A much needed new focus for SAPs
by Giovanni Andrea Cornia
Over the past two decades, a clear consensus, supported by a
growing amount of empirical evidence, has emerged on the growth and distributive
effects of appropriate health care, nutrition and education, and on the
rationality of investing in human resources to promote development in low-income
economies, and in sub-Saharan Africa in particular. It is now clear that public
and private rates of return on investment in primary and secondary education are
substantially greater than those in most industrial sectors. Similarly, it is
now well demonstrated that nutritional interventions in favour of workers at low
levels of food intake has a very large impact on labour productivity. Also,
provision of simple but adequate nutrition and health care at an early age has
been shown to spare substantially greater health outlays over the longer term.
While the number of these examples can be multiplied, the lesson is
overwhelmingly clear: investing in people is a sound development policy.
During the same two decades, a broad political consensus has
also emerged on the need to establish realistic targets in these areas and
mobilise adequate resources to achieve them. The latter decade, however, has
also been characterised by a widespread and acute recession which has reached
its lowest point in the African continent.
While there has been a general recognition of the link between
the deterioration of the social sector and Africa's economic decline, the
interpretations of the causes of this decline and how best to address them vary
The 1980s also witnessed the advent of structural adjustment, an
approach aiming at revitalising economies affected by severe crises and which
calls, amongst other things, for a reduction in public spending, including on
social services. With sharp falls in revenues, it was argued, African
governments could no longer afford to provide health and education free of
Governments were to encourage instead the role of the private
sector (since competition between state, private and non-governmental agencies
fosters efficiency and enhances quality) and make households bear a greater
share of the cost of publicly provided social services (as-it was
alleged-households had a strong 'willingness to pay' for better quality health
care and education).
In a debate characterised by an overwhelming concern for
expenditure reduction, little thought was given to protecting the provision and
effectiveness of basic health care and education for all. Nor was sufficient
attention paid to the mobilisation of additional revenue through taxation,
foreign aid or other measures Even the introduction of rationalisation measures,
which would have allowed for use and preservation of limited resources, received
scant attention. During this period, public policies often ignored a number of
well-known public finance arguments concerning market failures, externalities,
natural monopolies, etc. typical of basic health and education. It was all too
often forgotten, for instance, that the market systematically undersupplies
'quasi-public goods' like education, safe water, immunisation and collective
infrastructure (i.e. goods which produce benefits not only for consumers but
also for society as a whole); that incomplete or non-existent markets alone are
unable to provide complete insurance coverage against the risks of sickness and
disability; that the free play of market forces can generate levels of poverty
that are socially unacceptable, and that competition among many small providers
does not allow for the realisation of the economies of scale which only the
public sector can realise because of its size, administrative infrastructure and
revenue raising ability.
Greater and better government involvement
In this paper, I will attempt to point out the main distortions
in the social sector in sub-Saharan Africa and the major reforms, part of an
alternative policy paradigm to the standard approach focusing on large
expenditure reductions. My main line of argument is for more efficient and
renewed but, in most cases, greater rather than less government involvement. I
do not deny the important role of communities, nongovernmental bodies and
private providers of social services, but the retreat of the state, which has
been underway for several years in most of Africa, is palpably wrong: it may be
a good survival strategy, it is definitely not a good development strategy.
Any approach to revitalisation of the social sector in Africa
must aim at increasing both the quality and the relevance of services to the
life, health and productivity of the population; enhancing efficiency; reducing
distributive biases in the use of public expenditure, and improving the flow of
both public and private resources to the social sector. These sets of measures
are briefly discussed here.
Raising relevance, reducing unit costs and enhancing quality
Two basic initiatives are essential to the achievement of this
goal. Firstly, it is necessary to adapt the content of services to the demands
placed on the child, the adolescent and the family by the external environment.
Secondly, in doing so, it is necessary to provide minimum resources,
particularly recurrent inputs, for every unit of service offered.
Much can be done to improve the relevance of education in
Africa. A key element in this is the wider use of African languages in the
educational system. Admittedly such a step would probably encounter obstacles,
but the results of pilot projects in the continent and experiences in several
English speaking countries have been encouraging. The use of African languages
should be expanded throughout primary and at least part of secondary schooling.
And the curriculum of basic education should include basic health science,
agriculture, food and nutrition, basic technological science, ecology and
environmental studies as well as reading, writing, arithmetic, history,
geography and elementary civics.
