A. Introduction and background
B. Resources, costs and data
C. Alternative sources of education finance
AFFORDING THE UNAFFORDABLE: PLANNING AND FINANCING SUSTAINABLE EDUCATION SYSTEMS IN SUB-SAHARAN AFRICA
In many African countries, in spite of public expenditure restraint under various programmes of macroeconomic adjustment, there has been a steady increase in government expenditure on education. Although public expenditures on education expressed as shares of GDP have remained more or less constant, they have declined after debt costs are taken into account.1 At the same time average public expenditures per student are declining in real terms or are stagnant at low absolute levels because of population growth and increased participation. This 'educational stagflation'2 has long been recognised, as has the necessity of finding ways of using existing resources more efficiently and of augmenting total resources allocated to education derived both from tax revenues and 'private' contributions.3 There is a considerable body of literature which sets out the appropriate policy targets,4 but severe problems remain.
However, there has been insufficient attention paid to how policy advice is implemented, and one of the weakest - if not the weakest - link in the chain of implementation is the relation between planning and budgeting, including how budgets are made. Most African public sector budgeting procedures and formats have not changed significantly since colonial times, and they cannot cope with translating short and medium term adjustment policies into practice. Although the individual citizen's 'right' to education is now universally accepted, and although governments are obliged to respond to the 'social demand' of their populations for education, it cannot be denied that the provision of education services absorbs real resources which are limited at any given time and which can only achieve in most African countries a modest growth over time. There has been a tendency to put broad educational policy objectives on the one hand and the economic planning and management of resources on the other into two separate compartments, so that while there is no shortage of analysis of what needs to be done, the means of achieving given objectives are often unspecified. This has led to unfortunate and self-defeating tensions between those who propose policies in international financing agencies and in governments, and those who must manage the implementation of these policies.
This paper explores this problem and some of the solutions proposed to overcome it, with particular reference to Sub-Saharan Africa, although many of the observations and approaches are equally applicable to South Asian, Caribbean, and Latin American countries, as well as some of the former Soviet republics. It is essential that there be greater awareness of budgetary, fiscal and macroeconomic issues in the discussions of the critical shortage of finance in most African education systems. In this respect governments have a crucial role to play in the process of change even if in some aspects the 'market' will succeed where government planning has failed.5
The first part of this paper sets out briefly the basic issues, and discusses some current policy recommendations, including the introduction of school fees and the private provision of schools. The second part is concerned with approaches to strengthening and/or reforming planning and budgeting for education in African countries in order to provide a better base for the implementation of policies. These include the replacement of annual incremental planning and budgeting systems with approaches which may be more appropriate to current problems. I suggest that many attempts to undertake necessary reforms have not succeeded because they have been limited to interventions in the education sector ministries without reference to government budgeting and administration systems as a whole. Reforms should also take full account of the need to strengthen a potentially beneficial relationship between the state and the private sector. With these improvements better use can be made of external assistance which has, I argue, not always served those countries which benefit as well as it might have. The objective of the suggested changes is to enable countries to use their limited resources better and avoid stop-go educational development policies in order to achieve the capability of providing education which is both sustainable and affordable.
There are a number of factors which influence the planning and financing of sustainable education systems, and they include-,
(a) demographic factors(b) national economic performance, including the structure of the economy, degree of industrialisation and rural/urban economic activity, unemployment, the distribution of income, etc
(c) external assistance and levels of external debt
(d) patterns of previous provision and 'social demand'
(e) external advisers and external 'models'
(a) Demographic Factors
The growth of population in most African countries has resulted in an accelerating demand for basic education, and, as a natural consequence, for higher levels of education. In general this growing demand has exceeded the rate of growth of resources available to satisfy it. Although the ratio of the 0-14 age groups to the 15-64 age groups in low income countries is predicted to fall (from an average of about 40-50 per cent in 1989 to 30 to 40 per cent by 2025), implying an increase in the size of the population of working age, the ratio of taxpayers to non-taxpayers is likely to remain much lower than that found in developed countries for some time to come, while the structures of the growing education services in less developed countries which are mainly financed out of tax revenues are arguably comparable to those of the more developed countries. At the same time the rate of urbanisation will effect the delivery of education because of the different characteristics of urban schools.
