Uganda Note Note No. 5 – Education SABER Equity & Inclusion Program LEVELS AND EFFICIENCY OF SPENDING FOR EDUCATION Quentin Wodon, Elizabeth Ninan, Innocent Mulindwa, Mamy Rakotomalala, and Clarence Tsimpo, March 2016 KEY MESSAGES:  Due to population growth and higher attainment, enrollment is growing. Public funding for the sector has not kept pace, thereby shifting about two thirds of the financing of the education system to households. Cost-effectiveness is also low due to low internal efficiency.  Likely future increases in enrollment will require substantially higher budgets, as well as efforts to increase internal efficiency by reducing repetition and drop-out rates.  . Introduction Box 1: Note and Series Primer Due to population growth and gains in education Why a series of notes on education in Uganda? The Ministry attainment, the size of Uganda’s education system is of Education, Science, Technology and Sports (MoE) is expected to grow dramatically in the next dozen years in preparing a new education strategy. This note is part of a technical assistance task at the World Bank to help inform this Uganda. This will create pressure on scarce budget new strategy and the country’s efforts to ensure learning for all. resources. In order to accommodate larger cohorts of students and provide an education of quality, gains in What are the topics discussed in the series? The series internal efficiency will have to be achieved, but in addition looks among others at the importance of education for Uganda’s substantial additional funding will also be needed. development, education sector priorities including options and trade-offs, the efficiency and effectiveness of education This note discusses issues related to the level and expenditure and financing modalities, and implementation efficiency of spending for education in Uganda. After a challenges for education interventions. Additional topics, brief review of recent enrollment trends and levels of including the importance of early childhood development and gender gaps in education, are also considered. education attainment in the country, data on both public and private spending for education are provided. One key What is the question asked in this note? The question is: finding is public spending for education has not kept pace How much is spent on education in Uganda, is the system with enrollment. In part as a result, households currently efficient, and what may be spending requirements in the future? shoulder almost two thirds of public and private education spending in the country. Potential future trends in student How is the question answered? Statistical analysis of enrollment as well as budget implications are also household survey and administrative data as well as budge data discussed, suggesting that education spending as a share analysis and simulations are used to analyze trends in of GDP may need to increase substantially to achieve attainment, enrollment, expenditure, and future budget needs. universal upper secondary completion. Uganda Education Note  The number of students enrolled in school has 7.2 million in 2007 to almost 8.0 million in 2013. Annual increased over time, including at the secondary level growth is higher in upper primary (P5-P7) at 3.3 percent thanks to the universal secondary education policy. than in lower primary (P1-P4) at 1.0 percent. In Secondary Education, the number of pupils increased Uganda’s education system follows a 7-4-2 structure. from 0.95 million in 2007 to 1.36 million in 2013, with an Students complete seven years of primary schooling annual growth rate of 6.2 percent. There was a sharp (ages 6-12) which is compulsory and free under the increase in enrolments from 2007-2009 directly after the Universal Primary Education (UPE) policy. At the end of universal secondary education (USE) policy was primary, pupils take the Primary Leaving Examination introduced. Growth in enrolment slowed down after that. (PLE), which determines their eligibility to secondary school or technical school (vocational). PLE graduate opt for either four years of general secondary education or Figure 1: Enrollment in Primary 4,200,000 three years of vocational training at a technical school. At the end of this schooling/training cycle, pupils take the 4,000,000 national Uganda Certificate of Education (UCE) examination or Uganda Junior Technician’s Certificate 3,800,000 (UJTC) Examination. This is followed by two years of upper secondary school or more training at a Technical Boys 3,600,000 Institute. Students may also pursue tertiary education. Girls 3,400,000 The latest Education Sector Strategy Plan (ESSP: 2004- 2015) called for sustained investments in post-primary 3,200,000 education to provide future workers with the skills and 2007 2008 2009 2010 2011 2012 2013 competencies needed to enter formal labor market and Source: Compiled by the authors from MoES data. increase productivity. In recognition to this, in 2007, the Government of Uganda launched a comprehensive Figure 2: Enrollment, Lower Secondary reform program to provide Universal access to post primary education and training (UPPET) to embrace the 800,000 bulge from free primary education provided under UPE. The Government of Uganda created two sub-programs 600,000 under UPPET: (i) Universal Secondary Education (USE) to improve transition rates from primary to secondary 400,000 schools; and (ii) equivalent grades for vocational Boys education (schools/community/polytechnics). Through Girls UPPET, Uganda became the first country in Sub-Saharan 200,000 Africa to introduce Universal Secondary Education Policy. 