorities for that sector.

Figure 4: How donors allocated their resources in relation to Uganda’s health sector priorities in 2004-05

Source:Uganda Annual Health Sector Performance Report, Financial Year 2004/2005.

Together, these examples show how external assistance can complicate country planning and budgeting processes, and they could even reduce the impact of donor funding in achieving the desired health improvements. Some of the evidence (eg. from GAVI financial sustainability plans) also suggests that the recurrent costs of new GHP investments (human resources required to deliver new programmes, new treatment centres, costs of vaccines and life-saving drugs, etc) may not be sustainable for many low income countries.

While it is also fair to mention that donors and countries have begun to explore a number of new instruments and mechanisms for addressing the above problems (including the Paris Declaration on Aid Effectiveness which defined good practice in aid as well as indicators for measuring related progress), it is still too early to measure any lasting impact.

Concerns have also been raised about possible macroeconomic distortions that increased aid flows might cause for countries. The main risks in this regard include possible appreciation of the exchange rate, drop in exports and rise in inflation following large aid inflows (the so-called ‘Dutch Disease’), especially as a result of non-commodity components of such inflows. However, the evidence base for these concerns remains mixed.

Adam and Bevan (2003) found through a modelling technique that although such ‘Dutch Disease’ effects were likely in the short run, they could be reversed in the long run. Empirical studies by the IMF found inconclusive evidence for correlation between aid and exchange rate shifts. Other writers on the subject support the view that the negative short-run macroeconomic effects of aid can be reversed over the long run, so long as aid is invested effectively and enhances the productive capacity of the economy.

In the last section below, we will posit briefly a virtuous circle in which enhancing the productive capacity of the economy or economic growth will sustainably expand the fiscal space context in which African countries respond to the extraordinary health challenges they face.  In turn, improving health outcomes can also contribute to enhancing productive capacity. This, of course, is not a new observation but is perhaps a timely reminder in the context of the discussions regarding how to achieve the international health targets.

Economic growth and domestic resource mobilisation potential

Analysis of the various mechanisms and instruments available to African countries for expanding fiscal space in order to increase spending on health shows that there are limitations concerning the amounts that can be expected from   nearly all of these sources: taxation, improving efficiency, more aid, borrowing, debt relief, printing money, new health financing mechanisms, etc. For instance, though most African countries can be said to be in the low tax category, efforts to raise tax levels or to collect more tax revenues come up against technical, administrative, political and socio-economic bottlenecks in these countries. Improving efficiency would not increase the total resource envelope, though it would make existing resources go further and make more of a difference where those resources are most needed. Potential problems of aid funding have been discussed above.   The capacity to benefit from aid-financed investments is limited by the ability of the recipient country to bear the associated costs which depend on the economic prospects of the country. Borrowing to fund the investments depends on the existing level of aid dependence and capacity to meet the loan repayments and the interest on them, which again comes down to the country’s economic prospects.

The most trailed outcome of the G8’s Gleneagles Summit was the cancellation of $US40bn of the debt of 18 highly indebted countries, including 14 in Africa. This achievement sounded very impressive; but the economists at Goldman Sachs pointed out that the likely cash savings were actually quite small given that the annual debt service of the 14 targeted sub-Saharan African countries was just $500m. “This compares with the $4bn that sub-Saharan Africareceives each year, according to a 2003 World Bank report, from the remittances of migrant workers. The report noted that actual remittances were much higher as the statistics are heavily under- reported.”

Nevertheless, the Goldman Sachs team noted the key thing could be how the $500m a year was used. If it was spent in creating a climate more conducive to investment, growth and, in particular, if it encouraged governments to invest more on health and education, then this would be a good contribution to eliminating poverty.

Economic growth could unlock the potential of several of the mechanisms and instruments mentioned above for increasing fiscal space for allocating more resources to health. Tax receipts tend to rise with growth, and such growth will also generally improve the capacity to support higher levels of borrowing (both internal and external) and to pay for the recurrent costs of increased aid. New health financing mechanisms such as social health insurance will also tend be more viable with higher growth if this is translated into higher incomes.