While responding more appropriately to the life and
developmental needs of children and adolescents, these measures offer the
potential to improve substantially the cost effectiveness of the educational
system by reducing the number of drop-outs and repeaters, particularly if the
measures are accompanied by school-based nutritional support programmes and by
steps to ensure a steady, if basic, supply of teaching materials such as books,
chalk, slates, pens, exercise books and so on. Minimum expenditures for the
latter, estimated at $5-10 per child per year, have been shown to have a
profound impact on the quality and attractiveness of education.
Improving the relevance and quality of African secondary and
higher education will require an increase in the number of courses in science,
engineering, agronomy, management and other fields. Given the limited success of
vocational training and yet the profound need for this type of education, a more
flexible approach may be appropriate. Business associations and ministries of
labour, rather than ministries of education, for example, could organise the
type of vocational training that responds more closely to labour market
conditions. Much needs to be done to gear university research to local and
national needs. Rather than maintaining the model of independent residential
universities, faculties could interact more with and rely upon local
institutions and non-governmental organisations.
In the health sector, a more vigorous approach is necessary to
implement the 'Health for All' strategy. In particular, more attention must be
paid to the quality of the training imparted to village health workers, the
provision of minimum supplies of essential drugs, adequate supervision, the
mobilisation of communities and the strengthening of family planning services in
the first tier of the primary health care pyramid.
Greater efforts are needed to contain unit costs through the
mobilisation of community resources (such as materials and time) and the
introduction of 'product standards' and 'production techniques' more in line
with local factor endowments. With respect to capital expenditure, one major
source of potential savings is the substitution of imported materials with local
ones in the construction of schools, clinics and other public health facilities.
Through this approach, capital costs per student could be more than halved.
In health care, savings are also possible, in some cases, in
wages and other recurrent expenditures. The potential for savings through
increased efficiency in the procurement, management, storage, distribution and
prescription of pharmaceuticals is substantial.
And there is a need for more carefully articulated manpower
policies, with recognition of the cost implications of alternative skill mixes.
Many tasks carried out by senior professionals can be performed by auxiliary
personnel. In this regard, there is substantial scope for the application of
nominal fees as a tool to guide demand for health care and thus increase
Similar considerations are valid in the education sector.
Despite considerable declines during the 1980s in both French-speaking and
Eastern and Southern Africa, teacher salaries in the late 1980s were still
considerably higher than per capita GNP in several countries. While a cutback
may be possible in some countries, reductions below the 'efficiency wage' should
be avoided as they affect the quality of teaching.
There are, of course, other approaches to reducing unit costs
such as using teachers trained on crash courses and- paid salaries lower than
that of fully trained ones, multi shift education to increase teaching hours,
multigrade classes to increase class sizes and the creation of regional training
centres with donor support. In addition, in most of sub-Saharan Africa, all
university students receive fellowships to cover living expenses while at
school. In many cases these fellowships represent a substantial proportion of
total unit costs at the tertiary level. They add, in effect, to the regressive
nature of educational subsidies and should be suppressed for all except for
those who are least able to pay.
Substantial improvements in the distribution of benefits of
public expenditure could be obtained through its restructuring towards basic
services which benefit the poor more than they do the rich. This would mean,
typically, the reallocation of part of the expenditure on hospitals to primary
health care activities and part of the expenditure on universities to primary
education. It would also mean that a greater share of investments in water
supply and sanitation should be assigned to rural areas.
Correcting existing misallocation, however, is not simple, as
capital and labour resources are only partially mobile. This is particularly
true of capital expenditure which has already been invested in physical
infrastructure. Inter sect oral reallocation is easier for recurrent
expenditure, particularly when budgets are expanding, and it can lead to the
achievement of more efficient and distributive gains.
Government subsidies need to be improved through effective
targeting on 'merit goods' (such as food and shelter). Targeting is
distributionally and administratively more effective when it is accomplished
according to some objective criteria such as geography (the poorest areas), type
of commodity ( low quality foods not consumed by the rich), easily identifiable
population group (like pregnant women, young children, children in elementary
school in poor areas, etc.) or season. It is less effective when carried out by
means-testing or by targeting individuals.
The decentralisation of health care and education in the African
context is likely to improve the geographical distribution of services. Poor
communications and transport systems, population dispersion and the fact that
the supervision of personnel, resource distribution and operations is easier at
lower levels of administration, are all arguments in favour of such
decentralisation. Nonetheless, where it means loss of control over the
implementation of national policies (in curriculum development, manpower
training, drug procurement and so forth) and excessive additional financial
burdens on households and communities, decentralisation is potentially damaging.