0 The introduction of Universal Secondary Education (USE) 2007 2008 2009 2010 2011 2012 2013 policy remained a symbolic and political decision of the Government. However, tightening of fiscal space from Source: Compiled by the authors from MoES data. 2006 to 2012 resulted in shortage of funds to finance the projected costs of USE under ESSP. The allocation of Administrative data suggest a decline in the Gross resources since the revision of ESSP (2009/2010) has not Enrollment Ratio (GER) in primary education from 119.8 quite matched the projections. As will be discussed below percent in 2007 to 109.8 percent. This means that the in this note, annual education budgets in real terms have education system is able to absorb more than 100 percent remained fairly stagnant and shortfalls in budget of pupils ages 6-12. The slight decrease in the GER may allocations have undermined the ability of Ugandan be the result of gains in efficiency (fewer repetition) as Government to meet the policy ambitions of ESSP in all well as a reduction over time in the number of over-age sub-sectors, and even more so in the nascent USE policy. children entering the primary school system, a major Budget pressure are expected to continue for the medium issue in Uganda. The GER for secondary Education has to long term due to further increases in student improved modestly since 2007, essentially thanks to gains enrollment, as will also be discussed in this note. in 2007-2009 from the USE policy. In lower secondary, the GERs for boys and girls increased from 33.0 percent Looking back at the last half dozen years, due to both and 27.2 percent respectively in 2007 to 38.5 percent and population growth and higher education attainment on the 34.8 percent in 2013. GERs for upper secondary part of students, the number of students in primary and education did not improve much over that period, especially secondary education increased substantially. In increasing from 13.1 percent and 8.9 percent in 2007 for Primary Education, the number of pupils increased from boys and girls to 14.9 percent and 9.6 percent in 2013. Uganda Education Note  the same time, education attainment remains low. At rates also increased slightly from 10 percent to 13 percent For every student who starts primary school, only over the same period. Still, the issue of drops-out remains one in eight completes upper secondary education. a major issue at all levels, and have major implications for the internal efficiency and cost effectiveness of the Many children do not complete their education, not only at education system, as will be discussed below. the secondary level, but also at the primary level. An simple way to show this consists in providing a schooling Figure 4: Evolution of Key Education Indicators profile describes access at each grade measuring the 140% 131% proportion of children that have access to school and 114% proportion of pupils that stay in school. Using 120% 107% administrative data, cross-sectional schooling profiles are 100% constructed for 2008 and 2013 in Figure 3. There is a % Access decrease in the adjusted gross in-take rate from 130.8 80% 56% 59% 59% percent in 2008 to 107.3 percent in 2013. This is mainly 60% the result of a reduction in multiple-cohort entry induced 38% 41% 40% by the introduction of free primary education since 1997. 40% 30% 31% 24% Uganda experienced very high intake rates for several 20% 10% 10% 13% years after the introduction of UPE, but this multi-cohort phenomenon is now gradually fading away even If late 0% age of entry still remains an issue. 2007 2010 2013 Access P1 Primary completion Access S1 Complete S4 The school profile is also relatively stable (i.e., does not Complete S6 decline) from P1 to P4 implying fairly high retention rates in those grades. There is however a large drop in the Source: Authors, computed from Uganda EMIS data. profile after completing lower primary with only 59 percent students enrolled until P7, the end of the primary cycle. Figure 5 summarizes in a simple way the implications of The transition rates to secondary school are even lower the probabilities of transition through the system for an with only 40 percent children entering the secondary average student in 2013. For every 100 students who school cycle and 12.5 percent completing S6 which enroll in primary school, 59 complete the cycle. Of those, corresponds to upper secondary school. 40 transition into lower secondary school. Secondary. Only 31 students complete the lower secondary cycle, Figure 3: Schooling Profile by Age, 2008 and 2013 (S4) and only dismal 13 students complete the upper 140% secondary cycle (S6). 120% 100% For every 100 students who enroll in primary school, 59 complete the cycle. Only 31 students complete lower % access 80% secondary, and 13 complete the upper secondary cycle. 60% 2008 40% 2013 Figure 5: Education Pyramid for Uganda, 2013 20% 13 13 0% 13 12 Transition: 40 P1 P3 P5 P7 S2 S4 S6 11 31 Level 10 9 8 40 Source: Authors, computed from Uganda EMIS data. Transition: 68 Years 7 6 59 The school age profile did not improved substantially 5 between 2007 and 2013. As shown in Figure 4 primary 4 completion rates improved modestly from 56 percent in 3 2007 to 59 percent in 2013. Transition rates from primary 2 to secondary school remained essentially flat at around 107 1 40 percent. There are some gains in the secondary sub- 0 cycle. Lower secondary school completion rates 0 50 100 increased from 24 percent in 2007 to 31 percent in 2013, Percentage in large part thanks to the universal secondary education Source: Authors, computed from Uganda EMIS data. policy introduced in 2007. Upper secondary completion Uganda Education Note  High repetition and drop-out rates contribute to low Internal efficiency for a schooling cycle is defined as the levels of internal efficiency in the education system. total number of years of schooling in that cycle for