Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13

Private Secondary Education in Uganda: Implications for Planning

by W. James Jacob, Donald B. Holsinger & Christopher B. Mugimu - 2008

Purpose of Study: A fundamental question for educational planners and policy makers is which secondary school providers are most efficient in raising student learning for the most youth, given an available level of resources. Considerable attention has been devoted in recent years to the proposition that private providers offer efficient alternatives to government-financed and administered secondary schools. This study examines the rise of private secondary schools in Uganda and calculates a unit cost estimate for representative samples of secondary schools, both private and government.

Setting: A unit cost analysis was conducted among 28 randomly selected secondary schools in Uganda.

Research Design: A national sample of secondary schools was stratified in three ways. First, schools were selected according to the type of school, allowing half of the sample to be government schools and the other half private schools. A second stratification involved urbanicity, so both urban and rural schools were included in adequate numbers to allow subsequent comparison. Finally, the school sample was stratified by the four primary geographic regions in Uganda: Central (n = 10), Eastern (n = 6), Northern (n = 6), and Western (n = 6) Regions. All findings are based on analysis of sample survey data derived from randomly selected administrators, teachers, and parents of students. Using secondary school leaving examinations as the measure of effect, the article examines the relative cost effectiveness of private and government schools.

Results: Private schools in Uganda appear to be attractive, low-cost alternatives to government secondary schools. Per-pupil spending is significantly related to learning achievement, regardless of whether a student attends a private or government school. Thus, if higher performance is the desired ultimate student outcome, additional spending will be required. For their per-pupil cost, this article shows that private schools produce good learning gains—better, in fact, on a dollar basis, than government schools.

Conclusions/Recommendations: Unit cost data, when combined with reliable and valid measures of student learning achievement, are helpful to policy analysts and policy makers in determining the most efficient use of money. Although the comparatively high unit costs of government schools in this study are affected by the inclusion of capital costs, there is strong evidence to support the contention that private schools are a cost-effective alternative to government schools in the provision of general secondary education. Recommendations for future research include extending the analysis beyond the comparison of costs and examination scores by also considering the intangible component gains of schooling experiences in both private and government schools.


In a time of rapid expansion of secondary education in developing countries, leading government officials worldwide recognize the need for a thorough analysis of the private school sector prior to forming policy recommendations. Although this article focuses on one East African nation, Uganda, we believe that the problem addressed and the analytical approach in this research to be more generally applicable. Of the five broad policy priorities put forth by the Uganda Ministry of Education and Sports (MOES) in the period from 1998 through 2003, two involved the expansion of the private provision of schooling (MOES, 1998, 2000; Nsubuga & Lematia, 2000).

Literature on school expansion in Africa, as elsewhere, has for several decades drawn attention to the financial burden imposed by expansionist policies and the concomitant requirement to use resources efficiently. Our purpose here is not to review the substantial literature comparing government and private school providers. Numerous works by authoritative scholars have set a rich context for the debate on alternatives for public policy toward private education in international contexts (Aldrich, 2004; Anderson, 2002; Bangay, 2005; Kitaev, 1999; Levin, 1995; Lin, 1999; Tsang, 1995; Tsang, Wei, & Xiao, 2000). There are, as Estelle James (1994) noted, inherent ambiguities in the terms public and private when applied to educational institutions. Private schools, comparative newcomers in modern education systems, appear to have unique features that alter the supply and demand functions and endow them with a bewildering array of complex and diverse profile characteristics. The topic also invites comparison with a broader set of issues involving parental choice and the intensely debated topic of school vouchers that by now have found their way into Chilean education. This article is aligned more closely with a distinct subset of the above: namely, comparisons of the academic effectiveness of public and private schools.

The history of Uganda after independence is marked by political upheavals that stabilized somewhat after a rebel opposition group seized power in 1986. The economy was in a terrible state, devastated by mismanagement and successive wars. Heavily impacted by HIV/AIDS, Uganda had the world’s highest national HIV prevalence rate in the late 1980s and early 1990s (Morisky, Jacob, Nsubuga, & Hite, 2006). The conditions of the country were such that the government recognized that it could not rebuild its social and economic infrastructure without external financial assistance. Between 1987 and 2000, Uganda underwent seven major structural adjustment operations linked to both World Bank and International Monetary Fund credits. Among other things, these reforms supported the development of a more stable and “business friendly” environment for private sector expansion, as well as enhanced access to international markets for tradeable goods through liberalization of producer prices and currency markets. However, the dependency of the economy on “traditional” export commodities, especially coffee, continued, and the growth of foreign debt became an increasing burden to the country.

The dilemma posed by increased demand in times of fiscal stringency was only amplified when the government implemented universal primary education (UPE) in 1997 (Ablo & Reinikka, 1998; Cummings & Dall, 1995; Devarajan, Dollar, & Holmgren, 2001; Jimenez & Lockheed, 1991; Moumouni, 1968; Nielson & Cummings, 1997; Ssekamwa & Lugumba, 2001a, 2001b). Policy makers were left without a clear understanding of which sector, government or private, to support and at what level for future expansion. This study helps to fill this void by drawing on survey and test data to answer a fundamental question of great policy significance in Uganda as elsewhere: What is the relative cost effectiveness of government and private secondary schools? The answer to this question is remarkably difficult to obtain. We maintain that we have been able to do so through a careful study of unit costs of public and private secondary schools in Uganda and by combining them with national examination scores to form the basis for a comparative analysis of cost effectiveness. This is hardly a mere exercise in the application of unit cost formulae and statistical analysis;. rather, it is an examination of a very difficult and ever-present reality of schooling worldwide: How should schooling most efficiently be financed?

This article has two parts that combine to provide an analysis of secondary school providers in Uganda and an answer to the efficiency question posed. Part 1 provides the necessary conceptual and definitional framework for secondary education in Uganda and helps the reader understand the essential profile distinctions between private and government-aided schools. This part draws on survey data derived from interview schedules administered to a broad sample of secondary school administrators, parents of students, and teachers. Part 2 delves into the technical considerations required of a unit cost and cost effectiveness analysis. Data for this analysis were obtained in a careful compilation of the many “ingredients” that constitute the full range of the cost of secondary schools. By comparing costs with actual test score data for a representative sample of both government-supported and private secondary schools, a significant advantage for policy recommendation is achieved. The most important policy question is not simply whether public schools cost more than private ones (or vice versa), but what schools of each type cost in relation to their principal objective, student learning achievement.


