Achieving Universal Primary and Secondary Educationin Uganda

Access and Equity Considerations

B. Essama-Nssah, Phillippe G. Leite and Kenneth R. Simler[*]

Poverty Reduction and Social Protection Groups

The World Bank

Washington D.C.

December 17, 2008

Abstract

Education policy in Uganda has evolved over time from elitism to inclusiveness and is now at the center of the country’s Poverty Eradication Action Plan (PEAP). This paper uses the 1992 and 2005 nationally representative household surveys to assess the impact of the 1997 Universal Primary Education (UPE) and the 2007 Universal Secondary Education (USE) policies. Both policies entail the abolition of school fees for these two levels of education. We find that UPE did reduce school fees paid by families at the primary level of education. The reduction in school fees in turn led to a significant improvement in both access to and equity in primary schooling. However the policy did not increase the likelihood of completing primary schooling nor did it have any spillover effects for secondary education. Simulations from a probability model show that USE is likely to induce a big surge in secondary enrolment similar to the one associated with UPE at the primary level of education. This policy also has the potential of significantly improving the distribution of post primary education and training opportunities. The feasibility of these potential outcomes depends however on investment in space and other complementary inputs at levels commensurate with the expected influx of students.

1

Introduction

The Millennium Declaration (United Nations 2000) identifies income, health, education, shelter and governance as critical components of the living standard that every policymaker should promote for the population. Similarly, Becker (1995) argues that a country’s living standard is primarily determined by how well it succeeds in developing and utilizing the skills, knowledge, health and habits of its people. In this context, education plays a critical role in providing individuals and society with skills and knowledge to maintain and improve their standard of living. A key policy issue facing developing countries in particular is how to design and run their education systems in a way that effectively contributes to rapid socioeconomic development in a global and knowledge-based economy.

Formal education in Uganda comprises three basic levels. The primary level consists of seven grades (or standards) referred to as P1-P7 (where P stands for Primary). Children are supposed to start primary school when they are six years old. Successful completion entails an end of cycle examination, the PLE (or Primary School Leaving Examination). Pupils who have successfully completed primary school have two basic options at the secondary level namely general secondary, and vocational and technical education. General secondary education has two levels: the O-level and the A-level (where O stands for ordinary and A stands for advanced). The O-level or lower secondary takes 4 years and the corresponding grades are S1-S4 (S stands for senior[1]). At the end of this cycle, students sit for the Uganda Certificate of Education (UCE). Only UCE holders are eligible for upper secondary education at the A-level which covers forms S5-S6. At the end of this second cycle, students sit for the Uganda Advance Certificate of Education (UACE). The performance in this examination determines eligibility for university education. Both the UCE and the UACE are national examinations designed for certification and selection (Liang 2002). Total enrolment in primary school in 2007 was estimated at more than 7 million, (exactly 7,414,880) pupils (MoES 2007). For the same year, enrolment in the secondary education was estimated at 842, 683 students.

Parallel to general secondary education, there is vocational and technical education system offered in the so-called BTVET (Business, Technical and Vocational Education and Training) institutions[2]. This generally takes three years. Primary school graduates also have the option of entering four year teacher training colleges. Most of the BTVET institutions are public (87.1 percent in 2007) unlike secondary education where only 35percent of schools were government-owned in 2007. The share of private and community secondary schools in 2007 was respectively 47 and 18 percent (MoES 2007). A network of universities and colleges offer tertiary education.

A variety of socio-political crises in the 1970s and 1980s reversed the progress achieved by the education sector in Uganda since independence. It is reported for instance that, in 1985, the level of government expenditure on education was about 27% of that of the 1970s. Due partly to high costs to families, the gross primary enrolment rate stayed relatively constant for two decades after independence. It stood at 50 percent in 1980, the same rate as in 1960. A significant improvement was observed in 1985 when the gross primary enrolment rate increased to 73 percent. However, it remained at that level for about a decade until 1995 (Appleton, 2001).