a cohort of students divided by the years of schooling Simple statistics can be used to estimate the international needed to complete the cycle times the number of efficiency of an education system. As traditionally defined, students who complete the cycle. Due to high repetition internal efficiency for a given education cycle is defined as and drop-out rates, internal efficiency is low in Uganda. the number of years that should have been needed by those who completed the cycle divided by the number of Table 2 provides the estimates of internal efficiency. At years of schooling actually used not only by those who completed the cycle, but also by those who dropped out the primary level, thanks to slightly higher completion before completion. Two factors can thereby result in rates, there was a small increase in internal efficiency from 50.0 percent in 2007 to 55.4 percent in 2013. At the losses in internal efficiency: repetition and drop-outs. secondary level, internal efficiency remained essentially  Repetition reduces internal efficiency because flat at about 44 percent. As expected, most of the losses students take longer to complete a cycle than in efficiency come from drop-outs as opposed to they should. If 10 percent of students repeat each repetition. Overall, as a result of inefficiencies, it took 12.6 grade and then move on to the next grade, on years to produce a primary school completer in 2013, average students will need 1.1 years to complete which is only slightly better than the 14 years it took to each year in the cycle, resulting in an internal produce a completer in 2008. In secondary school as well, efficiency of only 91 percent (0.91=1/1.1). inefficiencies prevail, in large part because of low  Drop-outs also reduce internal efficiency, and transition to upper secondary. In order to improve typically more so than repetition rates because all efficiency, repetition rates would have to be reduced, and the years of schooling of those who drop out even more importantly, survival rates would have to be count towards the denominator of the internal increased so that a larger share of students starting the efficiency measures while none of those years primary and secondary cycles complete the cycles. count towards the nominator. Table 2: Internal Efficiency Estimates, 2007 and 2013 Table 1 provides the parameters used to estimate the 2007 2013 internal efficiency of Uganda’s education system for Primary primary and secondary education using administrative Number of years in cycle (1) 7 7 data from Uganda’s Education Management Information Completion rate for cycle (2) 42.3% 54.8% System. The first column indicates the grade. The next Required years for completers(1)*100*(2) 296.1 383.3 column indicates the survival rate from one grade to the Actual years w/o repetition 536.5 627.0 next. For example, for 100 students entering the first year Actual years with repetitions 592.5 692.1 of primary school in 2013, 54.8 percent are expected to Internal efficiency = 50.0% 55.4% complete the cycle. The next column indicates the share Impact of drop-outs 55.2% 61.1% Impact of repetition 90.6% 90.6% of students who repeat the grade. Due to data constraints, Secondary the same repetition rates are used for 2007 and 2013. Number of years in cycle 6 6 Completion rate for cycle 27.3 31.2 Table 1: Internal Efficiency Parameters, 2007 and 2013 Required years for completers(1)*100*(2) 163.7 187.0 Gr. Surv. (%) Repeats (%) Gr. Surv. (%) Repeats (%) Actual years w/o repetition 362.7 417.3 Primary, 2007 Primary, 2013 Actual years with repetitions 368.8 424.6 1 100.0 11.6 1 100.0 11.6 Internal efficiency 44.4% 44.0% 2 86.6 9.6 2 99.3 9.6 Impact of drop-outs 45.1% 44.8% 3 89.5 10.1 3 100.0 10.1 Impact of repetition 98.3% 98.3% 4 83.3 11.2 4 100.0 11.2 Source: Authors, based on Uganda EMIS data. 5 73.6 11.1 5 94.3 11.1 6 61.1 11.5 6 78.6 11.5 7 42.3 5.8 7 54.8 5.8 As a result of inefficiencies, it took 12.6 years to produce Secondary, 2007 Secondary, 2013 a primary school completer in 2013, which is only slightly 1 100.0 0.6 1 100.0 0.6 better than the 14 years it took to produce a completer in 2 74.7 1.1 2 91.2 1.1 2008. In secondary school as well, inefficiencies prevail, 3 68.5 1.7 3 87.9 1.7 in large part because of low transition to upper secondary. 4 63.4 4.3 4 78.2 4.3 5 28.9 1.1 5 28.9 1.1 6 27.3 2.1 6 31.2 2.1 Source: Authors, computed from Uganda EMIS data. Note: The same repetition rates are used for both years. Uganda Education Note  exact budget implications from improvements in survival Box 2: Reasons for Dropping Out rates depend on how the gains are achieved (i.e., they Repetition rates and especially drop-outs contribute to low levels depend on the reduction in drop-out rates by grade). of internal efficiency in Uganda’s education sector. A separate note in this series documents the factors that lead students to Public investments in education have not kept pace drop out of school. Cost remains a major issue for dropping put with the growth in enrollment and remain below the prematurely, but in addition, especially at the secondary school level, other factors also play a role, including child marriage and levels envisioned in the country’s education strategy. pregnancies for girls. The fact that many students drop out because they are not interested in pursuing their education also Public support for education is essential to ensure that all suggests issues of quality, which are also discussed separately children get the opportunity to receive a quality education. in this series of notes on the education sector. In Uganda, while the nominal budget for education has steadily increased, funding in real terms has remained Options could be considered to reduce dropping out rates. If relatively flat