This was the experience of Kenya when the purchase of drugs was decentralised at
the district level. It subsequently had to reverse this policy because of the
Increasing the flow of resources to basic social services.
Despite the considerable scope for protecting and, in some
cases, expanding the delivery of cost-effective and high-impact social services,
even during periods of declining resources, there are limits to what can be
achieved through the measures discussed above, particularly in countries where
the share in GDP of public health and education expenditure is already low.
Indeed in most African countries, the flow of resources to these sectors must be
vastly increased. The options available include:
-Increasing the share of government expenditure on key social
Top priority should be given to a radical restructuring of
government expenditure involving the shifting of resources towards those sectors
with high social rates of return. Expenditure on defence can be cut
substantially, while subsidies to chronically lossmaking public and private
enterprises and interest payments on debt (the relief effect on foreign-debt
interest payments of current provisions is utterly inadequate) should be
reduced. If realistic expenditure restructuring like that proposed here were
implemented, 1015% of overall public spending in many countries could be
redirected to social activities with high rates of return. One half to
two-thirds of these savings could be allocated to health care, nutrition,
training, education and welfare, thus raising the share of these activities to
about 30-35% of total expenditure.
Between the late 1970s and the mid-1980s, the average tax to GDP
ratio for sub-Saharan Africa fell from 18% to 16-17%. (Only recently are there
signs of recovery, but these are confined to only a few countries). This
resulted mainly from the recession affecting the region and from the limited
priority attached to revenue collection in most adjustment programmes
implemented during the first part of the 1980s. In these 'first generation'
programmes, the stabilisation of deficits was achieved mainly by cutting
expenditure and raising excise duties. Increases in income taxes and progressive
indirect taxes were far less frequent and land and property taxes were almost
While taking effect rapidly, the measures implemented led to
excessive drops in public expenditure on 'quasi public goods' and have had a
regressive distributive impact. It is now widely recognised that a more active
tax policy is needed. Such a policy should aim at improving revenue generation
and enhancing the efficiency and equity of the overall tax system. This would
permit a controlled rise in specific 'socially efficient' public expenditures,
particularly basic health, primary and secondary education and public
In the 1980s, a number of countries, including Burkina Faso,
Kenya, Malawi and Zimbabwe, initiated efforts to reform their tax systems. While
the choice of optimal tax instruments has depended on specific country
conditions, a few general principles of tax reform apply to all countries.
First, colonial type tax systems emphasising the taxation of exports (and to
some extent, of imports) should evolve towards systems which focus on income
taxes, progressive and efficient indirect taxes and increases in the prices of
such public utilities as telephone and electricity which benefit mainly the
wealthier sections of society. Second, purchases of goods and services consumed
by the poorest should be exempted from indirect taxes. Third, efforts should not
concentrate on raising marginal tax rates but rather on broadening the tax base
by reducing tax evasion and tax elusion, and fourth a low, uniform and
broad-based tax on wealth (such as on land or other assets) should be
A non-distortionary and equitably-spread increase in the tax
ratio equivalent to 2-3% of GPD is technically feasible in a good number of
African countries and would by itself add to the real flow of resources to
health care and education by about 15%, even assuming no growth and no shift in
priorities towards the social sector.
-Introducing 'earmarked' taxes or raising local ones.
The flow of tax resources to the social sector could be
augmented more readily if the new tax revenue thus generated, or at least part
of it, were directly earmarked for specific health and education activities (for
which there are guarantees that tax money is spent efficiently), or if these
taxes were raised directly by district and provincial authorities, who are
usually entrusted with responsibility for social services at the local level.
-Mobilising additional resources from households and
In sub-Saharan Africa, as in most other developing regions,
households have long borne a substantial share of the national expenditure on
health, education, water supply and so on by contributing resources (in kind,
cash and time) for the construction of health and education facilities, covering
direct and indirect costs (transport uniforms, fees, drugs etc.) associated with
the provision of public services and making payments to private providers.
Despite these already considerable contributions, user charges have been
introduced or raised on a large scale in recent years in many countries to
alleviate the burden on public expenditure imposed by the economic crisis.
However, while the small but significant contribution of user
charges to health and education budgets should be acknowledged, the negative
effects of the indiscriminate application of these charges should also be
emphasised. Indeed, in many African countries the need for health care,
education and other social services rarely coincides with the ability to pay.
The implementation of user charges can have adverse effects in three ways.