Over time, education policy in Uganda has been shifting from elitism to promoting inclusiveness. Considering that education has both intrinsic and instrumental value for development[3], the Government of Uganda has placed education at the center of its Poverty Eradication Action Plan (PEAP). In 1997 the government committed itself to the provision of universal primary education (UPE). The UPE policy is in fact an integral part of a broader reform program which can be traced back to the creation in 1987 of the Education Policy Review Commission (EPRC) to formulate policy recommendations for all levels of education. In their 1989 report to the government, the Commission set an ambitious goal for primary education and recommended that education policy should ensure that every child enrolls at the appropriate age and successfully completes the full cycle[4] (Grogan 2008). The first response to this recommendation came in 1993 in the form of the Primary Education and Teacher Development Project supported by the United Nations Children Fund (UNICEF) which sought among other goals to introduce a country-wide evaluation framework for the overall progress in education.

The sudden increase in the demand for public primary education created a series of challenges for the government. These challenges relate mainly to (1) financing the reform, (2) improving the quality of primary education, (3) ensuring equitable access to primary education, and (4) planning beyond primary school. To address these challenges, the government adopted in 1998 the Education Strategic Investment Plan (ESIP)[5]. How well did the government respond to these challenges within the ESIP framework? Two major factors help ease the financing constraint. First, there was an improvement in aid delivery mechanisms. The government and its partners adopted a sector-wide approach (SWAP) instead of a project-based one to deliver support to the education system via the Medium Term Budget Framework (MTBF). The effectiveness of this institutional arrangement was further enhanced by the development of an effective and reliable Education Management Information System or EMIS (Penny et al. 2008). Second, the Heavily Indebted Poor Countries (HIPC) initiative led to an increase in funds available to the education sector.

As for the need to improve the quality of primary education, the government has focused on five areas: (1) curriculum development, (2) provision of basic learning materials, (3) primary teacher development, (4) language of instruction for lower primary, and (5) establishment and maintenance of standards. There is evidence that reform in all these areas has not yet produced a significant improvement in the quality of teaching and learning in Uganda. Test results from the National Assessment of Progress in Education (NAPE) between 1996 and 2000 show deterioration in student performance in terms of numeracy, reading, science and social studies (Bategeka 2005). The key factor explaining this outcome is a lack of coherence and consistency within the whole system composed of policy and administrative reforms and changes in teaching and learning methods (Penny et al. 2008).

Overall education policy seeks to keep increasing in a sustainable manner equitable access to quality and relevant education. Thus, ten years after the introduction of the UPE program, the government launched (in 2007) Universal Post Primary Education and Training (UPPET) based on a partnership with the private sector[6]. The policy is also known as USE for Universal Secondary Education. This policy is particularly designed to address inequality in the distribution of PPET by deliberately targeting able primary school graduates who cannot afford the fees charged at the post primary level. In particular, user fees are abolished (through a scholarship or bursary program) for the targeted socioeconomic groups and conditional grants are made to participating institutions to help with their operational costs. The government is also engaged in a program of facility expansion.

The purpose of this paper is to assess the social implications of education policy reform in Uganda. The general evaluative issue of interest is the effects of the shift in policy regime (school fee elimination along with associated institutional reforms) on educational outcomesand the distribution of such outcomes. In particular we focus on the effect of the reform on enrolment, primary school completion and the cost of schooling. As far as equity is concerned, we consider how the policy effects vary with household income (poverty dimension), gender and area of residence (regional inequities). Our assessment relies on both the available literature on the effects UPE and our own econometric analysis of data from two nationally representative household surveys: the 1992 Uganda Integrated Household Survey (UIHS) and the 2005/2006 Uganda National Household Survey (UNHS).

The outline of the paper is as follows. Section 2 reviews some trends in the distribution of school attendance. Section 3 analyzes the effects of the 1997 reform on enrolment, primary school completion and the cost of schooling. The fact that we have repeated cross-section data imposes a constraint on the identification and estimation of the relevant policy effects and the comparison of outcomes across relevant socioeconomic groups. Section 4 focuses on the more recent USE policy and analyzes its likely impact on the propensity to enroll. Conclusions are presented in section 5.