over the last decade (see Figure 6 where cost is a major obstacle to schooling for the poor especially, budget data in real terms are in 2010 prices). After a interventions to reduce out of pocket and opportunity costs are gradual decline from 2002 to 2009, funding increased needed to enable those children to go to school and remain in school. Unfortunately, the coverage of these interventions in thereafter, but the level reached in 2013 was not Uganda tends to be low, and existing interventions are not very substantially higher than in 2002. well targeted. For those children who are too far from existing schools, expanding the school network or providing mobility This also implies that public funding has not reached the should be the priority. Finally, for specific groups of children, (ambitious) projections made under the current Education such as children with disability, additional specific interventions Sector Strategy Plan (ESSP). In addition, given would be needed. Potential interventions for other groups, such substantial increases in enrollment, as discussed earlier, as girls at risk of early marriage or an early pregnancy could also public spending per student has decreased, possibly be considered. Additional research would be needed to suggest contributing to lower levels of achievement. in more details cost effective option to reduce drop-outs. Figure 6: Education Sector Budget (UGX Bn) Improving internal efficiency would increase cost 1,800 effectiveness, but could also increase budget needs 1,600 as more students complete cycles (positive outcome). 1,400 In terms of cost efficiency, if one assumes that there is no 1,200 gain for children from only some primary schooling if they 1,000 do not complete the full cycle (a somewhat extreme assumption), the implication would be that the cost of the 800 education system in 2013 was 1.8 times what it should be 600 at the primary level (1.8 = 1/0.554). This high cost is born not only by the state (in terms of budgetary expenditures) 400 but also by families (in terms of out of pocket and 200 opportunity costs). At the secondary level, the cost of 0 service provision in 2013 was 2.3 times what it should 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 have been (2.3 = 1/0.440). Uganda’s education system is Education Budget (nominal) Education Budget (constant 2010 prices) clearly not cost effective, and from a cost point of view, a Source: Tsimpo and Wodon (2016). top propriety should be to improve internal efficiency. Public spending for education has declined as a share of The cost of the education system in 2013 was up to 1.8 GDP and the budget. In other words, relative investments times what it should have been at the primary level given in education through the budget have declined in students who completed the cycle. At the secondary level, comparison to other uses of available resources. In the cost of service provision was up to 2.3 times what it addition, public investments in education have not kept should have been. These inefficiencies result in high pace with the increase in student enrollment over time. costs not only for the state, but also for parents. Figures 7 through 9 provide more details on the spending Importantly, while improving internal efficiency would lead trend from 2009 to 2013, a period of moderate growth in to higher cost effectiveness, the implications for total the education budget after the decline observed in the expenditures for education are not the depending on how previous six years. In terms of allocations by sub-sector, the gains in internal efficiency would be achieved. While as shown in Figure 7, funding increased in real terms for lower repetition rates would reduce costs and improve pre-primary and primary education, as well as for internal efficiency, higher survival rates would increase secondary education and TVET (technical and vocational costs even though they also improve cost efficiency. The education and training), but not for tertiary education. Uganda Education Note  Figure 7: Education sector budget by sub sector Figure 9: Public Budget as Share of GDP and (constant 2010 values in UGX Bn) Education Budget as Share of Public Budget (%) 600 30.0% 500 25.0% 400 20.0% 300 15.0% 200 10.0% 100 5.0% 0 0.0% 2009 2010 2011 2012 2013 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 Year Year Pre-primary and Primary Secondary BTVET Tertiary Other Public budget as a share of GDP Education budget as a share of public budget Source: Tsimpo and Wodon (2016). Source: Tsimpo and Wodon (2016). Budget shares by sub-sector remained broadly stable, Table 3 provides more details on budget trends for except for tertiary education that experienced a decline, education overall and by sub-sector as well as variables as shown in Figure 8. Importantly, as shown in Figure 9, such as GDP (the resources available to the country), the the overall budget declined as a share of GDP, and the overall budget (a measure of the resources available to budget for education declined as a share of the overall the government), and the number of students by cycle (a budget. As a result, public spending for education measure of the needs to be met). The analysis is declined as a share of GDP. The relative investment conducted again for the last five years were the education made in education through the budget has been declined budget regained some of the losses of the previous six in comparison to other uses of available resources. years. The data suggest an increase in investments per student over the five years for primary and secondary Figure 8: Share of Education Budget by Sector (%) education (the base year 2009 was a year with low public 60.0% investments versus previous years, as shown in Figure 3). 