User Payments for Basic Education in Low-Income Countries

Contributor Notes

Authors’ E-Mail Addresses: hillman@mail.biu.ac.il; ejenkner@imf.org

Insufficient resources and inadequate public expenditure management often prevent governments in low-income countries from providing quality basic education free of charge. User payments by parents are an alternative means of financing basic education. This paper assesses how user payments affect educational opportunities and quality of education for children of poor families in low-income countries. Conditions are identified under which user payments can or cannot improve educational outcomes. User payments, whether taking the form of compulsory benefit taxation or voluntary user fees, are a temporary solution and second-best compared with free-access, publicly financed quality education that is consistent with macroeconomic stability.

Abstract

Insufficient resources and inadequate public expenditure management often prevent governments in low-income countries from providing quality basic education free of charge. User payments by parents are an alternative means of financing basic education. This paper assesses how user payments affect educational opportunities and quality of education for children of poor families in low-income countries. Conditions are identified under which user payments can or cannot improve educational outcomes. User payments, whether taking the form of compulsory benefit taxation or voluntary user fees, are a temporary solution and second-best compared with free-access, publicly financed quality education that is consistent with macroeconomic stability.

I. Introduction

Ideally all children, including children from poor households in low-income countries, should have access to free and publicly financed basic education. In many countries, revenue constraints or inadequate public expenditure management systems prevent governments from providing these basic services. As a result, user financing is being used to supplement public funds or to set up community schools where no public schools are available.

In theory, a strict distinction needs to made between compulsory user financing and voluntary payments. A government can oblige parents to pay compulsory school fees to finance the basic education of their children. In that case, if both user payments and the education itself are compulsory, then schooling is tax financed, according to a benefit principle. Alternatively, user payments can be voluntary, with noncompulsory school attendance and children being barred from school if the user payments are not made. Since payment is voluntary, such user payments are then user fees rather than taxes that reflect compulsory payment.2

Whether compulsory or voluntary, user payments can take different forms. Conventionally, school fees might pay for salaries of teachers and administrative staff, learning and teaching materials (such as pencils, textbooks, etc.), and maintenance of schools. In many low-income countries, user payments take the form of payments in kind, such as food for the teacher or labor provided for construction or refurbishment of the school. Compulsory school uniforms also entail a form of user payment, since parents are obliged to buy the uniforms if they want their children to attend school. User payments can also be used to improve the quality of education provided by public spending. Supplements to the teacher’s government-financed salary may improve the motivation of teachers and allow the recruitment of better-qualified staff.

However, requiring parents to pay for children to attend primary school, by law or by failing to provide adequate resources, is generally considered undesirable. For example, the nongovernmental organization Oxfam, in a briefing paper titled “Education Charges: A Tax on Human Development” (Oxfam, 2001), takes the position that “The evidence is indisputable. Success in achieving universal basic education depends on education becoming affordable to the poor, and this requires the abolition of education charges” (p. 15). There are two aspects to such a position. If education payments are compulsory taxes for compulsory education, poor parents may not be able to afford the taxes. If education charges are voluntary user payments, children can be excluded from school when parents are unable or unwilling to pay. In either case, the result is the same: children fail to receive basic schooling.

This study assesses how user financing, both compulsory and voluntary, affects the basic education of children in low-income countries. The study focuses on children from poorer households who would choose to attend quality, free-access government schools if such schools were locally available. The study is not concerned with high-income households in poor countries that can often afford and choose to send their children to expensive private schools financed through user payments.

The study proceeds as follows. Section II reviews school enrollment and primary completion data for low-income countries and distinguishes supply- and demand-side influences that may constrain educational attainment. Section III sets out the problems associated with user payments. Ancillary benefits derived when user payments finance schools are described in Section IV. Section V reviews evidence on the use of user payments and presents case studies demonstrating applications of user payments. Section VI concludes and sets out policy recommendations.

II. Schooling in Low-Income Countries

A. School Enrollment

At the beginning of the twenty-first century, school attendance for children in low-income countries is far from universal. There is also a clear gender bias: two-thirds of the children not attending school are girls. Primary completion rates are low and educational standards are often inadequate: teachers are poorly trained and paid and, as a result, are often not sufficiently motivated. Classrooms are overcrowded and basic teaching resources such as textbooks, blackboards, or pens and paper are lacking.3

The need to address these inadequacies is reflected in the Millennium Development Goals (MDGs)4 of eliminating the gender gap in primary and secondary education by 2005, and achieving universal primary education worldwide by 2015. The achievement of these goals, however, requires substantial improvements over the prevailing trends in many low-income countries.

Table 1 shows changes in net and gross primary school enrollment rates between 1980 and 1996/97.5 The data highlight significant regional disparities, with the lowest enrollment rates in sub-Saharan Africa and South Asia. Primary completion rates are an even more accurate indicator of educational attainment than enrollment, since enrollment does not guarantee completion. Furthermore, literacy surveys demonstrate that many adults who have completed fewer than five or six years of schooling remain functionally illiterate and innumerate.6 Table 2 provides an overview of primary completion rates by region in 1990 and 1999.

Table 1.

Primary Enrollment Rates

(In percent)

article image
Source: World Development Indicators 2002.
Table 2.