First, it may lead to a contraction in the demand for services. Second, in the
health sector, it tends to discourage the use of preventive services for which
potential patients may not see an immediate relevance (such as ante natal care
and immunisation). Third, it may adversely affect a household's ability to meet
other basic needs, such as the purchase of food, particularly if fees are
charged simultaneously for health, education and other services and if the
household's income declines. In addition to these problems, there is little
evidence that the revenue from user charges is actually reinvested to improve
the quality or expand the coverage of local services, as those who promote user
charges argue. Despite recent attempts at keeping a share of the user fees at
the local level (e.g. at the collecting facilities), in most cases revenue
merely flows to the public treasury. Moreover, user charges usually generate a
relatively small proportion of the total operating cost of the health care
sector, with gross yields at around 5%. If the costs of fee collection are
included, the yields are lower and, in some cases, negative.
However, health care, education and the provision of water
include different types of services, some clearly geared to the needs of the
poor and others catering more for middle and upper-income groups. While user
charges are not well suited to primary education, disease control programmes,
basic curative services and communal water supply, the same cannot be said of
university education and other 'income elastic' services. For the latter, cost
recovery measures would be advisable, although subsidised access for the
deserving poor must be assured. For the intermediate category, which includes
most curative health services and secondary education, selective cost-recovery
measures might be desirable, with low income individuals exempted or charged
only a nominal fee.
While direct user fees (levied on patients at the moment of
their treatment) are problematic from several perspectives, there are other
indirect ways to mobilise resources from communities (generally before episodes
of illness), such as various kinds of prepayment schemes, health insurance,
lotteries, mutual funds and the sale of produce from community fields, in
addition to their contributions in kind. These schemes have two main features
which are attractive for health-care: they shift the moment of payment, so that
individuals are not burdened with relatively more significant expenditure at the
time of illness, and they spread the burden of costs (e.g. for drugs) over a
-Increasing the volume of international aid.
There has been a considerable shift in official development
assistance (ODA) in favour of Africa over the past two decades. Africa's share
in total ODA rose from 19.4% in 1975/6 to 25.1% in 1980/1, and to 34.2% in
1988/9. However, the share assigned to human resource development has dropped,
while- debt-service obligations have often absorbed a growing share of ODA.
Greater flows are therefore needed. In the late 1980s, the World Bank estimated
that additional annual aid of about $1 billion was required during the 1990s to
halt social retrogression and ensure reasonable progress. But this flow of aid
has not materialised on any appreciable scale. Indeed the trend in the 1990s has
been towards an absolute decline in aid flows. The simplest and the most
cost-effective way to effect such aid transfer would be through debt relief,
although some increases in ODA are needed. Improvement in the quality of aid,
including substantial shifts to programme aid, greater participation in the
financing of recurrent costs and more focus on primary health care and education
and on rural water supply are also required.
What role for NGOs and private providers?
While private providers, including church missions, have always
been important in Africa, much of the market oriented literature tends to be
very optimistic about the scope for expanding private services in order to
replace government ones. Although a multiplicity of providers is desirable (and
frequently unavoidable), there are several reasons why such optimism should be
tempered. First, private sector services in Africa are typically concentrated in
urban areas, and because of their cost, cater for upper-income groups. Although
non-governmental providers (especially missions) are often located in rural
areas, and they generally provide valuable services, they are seldom able to
guarantee extensive coverage and in some cases may duplicate government
services. Second, private providers are not without costs to governments; they
often receive subsidies or preferential treatment in the allocation of foreign
exchange and typically offer higher salaries and better working conditions which
draw manpower away from the public sector. In Zambia, for instance, a teacher in
a primary school run by the Consolidated Copper Mines can earn more than a
lecturer at the University of Zambia. In countries with extremely limited
numbers of doctors, certified nurses, qualified teachers and so on, greater
privatisation would exacerbate the 'brain drain' from the public to the private
sector and the undersupply of 'quasi public goods' for the majority, leading to
harmful social differentiation. Third, quality is not necessarily an important
incentive in the private sector in developing countries. The view that private
institutions in health care and education are more efficient tends to be based
largely on analyses in developed countries. Fourth, the growing pressure for
pluralism in social service provision often gives scant attention to the
problems of central coordination and planning and the development of an adequate
'regulatory framework' for private sector activities. While playing an important
complementary role, the private sector and, to a lesser extent, the NGO sector,
are thus affected by important objective limitations. Indeed, ironically, their
proper functioning and social utility depends crucially on the existence of a
stronger, not weaker, state able to regulate, coordinate and ensure that basic
social services are provided - possibly through a multitude of public, NGO and
private providers-to all citizens of a nation.