2. Trends in the Distribution of School Attendance

Table 2.1 Attendance Rates by Quintile of the per capita Expenditure Distribution

Uganda 1992 and 2005

6 - 12 years / 6 - 8 years / 9 - 12 years / 13-18 years
Quintile / All / Girls / Boys / All / Girls / Boys / All / Girls / Boys / All / Girls / Boys
1992
Total / 66.5 / 64.2 / 68.7 / 55.6 / 53.1 / 58.0 / 76.8 / 74.6 / 78.8 / 12.0 / 10.9 / 13.0
1 / 50.2 / 45.7 / 53.9 / 37.2 / 33.5 / 40.5 / 62.3 / 58.0 / 65.7 / 2.6 / 2.4 / 2.8
2 / 62.2 / 58.7 / 65.4 / 50.5 / 47.3 / 53.4 / 73.6 / 69.7 / 77.2 / 6.2 / 4.9 / 7.4
3 / 68.5 / 66.5 / 70.5 / 54.7 / 51.5 / 57.9 / 81.6 / 81.0 / 82.2 / 9.2 / 9.1 / 9.3
4 / 73.7 / 70.9 / 76.3 / 65.2 / 62.9 / 67.2 / 81.8 / 78.1 / 85.6 / 14.5 / 11.9 / 17.5
5 / 84.2 / 83.1 / 85.5 / 79.0 / 77.6 / 80.4 / 88.8 / 87.6 / 90.2 / 27.5 / 24.1 / 31.3
2005
Total / 88.1 / 88.4 / 87.7 / 80.7 / 81.4 / 80.1 / 93.6 / 93.5 / 93.7 / 21.9 / 23.5 / 20.4
1 / 78.8 / 78.9 / 78.8 / 67.1 / 68.1 / 66.2 / 87.7 / 86.7 / 88.7 / 4.5 / 6.4 / 2.8
2 / 85.8 / 86.0 / 85.5 / 76.0 / 75.5 / 76.5 / 94.1 / 95.1 / 93.1 / 11.8 / 11.5 / 12.1
3 / 89.7 / 90.1 / 89.3 / 81.9 / 82.1 / 81.7 / 95.8 / 95.9 / 95.7 / 17.4 / 19.3 / 15.7
4 / 93.0 / 93.3 / 92.7 / 89.6 / 91.2 / 88.0 / 95.5 / 95.0 / 96.1 / 22.5 / 23.4 / 21.5
5 / 94.5 / 94.7 / 94.3 / 92.9 / 93.7 / 92.1 / 95.6 / 95.4 / 95.9 / 44.7 / 44.8 / 44.6
*All figures are percentages of respective age group.
Source: Computations from UIHS 1992 and UNHS 2005/2006 data

Our empirical analysis is based on the 1992 and 2005 nationally representative household surveys combined with two community surveys for the same years[7]. From the original datasets we retain a sample 16,607 individuals who were between 6 and 18 years old in 1992, and 14, 984 individuals of the same age group in 2005. In addition to the household/individual level information, we also merge the two surveys with a community level questionnaire that contains extra information about the level of development of the community that can be used as controls in our model. For example, we will use existence of primary and secondary (either public or private) school in the community and the average of fees paid in the community as controls later on in our analysis of correlation of fees paid and school attendance.