Primary Completion Rates, by Region, 1990 and 1999

article image
Source: World Bank, 2002.

Data are generally for 1999, or most recent year available.

The geographic profile of differences in educational attainment is accompanied by large income and gender disparities. Table 3 shows the proportion of children aged 6–14 years in school for a sample of low-income countries, distinguishing between children from poor and well-off families.7

Table 3.

Percentage of Poor 6–14-Year-Olds in School

article image
Source: Filmer (1999): poverty is defined with respect to ownership of assets.

While the rich-poor gap varies across countries, it is considerable in all cases with the exception of Uzbekistan, the only transition economy in the sample. The data confirm that low enrollment in low-income countries is primarily a problem for children of poor families: There is high or near universal school enrollment for the children of rich families. Finally, Table 4 illustrates the gender bias in education in a number of low-income countries. The bias is shown to be especially prevalent in the regions of North, Western, and Central Africa, and in South Asia. In some countries, on the other hand, the bias against girls is insignificant or marginally favors girls.

Table 4.

Percentage of 6–14-Year-Old Girls in School

article image
Source: Filmer (1999).

A World Bank study identifies 89 countries that are not on track to meet the Education for All (EFA) goals (World Bank, 2002). Of these countries, 29 are “seriously off track” and will require significant and unprecedented increases in school enrollment and retention rates to achieve the objectives of universal primary education and gender balance. These circumstances raise the question of how the required educational improvements can be achieved. A prerequisite to addressing this question is the identification of the reasons underlying low educational attainment. There are both demand- and supply-side aspects.

B. Influences on the Demand for Schooling

Low school enrollment can be the consequence of low demand for schooling. Poverty is often singled out as the main impediment to universal enrollment rates in low-income countries. Even when access to schools is free, parents face various opportunity costs of children attending school that the very poor cannot afford to incur. The opportunity cost can take the form of forgone income from child labor. The loss may also be through contributions provided by children in the household. Children may, for example, perform the household tasks of gathering and preparing food and tending to fields and animals. A high incidence of AIDS also reduces the demand for schooling when orphaned children take on the tasks of looking after younger children, or children stay home to attend to their incapacitated parents and other family members.

The demand for education may also be low because parents expect or perceive low returns from education for their children. There may be information failures, with parents being unaware of opportunities available to educated children. Parents may believe (or be aware) that in the society in which they live, family or tribal connections are more important than education in finding jobs and in personal advancement. Lack of social mobility, therefore, constrains the demand for education. Distance from urban labor markets, also, undercuts the demand for education. If more education brings no advantage in the local job market, families will not make sacrifices to send children to school. Finally, missing or inadequate credit markets prevent parents from borrowing to pay for their children’s education.

Social norms also influence the demand for education of children, through conventions. A convention that children do not attend school is adhered to because of social disapproval for acting inconsistently with norms.8 For example, the importance of social norms in influencing community education decisions regarding girls is highlighted by the regional gender bias in Table 4. Parents who place a low value on girls’ education are unwilling to pay either direct or opportunity costs to send girls to school. The opportunity costs of girls’ schooling tend to be higher than for boys when the social norm is for girls to contribute more labor to the household than boys, (McGee, 2000). Investment in girls’ education is not encouraged when the social norm is that girls marry young and when marriage markets do not provide a return to girls’ education.

C. Supply-Side Influences

Low school enrollment and low educational attainment can also be the result of supply-side influences, with governments supplying insufficient resources for primary education. Total government revenue may be inadequate because the tax base is constrained by a significant shadow economy and ineffective tax administration and collection. Reflecting these considerations, many countries that have been identified as unlikely to meet the universal primary completion goal have low shares of government revenue to GDP.9 Inflationary financing of social programs, in absence of sufficient revenue, is likely to hurt the poor.

Due to such resource limitations, public education may be rationed or only selectively provided. For example, government-financed schools may be available in urban areas but not in rural areas, or may vary in quality. Ineffective public expenditure management systems can also limit the resources available for basic education.10 Significant parts of budget allocations may leak from the expenditure disbursement system. In one example, tracking surveys in Uganda revealed that between 1991 and 1995 only 13 percent of nonwage recurrent expenditures allocated for primary education actually reached primary schools, on average (Reinikka and Svensson, 2001).11

Corruption also introduces a bias against social spending (see Gupta. De Mello, and Sharan, 2001; and Gupta, Davoodi, and Alonso-Terme. 2002). Corrupt officials tend to prefer spending allocations for defense or road construction, for example, rather than for education, because of the greater ease of arranging the payment of illegal commissions and greater flexibility in siphoning off funds for personal benefit (Mauro, 1997). This preference for budget allocations for capital spending rather than for the recurring expenses in schools, such as salaries or textbooks, is compounded by a similar bias in donor financing, which tends to cover a significant part of low-income countries’ budgets. Such imbalances within education budgets can depress the quality of education because schools are not maintained or adequately heated in winter, and basic educational materials are lacking.

The list of the impediments to public spending on education includes political-economy motives. Broadly available, quality public schools may not be of interest to the political and economic elites in low-income countries, who tend to send their children to private schools. Moreover, ruling groups may not have an interest in the emergence of an educated middle class that would, in the future, insist on more accountable and democratic government and upset the status quo (see Easterly, 2001).