50.0% Table 3: Budget for Education, Constant 2013 Prices 2009 2010 2011 2012 2013 40.0% Overall GDP 44,671 49,823 46,993 52,515 56,287 30.0% Budget 6,764 7,305 8,870 10,156 10,903 Budget/GDP 15.10 14.70 18.90 19.30 19.40 Education 1,143 1,287 1,494 1,494 1,593 20.0% Ed./Budget 16.90 17.60 16.80 14.70 14.60 Primary 10.0% Budget (Bn) 589 624 758 741 795 Share (%) 51.50 48.40 50.70 49.60 49.90 0.0% Students 7,420 M 7,670 M 7,326 M 7,426 M 7,507 M 2009 2010 2011 2012 2013 Year Unit cost (M) 79 81 103 100 106 Primary Secondary BTVET Tertiary Other Secondary Budget (Bn) 331 418 458 447 500 Source: Tsimpo and Wodon (2016). Share (%) 29.00 32.50 30.70 29.90 31.40 Students 550,631 667,396 667,796 656,657 669,225 Concerns have been expressed about trends in public Unit cost (M) 602 627 687 681 747 TVET financing for education. The introduction of Universal Budget (Bn) 60 59 78 103 91 Secondary Education (USE) was a major step to expand Share (%) 5.30 4.50 5.30 6.90 5.70 access to secondary education for the poor. At the same Students 24,598 23,536 34,226 34,380 42,674 time, a fiscal crunch resulted in lower levels of funding Unit cost (M) 2,464 2,462 2,314 3,006 2,132 than anticipated, probably leading to lower quality as well. Tertiary Budget (Bn) 137 157 161 201 127 Share (%) 12.00 12.20 10.80 13.50 8.00 Students 169,476 174,375 179,569 198,066 201,376 Unit cost (M) 810 901 899 1,017 631 Source: Tsimpo and Wodon (2016). Uganda Education Note  materials, transport to schools, possibly lodging at Box 3: Spending and Education Outcomes higher levels of schooling, etc. The literature on education suggests that more spending does  Almost two thirds of private spending goes to not necessarily lead to better learning. There are several private schools. Within private schools, most reasons for this. First, factors not related to school inputs, such private expenditure are paid to private secular as the socio-economic background of children, often have a stronger impact on how well children lean than school inputs. In schools as opposed to NGO/religious schools. addition, the composition of spending matters at least as much This is because the market share of private as the level of funding. For example, teacher salaries account secular schools is much larger, but also because for a large majority of public education spending. But increasing the unit cost per student is higher in those spending for other inputs – including teacher training – can be schools versus NGO/religious schools. more efficient to improve learning than raising salaries. Finally,  While basic education in public schools is in how resources are allocated between schools, and whether they principle free, given that fees have been reach vulnerable children is also important. This matters as eliminated under the universal primary and gains in learning among vulnerable groups from many types of secondary education policies, households must interventions are often larger than for other groups. still pay for other types of education expenditure. At the same time, for schools to function, budgets are needed, For the poor, especially at the secondary level and when both attainment and learning are taken into where the average cost for households of sending consideration, the relationship between costs and performance a child to school is ten times the cost observed for is stronger. While the availability of funding does not guarantee primary schools, these expenditure may not be a quality education, without adequate funding a quality affordable and thereby contribute to drop-outs, as education cannot be provided. discussed in a separate policy note in this series. Overall, concerns have been expressed about trends in Households shoulder two thirds of education expenditure public financing for education in Uganda. The introduction in Uganda, with public accounting for only one third. Most of Universal Secondary Education (USE) was a major private spending are for enrollment in private schools. decision to expand access to secondary education for the Primary and lower secondary education in public schools poor. At the same time, a fiscal crunch resulted in lower is in principle free, but households still face substantial levels of funding than had been anticipated. Funding for costs for sending their children to public schools. This is education as a share of the budget has been lower than especially the case at the secondary level, which anticipated in the education strategy. Instead of contributes to affordability constraints for the poor, but increasing, spending levels have as noted earlier public-private partnerships have helped (see Box 3). decreased as a share of both GDP and the budget. In part because of low public spending for education, Table 4: Private Expenditure on Education (UGX Bn) two thirds of education expenditure are currently paid Primary Secondary Tertiary All for by households, which has implications for equity Cost per student Public school 55 522 1,808 - and affordability. Private school 272 940 1,602 - NGO/rel. school 142 857 2,041 - The fact that public spending for education is relatively Average 115 717 1,718 - low relates in part to a high market share of the private Total cost sector, especially at the secondary level where almost Public school 430.4 546.3 210.6 1187.3 half of the students attend private schools (detailed data Private school 800.8 829.6 194.5 1825.0 on the market share, benefit incidence, and costs of NGO/rel. school 18.3 52.3 23.8 94.4 private schools are provided in a separate not in this Total 1249.6 1428.2 429.0 3106.7 series). This also means that a large share of investments Source: Tsimpo and Wodon (2016), based on 2012/13 UNHS. in education are made by households as opposed to the state. Data from household