(b) National Economies
The level of industrialisation in different countries is an important aspect of the nature of the relationship between education and training systems and the employment markets. In this sense 'levels' of industrialisation refer to a continuum with pre-industrial countries (Africa, South Asia) at one end, moving through light and heavy industry to high technology industry at the other end. For example, countries such as Korea with high and growing quality control in industry require a very different approach and education content from countries with negligible industry and little drive for quality control. Countries dominated by low skilled rural employment differ from those with high skilled urban employment. Countries where school and university leavers are likely to be employed within a reasonable time after graduation have very different relations between education systems and labour markets from those with high rates of unemployment. At all levels of development trainability is a prized factor among employers, but in more industrialised countries basic school education and trainability are extremely important: without these attributes new entrants to the labour force can quickly become marginalised.6 In pre-industrial countries marginalisation may take place irrespective of education and training simply because of insufficient employment opportunities in relation to the number of qualified applicants.
The issue of job creation is crucial. Increased public investment in the expansion of education systems may mean less resources devoted to job creation: the Pacific rim countries have demonstrated the strong potential role of governments in this area.7 Expansion of education in economies with restricted employment opportunities may not always be the most appropriate use of government resources.8 Industrial growth in Africa will be limited, and while formal education will play a direct part in this growth, it will be also significant in terms of indirect influences such as on the redistribution of income,9 and as a base for specialised training. Although some10 insist that the state should not play a direct part in providing finance for training, experience from Asian countries suggests that there is a case for the state doing so.11
(c) External Assistance and Public Debt
Countries in Africa, South Asia, the Caribbean and Latin America which have recently experienced a decline in their economies suffer from falling or stagnant tax revenues, and therefore from a decreasing ability to provide additional public domestic finance for growing public services.12 In order to support the increased claims on public spending, in some countries foreign aid has assumed such significant proportions that education systems are largely supported by it.13 The data conceal the volumes of external aid supporting education systems, which may be proportionally larger than in other sectors, and also conceal the varying applications of aid within different parts of education systems.
Donors have put considerable pressure on African countries to accept financial and technical assistance to develop their education systems.14 Yet the relation between external funding and domestic capacity to pick up the resultant capital repayment and interest costs as well as the associated recurrent and capital replacement costs has consistently been ignored, or if not ignored, it has been assumed that economic growth will take care of the problem. There have been many attempts to quantify the finance gaps, from the millions of dollars needed in the 1960s to reach 'takeoff' to the UNICEF attempt to show how much is needed to provide basic education for all.15
In principle, the argument runs, the combination of the extension of education provision and other externally financed investments will accelerate growth and development and also fulfill basic requirements and entitlements, justified on moral grounds. However, whatever the longer term effects, the expansion of education systems has demanded a rapidly growing share of fiscal resources while tax revenues cannot keep pace with the increased demands. The resultant decline in the quality of public provision has meant that it is no longer possible to take indices such as examination results and enrolment rates as indicators of returns to this public investment. For example, a common structure of an education system is a primary-secondary-tertiary cycle which may cover 15 years altogether. If a six year primary education cycle is characterised by illiterate school leavers it must be concluded that there is something inappropriate about such a system. If schools lack books, equipment and decent buildings and furniture, the reasons for their very existence is in question.
Encouraged by the Jomtien conference, most of the current literature which considers education finance appears to start from the premise that schooling for all (SPA) by the year 2000 is an axiomatic objective.16 There are parallels with the commitment to Universal Primary Education (UPE), which was promoted often with little real regard for resource requirements. The global objectives are endorsed by governments, which are encouraged to aim for them and are promised external finance to support their aims. Governments themselves seem to have had little real confidence that they could make sufficient resources available. There should by now be sufficient experience to realise that these targets are not attainable, and, more importantly, that trying to attain them through heavy doses of foreign aid leaves major problems of sustainability.
The amount of public finance available for spending on education has been severely affected by the volume of the interest costs of government debt which have first claim on fiscal revenues. However, over the period of adjustment in many countries the volume of government expenditure has risen slowly in real terms. Social sector spending has also risen,17 but because of the requirement on governments to service debts the rate of increase of public expenditure is limited.
The significance of the combined influences on education of the growth and distribution of population, recession, and rising interest payments, has not only reduced the extent to which tax derived public expenditures can support the growing demand for services: they have also affected the availability of private finance, that is, the total resources available. On the government side, the debt burden combined with population growth has rendered many governments' attempts to cope with the social sectors a Sisyphean task. One of the difficulties which face policy makers and sectoral aid donors in Africa in the near term is how to judge the debt problem: much foreign aid would be unnecessary were debt to be cancelled because it would allow more domestic finance to be reallocated for education.18 On the private side, as public sector provision, which is a function of economic capacity, declines, households are forced to contribute more of their own resources through fees and other types of tax, though in many cases there may be very little surplus in the household budgets.