D. Conditions Determining Whether User Payments Can Increase Enrollment

When demand-side influences—through opportunity costs, low returns due to social and geographic immobility, and social norms, in particular biases against girls—are the reasons for low school attendance, free access to publicly financed education may not be sufficient to induce parents to send their children to school. Under such conditions of low demand, user payments are ineffective as a means of financing basic education, because of the inability or unwillingness to pay.

User payments can, however, provide resources to increase the quality of education. In that case, the increased quality can increase the demand for schooling by overcoming opportunity-cost impediments that are present when the quality of education is low. The relation between user payments and demand can therefore be positive, because of the intervening effect through quality improvement.

Where demand is present, but low enrollment and low educational achievement are due to supply-side limitations, user payments are a means of removing such constraints. Compulsory user payments oblige parents to provide resources for schools; with voluntary user fees parents can choose to pay in order to provide improved education for their children. There are, however, fundamental problems that make the financing of children’s education through user payments undesirable. The following section reviews these problems.

III. Why User Payments May Be Undesirable

A. Regressive Taxes

In general, it is preferable that schools be financed through general budgetary spending based on progressive taxes, or taxes that appropriately trade off equity and efficiency. When schools are financed through general progressive taxation and public expenditure, higher-income groups subsidize the education of children from lower-income families. Also, taxpayers without children or with a small number of children may subsidize the education of children from larger families. Compulsory user payments are usually regressive taxes. When voluntary user fees finance children’s education, not all parents may pay, but, for those parents who do pay and who send their children to school, the user payments are again regressive.

B. Voluntary User Payments and Demand

Voluntary user fees, like all prices, exclude people who are unwilling or unable to pay the market price. When parents are unwilling or unable to pay, financing schools through voluntary user payments can therefore prevent children from receiving a basic education. A negative price elasticity of demand (other things being equal) for children’s education is confirmed by studies of schooling decisions by poor households in low-income countries (for example, Birdsall and Orivel, 1996; and Gertler and Glewwe, 1989). In Ghana and Côte d’lvoire, primary school enrollment declined after the introduction of fees (World Bank, 1993). Primary school enrollment increased after the abolition of fees in Indonesia, Ghana, Kenya, Malawi, Uganda, and Tanzania (World Bank, 1995a; Lockheed and Verspoor, 1991; Bray and Lillis, 1988; and Oxfam, 2002).

The price elasticity of demand is related to parents’ income. Several studies have shown that the price elasticity is far higher for poor households than it is for the rich (Morrisson, 2002). Even low user fees can be significant for poor households in low-income countries. For example, user fees in Tanzania of between US$8 and US$16 a year (depending on the grade) were equivalent to one to two months of agricultural wages (Oxfam, 2002). For a household with several children, such user fees are unaffordable. Parents are then placed in a position of having to choose among their children, or they may not be able to send any children to school at all.

Gender biases can make the price elasticity of demand for education for girls greater than for boys (Gertler and Glewwe 1989, 1992). In Kenya, for example, girls were twice as likely to be taken out of school than boys when school fees increased (World Bank, 1995b).

C. Exclusion

A fundamental problem with user payments is the exclusion from school of children whose parents do not pay. Figure 1 demonstrates how exclusion can arise when user payments finance schools.12

Figure 1.
Figure 1.

A Self-Financing User-Pricing Solution for Education

Citation: IMF Working Papers 2002, 182; 10.5089/9781451859270.001.A001

In the absence of outside funding, user payments need to cover educational costs, enabling schools to finance themselves. In Figure 1 DD shows demand by parents for education of children. With parents’ deciding to send children to school or not (so payments are voluntary), parents of n children are ranked along the demand function DD by declining ability or willingness to pay for schooling. As the user fee falls along DD, more parents pay, and send their children to school. In the case of the demand function in Figure 1, a zero user fee is required for all parents to choose to send their children to school.

The total costs of schooling are shown in Figure 1 as an approximation, and as fixed (the costs are for the class requirements of the blackboard, the salary of the teacher, maintenance of the school building, etc).13 The function AC shows the declining average cost as the number of children in the class increases (AC is a rectangular hyperbola given by C/n, where C is the fixed cost).

The lowest self-financing (or cost-covering) user fee in Figure 1 is Puse. At this user fee, the number of children attending school is n1 This number of children is smaller than the total number of children n who would attend school if compulsory attendance were financed through public expenditure. User payments, therefore, result in the exclusion of (n-n1) children from school, because of their parents’ inability or unwillingness to pay a user fee that covers the average cost.

With costs fixed, the marginal cost of admitting an additional child to the class is close to zero. Efficiency, therefore, requires that no child be excluded from school.

At the same time, efficiency is not the only objective. Excluding children from school is also unfair or socially unjust.

Exclusion can be avoided through exemptions or personalized payments to accommodate parents who would otherwise not pay for the schooling of their children (see Chu and Hemming, 1991). Discriminatory user-fee schemes are, however, costly to administer and often not administratively feasible in low-income countries. There are also moral hazard problems: some parents might declare unwillingness to pay user payments for their children in the knowledge that the community will not allow the children to be excluded from school. In practice, attempts to implement discriminatory user fees to avoid exclusion have not been very successful (for a survey of the evidence, see Reddy and Vandemoortele, 1996).