For these reasons, it comes as potentially good news that economic growth has been on the rise in Africa in recent years. According to the OECD, African economies as a whole are expected to grow by 5.8 per cent this year and another 5.5 per cent next year, with oil-exporting countries far outpacing the others by a substantial margin. 

The OECD report continues that overall, the outlook for much of Africa continues to be more favourable than it has been for many years. Continued global expansion means sustained demand and higher prices for African oil and other industrial raw materials, while an increase in official development aid and improving macroeconomic stability have also contributed to Africa’s positive economic outlook.

This new prosperity has been helped by the boom in natural resources, but countries without them are also thriving. For instance, in Kenya, despite the drought in the north of the country, growth last year was close to 6%, helped by a 13% growth in tourist receipts, and is expected to be higher this year. And what would make an even greater difference would be the elimination or reduction of barriers to agricultural trade. Goldman Sachs estimated that liberalisation of trade would increase growth by 1% across the African continent.

According to the campaign group, Data, an expanded and fairer trade system would allow African countries to earn the resources they need and not only achieve the Millennium Development Goals but reach a more important goal for Africa—self-sufficiency. The group argues that in 1980, Africa had a 6% share of world trade. By 2002 this had dropped to just 2% despite the fact that Africa has 12 % of the world’s population. If Africacould regain just an additional 1% share of global trade, it would earn $70 billion more in exports each year – more than three times what the region currently receives in international assistance.

However, a focus on growth alone would not automatically lead to the desired health outcomes, unless there is a simultaneous attention on eliminating the major causes of poverty, and on promoting greater equity – which in turn requires a focus on increasing the average incomes of the poor, and investing in health, education and other social and infrastructure services. The high-income Asian countries and the high-income Latin American countries offer perhaps some salutary, but contrasting lessons on how emphasis on growth alone does not necessarily lead to the desired outcomes in the social development indicators.

Conclusion

Health spending levels in Africa are generally low, and although the Abuja target may serve as a way to gauge African countries’ commitment to increasing resources for health, that target can also be misleading if the policy makers and peoples are led to believe that 15% of their national budgets is all it would take to get their health services to perform as they should. The CMH recommended target of $34 per capita seems a better approach but it also has its limitations, chiefly due to its generic character. We suggest the less headline-grabbing approach of country-specific target setting, in consultation with all the country’s stakeholders and partners.

The resources required to achieve the international targets such as the MDGs are well beyond what low income countries can afford from their own resources alone. It is clear that these countries will continue to draw on foreign aid to fund part of their services for sometime into the future. However such aid also has its potential drawbacks,  The only sustainable way to provide resources for the health sector in the long run is to focus on higher economic growth. But such growth will not automatically translate into better health outcomes unless there is also a sustained focus on eliminating poverty, by increasing average incomes of the poor and investing in social sectors such as health and education. Therefore, there is still a need to keep making the case for the required allocations to the health sector in order to achieve the targets for priority interventions. In the end, this is the case for a virtuous circle, where more health spending and better economic growth both lead to better standards of living as well as a healthier populace.

References

• Adam and Bevan 2003 – see Lewis 2005.Bulir A. and Hamann A J 2005. “Volatility of development aid: From the frying pan into the fire?” International Monetary Fund Working Paper. Cited in Lewis 2005.
• DATA . undated.– “Trade for developing countries: essential in the fight against poverty”.Summit of African Union Heads of State and Government on HIV/AIDS, TB and Malaria, held at Abuja, May 2-4, 2006.
• Foster, M. 2005. “Improving the Medium- and Long-Term Predictability of Aid”. World Bank, Washington, D.C. Cited in Gottret and Schieber (2006).
• Gottret, P, and Schieber, G. 2006. Health Financing Revisited: A Practitioner’s Guide. The World Bank, Washington, D.C.
• King, Max. Jul 1 2006. “Growth helps Africa more than Gleneagles”. Article on the icwales.co.uk web site.
• Lewis, M. April 2005. “Addressing the Challenge of HIV/AIDS: Macroeconomic, Fiscal and Institutional Issues.” Center for Global Development, Working Paper Number 58.
• OECD Development Centre and the African Development Bank. African Economic Outlook 2005/2006. 16 May 2006. Published by the OECD Development Centre and the African Development Bank, with financial support from the European Commission.