Data supporting the description and analysis of private and government school distinctions and modal characteristics are from a 2001 survey of 97 students, 78 teachers, and 18 secondary school administrators interviewed from private and government secondary schools randomly selected in Kampala and Mukono Districts. The sampling procedures followed strict probability estimates. Interview schedules were pretested and revised following preliminary statistical analysis.

For the purposes of this article, the term private refers to all schools in Uganda that do not receive government support. Private schools receive no direct government support; there is no public subsidy of costs associated with infrastructure or teacher salaries. The government bears the cost of the general and professional teacher education, even for those hired ultimately by private schools. The term public refers to a government-aided school; we use the terms public and government interchangeably in this article. There are four types of private schools in Uganda: (1) mission schools, (2) dismissed teacher schools, (3) for-profit institutions, and (4) progressive or humanitarian schools. J. C. Ssekamwa (2000) provided an overview on the history of private school types in Uganda. Reader comprehension is aided by a brief discussion of each type of school.

Missionary, religion-based schools. Missionary schools were the backbone of the education system in Uganda from 1877, when missionaries were first invited into the country by Muteesa I, to the time when Uganda became independent from Britain in 1962. At the time of independence, the government of Uganda declared that all religion-based and discriminatory schools would be taken over and administered by the government. The primary reason for this move was to unify the country because religious schools had hitherto been viewed as discriminatory and disruptive to unity. They were considered discriminatory because only official church members who adhered to the codes and values of the respective churches were allowed to attend. Thus, marginalized groups and members of other faiths were overlooked.

For-profit schools. These institutions were set up by either a founder, group of founders, or sometimes a foreign investor who essentially ran the school as a business for financial and capital gain. For-profit schools have existed since the 1920s and have been periodically discouraged by the government and religious schools. With the advent of UPE in the early 1990s, for-profit institutions proliferated at an astounding rate, and this subsector is continuing to grow to this day.

Progressive/humanitarian schools. Progressive-minded teachers organized a subsector of private schools that operate like humanitarian institutions. Humanitarian schools accommodate students who want to further their education but either do not meet the academic requirements to attend the premier government schools or do not have the financial ability to pay for a quality private education in one of the more prestigious private schools. Many teachers and administrators working at these schools often go months with little or no wages.

Dismissed teachers schools. These schools arose from teachers who had been expelled from the missionary, religion-based schools because of an inability to comply with religious codes and strict moral values. Teachers who were not able or had no desire to abide by the statutes of the church-operated schools broke off and formed their own schools. Although the mission, for-profit, and progressive/humanitarian schools remain today, the weakening of the missionary school system has obviated the need for dismissed teacher schools, which faded from the private scene shortly after independence.

It is sometimes difficult to group a private school into one or more of the four types of schools listed above. In some cases, it is difficult to distinguish between for-profit and humanitarian schools. This overlap of definitions occurs because some for-profit schools offer scholarships to needy and orphaned students to help provide a way for them to gain an education. However, it is widely recognized that the private sector cannot absorb the national demand for such humanitarian education services, as demonstrated over the past decade as millions of children in sub-Saharan Africa have been orphaned due to the AIDS epidemic (Nsubuga & Jacob, 2006). The Joint United Nations Programme on HIV/AIDS (UNAIDS) and the World Health Organization estimate that Uganda has among the highest number of AIDS orphans in the world at 940,000 (UNAIDS/WHO, 2004). New avenues will have to be developed by both government and private sectors to address this tremendous need (Jacob, Smith, Hite, & Cheng, 2004).


Private schools are administered directly by a senior or head teacher. School policies, including the pay scale of teachers and fee structure, are determined by a board of directors that includes the founder or owner, who figures preeminently in decision making. An important distinction between government and private school administration, and one that bears on their relative cost effectiveness, is their different methods of procuring teachers. All government teachers are hired on a full-time salary basis. Private school teachers are made up of both full-time and part-time teachers who are paid on an hourly or per-class basis. This allows high-demand teachers to teach part time for two or more schools. It is not uncommon for a quality government school teacher to teach full time at a government school during the day and then teach part time at one or more private schools in the evenings or on weekends. In this way, quality teachers maximize their income. This arrangement is also advantageous to private schools because they are able to advertise to parents and students that they have on their teaching staff the best math and science teachers in the district, even though they may only teach one or two classes per week at their school. Although some students and parents consider the part-time teaching dynamic that exists in many private schools a disadvantage to students in establishing close relationships with their teachers, others provide an alternative viewpoint. They argue that having part-time teachers allows students the opportunity to associate with and learn from some of the district’s top teachers. This opportunity would not be possible if only full-time teachers were employed at private schools, largely because most private schools cannot afford to hire the best teachers on a full-time basis.


Private providers, like government providers, are required to maintain clean schools to meet a range of health and sanitation standards, to have adequate buildings and furniture, and to comply with curriculum and teacher qualification requirements. Unfortunately, like many ministries of education in developing countries, the Uganda MOES does not have the resources to ensure that these standards are met (Tooley & Dixon, 2005). Private and public schools alike function largely unsupervised.

Theoretically, the MOES also requires schools to submit financial portfolios each year, though this also is seldom done. If asked to produce such records by the government, schools are required to comply, though it may take some time because of inadequate or unqualified accounting staff. Some founders in the early days of the for-profit private schools collected student tuition and then disappeared with the money. Fraudulent behavior exists today on a limited scale in for-profit schools. Unfortunately, these incidents have given the for-profit sector its popular image of exploitation.

The government also regulates student acceptance into all public schools. Students have the option to attend a public school if admitted, but they are not allowed to attend the highest ranked government schools without adequate Primary Leaving Examination scores. Private schools usually recruit and absorb the remainder of the students who failed to qualify for the government schools. Government-supported schools tend to cater to the masses, whereas most private schools focus on individual needs. This is especially the case for students who tend not to fit the typical secondary education student norm, either in terms of academic performance or behavioral discipline (de Regt & Weenink, 2005). Only certain registered schools or organizations are able to administer the Uganda National Examinations Board (UNEB) examinations. Many private schools, especially newly established ones, are not yet registered or qualified to offer such exams and thus must send their students to a school that is registered to administer the exams.