The social benefits from externalities call for subsidies for education (see, for example, Hillman, 2003, Chapter 4). The costs imposed on parents through user payments are the precise opposite of the subsidies appropriate for achieving socially desirable enrollment and educational attainment rates. Exclusion by means of user fees denies children their personal educational entitlement, but broader social benefits are also lost when children do not go to school because parents do not pay (Jimenez, 1989; Birdsall and Orivel, 1996).

D. Self-Financing User Payments May Not Exist

A self-financing solution, whether through compulsory or voluntary payments, may simply not exist. In Figure 1, the shapes of the demand and cost curves provide for a self-financing user fee. In Figure 2, the combination of demand and costs does not yield a user fee that can covers costs. In Figure 2, the willingness or ability to pay (shown by HR) of any group of parents is always below average cost AC. There is therefore no user fee that, if paid by some groups of parents, allows costs to be covered.

Figure 2.
Figure 2.

No Self-Financing User-Payments Solution Exists, but the Project Satisfies the Test of Cost-Benefit Analysis

Citation: IMF Working Papers 2002, 182; 10.5089/9781451859270.001.A001

Yet in Figure 2, the total benefit from providing schooling for all n children (and so excluding nobody), measured by the parents’ personal valuation of the education to the children, exceeds total cost. The total benefit is the area HRO. The total benefit exceeds the total fixed cost, which is equal to OJER (because the area JAH exceeds the area REA). A cost-benefit calculation based on parents’ own evaluations of the benefits of education of their children therefore calls for schooling to be provided for everybody. Since, however, no user fee that covers costs exists, the only way to finance schooling is through government expenditure financed through general revenue taxation.

Figures 1 and 2 demonstrate the fundamental theoretical problems with user payments. A self-financing user fee will, in all likelihood, exclude some children from schooling. Such exclusion is neither efficient nor socially just. On the other hand, as the case in Figure 2 shows, there is also no assurance that a self-financing user fee will always be feasible, even if the intention is to finance education through user payments.14

E. Special Education

It is estimated that in Africa only 5 percent of children with learning disabilities requiring special education attend school, whereas 70 percent of these children could attend school. In high-income countries, governments acknowledge social obligations toward children with learning disabilities by providing for children’s needs through public spending. In low-income countries, the poor households’ children requiring special education may not attend school because there are no schools provided by the government for them to attend. The supply-side problems that may affect schooling in general are, moreover, compounded for children who may require special resources and suitable access to classrooms.

Society and parents in richer societies face beneficial incentives to pay for education of children with learning disabilities, since there are employment opportunities consistent with such children’s attainments and abilities. In low-income countries, the personal incentives of parents are, unfortunately, often different. Disabled children may be sent out to beg. The opportunity cost of educating a disabled child is hence the loss of income from begging.15 There is, therefore, a demand-side constraint to educating disabled children, and user payments can only make the education of disabled children even more unattractive.

F. User Payments as Preempting Public Spending

There is a potential problem when user payments do successfully self-finance children’s schooling. Once schooling is financed through voluntary or compulsory user payments, a government may feel that it has been absolved of its responsibility to provide free-access basic schooling financed through public expenditure. User payments should, however, not preempt public spending. User payments are a temporary, stopgap means of providing resources for schooling, until underlying revenue and public expenditure management issues have been addressed and governments can fulfill the responsibility of providing quality free-access education financed through general purpose taxation or donor resources.

IV. Ancillary Benefits Of User Payments

When user payments finance children’s schooling, there are ancillary benefits. That is, although children should attend free-access quality schools rather than schools financed by user payments, the user payments nonetheless provide benefits.

A. Mobilization of Resources

The most direct purpose of user payments is to mobilize resources for education when public resources are unavailable or inadequate. As has been noted, for government subsidies and public spending to be effective, expenditure management capacities, political will, and good governance are required. When conditions prevent adequate resources from reaching schools through government, user payments allow parents themselves to provide the necessary resources (see Jimenez, 1987 and 1989).

B. Market-Related Incentives and the Quality of Education

The resources mobilized through user payments can improve school quality. Poor education can leave children illiterate after years of schooling and depress parents’ demand for schooling for their children because of low expected returns. On the other hand, improved quality can increase demand. Poor education may be the consequence of unqualified and unmotivated teachers, and teacher absenteeism. Government spending is ineffective when funding intended for schools is appropriated or diverted. However, in the absence of monitoring and supervision, the funds that do reach schools or teachers may not have the desired effects either.16

User payments address this problem: when parents pay directly, they tend to become actively involved to ensure that they receive benefits in return. The monitoring by parents that accompanies user payments can thereby improve both the quality and cost effectiveness of education.17 While the amounts that parents can afford to pay through user payments may be low, supervision, monitoring, and the sense of ownership by parents increase, which in turn enhances the productivity of resources provided.