Table 2.2 School Attendance and Reasons for Non-Attendance by Region

Uganda 1992 and 2005

Uganda / Rural / Urban / Center / East / North / West
1992
School attendance
6-12 year olds in primary school (%) / 66.5 / 64.6 / 82.2 / 78.8 / 67.1 / 51.6 / 65.0
6- 8 year olds in primary school (%) / 55.6 / 52.7 / 79.2 / 72.2 / 52.9 / 37.7 / 55.4
9-12 year olds in primary school (%) / 76.8 / 75.7 / 85.1 / 85.0 / 80.3 / 65.1 / 73.8
13-18 year olds in primary school (%) / 43.6 / 45.0 / 34.9 / 38.5 / 45.1 / 50.8 / 42.1
13-18 year olds in secondary school (%) / 12.0 / 9.1 / 29.0 / 18.5 / 10.8 / 8.0 / 8.5
Reasons for never attending/dropping out*
Calamity in family, pregnancy, disabled / 1.2 / 0.8 / 6.6 / 3.2 / 0.8 / 0.5 / 1.0
Lack of interest / 49.2 / 49.5 / 45.2 / 44.9 / 51.5 / 50.3 / 48.6
Need to work / 1.7 / 1.7 / 2.1 / 0.3 / 0.8 / 4.2 / 0.7
Cost of attendance / 42.1 / 42.0 / 43.5 / 45.5 / 42.4 / 36.1 / 46.7
Transport/distance / 3.3 / 3.4 / 0.7 / 3.7 / 2.3 / 5.0 / 1.8
Poor quality of school / 0.3 / 0.3 / 0.0 / 0.6 / 0.5 / 0.2 / 0.2
Other / 2.2 / 2.2 / 1.9 / 1.8 / 1.7 / 3.8 / 1.0
Number of individuals (6-18 years) / 16,607 / 11,027 / 5,580 / 4,427 / 4,172 / 4,089 / 3,919
2005
School attendance
6-12 year olds in primary school (%) / 88.1 / 87.4 / 92.1 / 92.7 / 89.7 / 79.2 / 88.4
6- 8 year olds in primary school (%) / 80.7 / 79.5 / 89.2 / 89.4 / 81.5 / 67.7 / 81.2
9-12 year olds in primary school (%) / 93.6 / 93.6 / 94.1 / 95.0 / 96.2 / 88.4 / 93.7
13-18 year olds in primary school (%) / 57.1 / 61.2 / 35.7 / 45.1 / 62.4 / 64.9 / 60.8
13-18 year olds in secondary school (%) / 21.9 / 18.4 / 40.0 / 31.7 / 22.4 / 10.3 / 18.1
Orphans (%) / 5.5 / 7.3 / 5.2 / 4.9 / 7.5 / 3.2 / 6.9
Children with one alive parent (%) / 16.7 / 16.4 / 18.6 / 19.8 / 14.9 / 15.8 / 15.6
Reasons for never attending/dropping out*
Calamity in family, pregnancy, disabled / 7.5 / 7.5 / 8.0 / 11.4 / 9.5 / 5.1 / 6.8
Lack of interest / 56.3 / 58.8 / 26.9 / 29.9 / 73.7 / 57.0 / 57.6
Need to work / 8.9 / 9.3 / 4.2 / 3.8 / 2.5 / 17.9 / 4.0
Cost of attendance / 12.8 / 10.1 / 44.7 / 36.1 / 6.5 / 4.7 / 15.4
Transport/distance / 5.8 / 6.3 / 0.0 / 5.4 / 2.6 / 4.6 / 11.0
Poor quality of school / 1.2 / 1.3 / 0.5 / 4.1 / 0.2 / 0.7 / 1.1
Orphaned/displaced / 1.1 / 0.9 / 3.8 / 0.6 / 1.6 / 1.4 / 0.5
Other / 6.3 / 5.8 / 11.9 / 8.7 / 3.5 / 8.6 / 3.5
Number of individuals (6-18 years) / 14,984 / 11,895 / 3,089 / 3,975 / 4,087 / 3,238 / 3,684
Source: Computations from UIHS 1992 and UNHS 2005/2006 data
* Percentages for reasons for non-attendance refer to children 6-12 years old

Table 2.1 presents attendance rates by quintile of the distribution of per capita expenditure for both 1992 and 2005. For each year and each socioeconomic group considered here, the data clearly show that attendance is an increasing function of household income.

Focusing on 1992, prior to the introduction of UPE, we note a big gap between the poor and the non-poor. The attendance rate for the poorest quintile in the 6-12-year old group is about 50 percent compared to 84 percent for the richest. A similar gap is observed for the other age groups as well. The data also reveal a gender bias for all age groups and income levels. Attendence rates are consistently higher for boys than for girls.