In many countries, people often associate private with greater prestige, but this is not necessarily the case in Uganda. There are at least four reasons that some private schools may not attain the reputational status of prestigious government schools. First, teachers do not necessarily meet certification requirements. In many private schools, especially among the humanitarian schools, teachers tend to be underqualified or not qualified at all, and the results are evident in student performance on the Cambridge School Certificate Examinations. Second, the majority of teachers are underqualified and rely on one or two stellar teachers to attract students to their school. Third, in poorer rural regions of the country, some private teachers and staff go unpaid for months. And finally, the quality of instructional materials used in private schools does not typically differ from materials found in public schools; materials come from the same publishers. However, private schools with high fee requirements may enjoy the advantage of being able to pay for newer materials or materials in more adequate quantity for their students’ use.

Those public and private schools enjoying the highest level of popular prestige invariably have modern facilities (computers, science laboratories and equipment, vocational machinery, electricity, running water, and so on), current texts, and excellent teachers and administration. Unfortunately, the opposite is also true when considering the poor-quality end of the prestige continuum.


Postindependence secondary education reforms focused primarily on minor curricular changes, such as allowing students to take African history in place of the traditional history of the British Empire and introducing Swahili/Luganda literature and languages (Bagunywa, 1980). Although the MOES regulates textbooks, requiring that they be loosely related to the official government-prepared curriculum, they are produced by the private sector. MK Publishers employs local teachers, professors, and government officials to write their texts offered for primary and secondary school use. Their books are marketed to every district in the country, though only schools with sufficient funds to afford the texts are able to remain current with the latest offerings.

Private schools tend to have greater autonomy and flexibility in deciding what is taught and how to teach it. In addition, the schedule of their classes is generally more flexible to meet the needs of both the students and the teachers.

Because the Ugandan education system is patterned after the British colonial model, the curriculum is oriented toward a final examination to test overall learning. Thus, both private and public schools teach to the test to ensure that their students are prepared to pass the exam. The proportions of students passing the exam is a principal means of prestige enhancement and hence the success of future student recruitment. Questions of curriculum relevancy to anything other than exam performance often arise when the primary objective for learning is based on an examination. We recognize the inherent limitations in using the results of a single final exam to measure student learning achievement, but these are the only systematic data available in a country such as Uganda.


For the most part, parents prefer the traditional government boarding schools for their children. Nevertheless, students unable to gain admission to these prestigious secondary schools eagerly enter private schools instead. Having made this decision, parents often cite good teacher pay, school autonomy, and teacher commitment as advantages that private schools enjoy over their government counterparts.

Parents who are willing to pay enough money can find an excellent private school for their children. Some private schools do not charge fees to disadvantaged, poor, or orphaned students. This humanitarian perspective on education, unique to the private sector in secondary education, provides many students with the opportunity to further their education that they would not have otherwise. Still, only a small percentage of qualified students actually advance to the secondary level, and inability to pay is a significant causal factor. Private schools may accept alternative ways to pay tuition. For instance, some students’ parents will provide firewood for cooking during an entire term or year to help offset tuition costs. Others may bring much-needed food supplies. Some students work for the school, doing such tasks as cleaning the compound or helping to build a new classroom, and receive a scholarship-like waiver of their tuition in return for their labor. Private schools desiring to expand their infrastructure may require students to provide resources, such as a bag of cement, in partial payment of their tuition. Students at private schools are often required to give one ream of paper each term.

The predominant fee structure is on a term-by-term basis. Generally this is broken down depending on whether the student is at the Ordinary or O level (i.e., consisting of Senior 1–4 grade levels, which are comparable to Grades 7–10 in U.S. schools) or at the Advanced or A level (i.e., Senior 5 and 6 grade levels, which are comparable to Grades 11 and 12 in U.S. schools). Additional fees are required for certain subjects as well, such as science, art, and home economics.


Determining who pays the true cost of private education is not an easy task, given in-kind payments and government subsidies for teacher training, curriculum development, and so on. The wealthy are able to pay for the best private schools if their children are unable to enter the most prestigious government schools. Such parents do not hesitate to pay the associated boarding fees, ensuring that their children progress through the traditional academic pipeline. Many poor parents are simply unable to pay. And if they are also unable to enter a public institution, they are forced into either attending a humanitarian school or, more often, simply entering the job market at the lowest end of the pay scale.

School finance in Uganda, as elsewhere in Africa, is a huge problem (Anderson, 2002; Brock-Utne, 2000). The government of Uganda has long recognized its inability to absorb the entire cost. Communities, individual households, students, nongovernmental organizations, religious organizations, and foreign donors all contribute to financing Ugandan education. In this section, we examine various ways in which private schools are financed.

Cost sharing is ubiquitous in Ugandan education, meaning that parents of students and the communities jointly finance the education of young citizens, together with the government. Even government schools are not entirely financed by the government. In fact, many so-called government schools were once private schools. The government pays what it can, and the rest of the education costs are absorbed by local community members and parents. Parent Teacher Associations (PTAs) have been established throughout Uganda and now exist in most communities and villages. These organizations contribute as a primary or secondary source of financing.

Prior to the 1963 Education Act, religious organizations were the predominant contributors to secondary school finance. Since that time, their influence has diminished. In recent years, religious organizations have been allowed to establish schools, but such schools are no longer based on racial or religious segregation, which is discouraged and often prohibited by the government. Nongovernmental organizations (NGOs), which include all organizations not entirely supported or controlled by the government, are a common source of financial support for private secondary schools. Of the 28 schools in our unit cost study, three private schools had official exchange partnerships established with international-based NGOs. Local and international-based NGOs, embassies, and other foreign agencies quite often offer grants in the form of building materials, books, technical support, computers, and so on. Several of the schools in our study had donations of one sort or another from international-based NGOs or local religious organizations. Three of the private schools in our study had received entire libraries, computer labs, and other miscellaneous in-kind donations contributed by international NGOs. In-kind donations on a regular basis give certain private schools a strategic competitive advantage over schools with no access to such contributions.

Private schools rely on personal initiatives to raise funds from their supporters within the local communities as well as international bodies. Government-aided schools often receive foreign aid through the MOES, particularly for building infrastructure (classrooms, laboratories, teachers’ houses, laboratory equipment, textbooks, and so on). Government-aided schools also receive capitation grants from the state budgets. A few private schools receive some government-sponsored teachers.