surveys can be used to Table 5 shows the types of expenditure incurred by estimate the magnitude of private spending for education households in terms of the categories of spending. The in Uganda. Table 4 provides both average household largest costs are paid for school and registration fees, spending per student and total spending for all students followed by boarding costs, the “other” category and the by type of school. A number of findings stand out: cost of books/school supplies. The cost of uniforms and  Total private spending for education in 2012/13, sports clothes is lower, but for the very poor, these costs at more than 3,100 billion UGX, was twice the are often high. Costs are as expected much higher on level of public spending, at 1,593 billion in 2013. average at the secondary and tertiary levels than at the Private spending includes not only school fees primary level. Note that the average costs per student per and PTA dues, but also other types of year displayed in table 5 include students in both public expenditures such as for uniforms, school and private schools, and the total cost in table 5 corresponds to the average cost displayed in table 4. Uganda Education Note  The largest costs paid by households are for school and Due to population growth, enrollment will continue to registration fees, followed by boarding costs, the “other” increase. Together with expected gains in attainment, category and the cost of books/school supplies. The cost especially at the secondary level, this will add of uniforms and sports clothes is lower, but for the very significant pressure on public and private budgets. poor, even these lower costs are often substantial. The pressure from larger student cohorts due to demographic growth and higher attainment will increase Table 5: Average Cost by Spending Category (UGX) over time, with implications for both the state’s budget and Level of education All household expenditure for education provision. The Prim. Sec. Tertiary School fees 62,691 387,359 1,128,214 132,171 pressure from demography stems from the fact that Boarding 8,975 114,950 187,304 28,432 Uganda’s the total fertility rate (TFR1) remains very high at Uniforms/clothes 9,519 29,371 16,594 12,663 5.96 live births per woman in 2012, contributing to high Books/ supplies 17,087 76,598 100,894 27,703 population growth (3.3 percent per year). As shown in Transport 3,361 29,402 108,587 9,310 table 6, the country’s population is estimated to be 39 Other 13,104 79,707 176,633 26,310 million in 2015. This may increase to 62 million by 2030, Total 114,737 717,386 1,718,225 236,588 102 million by 2050, and 203 million by 2100 according to Source: Tsimpo and Wodon (2016), based on 2012/13 UNHS. the most recent estimates from the United Nations (2015). Box 4: Cost Savings through Public-private Partnerships Table 6: Population Statistics and Projections Estimates Uganda was one of the first countries to make education in Current population, 2015 (million) 39.03 public schools free at the secondary level in sub-Saharan Africa. Total fertility rate, 2012 5.96 Ass discussed in more details in a separate note in this series, Population growth rate, 2013 3.3% the 2007 Universal Secondary Education (USE) and Universal Expected population (million) Post Primary Education Training (UPPET) policy led to a large 2030 61.93 increase in secondary school enrolment. 2050 101.87 2100 202.89 Apart from eliminating fees in public schools, Uganda adopted a Source: United Nations (2015), except TFR and population public-private partnership (PPP) policy with capitation grants to growth from World Bank Development Indicators. selected private secondary schools. This PPP policy made sense given the increase in enrollment that followed the USE policy and absorption capacity constraints in public schools. The Uganda’s population is expected to reach 62 million policy is still in effect today. Eligible private schools receive people in 2030, 102 million in 2050, and 203 million in 47,000 USHs per student per term to enroll as many students as 2100. This will lead to much larger cohorts of children in they can accommodate without charging additional fees for the age of schooling in the future, and thereby higher costs. students enrolling under the program. In other words, participating private schools are also free of registration fees for students who enroll under the program. The schools may In a recent study on teachers for the Ministry of Education however also enroll non-USE “private” students for a fee. The and Sports (2014), the UNESCO Pôle de Dakar team management of the schools in terms of budgets, expenditure suggests that student enrollment may increase rapidly as allocation, and hiring/firing of teachers remains private. a result of both demographic growth and higher attainment. As discussed in a separate note in this series, Between 2007 and 2010, more than 600 private schools two scenarios for simulating teacher needs are analyzed. participated in the program. An evaluation of the policy by Barrera-Osorio et al. (2014) suggests that the PPP policy has The first “low” scenario is based on the hypothesis that been successful, including in terms of cost savings. The the system will continue to expand according to recent MoEpays UGX 47,000 per student in private USE schools and trends; the second “high” scenario is more ambitious and UGX 85,000 in private UPOLET schools. While this may seem is based on the hypothesis that education actors will take similar to the capitation grants provided to public schools, in the all possible actions to achieve the education sector case of public schools these grants are on top of the operating objectives adopted by the government, especially the costs of the schools. When operating costs are taken into targets set under the UPE and USE/UPPET policies. account, the cost of a secondary school student in a public school is more than four times higher than the cost of a student in a PPP school. Therefore, if the assumption is made that a large share of students would probably enroll in public schools without the policy, the policy can be argued to provide substantial budget savings for the government. 