(d) Advisers and External 'Models'
Educators from developed countries have generally not accepted that education systems may develop through 'stages', and that these 'stages' reflect the capacity of a country to provide a reasonable quality of education to its citizens within its resource constraints.19 A aucial factor in many countries has been the role of foreign advisers whose advice has been supported by large sums of foreign aid.20 Their influence has been compounded by the demonstration effects of the education systems of developed countries: these systems evolved over many years largely on the basis of their close relationship to employment markets. Countries which are 'late developers' have imported wholesale foreign techniques and institutions, including examination systems which have become the main linkage between schooling and restricted employment markets, dominating both.21 Although there were many who were well aware of the growing tendency of foreign aid to promote the 'western' norms of the quantitative evaluation and assessment of extrinsic educational characteristics,22 there has over the years been a tendency to promote the expansion of education systems by making them more complex and 'modern' without concomitant attention to assessment reform. This is analogous to the heavy promotion of better seed varieties and husbandry practices for peasant farmers while ignoring the issues of crop pricing and marketing.
(e) Patterns of Previous Provision
One of the main constraints on reform policy is the expectations and aspirations generated by the provision of education and training in the recent past. The determinants of this provision are complex and include cultural and social, as well as political and economic, factors. This paper does not explore this aspect in detail, partly because of its complexity, and partly because there are significant variations between countries. A principal manifestation of popular aspirations to education is in the sensitivity of political authorities to current 'social demand' and the resulting reluctance of governments to take difficult decisions on the basis of the long view rather than of short-term concerns. There are wide differences in patterns of previous provision between countries. Some have had large university sectors (at least in terms of their budget shares), while others chose not to develop post-primary education. Attitudes to girls' education vary between countries, as do local communities' attitudes to participation in schools. The status of teachers also varies. Some countries have over the years developed strong elements of private provision. In response to 'social demand', existing government budgeting and planning further consolidate short term incremental planning, reflecting a seemingly inexorable and immutable progress on the basis of past trends and views. The changes that need to be made are in many cases so large that it can be difficult to see where to start and how they can be implemented. Major differences between countries in this respect dilute the value of 'global' prescriptions.
I have, in this introduction, sought to suggest that there is little reason to assume optimistically that the approaches of the last two decades will be appropriate for the next decade, a position which is directly contrasted to that of the Jomtien conference and the view that resources will be made available through the combination of efficiency measures and foreign aid in a period of growth in the nineties. The 1990s are now a third completed, and there is little evidence that the necessary steps are being taken quickly enough, if at all. The cumulative effect of the factors summarised in this section is to force the conclusion that the environment for education development in many countries is hostile to progress, and likely to remain so.
The main differences between education policies which were most widely advocated in the 1970s and the first half of the 80s, and those now widely proposed by international agencies and others, arguably relate to the role of the 'market'. Much of the literature on reform is based on assumptions about 'the market', and on how reducing the role of government can create competition and therefore efficiency. Cost recovery and privatisation are often regarded not only as means of augmenting resources, but also as sufficient conditions for greater efficiency. There is at the same time a general acceptance of the desirability of Schooling for All (SFA), just as there was of Universal Primary Education, and (SFA), is widely enshrined in public education plans. These are essentially and inevitably long term objectives and do not address the problems which we currently face.
This paper tries to steer a way between the Scylla of the market and the Charybdis of Jomtien. I express scepticism both about market solutions and about the feasibility of significant expansion of education systems, at least in their present forms. While the Jomtien objectives are desirable in principle, the first step must be to do what we can to ensure that governments are capable of managing the reforms in many African countries. This in itself will take up the rest of this decade.
It is important to distinguish between educational expenditures and educational costs, and to relate them both to a given product. Analysis of education systems can be misleading because of the failure to make these distinctions. This is exacerbated by the inadequacies of data. In that policy prescriptions can have wide-ranging effects, it is most important that they are made on as full an information base as possible. The purpose of this section is to define briefly the appropriate terms for subsequent analysis.
(a) Education Expenditures
Budgeted and Actual Expenditures. Expenditures are actual money outlays. It is necessary for the analyst to distinguish between budgeted expenditure on the one hand and actual expenditure on the other: actual expenditures include household and other non-government expenditures. A common error in education finance analysis is the assumption that budgeted expenditures equate to actual expenditures. This is partly because the standards of public accounting in many countries are very poor, and only budgeted expenditures are available in the public domain. Publication of actual expenditures, which are often incomplete, is usually at least two years after the end of the fiscal year,23 and budget estimate books commonly have three types of analysis: actual expenditure (present minus two years), revised budget estimates (last year) and budget estimates for the current year. In that the expenditures can be as much as 30 per cent over or under the estimates, the use of budget data as a proxy for expenditure is particularly misleading.