Many private schools in our study had established student work programs, but only a few offered a program that offset fees. Community participation has been very significant in the construction of buildings for private schools and some government schools. It has been the government’s strategy to encourage parents and communities to support school construction through brick making, carpentry, and land donation. This strategy reduces cost and provides an avenue for expanding the education sector at a much faster pace. Teachers in private schools tend to carry larger teaching loads than their counterparts in government schools, thus reducing the instructional costs of private education.

Government schools share many benefits over private schools. Many government schools are able to provide housing to teachers and administrative staff. All government employees receive a set salary based on fixed rates determined by qualifications and experience, yet both public and private schools charge fees to students. In some instances, the fees charged by government schools may be comparable to or higher than they are at private schools. Additional government school benefits include the annual capitation grant based on student enrollment and access to laboratory equipment and textbooks. These stable benefits attract good teachers and encourage them to remain at government schools.

The private sector also offers many benefits to teachers in secondary education, including (1) a flexible schedule, (2) smaller bureaucracy that allows easier communication between the administration and teachers, and (3) individual attention. Our survey showed that the primary reasons that teachers gave for preferring private schools were timely payments for services and good relationships with the school administration. However, in a few private schools, teachers were dissatisfied with school leadership and only remained teaching in the private school because they had no alternative. Follow-up interviews added an additional perspective. The timely payment advantage applies primarily to for-profit and religious schools because many teachers at humanitarian schools had not been paid for periods of up to 6 months or more. In these cases, mostly in the extreme rural regions of the country, government schools do offer a more stable and regular income than the humanitarian schools.


Teachers in government secondary schools generally receive higher salaries compared with their counterparts in private schools. This disparity is amplified when considering other benefits offered by government schools, such as job security, housing facilities, gardening space, transportation, electricity, security, water, and utilities. Of course, most of these fringe benefits are lacking in most private school employment.

Teachers in nine surveyed private secondary schools of Kampala and Mukono districts expressed satisfaction with the mode of payment in private schools. Most teachers said that their salaries are paid on time and that it is much easier to negotiate terms with school administrators and school owners at private schools than it is at government schools. Most private schools pay according to work done; thus, government teachers find it convenient to take on some classes in addition to their regular teaching load in government schools. Although the average starting salary for teachers who hold only a teaching certificate qualification at government schools is about US$100 per month, teachers with a bachelor’s degree qualification earn approximately US$150 per month. However, actual income depends on how much additional salary a teacher receives from the local PTA and other sources of funds. Full-time teacher earnings and allowances based on educational training and qualifications in private schools are comparable with those of government schools. But it is important to consider that to maximize teaching hours taught, most private school administrators do not hire teachers on a full-time basis.


Most government schools have a quota on the number of students allowed to enroll each year. This restriction does not apply for most private schools. Depending on the type of private school, many owners and administrators will admit as many students as they can accommodate. Thus, many private schools tend to maintain higher student/teacher ratios compared with government schools, which is a frequent criticism of those who oppose the private school movement. Many critics believe that for-profit private schools will admit as many students as they can to boost net earnings, regardless of the student/teacher ratio. For-profit and humanitarian private schools counter this claim with the perspective that students desiring a secondary education should not be denied one, despite large class size. Moreover, many for-profit schools defend their quality through outcome measurements. Our study lends additional support to this viewpoint.

Parents who can afford to send their students to boarding schools generally choose to do so. The perception among most parents in Uganda is that boarding schools offer students the best environment for learning. This may be true, because the majority of large, traditional secondary schools also tend to be boarding schools. In the colonial era, boarding schools were one means of socializing people to a new way of life, religion, and set of cultural values. But boarding schools are expensive, costing each student nearly five times on average what it costs to finance a day student. Secondary students in remote rural regions rarely have the financial backing to pay for the finest boarding schools. It is not uncommon, however, for students in rural areas to live with extended family members in urban areas during the academic year. This trend exemplifies the strong familial ties and the sense of obligation that family members feel to offer educational opportunities to their kinsmen. Boarding schools produce a new type of citizen. The teaching of values, culture, and many practical skills was once the responsibility of parents and extended family members, but boarding schools place this responsibility on dorm parents, teachers, and administrators.


The preceding examination of the private/public school debate in Uganda is meant to provide a historical and socioeconomic context for a more focused analysis of relative unit costs. It is by itself an inadequate basis on which to provide specific guidance on important education resource allocation issues. The usual way of doing this is a two-stage process in which fiscal and private efforts are calculated and then divided by the number of government and private students. Although relative per-pupil costs, or unit costs, are a useful measure by which to compare system efficiency, they still do not address a more fundamental issue concerning human capital formation: namely, how much students learned in relation to expenditures. By comparing costs with actual test score data for a representative sample of both government-supported and private secondary schools, a significant advantage for policy formulation is achieved. The most important policy question is not simply whether public schools cost more than private ones (or vice versa), but what schools of each type cost in relation to the quality of their principal objective: student learning achievement. This research follows lines traced in Gerald Lassibille, Jee-Peng Tan, and Suleman Sumra’s (1998, 2000) cost effectiveness analysis of private secondary schools in Tanzania. This persuasive article provides a literature review of trends in private secondary enrollments that is not necessary to repeat here. Although both studies attempt cost effectiveness estimates, our efforts to provide valid unit cost data are very different from those of the Tanzania study and significantly add to the richness and reliability of evidence on private secondary schooling.


In a pilot study, we selected one private and one public school to determine the range of inputs (ingredients) that would necessarily constitute the complete cost picture. This experience convinced us that the process was indeed complex and labor intensive. But the results would permit a clear comparison of government and private schools and would contribute to the policy dialogue concerning the relative efficiency of the two most common providers of schooling in Uganda. We later expanded our sample of schools to include 26 additional schools scattered over the four traditional geographic regions of Uganda: Northern, Central, Eastern, and Western Regions (see Figure 1). At the time of the sample selection, there were 46 districts in Uganda. Mugisha Odrek Rwabwoogo’s Uganda Districts Information Handbook (1998) lists the 46 districts that existed at the time of selection. Our sample includes two districts from each of the Eastern, Western, and Northern Regions, and three districts from the Central Region. Decentralization efforts have led the government of Uganda to expand its number of districts to 70 as of 2006. To ensure the reliability and accuracy of our data, we conducted follow-up visits to some of the participating schools in our sample in 2001 and 2002. The follow-up data showed much consistency with the data collected in 2000.