Evidence from low-income countries supports the link between user payments and the enhanced quality and cost-effectiveness of education.18 Jimenez and Paqueo (1996) studied 586 primary schools in the Philippines and concluded that schools relying more heavily on local funding from municipal government and parent-teacher associations tended to be more cost effective. A study of El Salvador’s EDUCO program similarly confirms that decentralized administration improves the quality of schooling and enhances educational outcomes (Jimenez and Sawada, 1999). Gershberg (1999) concluded that increased accountability by teachers and administrators to parents, with associated financial incentives, was instrumental in the implementation of Nicaragua’s Autonomous School Program (ASP). In general, parents have also shown a willingness to pay for improved education. Mingat and Tan (1986) found that parents in Malawi were prepared to pay user payments in return for higher-quality education. Birdsall and Orivel (1996) studied rural Mali and discovered that, although fees for primary schools (other things being equal) reduced demand, improvements in quality and proximity to schools more than offset the negative effect of user payments on school enrollment. Gertler and Glewwe (1989) report that in rural Peru, households were willing to pay user payments sufficient to cover the operating costs of new local schools that reduce travel time for their children.

The direct involvement of parents that accompanies user payments can also result in quality improvements that increase school attendance by girls: For example, the construction of separate latrines for girls had a positive impact on female enrollment in many African primary schools.

C. Cooperation, Mutual Insurance, and Governance

Decentralized financing and administration through user payments can also help overcome problems of exclusion and governance.

Although moral hazard makes schemes to avoid exclusion through personalized payments difficult to implement, information available to members of a small community—and the repeated personal interactions within the community—can make personalized user payments and rudimentary community-insurance schemes feasible. For example, it will be evident within a community if a particular household has fallen on hard times, and fees can be adjusted accordingly. Mechanisms of local collective action may therefore spontaneously arise to deal with the problem of exclusion and overcome moral hazard problems. Anecdotal evidence from Chad confirms this possibility. Oxfam (2001), on the other hand, points to studies of partially user-financed government schools where children are excluded from schools and from exams, and parents are confronted with court proceedings because of nonpayment or insufficient payment of school fees.

User payments may attract the attention of local government officials, who may view the local organization of parents to pay user payments as part of their domain of governance. The problems of corruption that give rise to user payments in the first place can then recur at the local government level. Furthermore, if education providers are corrupt, user charges may simply be a means of rent extraction by public sector employees, without improving education. For user payments to be beneficial, the local mechanisms for organizing collective action by parents have to be able to withstand such local governance problems.19

There is some evidence that governance issues are less severe at local levels than at central levels of government (Fisman and Gatti, 2000). This is not necessarily the case in all circumstances, however. As we shall see in the case study of Mexico’s Progresa program, when the government wished to ensure that funds targeted to local communities were properly disbursed, the central government took measures to ensure that local government officials could not access the funds and Progresa officers could not develop long-term relationships with the program beneficiaries.

An advantage of user payments is the proximity of parents to local officials. When user payments are voluntary, parents can also choose not to pay. The threat of withdrawal of payment can act as a disciplining mechanism for governance problems. When a committee of parents and teachers takes responsibility for collecting and allocating payments, problems of political intrusion are either more limited, or do not arise.

D. Distribution of Benefits from Public Spending

User payments have an advantage in allowing parents to overcome political economy problems that can impede public spending for education. Public spending may selectively favor certain groups and fail to benefit other groups.

User payments, in particular through voluntary collective action in organizing and financing schools, allow children in marginalized groups or peripheral areas to receive entitlements to basic education. Voluntary collective action by parents to organize and finance schools may be seen as a response to governments’ not providing the desired resources for schools.

Over the longer term, democratic institutions should ideally emerge to enfranchise poorer parts of the population and to put a check on corruption; eventually, democratic institutions can result in replacing user payments with public spending. When corruption has been eliminated or relegated to a marginal phenomenon, the bias against public spending on education and in favor of capital-intensive projects (such as defense and construction) can also be expected to diminish. Until such institutional changes take place, user payments allow self-financing, so that the children of the poor can receive an education that would otherwise not be available to them.

Furthermore, the education made possible by user payments is in itself an impetus for democratic institutional change. As part of the beneficial externalities from education, a more educated population will seek greater democratic participation and greater accountability from the government, and will also insist on more benefits through public spending. We have observed that one reason why governments may be reluctant to provide public spending for children’s education is precisely to avoid such future calls for democracy and political openness by a more educated population.

E. User Payments and Taxes

The personal benefits from taxes paid to a central government may not be clear, and the absence of a direct link between tax payments and benefits can be the reason for a social norm of tax evasion. Voluntary payment for children’s schooling can create social norms that result in the vast majority (or all) children’s attending school. This occurs when norms of evasion of tax payments to the central government are replaced by conventions of voluntarily paying for children’s schooling, because of the directly perceived benefits. A demonstration effect can affect behavior through changed norms. Parents who do not pay confront questions from their children who do not attend school about why other parents are prepared to pay for their children to attend school. The norm or convention of sending children to school (because others do) can thereby influence the decisions of parents who are not sending children to school. With a social norm of educating children in place, the norm can continue if, in the future, the government provides free-access public education.