Total Expenditure Estimates. Another common feature of education analysis is the presentation of tables showing 'total expenditure on education', which normally, of course, show only total government (often budgeted) expenditure, sometimes of the Ministry of Education only. Firstly, household expenditure is a major component of total expenditure: at the primary level, for example, household direct expenditure on education can equal and indeed surpass government expenditures. Secondly, most education budget estimates fail to include the interest costs of capital and project expenditures, whether financed by loans, grants, or 'own' funds. Thirdly, in many countries external assistance, which may not be included in the government budget, can substantially increase total expenditures. Fourthly, it is well known that public expenditures on education and (more so) on training are made in many other ministerial sectors apart from education. These omissions or inaccuracies can result in substantial underestimates of actual expenditures.
(b) Education Costs
While costs to the economist are a measure of what has to be given up to achieve something, to the accountant they are money outlays (except for depreciation). Economic costs are made up of direct and indirect costs. Direct costs are usually money expenditures. Where average, or unit, direct costs are analysed, they often refer only to public expenditure divided by the appropriate unit, usually enrolments, at a given point in time for a given level of educational provision. The total direct costs of education exceed the levels of expenditure by governments, and the excess is borne by households and firms. To define the direct costs it is necessary to specify at the same time what is actually being purchased.
The difference between the actual expenditures and the (estimated) direct costs of providing a reasonable quality and quantity of education is the level of under-funding.24 For example, the costs of production of books can be more or less estimated, a 'reasonable' level of teachers' salaries can be proposed based on various measures and indices, building costs are usually known within, say, a 20 per cent variation for any given design, and the maintenance of pupils, teachers, materials, equipment and buildings can be estimated within an acceptable level of tolerance. If we then buy what we understand as 'education' but in fact receive something different, we have purchased a different product, and its costs are not those of 'education'. If we say that we are buying 'education' we expect that what we are buying will serve its intended purpose. In this sense education expenditures do not equate to education costs in this particular time period.
Total cost, however, is a wider concept, and includes the costs to society as a whole, usually termed 'social costs'.25 The measurement of what has to be given up to purchase education is complex. The most widely analysed indirect cost is the 'opportunity cost' to individuals and households, that is, the alternative uses to which they could have put the resources had they not invested them in education. Less widely analysed are the opportunity costs of government expenditure and of external assistance. External assistance can promote inefficiency, as in the case, for example, of 'tied aid', where recipient governments are obliged to purchase goods and services from the 'donor' country even though these goods and services might be more expensive than suitable alternatives available from other sources. Another social cost which is little analysed the cost of tax finance. Raising and administering taxes require expensive bureaucracies. In most countries tax systems discriminate against the poorer sections of the population, or they encourage the inefficient use of resources. Finance raised through such systems has significant social costs. While we are mainly concerned with government expenditures and the planning of the education sector, it is necessary to bear in mind the nature and scope of costs of provision of education, as we will see later.
(c) Data
The education system analyst requires three categories of basic data, 'physical' data relating to enrolments and other quantities, financial data relating to costs and expenditures, and economic data relating to the economy as a whole. Interpretative data are derived from these basic data, such as enrolment rates, average expenditures and other ratios
National Income Accounting. At the apex of the data structure national income accounts are frequently drawn upon for policy prescriptive purposes. The unreliability of national accounts is well known, but despite this they continue to be widely used, for example, in comparing government education expenditure as a ratio of GDP between countries, and in advocating that expenditure on education reach a given level, say 5 per cent, of GDP. In these cases the numerators and the denominators of the ratios are so unreliable that the resulting ratios are of little use, especially for comparative purposes.26
Public Budgets. The next commonly used aggregate financial interpretative data are educational expenditures expressed as proportions of public expenditure. The most frequent failure in considering these data is the need to distinguish between discretionary and non-discretionary expenditure. Interest payments on government debt are a priority obligation in government expenditure. Sectoral budgeted expenditures are therefore secondary claims, and government has discretion on allocation. Total discretionary budgeted expenditure is therefore total budgeted expenditure after provision for local and foreign government debt repayments is made.27 Many analysts cite ratios which use total budgeted expenditure as the denominator, and it is not always clear from published statistics whether debt obligations are included or not in the total recurrent budget estimates. Some countries have been expected by international institutions to raise sectoral shares because of confusion of discretionary and total expenditures. If the 'gross' ratio of education expenditure to total recurrent expenditure (including debt repayment) falls over time it is difficult to determine whether the fall is because of increased debt repayments or because government policy is to reduce expenditure on education. The debt issue cannot be overemphasised and is discussed further below in relation to capital budgeting and foreign aid.