Figure 1. Map of Uganda showing the number of schools by region in our sample

click to enlarge

The final sample consisted of 28 secondary schools in Uganda that were randomly selected from the official MOES Register of Secondary Schools. The school sample was determined by a randomized procedure using three strata: (1) by geographic region, (2) according to urbanicity, and (3) by falling into either government or private school classification. Recognizing that it is difficult to come to a consensus on the precise definition of these sampling criteria, we consulted extensively in Uganda with social scientists, education researchers, and officials from MOES. These consultations led to the following definitions.

A. Urban and Rural Schools

We stratified our sample of schools by urbanicity, ensuring an equal number of rural and urban schools. The term urbanicity was coined in 1974 by W. Allen Martin (1976). Unlike other terms that deal with urban issues (i.e., urbanism or urbanization), urbanicity includes the ecological context of a geographical location, considering the social and economic dominance in the spatial area (see also Martin & Spurgin, 1983).

Urban school. Any school located within a 15-kilometer radius of the district capital city with a population greater than or equal to 300,000. An urban school was considered urban if it was located within a 10-kilometer radius of a district capital city/town with a population less than 300,000.

Rural school. Any school located beyond a 15-kilometer radius of the district capital city, with a population greater than or equal to 300,000, was considered a rural school. Likewise, a school qualified as rural if it was located beyond a 10-kilometer radius of a district capital city/town with a population less than 300,000.

B. Geographic Region

Uganda is divided into four geographic regions—Central, Eastern, Northern and Western. The majority of the population resides in the Central Region, where the capital city Kampala is also located. In this study, we stratified selection according to geographic region with aggregate total of 28 schools nationally: Central (n = 10), Eastern (n = 6), Northern (n = 6), and Western (n = 6).

C. Government and Private

Three factors are essential for a school to be considered government aided and thus categorized as a government school. Schools not meeting these three requirements were classified as private schools. Government and private schools were evenly stratified at 14 schools apiece.

(1) If the school is jointly or completely owned by the government. Regardless of the percentage breakdown, if a school is partially supported by the government (i.e., government pays 10% of total expenditures), it would be considered government aided or recognized as a government school by the MOES.

(2) If government pays the salaries of school teachers/staff and if the school receives a capitation grant based on student enrollment. (3) If the school receives the Capital Development Grant under which government supports the development of school infrastructure (facilities, buildings, and so on) and/or rehabilitation of existing facilities.


There are basically two ways to compute unit costs. The first is to divide aggregate spending by the number of students. The second is by building up from the constituent parts of costs. In this second approach, “the cost components are identified, evaluated, and then aggregated to obtain the desired estimates” (Jarousse, Mingat, & Tamayo, 1996, p. 2). Almost invariably, most unit cost estimates are based on the aggregate (official) spending calculations because of the simple and compelling facts of overwhelming simplicity, data availability, and ease of calculation. But although this is the common approach, it is not the most accurate. Government expenditures data for a given subsector of education do not and cannot easily capture the subtle “leaks” that drain away resources before they reach the school level. On the other hand, government expenditure data do not reflect household expenditures, and these can be a considerable portion of the total cost structure of schooling. For these and other reasons, we decided to focus our efforts on the constituent parts method.


Rather than simply taking the aggregate spending in a given year and dividing by the number of students, as shown in Figure 2, we followed Jarousse et al.’s (1996) approach whereby the first equation is disaggregated into its composite parts, as represented in Figure 3.

Figure 2. Traditional unit cost equation


Figure 3. Disaggregated unit cost equation


The components of our unit cost (UC) analysis are capital (C), personnel (P), student fees (F), other recurrent costs (OR), and total number of students (N). Personnel costs are disaggregated further into teachers and administrative and support staff costs, and capital costs consist of both buildings and land. This approach is complicated because some school buildings in Uganda are more than 50 years old, and others were built in the past few months. Our approach to capital costs was based on the replacement value of school infrastructure. In other words, what would it cost in terms of both land and construction to duplicate a facility of the same size today?


Capital costs: Land and site improvements (sewer lines, fences, and so on), buildings, furniture, and machinery that give service for more than one fiscal year (Coombs & Hallak, 1987).

Recurrent costs: Human services and physical supplies that the educational process consumes within a single budget year. Because these typically amount to 80% or more of total educational costs, they merit major attention in any cost analysis, whatever its particular purpose (Coombs & Hallak, 1987). We have broken down recurrent costs into three variables: personnel, student fees, and other recurrent costs (encompassing all other recurrent costs).

As mentioned, each of the four components from the second equation can be disaggregated further. The two components that have greatest variability between private and government secondary schools in Uganda are UCf and UCp. Because of their importance in our unit costs analysis, let us now examine these two variables more closely. First, we present the breakdown of UCf as shown in Figure 4.

Figure 4. Breakdown of the disaggregated unit cost equation for student fees


FB and BD are, respectively, the average annual fees of boarding students and day students; NB and ND are the number of boarding and day students; and UCfb and UCfd are the cost per student for boarding and day students.

Next, we can break down UCp (see Figure 5) where PA, PT, and PNT are the average annual salaries (including benefits) of administrators, teachers, and nonteaching staff; NA, NT, and NNT are the number of administrators, teachers, and nonteaching staff; and UCpa, UCpt and UCpnt are, respectively, the cost per student in salaries for administrators, teachers, and nonteaching staff.

Figure 5. Breakdown of the disaggregated unit cost equation for capital costs


Each of the components in Figures 4 and 5 can be further disaggregated to support Coombs and Hallak’s (1987) claim that “the further you break down the costs into meaningful subdivisions, the more versatile, useful, and accurate your analysis will be” (p. 55).

The detailed and tedious cost analysis followed in this study yielded results showing simply that private schools have substantially lower overall capital costs and somewhat lower recurrent costs compared with government schools in our sample (see Table 1).