V. User Payments in Practice

A. Prevalence of User Payments

The question of whether user payments should finance the basic education of children in low-income countries would be hypothetical if user payments were in fact uncommon. A comprehensive World Bank study (Burnett and Bentaouet-Kattan, 2002) reveals that, in 77 out of 79 countries surveyed, user payments were present in some form or other. Often payments by parents supplemented government spending or took on the from of uniform requirements or payments-in-kind.20 In some cases the payments by parents are compulsory; in other cases schooling for children is based on “voluntary” payments that are nonetheless a precondition for school admission.

There are both demand- and supply-side effects on the education of children. Anecdotal evidence of demand-side effects (for example, Oxfam, 2001) points to parents who are too poor to afford even small school fees, or for school uniforms or textbooks. Other demand-side impediments that can be present are perceived low returns for education, and bias against girls and against children with disabilities.

On the other hand, evidence indicates that parents are willing to pay for their children’s education, and for improved quality or closer proximity of schools (Gertler and Glewwe, 1989; Mingat and Tan, 1986; Birdsall and Orivel, 1996; Morrisson, 2002). The prevalence of cases where poor parents are willing to pay voluntary user fees for schooling indicates that poor parents value education for their children, and confirms problems with government supply. In Haiti, which is the poorest country in the Western Hemisphere, for example, some 65 percent of children are enrolled in the fee-paying private sector in the face of few public alternatives.21 Community schooling based on voluntary user payments also occurs in rural areas in parts of Africa and Latin America (e.g., El Salvador, Honduras, Chad, or Somalia). Prohibition of user payments in these cases would deny parents the right voluntarily to finance schooling for their children when governments have not provided schools or when resources provided by governments have been inadequate.

Case studies illustrate the different circumstances that can be associated with user payments for schools. Evidence from Chad shows how voluntary community involvement provides resources for education when public resources are insufficient. Mexico’s Progresa program demonstrates a case where the government paid parents for sending their children to school, so that user payments were negative. Experiences in Uganda, Malawi, and Tanzania show the consequences of attempting to replace user payments with free-access, publicly financed schools.

B. Spontaneous Community Organization and User Payments in Chad

Chad has an estimated per capita GNP of US$215.22 Although public spending on education is 20 percent of the total domestic revenue, the total domestic revenue is only 8 percent of the GDP, and the government has had only limited success in providing free-access, publicly financed universal basic education.23 Net primary school enrollment is about 50 percent; the primary completion rate is 20 percent; and the illiteracy rate is in excess of 60 percent, and is high even amongst school graduates. There are also wide disparities in schooling among regions and along gender lines.

The case of Chad nonetheless demonstrates the value that poor parents in low-income countries can attach to education. Schooling has been financed by user payments collected through spontaneous community organization when government spending has been ineffective.

Parental involvement in primary education, both administratively as well as financially, has a long history in Chad. Community-managed schools emerged during the colonial period. Against the background of political and economic instability, and the inability of the central government to provide even the most basic education for the majority of children in rural areas, parents’ associations took full responsibility for the management of schools, including schools abandoned by the government as well as for the construction and operation of new schools. Some 20 percent of students are currently still enrolled in community schools that are run without governmental involvement.

Government schools, where 75 percent of children are enrolled, are also significantly supported by parents, who hire and pay the salaries of many teachers.24 Excluding in-kind expenditure such as the supply of books and volunteer time, parents in Chad spend the equivalent of USS2 annually per child. Reports from the field suggest that children are not excluded from attending school, even if parents cannot afford to pay this moderate user-price. Cooperation through mutual insurance prevents exclusion.

Regional and ethnic divisions have affected education in Chad. Centrally trained teachers refuse to teach in remote regions; furthermore, when and where they do teach, the teachers are poorly supervised by the government. Currently only 46 percent of teachers are government civil servants. The rest are community teachers, who are literate volunteers teaching in their own communities after minimal training. With support from the World Bank, the government covers 80 percent of the salary these teachers receive (about a third of “civil service” teachers’ salaries), with the community’s paying the additional 20 percent. Parent-teacher associations are responsible for hiring and supervising the teachers.

Consistent with the earlier observations on local accountability, there are close ties between parents and the community teacher. The close ties are less likely when civil service teachers instruct in the schools, because ties are also affected by Chad’s ethnic diversity.

The poor qualifications of teachers, combined with poorly supplied schools, have kept educational attainments low.25 Only 47 percent of adults can read fluently after six or more years of schooling, in contrast to 90 percent in Rwanda or Burundi (Mingat and Rakotomalala, 2002).

In addition, the government has expressed its intention to increase public spending for education, as resources will become available through debt relief under the HIPC Initiative. Also, revenue from new oil discoveries will be available in the future. In the meantime, however, financial and human resource constraints make user payments the only means of financing schooling for many children. Thus, eliminating user payments in Chad under prevailing institutional conditions would deprive a large part of the school-age population of the opportunity to go to school.

C. An Incentive-Based Welfare Program in Mexico

Under the Progresa incentive-based welfare program in Mexico, poor families receive income transfers from the central government, conditional on the regular school attendance of their children.26 In villages targeted by Progresa, schooling increased from 6.8 years on average to 7.5 years (Schultz, 2001). Transitions to secondary school increased by nearly 20 percent, with a more significant increase for girls, who were targeted through higher transfers.