The next key issue is the composition of education budgets. Different countries include different items, and Ministry of Education budgets do not include all budgeted expenditures on education in many countries, particularly those in which education management responsibilities are devolved on local government or federal state ministries. It is often difficult to determine the total public budgeted expenditure on education because of the variety of data sources that must be consulted.28 Published and easily available data within countries rarely provide information on education finance which is useable for detailed analysis. Few government Statistical Abstracts and Digests include budgetary data.
Capital Budgets. A major data weakness is capital budgets, which are often very misleading since many contain recurrent elements and omit foreign grant and loan contributions. The servicing of foreign loans is to be found in a consolidated budget, and does not form part of sectoral budgets, although there is a case for them to be recorded there. Capital budgeting is discussed later.
Educational Cost Benefit Analysis. A final point should be made about rate of return analysis. The use of internal rates of return to education has been vigorously promoted over the last fifteen years, and has had a great influence on education policy, although the reported historic 'high' rates of return have not been reflected in many cases in overall national economic performance. The technique has many serious flaws, most of them being well known (such as its inability to take quality into account29, and the equation of differences in earnings with the effects of education on productivity), some less so (such as the true extent of costs and benefits, including social costs and benefits30, and externalities31). Internal rate of return to 'education' calculations are without doubt subject to serious upward and downward bias (which may not, contrary to the suggestions of some writers, cancel out), yet the findings, surprisingly calculated to a single decimal point, are invariably used to justify the emphasis of public investment on primary education. The use of cost benefit analysis for education policy making is not the subject of this paper, but it has had important effects on education planning and budgeting. In reality, it takes many years to redirect public expenditure from one level of the education system to another, during which time relative internal rates of return may alter between levels for several reasons, including their dependence on average wages of school leavers which may change rapidly in a short time.32 As such a redirection may imply the introduction of fees and private education, of which there is little experience in many countries, there are major political and social issues at stake. While cost-benefit analysis in principle provides important policy data, in practice the potential for significant error suggests that evidence of internal rates of return to schooling should be used with circumspection, if at all, particularly in countries with low school quality.
Although it is by now universally recognised that public education systems in most African, South Asian, Caribbean and Latin American countries are severely underfunded in relation to what they are trying to achieve, the mechanisms for improving efficiency are not in place. Progress can only be made if planning and budgeting are improved, and this cannot happen while ancient and inappropriate government financial and planning approaches and systems remain. In general, governments must follow one or both of two types of policy to improve the quality and level of provision of education. On the one hand existing resources must be used more efficiently, and on the other resources must be augmented. A necessary condition for both must be the development of realistic budget and expenditure control systems for all government expenditures which work hand in hand with better planning.
The necessity of a public subsidy to education is justified by the presence of constraints on private credit (individuals often cannot borrow to finance education); imperfections in the availability of information (parents may systematically underestimate the value of education because of lack of information); externalities (such as the apparent relation between length of primary schooling and reduced fertility rates); and the public good aspect of education. Public goods are goods which, because they cannot be withheld from one individual without withholding them from all, must be supplied communally. Access to basic education might qualify as a public good. The presence of any of these conditions indicates 'market failure', and as they are all to a greater or lesser degree applicable to education, subsidies to the provision of education services can be justified on economic grounds. At the same time, funds for education, whether from tax or non-tax sources, are not infinite, and will always be rationed. The case for providing a fully subsidised service is thus undermined by the inability of net fiscal sources (ie taking also into account direct and social costs incurred in providing subsidies out of taxation) to satisfy the social optimum, taking into full account the various external benefits to education. Even were sufficient funds to be available to support a fully subsidised system, bureauaatic inefficiency in unaccountable education services leading to 'government failure,'33 would have to be taken into account. A balanced approach is therefore needed.
Efficiency. Much of the literature considers the option of reducing unit costs.34 But improvements in efficiency can take place, with or without reductions in unit costs. If we use the average cost per student as a measure of efficiency, then clearly costs may be reduced by reducing the quality of the product. It is difficult to see how average costs can be reduced in many African countries without further deterioration in the system. Some components of average expenditures should be reduced, but should result in transfers to other components in order to increase average expenditures per pupil or teacher. This is contrary to the view of those who believe that the greater role of the market permits governments to save money.35 The problem for the public sector is to raise average expenditures within the constraint of existing total public expenditures: and for the education system as a whole, total and average expenditures must be increased within the constraint of net additional private resources.