Table 1. Mean capital and recurrent costs expressed in CY2000 U.S. dollars for private and government sample schools

Cost Category

Private Schools

Government Schools










In this study, we compared the O-level (Ordinary) examination scores for all students who took the examination in 2000. The mean scores for students from private schools are slightly higher than for students of government schools. Using the “ingredient” method, we calculated unit costs for each school in the sample. Whereas the total costs in both recurrent and capital categories are lower for private schools, their unit costs are moderately higher. This is explained by the fact that private secondary schools have fewer students per school than do the government schools. The average O-level examination score for private schools is slightly higher for private than for government schools. This is remarkable because government schools derive a substantial advantage in the overall quality of their incoming students in terms of the Primary Leaving Examination (Primary 7, or P7) scores. Finally, this study estimates the dollars associated with one point on the O-level exam and finds that the expenditure of government schools to produce an O-level exam point to exceed that of private schools by 26%. By holding constant the primary leaving, or P7, scores, the advantage of the private school in terms of points by dollar of unit cost increases.

We are not surprised that the capital costs of private schools are lower than government schools, because much of this difference is accounted for by the relative amount of land owned by each. Government schools historically have had access to large amounts of land, whereas the newer private schools typically cannot afford much more land than is required by the buildings they occupy. Not surprising either is the more equal recurrent cost means. Schools all compete for the same pool of teachers, buy the same books, pay the same utility rates, and so forth. That the private schools do so at lower overall costs is an interesting and important area in need of further investigation. Before moving on to the more telling unit cost picture, we pause to point out that the correlation between capital and recurrent costs is positive and high at r = .76.


Table 2. Unit cost in U.S. dollars in 2000 for sampled private and government secondary schools in Uganda


Private Schools

Government Schools

Total cost in U.S. dollars



Total enrollments



Unit cost in U.S. dollars



The unit cost calculation as portrayed in Table 2 is revealing. Whereas the total costs in both recurrent and capital categories are lower for private schools, their unit costs are moderately higher. This is chiefly because they have fewer students per school than do the government schools. This is not the whole story, because human capital is not merely the sum of all students who have accumulated “seat time” in school; it is a function of what is learned, both cognitive and noncognitive. Student learning is the unquestioned product of schooling (for more information on the noncognitive outcomes of schooling, see Holsinger, 2000). Even though the government schools appear to be more efficient on a per-capita basis, what happens when we estimate the amount each student learns from a dollar invested?

The O-level exam is required of all secondary students who complete the first of two secondary school cycles. In Uganda, not all secondary schools offer the A-level (Advanced) cycle that represents an additional 2 years of study. Although the O-level examination is not a standardized test and therefore cannot be used to measure changes over time, it is uniform for all schools in a given year. We are aware of the limitation associated with a one-shot examination (see for instance, Hoyle, Harris & Judd, 2002; Neill, 1997). We are also aware that a major determinant of examination performance is prior learning. Any attempt to explain variations in O-level examination results by reference to the kind of school a student attended without also taking into consideration the prior learning of students would be to ignore a major source of variation and to confound the analysis. A numerical summary of the unit cost results and the O-level exam results is presented in Table 3.

Table 3. Comparison (means and SD) of private and government secondary schools with respect to O-Level and Primary Leaving Examination (P7) scores and cost data











(2 tail)

Mean P7 Score






Mean P7 standardized with 100 as high






Mean O-level exam score






Mean O-level standardized with 100 high






Total capital cost in U.S. dollars






Total recurrent cost in U.S. dollars






Total (capital + recurrent) cost in U.S. dollars






Unit cost in U.S. dollars (2000)**






O-level points per dollar of unit cost (all schools)






O-level points per dollar of unit cost (top P7 intake schools)






O-level points per dollar of unit cost (bottom P7 intake schools)






Dollar cost of one O-level point






* Significance level was determined by independent samples t test. Equal variances were not assumed. Significance level not computed for constructed variables and where not all 28 schools are represented.

A low score denotes high performance in the P7 examination.

The low O-level scores indicate the higher performance.

Standard score scale for which O scores run from 0 (low) to 100 (high).

** The 2000 exchange we used was 1,644 Uganda shillings to the U.S. dollar.

The private schools, as predicted by many, have a modestly lower unit cost. If these costs, expressed in dollar terms, seem too high, it is important to remember that we are including here capital and recurrent costs. Thus, strictly from the annualized cost of schooling an O-level student, Uganda’s private schools come out as clear winners.

Although government schools by and large have more land than private schools, this fact alone is not responsible for the higher unit costs of government schools. The bivariate correlation between recurrent and capital cost is a high r = .76. Land ownership confers upon a school the possibilities of sports, playground, and other recreational pursuits (cricket, soccer, and so on) that a private school with little or no schoolyard cannot offer to its students. Land also confers a certain impressive visage, an intangible prestige, giving the school with abundant land a competitive advantage. Although government schools have access to subventions that facilitate the financing of schooling (e.g., capitation grants, teachers’ salaries, science laboratory equipment) and that are not available to private providers, who must find alternative sources to finance essentially the same costs, we draw attention to the fact that the source of funds does not alter their inclusion in a unit cost calculation. Hence, government schools are not more cost effective merely because they receive direct government support. Such support may affect the fee structure of schools, but not their costs.


The exam score picture is more difficult to interpret. Schools in Uganda, as elsewhere, compete for the best students. Traditionally, the students with the highest P7 scores choose the best or most prestigious schools, which are invariably boarding schools in urban areas. To have a reasonable basis for comparing the performance of students in government and private schools, we needed to control for the effects on O-level exam results attributable to the quality of student cohorts entering the secondary school stream following completion of the primary cycle.

In our sample of secondary schools across Uganda, the government schools do in fact follow this pattern despite the considerable inroads that the private sector has made in attracting good students. As seen in Table 3, the average P7 score for all government schools is 14.0, compared with the lower (numerically higher) 16.5 for the private schools in our sample. So with respect to the quality of the entering student cohort, government schools are able to attract a somewhat more highly qualified student body. Although we were dealing with cross-sectional data and therefore unable to analyze the test score performance of the same students over time, we have no reason to believe that the pattern of student preference for government schools has changed in the past decade. If anything, we would expect that the entering class of government secondary schools of 3 years ago would be even more different in terms of average P7 scores compared with private schools than the difference between government and private school students today. This average difference in favor of government schools should manifest itself in overall better O-level performance, but our data have not supported this conclusion.