1 The TFR refers to the average number of live births a woman would have if she were subject to the current age-specific fertility rates throughout her childbearing years (15-49 in the DHS). Uganda Education Note  Table 7 provides the main results in terms of expected Given that teacher salaries and benefits represent the student enrollment. Under the high, more optimistic bulk of the budget for the education sector (as well as the scenario in terms of gains in education attainment from largest cost of operating private schools), and given that the USE/UPPET policies, total enrollment at the primary many other costs, such as the cost of providing textbooks, level is expected to grow by 70 percent between 2010 are also proportional to student enrollment, this expansion (the base year for the projections in the report) and 2025. in enrollment would probably require an increase in public Enrollment in lower secondary school is expected to triple, expenditure for education of the same order of magnitude while enrollment in upper secondary school is expected to for each of the education levels considered. Since unit quadruple. For business, technical and vocational costs of service provision are substantially higher at education and training (BTVET), enrollment is expected to higher levels of schooling, the overall increase in the increase six-fold. In the low scenario, increases in budget would be larger than the total increase in school enrollment rates at the secondary and BTVET level are enrollment. The implications for private household expected to be lower, but still significant. expenditure would be dramatic as well. Some savings could probably be achieved, for example through lower What does this imply for education expenditure, especially repetition rates, but the increase in expenditure would still for the public sector? Expected increases in enrollment remain very large even if savings were indeed achieved. are higher for public than for private schools (given the lower cost of schooling in public schools at the secondary If the ambitious targets adopted under the UPE and level under the USE/UPPET policy). This suggests that USE/UPPET policies were to be achieved, enrollment at the impact of the increase in student enrollment will be the primary level could grow by 70 percent between 2010 proportionately larger for the public budget than for and 2025. Enrollment in lower and upper secondary household budgets. It seems reasonable to assume pupil- school could triple and quadruple, respectively. Finally, teacher ratios should not change dramatically. They BTVET enrollment could increase six-fold. This would should actually be reduced to improve quality at the require large increases in education expenditure. primary level, which would lead to higher costs. Table 7: Projected Number of Students Enrolled in School under Low and High Scenarios, 2010-25 Low Scenario High Scenario Government Private Government Private Total Total Schools Schools Schools Schools Primary 2010 (base) 7,171,690 1,515,060 8,686,750 7,171,690 1,515,060 8,686,750 2025 (projection) 11,911,920 2,977,980 14,889,899 11,843,353 2,960,838 14,804,191 Increase 2025 vs. 2010 66.1% 96.6% 71.4% 65.1% 95.4% 70.4% Lower Secondary 2010 (base) 518,284 834,988 1,353,272 518,284 834,988 1,353,272 2025 (projection) 1,753,191 2,629,787 4,382,979 2,784,980 2,784,980 5,569,961 Increase 2025 vs. 2010 238.3% 214.9% 223.9% 437.3% 233.5% 311.6% Upper secondary 2010 (base) 67,514 118,659 186,173 67,514 118,659 186,173 2025 (projection) 356,642 356,642 713,285 462,491 462,491 924,982 Increase 2025 vs. 2010 428.2% 200.6% 283.1% 585.0% 289.8% 396.8% BTVET BTVET BTVET BTVET (Lower level) (Middle level) (Lower level) (Middle level) 2010 (base) 11500 12000 11500 12000 2025 (projection) 87660 49930 111399 83248 Increase 2025 vs. 2010 662.3% 316.1% 868.7% 593.7% Source: Ministry of Education and Sports (2014). See also Coury et al. (2016). Uganda Education Note  Other projections of potential future costs for Uganda’s Still, as an illustration, the requirements by education level education sector also suggest the need for dramatic as a share of GDP are visualized in Figure 10, with the increases in support to the education sector over time. abbreviations US standing for upper secondary, LS for For example, projections have been prepared for the lower secondary, P for primary, and PS for preschool pre- latest Global Monitoring Report of UNESCO in order to primary. Both recurrent and capital costs are taken into assess the cost of achieving universal school enrollment account. The estimates suggest that Uganda will need to by 2030 at the preschool, primary, and secondary levels allocate much more resources to the education sector in (Wils, 2015). One limit of the UNESCO analysis for future the future, and that it will also need support from donors in scenario is that estimates of the cost of reaching universal order to be able to finance the expected expansion in education are likely to be higher than what is actually student enrollment that would result from demographic spent, as many countries are not likely to reach universal growth as well as gains in education attainment. education targets up to the secondary level by 2030. Still, the model provides highly valuable insights. Conclusion Public spending for primary and secondary education