Comparing 1992 attendance rates with 2005 reveals a significant increase in overall attendance among primary school age children (6-12 years old) and that of the poor. At the bottom quintile, attendance rate has increased by more than 28 percentage points (from 50.2 percent in 1992 to 78.8 percent in 2005). However, the gap between the bottom and the top quintile does not seem to be narrowing. The gender difference in attendance seems to have disappeared in 2005. As noted in the introduction, one of the education policy objectives in Uganda is to get students to enroll at the right age. The age of enrolment is indeed a key determinant of dropout rates. Focusing now on the 6-8-year age group, we note that they have the lowest attendance rates both in 1992 and 2005 compared with the 9-12-year age group. However, the enrolment rate seems to have gone up significantly, jumping from 37 percent in 1992 to about 81 percent in 2005.

Attendance rates for secondary school age children (13-18 years old) did also increase between 1992 and 2005. The increase, however, does not seem as dramatic as for the case of primary school age children. While there was a gap between boys and girls in 1992, this seems to have disappeared in 2005. Overall, the enrolment rate for girls is somewhat higher than that of boys (23.5 percent versus 20.4 percent).

Table 2.2 presents information on attendance and reasons for non-attendance by region. There is a bias infavor of urban residents both for the 6-12-year and the 13-18-year age groups. For the primary school age group this bias has shrunk significantly between 1992 and 2005. The difference between enrolment rate for urban children and that of rural children decreased from about 18 percent in 1992 to about 5 percent in 2005. For the secondary school age children, this difference remained at about 20 percent[8]. Looking at the regional distribution of attendance, we note that for both years, the Central Region has the highest attendance rates for all age groups while the Northern Region has the lowest.

The information about the reasons for non-attendance or dropping out reveals two key determinants of this outcome: lack of interest and the cost of attendance. However, the importance of cost as a reason for not attending school has significantly declined. The overall percentage of children non-attending school for cost-related reasons decreased form 42 percent in 1992 to about 13 percent in 2005. The urban-rural comparison shows that the level of this constraint has not changed in the urban areas and has not declined much for the Central Region (going from about 46 percent in 1992 to about 36 percent in 2005). Lack of interest remains a severe constraint to attendance. The overall score for this factor increased from 49 percent in 1992 to 56 percent in 2005. Looking at the regional distribution we note that it decreased only for the Central Region, moving from 45 percent in 1992 to 30 percent in 2005

These observed trends in enrolment suggest that the UPE policy may have induced a dramatic increase in primary school attendance, but (as noted earlier) may have also led to a decline in the quality of education. For policy analysis it is important to ascertain that this is not due to some diffuse change in the overall socioeconomic environment. We need to determine the extent to which the outcome can be attributable to the policy shift. In the next section, we present some econometric evidence that ought to increase our confidence, to a significant extent, that the increase in primary enrolment is due to the 1997. We will also demonstrate that the policy shift had no significant spill-over effects on secondary school outcomes. Furthermore, we will assess the equity dimension of these results by considering the impact across various socioeconomic groups.

3. The Impact of UPE

As noted earlier, the UPE policy is in fact an integral part of a broader reform program. But it is commonly characterized by the elimination of fees at the primary school level. This is accomplished through two types of grants designed to increase equitable access to primary education: the capitation and school facilities grants (SFG), the purpose of the capitation grant is to shift some of the burden of school fees from the parents to the government[9], and to provide schools with resources necessary to run the school and support teaching and learning. The capitation grant is meant to cover tuition fees only. Families remain responsible for writing material, uniforms and lunches. This grant is paid to schools in nine monthly installments at an annual rate of 5000 Ugandan shillings (about $3.00) per pupil in P1 to P3 and 8100 Ugandan shillings ($4.75) per pupil in P4 to P7 (Penny et al. 2008)[10]. The SFG is designed to assist schools in the neediest communities to build classrooms, latrines and teachers’ houses, and to procure furniture. After a favorable review in 1999, it became the only mechanism through which public funds are channeled to the construction of school facilities. By 2004, a total of 29,000 have been built. Both grants are conditional to the extent that funding is given to Districts or municipalities under strict guidelines and regulations and the supervision of the Ministry of Education and Sports[11].