Private school students who took the O-level examination in 2000 (the year these scores were reported) outperformed their government counterparts by a slim margin. The average O-level score for the private schools was 31.3, a slightly better overall performance than for the government schools, whose average O-level score was lower (numerically higher) at 32.2. This is not a large difference, but it does demonstrate that although the private schools were working with students of an overall lower intake quality, they still managed to outperform the government schools just a few years later. This is a remarkable achievement, one worthy of additional study. What is it about the private secondary schools that make them more effective where the O-level exam score is the measure of effect?


This is the “bottom line” as far as school efficiency is concerned. The relative cost effectiveness of different types of schools in relation to student learning achievement is rarely presented for the simple reason that it is so difficult to obtain. To aid interpretation of this analysis, we converted the P7 and O-level scores to standard form and inverted the results such that a high number represents the best performance. We also converted the scores to a scale in which 100 represented the theoretically highest possible score. The private school cost effectiveness (with the school mean O-level exam score as the measure of effect) figure is higher (16.0 private and 13.3 government) than that of the government schools. But this is not the whole story because the quality of the initial student intake is not the same, with the advantage going to the government schools.

We further divided the sample into two additional halves corresponding to private and government schools with P7 scores below the median and schools with P7 scores above the median (see Table 4).

Table 4. O-level points per dollar of unit cost for private and government secondary schools controlling for the Primary 7 Leaving Examination (P7) intake

School category

Category mean

P7 Score

Below median of P7

Above median of P7



(n = 12)


(n = 8)


(n = 4)



(n = 12)


(n = 1)


(n = 9)

Note: Table 4 presents n indicating number of schools for which O-level scores were available and number of schools for which both O and P7 scores were available.

In this analysis, we were able to compare private and government schools with roughly comparable quality of student intake, although there was only one government secondary school in our sample with a below-median-average P7 score. Nevertheless, the results are quite dramatic, with the private schools outperforming the government schools by a substantial margin in terms of O-level scores per dollar of unit cost.


Although these findings are not different from those of other researchers who have compared private and government schools, they are, generally speaking, more accurate. For one thing, we have calculated unit costs using the complex “ingredient” method, as opposed to the commonplace method of dividing aggregate spending by the number of students. For another, we have compared those costs against actual student test scores. And finally, we have the ability to hold constant initial student ability where the P7 is a proxy for ability.

Our attempts to build a multivariate model to predict O-level examination scores was met with moderate success. Because our primary interest here is to present our findings to the largest possible audience, including policy makers in Uganda, we will not give much space to modeling. A simple regression model in which two independent or predictor variables are allowed to enter produced an R square coefficient of .71. The common interpretation of the coefficient is the total explained variation in the O-level scores. A very substantial and significant 72% of variation is explained by just two variables: (1) students’ Primary Leaving Examination and (2) the unit cost. Controlling for the effect on O-level exam performance by including the P7 score in the equation, the unit cost variable is still highly significant, as seen in Table 5.

Table 5. Regression model: Dependent variable: O level examination score. R squared is .71 and SE of the estimate is 6.075









Std. Error




Unit cost in U.S. dollars

P7 score in standard form

Private (1) / Gov’t school (2)























There are several policy messages in these findings, and some are paradoxical. For example, on the one hand, unit cost data are helpful to policy analysts and policy makers in determining the most efficient use of money. In this case, private schools appear to be attractive, low-cost alternatives to the longstanding traditional government secondary schools. On the other hand, per-pupil spending is significantly related to learning achievement. If higher performance is wanted, perhaps more spending will be required. Put another way, for what private schools cost per pupil (somewhat less), they produce good learning gains—better, in fact, on a dollar basis, than government schools. At the same time, the more schools spend per pupil, whether government or private, the more pupils appear to learn. It may be the case that private schools are more efficient because they “buy” more of the input mix that actually influences student-learning achievement than do government schools. These may include, for example, better teachers who become affordable because they teach part time for private institutions that are not responsible for their pensions or health care. Private schools also buy less of things not highly related to learning, such as land.

The motivation of private school students, parents, teachers, and school administrators to succeed is considerable. Costs of private schools are held down through the consistent use of part-time teachers, often recruited from among the most talented teachers in the public schools. We will have more to say on the reasons for the remarkable performance of Ugandan private secondary schools in future publications. At this time, we conclude that private schools enjoy a considerable cost effectiveness advantage over government schools. Potentially, this research creates the need to extend the analysis beyond the comparison of examination scores and costs by also considering the intangible component gains of schooling experiences in both government and private schools.


Ablo, E., & Reinikka, R. (1998). Do budgets really matter? Evidence from public spending on education and health in Uganda (Policy Research Working Paper No. 1926). Washington, DC: The World Bank.

Aldrich, R. (2004). Public or private education? Lessons from history. New York: Routledge.

Anderson, S. E. (Ed). (2002). Improving schools through teacher development: Case studies of the Aga Khan Foundation Projects in East Africa. Lisse, The Netherlands: Swets and Zeitlinger.

Bagunywa, A. M. K. (1980). Critical issues in African education: A case study of Uganda. Nairobi, Kenya: East African Publishing House.

Bangay, C. (2005). Private education: Relevant or redundant? Private education: Decentralisation and national provision in Indonesia. Compare, 35, 167–179.

Brock-Utne, B. (2000). Whose education for all? The recolonization of the African mind. New York: Falmer Press.

Coombs, P. H., & Hallak, J. (1987). Cost analysis in education: A tool for policy and planning. Baltimore: Johns Hopkins University Press.

Cummings, W. K., & Dall, F. P. (1995). Implementing quality primary education for countries in transition. Amman, Jordan: Al Kutba/United Nations Children’s Fund.

de Regt, A., & Weenink, D. (2005). When negotiation fails: Private education as a disciplinary strategy. Journal of Education Policy, 20, 59–80.

Devarajan, S., Dollar, D. R., & Holmgren, T. (Eds). (2001). Aid and reform in Africa: Lessons from 10 case studies. Washington, DC: The World Bank.

Holsinger, D. B. (2000). Positioning secondary-school education in developing countries: expansion and curriculum. Paris: International Institute for Educational Planning (IIEP).

Hoyle, R. H., Harris, M. J., & Judd, C. (2002). Research methods in social relations (7th ed.). Belmont, CA: Wadsworth.

Jacob, W. J., Smith, T. D., Hite, S. J., & Cheng, S. Y. (2004). Helping Uganda’s street children: An analysis of the Model for Orphan Resettlement and Education (MORE). Journal of Children and Poverty, 10, 3–22.