Two features of the program merit special attention. First, cash transfers are given only to the female head of household.27 Second, in order to circumvent corruption, resources are channeled from the central administrative office directly to Progresa officers in eligible villages without the involvement of local government officials. These program officers are moreover rotated regularly, to minimize the familiarity that may become the basis for collusion with local beneficiaries.

In the case of Progresa, user payments for basic education are negative. In principle, the income subsidies could change social norms regarding education and child labor decisions. While average schooling duration has increased, demand for schooling has, however, been quite inelastic: the school subsidy has cut the cost of attending school by more than half, but enrollment has increased by only 10 percent (Schultz, 2001). The limited response indicates demand-side constraints.

Progresa in Mexico benefited from generous funding and commitment from the federal government. The poorest households in a politically unstable region were targeted. There was strong political will for the program to succeed, hence the payments to female heads of households and the concerted measures against corruption by ensuring that local government officials were not involved in the program.

The Progresa program has been copied elsewhere, but the special circumstances of the program in Mexico may limit replication. Resource considerations apart, if inadequate public spending for education is due to political-economy problems, governments may lack the political will to counter corruption and to target resources effectively.

D. Abolition of User Payments: Uganda, Malawi, and Tanzania

Malawi in 1994, Uganda in 1997, and Tanzania in 2001 implemented universal education initiatives that included the abolition of user payments for children’s education. Tanzania’s experience remains too recent for conclusions, while the experiences of Malawi and Uganda do suggest some observations.

User payments were abolished in Malawi and replaced by a policy of Free Primary Education (FPE). Primary education was made a priority in public spending on education: 65 percent of spending in 1997 was for primary education and total spending on education was 24 percent of the total current expenditure (see Kadzamira and Rose. 2001). A democratic election had therefore resulted in more public spending focused on the poor (Castro-Leal, 1996). The higher budgetary allocations for education nonetheless proved inadequate. Although school enrollment increased by over 60 percent, the quality of education declined and dropout rates increased. In 1999. the primary completion rate was only 50 percent, despite a gross enrollment rate of 117 percent.

Quality deteriorated because of crowding: there were inadequate classrooms and trained teachers for an additional one million students. Also, it appears that teacher performance deteriorated as a result of reduced accountability vis-à-vis parents, who felt less compelled to monitor teachers, given the parents’ reduced personal financial involvement (Kadzamira and Rose, 2001). Donor funding, which had previously provided for about 40 percent of the total primary education budget, was delayed (Bernbaum and others, 1998). Low educational attainment increased the effective cost of schooling for parents, and reduced the perceived returns from education. Moreover, although school fees were abolished along with school uniform requirements, parents were still expected to contribute labor and materials to school construction and to buy school supplies and clothes. These costs were additional to the opportunity cost through loss of income from child labor, leaving the total cost of education as significant for some parents. Gender biases persisted, and some teachers continued to regard girls as less intelligent than boys.

In Uganda, tuition fees for primary schools were waived for four children from each household under the Universal Primary Education (UPE) initiative in 1997. Education did not. however, become “free.” While tuition fees were waived, households were still responsible for paying for uniforms and school materials, and for contributing to school construction and maintenance, and there were also fees for primary final exams (McGee. 2000). In contrast to the experience of Malawi, the government in Uganda was careful to prepare for the anticipated increase in enrollment rates. An increase of 70 percent in school enrollment was matched by a doubling of the share of recurrent spending targeted to primary education in the government budget. External aid also assisted in training additional teachers, building classrooms, and providing teaching materials.

The shift of resources to social sectors and infrastructure projects continues under the Poverty Action Fund (PAF). Nonetheless, between 1997 and 2000, net school enrollment declined from 85 percent to 77 percent, despite the increases in gross enrollment rates.

Regional gender biases also persist (Uganda, 2001; and McGee, 2000). The low quality of education due to high pupil-teacher and pupil-classroom ratios and the inadequate educational materials has tended to depress demand for schooling (Uganda, 2001b).28

In Tanzania, the abolition of user payments in 2001 appears to have substantially increased the demand for schooling (Oxfam, 2002). School enrollment rates had been low, with net enrollment below 50 percent between 1994–97 and gross enrollment in 1999 below 66 percent. In cooperation with various donors, the government developed a comprehensive basic education strategy to enhance service delivery, in conjunction with efforts also to improve the public expenditure management system. Total spending on education increased from 2.6 percent of GDP in 1999 to 3.3 percent in 2000, and to 4.1 percent in 2001. While it is too early to judge the consequences of the abolition of user payments in Tanzania, the evidence suggests that the public education system is experiencing difficulties in coping with the large increase in demand. There are claims that user payments are being reintroduced through the back door (Oxfam, 2002).

While the circumstances and outcomes in these three African countries vary, general conclusions can be drawn. First, universal primary completion rates and true improvements in educational attainment cannot be achieved through higher gross enrollment rates alone, as particularly the case of Malawi demonstrates. Quality standards are critical and depend not only on spending levels but also on policy planning, implementation, and monitoring. Second, despite announcements of the elimination of user payments, primary education can still remain far from free. Substantially greater commitments of resources are required in these countries, perhaps with Progresa-style income transfers, to relieve poor parents of all costs associated with education. To illustrate this point, World Bank simulations have shown that sub-Saharan African countries alone face a financing gap of US$ 2.1 billion to achieve universal primary education (World Bank, 2002).29 The experiences of Uganda and Malawi appear, moreover, to indicate that universal primary completion rates and gender equality cannot be achieved solely by reducing the costs of schooling. Although the evidence is weaker than in the case of Progresa, where even opportunity costs were offset to a degree, observers have noted that other complementary and targeted policies are needed to improve educational attainment. Cultural barriers remain, prejudices have to be overcome, and higher opportunity costs for girls than for boys have to be addressed (McGee, 2000).