is Due to population growth and gains in education currently below three percent of GDP. In order to achieve attainment, the size of Uganda’s education system is universal upper secondary school completion by 2030 growing, and is expected to grow faster in the future. If funding needs could reach almost nine percent of GDP. one were to assume that the targets of the USE/UPPET While lower budget needs would be required under less policies will be achieved, total enrollment at the primary ambitious scenarios, the exercice helps illustrate how level is expected to grow by 70 percent between 2010 budget needs for the sector will increase in the future. (base year for the projections) and 2025. Enrollment in lower secondary school could to triple, while enrollment in upper secondary school could quadruple. BTVET It was mentioned earlier in this note that public spending enrollment could increase six-fold. This would require a for primary and secondary education is currently below dramatic increase in public and private spending for three percent of GDP. According to the UNESCO model, education, putting additional pressure on the budget. the level of funding under the budget that would be Internal efficiency gains in the education sector are required to achieve universal upper secondary school necessary, but would not fundamentally change budget completion by 2030 – clearly, a very ambitious target for a needs even if a reduction in repetition rates would country such as Uganda, could reach almost nine percent generate substantial savings. The analysis in this note of GDP by 2030. Of course, less ambitious targets would suggests the need to prepare detailed long term budget relieve some of this budget pressure. scenarios in order to anticipate ahead of time the budget implications of rising enrollment. Various simulation tools Figure 10: Potential Future Budget Needs for are available for this purpose. Education as a Share of GDP, UNESCO Model (%) 10.0% 9.0% 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% 2.0% 1.0% 0.0% 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029 2030 US Margin. Support US Capital US Recurrent ALL BASIC ED Youth literacy training LS Margin. Support LS Capital LS Recurrent P Margin. Support P Capital P Recurrent PS Margin. Support PS capital PS recurrent Source: Authors’ simulation using Wils (2014) model. Uganda Education Note  References Barrera-Osorio, F., P. de Galbert, J. Habyarimana, I. Annex: Notes on Education in the Uganda Note Series Mulindwa, and S. Sabarwal, 2014, Public Private (As of March 2016) Partnerships in Education, Evidence from a Randomized Controlled Trial in Uganda, Report Developed for The As mentioned in Box 1, this note is part of a series on Uganda. Ministry of Education and Sports, Republic of Uganda. Cross-references are made in this note to some of the other notes in the series. The list of notes focusing on the education sector available in the series as of March 2016 is as follows: Coury, D., P. Nkengne, O. Pieume, and Q. Wodon, 2016, Teacher Needs, Background Paper for the Uganda Note 1: Education in Uganda: Summary Assessment. Education Note Series, Washington, DC: The World Note 2: Importance of Investing in Education. Bank. Note 3: Improving Education Attainment. Note 4: Improving Education Achievement. Ministry of Education and Sports, 2014, Teacher Issues in Note 5: Levels and Efficiency of Spending for Education. Uganda: A Shared Vision for an Effective Teachers Note 6: Benchmarking Public Spending for Education. Policy, Kampala: Ministry of Education and Sports. Note 7: Private Schools and Public-Private Partnerships. Note 8: Future Needs for Teachers and Teacher Policies. Note 9: School Facilities and Challenges in Construction. Tsimpo, C., and Q. Wodon, 2016, Public and Private Note 10: Investing in Early Childhood Development. Spending for Education, Background Paper for the Note 11: Child Marriage and Education. Uganda Education Note Series, Washington, DC: The World Bank. In addition to the above notes, edited volumes with more in- depth analysis as well as technical background papers are being United Nations, 2015, World Population prospects: Key prepared. For more information, please contact the authors at Findings and Advance Tables, 2015 Revision, New York: the World Bank. The edited volumes will be made available United Nations. through the World Bank’s Open Knowledge Repository at https://openknowledge.worldbank.org/. Wils, A., 2014, Reaching education targets in low and lower-middle income countries: Costs and finance gaps to Acknowledgment and disclaimer: This note was prepared by a World 2030, Background Paper for UNESCO Global Monitoring Bank team as an input into a technical assistance task for Uganda’s report, Paris: UNECO. Ministry of Education, Science, Technology, and Sports. The objective of the notes series is to convey research findings in a rapid and more accessible way than though technical papers. The series benefitted from funding from the Global Partnership for Education as part of a global research task on out of school children and disadvantaged groups. Support under the SABER (Systems Approach for Better Education Results) Equity and Inclusion work program is also gratefully acknowledged. The authors benefited from comments and suggestions among others from Melissa Ann Adelman, Juan Baron, Sajitha Bashir, Luis Benveniste, Andreas Blom, Christina Malmberg Calvo, Meskerem Mulatu, and Lianqin Wang. The opinions expressed in the notes and related background studies are those of the authors only and need not represent the views of the World Bank, its Executive Directors, or the countries they represent, nor do they represent the views of the Ministry of Education, Science, Technology, and Sports.