James, E. (1994). The public-private division of responsibility for education. In T. Husén & T. N. Postlethwaite (Eds.), The international encyclopedia of education (pp. 4831–4836). Oxford, England: Pergamon Press.

Jarousse, J.-P., Mingat, A., Tamayo, S., & Tan, J.-P. (1996). Cost analysis in education: A three-part hands-on training module. Washington, DC: The World Bank.

Jimenez, E., & Lockheed, M. E. (1991). Private versus public education: An international perspective. International Journal of Educational Research, 15, 353–497.

Joint United Nations Programme on HIV/AIDS & World Health Organization. (2004). Uganda epidemiological fact sheets on HIV/AIDS and sexually transmitted infections: 2004 update. Geneva, Switzerland: Author.

Kitaev, I. (1999). Private education in sub-Saharan Africa: A re-examination of theories and concepts related to its development and finance. Paris: IIEP/UNESCO.

Levin, H. M. (1995). School finance. In M. Carnoy (Ed.), International encyclopedia of economics of education (pp. 412–419). Oxford, England: Pergamon Press.

Lin, J. (1999). Social transformation and private education in China. Westport, CT: Praeger.

Lassibille, G., Tan, J.-P., & Sumra, S. (1998). Expansion of private secondary education: Experience and prospects in Tanzania. Washington, DC: The World Bank.

Lassibille, G., Tan, J.-P., & Sumra, S. (2000). Expansion of private secondary education: Lessons from recent experience in Tanzania. Comparative Education Review, 44, 1–28.

Martin, W. A. (1976). The conceputalization and measurement of urbanization. Unpublished doctoral dissertation,University of Texas, Austin.

Martin, W. A., & Spurgin, J. H. (1983). Urbanization and political party competition: A reconceptualization. Texas Journal of Political Studies, 5, 58–68.

Ministry of Education and Sports. (1998). Education strategic investment plan 1998–2003. Kampala, Uganda: Sagimex Enterprises.

Ministry of Education and Sports. (2000). The contribution of non-government schools to primary and secondary education in Uganda. Kampala, Uganda: Education Planning Department.

Morisky, D. A., Jacob, W. J., Nsubuga, Y. K., & Hite, S. J. (Eds.). (2006). Overcoming AIDS: Lessons learned from Uganda. Greenwhich, CT: Information Age.

Moumouni, A. (1968). Education in Africa. New York: Praeger.

Neill, D. M. (1997). Transforming student assessment. Phi Delta Kappan, 79, 34–41.

Nielson, H. D., & Cummings, W. K. (1997). Quality education for all: Community-oriented approaches. New York: Garland.

Nsubuga, Y. K., & Jacob, W. J. (2006). A multisectoral strategy for overcoming AIDS in Uganda. In D. E. Morisky, W. J. Jacob, Y. K. Nsubuga, & S. J. Hite (Eds.), Overcoming AIDS: Lessons learned from Uganda (pp. 15–42). Greenwhich, CT: Information Age.

Nsubuga, Y. K., & Lematia, R. M. O. (2000). Strategic plan for post-primary education. Kampala, Uganda: Ministry of Education and Sports.

Rwabwoogo, M. O. (1998). Uganda districts information handbook. Kampala, Uganda: Fountain Publishers.

Ssekamwa, J. C. (2000). History and development of education in Uganda (2nd ed.). Kampala, Uganda: Fountain Publishers.

Ssekamwa, J. C., & Lugumba, S. M. E. (2001a). Development and administration of education in Uganda (2nd ed.). Kampala, Uganda: Fountain Publishers.

Ssekamwa, J. C., & Lugumba, S. M. E. (2001b). A history of education in East Africa Kampala, Uganda: Fountain Publishers.

Tooley, J., & Dixon, P. (2005). An inspector calls: The regulation of “budget” private schools in Hyderabad, Andhra Pradesh, India. International Journal of Educational Development, 25, 269–285.

Tsang, M. C. (1995). Public and private costs of schooling in developing countries. In M. Carnoy (Ed.), International encyclopedia of economics of education (pp. 393–398). Oxford, England: Pergamon Press.

Tsang, M. C., Wei, X., & Xiao, J. (Eds.). (2000). Economic analysis of educational policy. Beijing, China: People’s Education Press.

Cite This Article as: Teachers College Record Volume 110 Number 4, 2008, p. 867-893
https://www.tcrecord.org ID Number: 14626, Date Accessed: 5/26/2022 10:17:28 AM

Purchase Reprint Rights for this article or review
Article Tools
Related Articles

Related Discussion
Post a Comment | Read All

About the Author
  • W. James Jacob
    University of Pittsburgh
    E-mail Author
    W. JAMES JACOB is associate director, Institute for International Studies in Education (IISE). He is Visiting Assistant Professor, Department of Administrative and Policy Studies at University of Pittsburgh's School of Education. His primary scholarly focus is on the planning, development, and evaluation of international organizations and programs with geographic emphasis in sub-Saharan Africa, China, and the Pacific Rim. His most active current research interests involve HIV/AIDS organizations; higher education institutional change; social change and development; organizational development; and organizational effectiveness. Recent publications include Overcoming AIDS: Lessons Learned From Uganda (Information Age Publishing) and a 2006 article titled, “Social Justice in Chinese Higher Education: Issues of Equity and Access” (International Review of Education).
  • Donald Holsinger
    Brigham Young University
    DONALD B. HOLSINGER is professor of education and development at Brigham Young University in Provo, Utah. He has recently returned from spending 18 months in Cairo, Egypt, for USAID, where he was senior education advisor on leave of absence from BYU. Holsinger has served as a senior education specialist with the World Bank in Washington, D.C., where he was for a period of years responsible for that institution’s secondary education policy research program. Professor Holsinger has served as president and on the board of directors of the Comparative and International Education Society. He has held two Fulbright Research Fellowships to Brazil and Vietnam and has a long record of publications in leading education and development journals, as well as several books. Holsinger was a student of Alex Inkeles at Stanford University, where he received his PhD in 1972.
  • Christopher Mugimu
    Makerere University
    CHRISTOPHER B. MUGIMU is a lecturer at Makerere University’s Department of Curriculum, Teaching, and Media.
Member Center
In Print
This Month's Issue