Embedded in the proposition that unit costs can be reduced is the inescapable need for the scope and volume of publicly financed activities to be reduced. 'Efficiency' measures will yield some transferable funds in the form of available cash. This is important. For example, many education projects have had as their stated aim the increase in pupil-teacher ratios in order to reduce the total number of teachers, the resulting 'savings' then to be applied to other parts of the sector. In reality, under the usual budgeting systems, what savings that are made are rarely identifiably transferred in the form of extra cash for, say, books. They are translated into a reduced rate of growth of the education budget: increments on non-salary charges are held at the 'normal' rate. In the context of extreme underfunding it is not clear how significant such transfers will be. Sources of additional funding will tend to be found outside the normal tax system in the foreseeable future.
The 'liberalisation' of the education market, the encouragement of private schools and the formalisation of cost-recovery schemes are expected to go a long way to bridge the finance gap. Most countries expect a growing proportion of total expenditure on education to come from non-fiscal sources. The main point to be made at the policy level is how far user fees can and should augment, and how far substitute for, public resources, assuming, of course, no deterioration in the quality of the product. There is relatively little evidence on which to base confident predictions of the extent to which expenditures on education can be increased through allowing a market in education service provision to develop. The development of such a 'market' would be through (a) the introduction and expansion of 'user fees'; (b) the expansion of the private education sector; and (c) the greater use of non-government sources of specific goods and services.
(a) User Pees end 'Cost-Sharing'
It seems to be widely accepted that subsidies should be focused on primary education, and that the higher up the system the lower the subsidy should be.36 However, because of higher costs, the ratio of subsidy to private contribution generally increases the higher up the education pyramid a student climbs.37 Whereas direct parental contributions finance up to half or in some cases even more of the total expenditure at the primary level, this share falls rapidly at the public secondary level. At the university level direct and indirect parental contribution as a proportion of total expenditure is often very slight. In order to reduce the burden of subsidy at higher levels many countries have introduced compulsory fees for secondary and tertiary education.38
User fees may replace government subsidies to education, they may augment them, or they may partly replace and partly augment at the same time. Most proponents of fees assume that fees should and will augment total expenditures. In general it seems that the approach proposed by Thobani39, which has influenced the thinking of international agencies involved in education, is most widely accepted: fees should be increased so long as there is 'excess demand' for the service. As I have noted, however, deterioration in the education system changes the product and 'excess demand' has little meaning. Excess effective private demand for 'education' as a service which will clearly benefit its 'users' by enabling them to learn will not be observed because of this. In a number of African and South Asian countries parents are withdrawing their children from school because of poor quality and because they do not consider schooling relevant to their needs.
Clearly, where there is substantial underfunding combined with deteriorating quality, gains from increased user fees may generate resources for expansion and quality improvement, and add significantly to total resources expended on education. However, it is not at all evident that sufficient attention is given to substitution effects. If household budget ceilings are fixed, the requirement to pay fees could well mean that money which had previously been spent on children and schools (for example, on books and construction), would be switched to fee payments, and parents would expect a greater level of public provision to provide what they themselves had previously provided. This means that no augmentation of total expenditure on education would take place. Similarly, little is known about the opportunity costs of fees paid by the 'rich' for higher education, though much is written about the need to lower taxes to promote investment by the 'rich'.
It is thus necessary to distinguish between compulsory fees and voluntary contributions, and to understand the degree of substitutability between them. Where compulsory fees are concerned, a distinction must be made between fees collected by the school and used in the school, and those collected and administered centrally. Compulsory fees are in effect 'earmarked' taxation, that is, taxes collected for a specific purpose,40 where they are for compulsory education. As I have noted, taxation has its own associated costs, and therefore the result of charging fees which cause households to transfer expenditure from direct voluntary contributions to compulsory fees could result in a net decrease in education expenditure due to the costs of fee collection and administration.