VI. Conclusions

User payments for basic education are controversial because of the burden placed on parents, particularly in low-income households. The social benefits from education and the entitlement of children to an education suggest that, ideally, governments should rather provide quality educational opportunities for all children in free-access schools financed through general taxation. The evidence shows that enrollment and primary completion rates are low in many low-income countries, and that quality, free-access education financed by public spending is not being provided. Also, user payments are prevalent in many poor countries.

Where user payments are compulsory, in effect regressive benefit taxation has been chosen in place of budgetary spending financed through general taxes. Where user payments are optional, and schooling is correspondingly also optional, the financing of schools is again regressive; in addition, children may be excluded from school because of parents’ inability or unwillingness to pay.30 Both compulsory and voluntary user payments are socially inferior means of financing children’s education in comparison to public expenditure, financed by general purpose taxation. Moreover, user payments by poor parents do not displace public expenditure financed through general-purpose taxation. On the contrary, it appears that user payments by parents are rather a response to the absence of adequate budgetary resources for their children’s basic education.

Clearly, user payments have undesirable attributes: they are regressive, and they exclude children from educational opportunities where compulsory attendance is not enforced. Exclusion because of failure to pay can be expected disproportionately to affect the very poor.

Practically speaking, voluntary user fees introduce a dilemma in the choice between efficiency and equality. A decision regarding user payments may, for example, require a judgment about whether 45 children attending school and 5 children excluded is preferable to 50 children not attending school.31 The usual criteria of efficiency and social welfare suggest that the more children in school the better, and that exclusion of some children is preferable to an outcome of no schools at all because the means of financing basic education is absent. A strong preference for equality could nonetheless lead to a judgment rejecting financing through user payments, even when public expenditure cannot provide reasonable schooling, so as to avoid the unequal exclusion of some children. Equality may then mean that all children are functionally illiterate.

Case studies have shown that user payments can be important for providing basic education in poor countries, and that the replacement of user payments with public finance requires careful planning and preparation through resources to meet increased demand. The case studies also show that the reduced quality of education when enrollment increases (if not matched by additional resources) can decrease demand for schooling even when education is free, because of the opportunity costs of sending children to school. The case of Progresa in Mexico illustrates how constraints on education need not be only on the supply side, but can be on the demand side.

Box 1 provides a quick overview of the different circumstances that may or may not make user payments a success. When demand-side problems are the reasons for low enrollment and educational attainment, we do not expect to see the financing of education through user payments. Demand-side problems arise because of poverty or deeply-rooted social norms, which result in the inability or unwillingness to pay. The introduction of user payments will then not solve the problems of low enrollment and educational attainment. There are, however, qualifications. The improvement in educational quality facilitated by user payments can increase demand for schooling. The monitoring and accountability that accompany user payments are another source of improvement in educational quality, and can also increase demand. User payments may also change social norms by changing the convention that children do not attend school. That is, there are ancillary benefits of user payments that can increase demand for schooling.

Box 1. Quick Guide for Policymakers

Circumstances under which user payments at the local level can be beneficial:

Insufficient public resources

Inadequate public expenditure management (PEM) systems

Governance/political-economy problems

Poor-quality service due to inadequate monitoring

Circumstances under which any user payments are inappropriate:

Poverty or social norms are the main impediments to enrollment Governance problems extend to those administering user payments Resources available to the budget, from both domestic and external sources, are sufficient

Where impediments to free-access, publicly financed schooling are more deeply rooted in political-economy aspects of policy decisions, user payments may be the only means whereby poorer parents can provide an education for their children.

None of the benefits that can accompany user payments makes placing the burden of financing of basic education on parents preferable to free-access quality education financed through budgetary public expenditure. User payments, whether taking the form of compulsory-benefit taxation or voluntary user fees, are a temporary solution, and second best to free-access, publicly financed, quality education. In some instances, achieving the latter may only require a reallocation of resources in the existing expenditure.

Postscript

In a special initiative, donors will grant resources to 23 “fast-track” countries that are at serious risk of not achieving the EFA Millennium Development Goals. The donor community pledged an additional US$12 billion a year to help achieve universal primary completion rates by 2015 at the UN Conference on Financing for Development in Monterrey. However, as this paper has emphasized, constraints need not be only of a financial nature. Apart from improved expenditure capacities and mechanisms ensuring minimum quality standards, political commitment to effective implementation is also required, as is change in social norms regarding education.

User Payments for Basic Education in Low-Income Countries
Author: Ms. Eva Jenkner and Mr. Arye L. Hillman
  • View in gallery

    A Self-Financing User-Pricing Solution for Education

  • View in gallery

    No Self-Financing User-Payments Solution Exists, but the Project Satisfies the Test of Cost-Benefit Analysis