In addition, many fee proposals may not have sufficiently considered the effect on households of the combination of user charges and the tax system. The collection of fees and non-fee contributions for compulsory education outside the tax system, either through coercion or through deliberate starvation of subsidies forcing up private contributions, can increase the regressive nature of the tax system.41
Two broad justifications are commonly advanced for 'cost-sharing' policies. The first is that governments have insufficient revenue expenditure to finance education services fully; and the second is that a decreased reliance on 'government' revenues will promote competition and therefore efficiency. In so far as 'cost-sharing' in education has any meaning at all it refers to the sources of finance for education. Sources may be discretionary or non-discretionary for both household and government budgets. Both governments and households have limitations on what resources they can budget for any given activity. For individuals and households there are three types of 'cost-sharing': (a) voluntary contributions; (b) obligatory charges which do not go into government revenues but are retained, for example, at the school; and (c) obligatory charges which go into government revenues. Voluntary contributions may be correctly considered as cost-sharing while obligatory charges must be considered as taxes. However, where government allows for a certain level of 'voluntary' contribution by deliberately reducing subsidies, even voluntary contributions become obligatory.
The basic question is what affects the relative levels of voluntary and obligatory expenditures. Taxes and compulsory fees paid by individuals are obligatory (non-discretionary) they must be paid - and hence there is no essential distinction to be made between them. Where fees are used directly for a specifically identified purpose, unlike most taxes which go into a common pot, they may be considered as earmarked, but they are nevertheless obligatory. 'Cost-sharing' does not therefore signify a division of financing responsibility: revenues still derive from citizens. Rather it signifies a redistribution of financing shares.
The basic relation between public expenditures on education on the one hand and direct payments households may make apart from their normal tax obligations on the other is that the payments that households make are generally residual, that is, after public subsidies are taken into account citizens are asked to-make up any shortfalls. There is therefore a clear relationship between public sector efficiency and the level of 'private' contribution: parents may be required to pay more to support government or private sector inefficiency. How can I, as a parent, influence the costs of the education which I am obliged to cover, directly or indirectly, whether for public or private education, if I want my children to be educated? Can I say 'reduce the number of subjects so at least some are well-resourced'; or 'in our school we can't afford all the teachers we are allocated'; or 'change the school year and modify the examination system to fit into the fishing calendar so my children can help in the fishing season'? I can't, of course. The whole construct of the 'market' is that it creates competition and choice, but it is obvious to nearly every parent, particularly those without much money, that in reality there is not much choice, and hardly any at all where the determinants of the costs of education are concerned.
There is a major need for research into the relation between the costs of schooling and how people can pay them. At the household level expenditures may be made from current income, from borrowing, and from sales of assets. Nearly all surveys of household expenditure assume expenditures to be from current income. But if the sum total of current income is insufficient to cover household expenditures, whence comes the income to make up the difference? It must come from debt or from sales of assets. If it comes from other members of the extended family it may come from current income, or from debt, or from sales, but in aggregate the total incomes and expenditures must balance.
Assets such as cows or taxis earn income. It may make sense to sell them if they finance higher yielding assets, rather than if they finance consumption. If households sell assets to finance education, assuming education to be investment and not consumption (many, if not most, surveys treat education expenditure as consumption expenditure), it is, in effect, a switch of investments out of cows or taxis into human capital. Thus, where individuals are required to purchase education, it is necessary to understand the sources of finance used for the purchase. If I believe my children will earn more from having attended school, the additional future earnings might enable me to buy two cows or two taxis, and my short-term hardship might be justified. If in fact they do not earn sufficient to have made my sacrifice worth while, I have made a serious loss. Moreover, society has also made a loss.
In the past budget deficits were financed by government borrowing and foreign aid. In principle there is nothing wrong with a deficit as long as people are willing to finance them. As governments are no longer able and/or willing to finance deficits, households and firms are being asked to do so. If individuals wish to receive education and an academic certificate they will have little choice but to pay increasing levels of fees, regardless of the efficiency of government provision. This is the logic of the Thobani 'rule'.
As I have mentioned, the argument is that at the primary level parental contributions should in general augment public resources, while progressively at the higher levels they may to some degree substitute for public resources, but in order to redirect the resources to basic education. In reality, this is hard to achieve. Higher government expenditure on primary education might, for example, enable poor households to release money used to finance education for other equally important uses. Furthermore, in countries where equity considerations are important in policy making, it is possible that focused subsidies such as scholarships will reduce net revenues from fees, as the costs of subsidies are balanced against the extra revenue from fees.42 Where fees substitute for public finance, it is assumed that the public finance thus released will be returned to the education sector in other areas. This, as I have noted, is one of the basic justifications underlying fee proposals such as those found in the cited World Bank literature,43 but it is by no means clear from this literature that it actually happens. I suggest that it is unlikely to happen in the absence of appropriate planning and reformed budgeting systems.
(b) Private Educational Provision
Much of Africa's 'modern' education system up to independence was private in the sense that it was run by missionaries. The secularisation of education and the perceived need for nation-building, as well as in some cases political activities of churches, led to an almost total take-over of educational institutions by governments. Missionary schools were not profit-making, but aimed for self-sufficiency, depending on local and foreign contributions.
The reasons for encouraging the development of private schools in less developed countries generally lie with the inability of governments to provide sufficient student places from tax finance. Like fees, it is a way of capturing more private finance outside the tax system. This is deemed to be more efficient in the sense that it avoids the costly public bureaucracy necessary to administer the public system. However, it is necessary to take into account the costs of regulation.44 At the same time private schools, particularly in that they operate for the most part within education systems in Africa which are geared to getting pupils through examinations, may, if left sufficient flexibility, serve as centres of innovation, even to the point of shortening the length of the primary or secondary school cycle with no loss of examination success. The European experience, particularly in Scandinavian countries, provides interesting examples of this phenomenon.
There is an important distinction to be made between profit-making and non-profit private schools. In the case of private profit-making schools there is no reason to suppose an equally proportional relationship between rises in household spending on schooling and total expenditure on education because of the profit which school owners take out. It is thus by no means axiomatic that the replacement of public schools by private schools will result in an equal augmentation of resources, particularly when regulation costs are taken into account.45
The privatisation of education may not necessarily favour the better off by the creation of elite private schools access to which is restricted by price, as is often believed. There may be significant incentives to compete for subsidised school places, which are allocated on entry requirements which include academic performance as measured by examinations. Examinations have distorted education systems and equated examination passes with education. To the extent that there is a relation between examination success and relative social advantage, examinations become a rationing device for future study, and favour the better off in their search for publicly subsidised school places. More importantly, whatever the economic status of students, those with lower academic achievement, which is not the same as saying those with less academic ability, may have less access to good schools.46
There is undoubtedly an important role to be played by private schools in African countries. How important remains to be seen. Private schooling should not be seen as a panacea, and its potential as a future demand on government obligations under multi-part systems must be borne in mind: where private schools may suffer financial problems there may be strong calls for government help.47 Private systems are often subsidised directly or indirectly, and may account for a proportionately larger percentage of total expenditure than of total enrolments. Indeed, the real issue is how subsidies to 'private' schools are operated, as in most countries they operate through the tax system or directly though grants. In many ways it is unhelpful to be constrained by the term 'private', and perhaps we should consider the issue more in terms of diverse forms of state funding. To realise the possibilities that exist significant changes would be required in how government budgets for education are made up and regulated.
(c) Alternative Provision of Goods and Services
Many countries have now contracted the publishing of school books to private sector publishers, and are contracting teachers and others to work on curriculum development, in contrast to previous reliance on Ministry of Education units to do this work. There has been less experimentation in some of the other parts of the education service, particularly in high cost areas such as the training of teachers. As teachers make up the single greatest cost component of primary and secondary education, and, at the same time, in nearly all countries perceive themselves to be underpaid and working in poor environments, a prime concern must be to improve their conditions. To achieve this they must (a) have more capital to work with (better buildings, books and equipment, etc); and (b) in most countries have improved salaries. Yet because teachers are in most cases only supplied by government which is constrained by public expenditure ceilings, neither of these improvements are likely to be achievable in most Sub-Saharan African countries in the near future.
In addition, the cost of teachers to schools is fixed by centrally determined norms, expressed both in salary scales and levels of qualification. In many countries the problems of poor quality in schools are also found in teacher training institutions, and the professional capability of teachers suffers. Put another way, although teachers may possess appropriate paper qualifications attesting to years of training, they are not axiomatically well-trained teachers. This would explain to a large degree the findings of researchers which suggest that the level of training of teachers may not be significantly associated with pupil performance in many countries, particularly in primary schools.48
Any discussion of education fees and private schooling in Africa is likely to end indeterminately as most countries have little experience of the complex interaction between household and government expenditures and the quality of educational provision. The relevance of the issue to public budgeting is the need for planners to determine the amount of private finance available to supplement government finance at all levels, and to determine the effects of the distribution of subsidies on the availability of education to all income groups. The foregoing brief discussion of approaches to augmenting the resources which are available to support the development of educational systems highlights the problems facing planners in the implementation of the proposed 'new priorities'. While several key policy targets are proposed, they are often based upon assumptions that have not yet been tested and which have a number of practical difficulties attached to them. Countries should not accept such policy advice without giving careful thought to how it should be initiated and tested. This involves setting place the appropriate mechanisms to allow them to do so effectively.
