EDUCATION FINANCE WATCH 2024 Introduction To achieve their national and international education goals, many countries need to invest more and better in their educa- tion systems. During the last decade, total education spending by governments, households, and donors globally has increased steadily, but this has not led to significant increases in alloca- tions per child,1 especially in poorer countries with their grow- ing populations. Total education spending per child has either decreased or stagnated globally. Additionally, the combination of the financial repercussions of the COVID-19 pandemic plus escalating global debt is likely to be limiting the ability of coun- tries to augment their investments in education. Moreover, the strain on public education budgets in recent years has coincided with a 4 to 8 percentage point decline in minimum reading and math proficiency among 15-year-olds compared to 2018 pre-COVID levels in middle-income countries (OECD, 2023). In low-income countries (LICs), where data on educational outcomes are scarce, simulations suggest that the incidence of learning poverty, which was already high before 2020, is likely to have risen in the aftermath of the COVID-19 pandemic (World Bank et al., 2023). The current challenge in education finance is the need to mobilize more resources while at the same time increasing the adequacy, efficiency, and equity of funding in the face of tight fiscal space and competing priori- ties. Tackling the spending inefficiencies and inequalities that © Bart Verweij /World Bank are common to many education systems will be vital to enable countries to make better use of their resources and strengthen the link between spending and education outcomes. for Statistics (UIS). It summarizes available information on patterns and trends in education financing around the world. The Education Finance Watch (EFW ) is a collaborative To do so, the EFW draws on various sources of education, effort between the World Bank, the Global Education economic, and financial data from the World Bank, the UIS, Monitoring (GEM) Report, and the UNESCO Institute the International Monetary Fund (IMF), the Organisation for 1 EFW2024 uses “per school-age individual” and “per child” interchangeably, with school age referring to those aged between 5 and 24 years old. Education Finance Watch 2024 © Dominic Chavez/World Bank Economic Co-operation and Development (OECD), and the of EFW2023 was prepared with a focus on education spending United Nations (UN). This Education Finance Watch 2024 in Africa for the African Union Year of Education 2024 (World (EFW2024) report provides a comprehensive overview of the Bank and UNESCO, 2024). This year, EFW2024 explores key trends and sources of global education financing that can the interplay between education financing and rising trends in be used as a foundation for further analysis and future policy international debt. dialogue. Researchers and policymakers seeking to use the report’s analytical underpinnings and data for further study and policy dialogue are invited to read the EFW’s accompa- Key findings nying technical note, which provides additional details about the analytical methods and terminology used in the report. 1. Total education spending by governments, households, and donors has increased over the past decade, but funding Each year, the EFW is dedicated to a special topic of inter- in LICs is insufficient to overcome their learning deficits. est that highlights critical issues in education financing. The Total education spending across the globe has been on an first volume of the EFW report (EFW2021) documented the upward trajectory over the past decade, signaling a commit- continuous increase in global education spending in absolute ment by governments to enhancing learning opportunities terms over the previous decade and concluded that the COVID- for their populations. Both LICs and lower-middle-income 19 pandemic was likely to have slowed this trend.2 EFW2022 countries (LMICs) have experienced a more rapid annual shed light on the pandemic’s impact on global education spend- increase in education spending than wealthier nations. ing in 2020, its first year, and revealed that half the analyzed However, in many LICs, even those that have reached their countries had reduced their annual education spending in real education spending targets for countries at their GDP level, terms. EFW2023 spotlighted demographic changes in school- their absolute levels of funding are too low to guarantee age populations and projected the financial implications for adequate student learning. As of 2022, annual expendi- selected countries over the following ten years. A special edition ture per child in LICs is insufficient to ensure adequate 2 EFW2022 and EFW2023 initially suggested no significant change in total global education spending in 2020. Nevertheless, subsequent data presented in EFW2024 has indicated an actual rise in total global education expenditure in 2020 compared to 2019, adjusted for inflation. This increase can mainly be attributed to higher government spending globally than had earlier been estimated. For more details, please see the accompanying technical note to EFW2024. 2 Education Finance Watch 2024 student learning, amounting to no more than US$55 (or PPP$172). Globally, most education financing comes from government expenditures, which account for approximately three-quarters of the total. Most of the remaining quarter comes from household contributions. 2. To overcome the global learning crisis, LICs and LMICs must focus on increasing the adequacy, efficiency, and equity of their educational expenditures. Although total education expenditure has increased since 2010, education spending per child has largely stayed the same, reflect- ing global demographic shifts. There is a clear correlation between increased financial investment in education per child and improved educational performance, especially in LICs.3 Nonetheless, LICs and LMICs often face chal- lenges in trying to allocate educational funds efficiently, © Dominic Chavez/World Bank which can undermine the impact of their spending. To improve educational outcomes, governments should prior- itize enhancing the efficiency of their current spending fiscal challenges are preventing some countries, especially by optimizing public financial management, improving in Africa and South Asia, to allocate sufficient funds to school management and teacher performance, strengthen- education. As developing countries struggle to manage ing governance, and channeling resources to cost-effective their debt, there is less direct government financing avail- policies and programs. able for education. Some countries are exploring inno- vative financing mechanisms for short-term relief, such 3. The amount of aid provided for education in LICs is as debt restructuring, debt swaps, debt-for-development high, but the proportion of aid allocated to education agreements. However, these measures must be comple- has declined. Globally, total education aid or donors’ offi- mented by sustained domestic resource mobilization, effi- cial development assistance (ODA) from donor coun- cient spending, effective public financial management, and tries reached a record high of US$16.6 billion in 2022, up robust economic growth to ensure that their populations from US$14.3 billion in 2021, a growth in real terms of 16 can receive quality education. percent year on year. Nevertheless, the share of total ODA allocated to education decreased from 9.3 percent in 2019 5. To maintain a clear global picture of education financing to 7.6 percent in 2022. This shift reflects a significant real- trends, it is imperative for countries to report their educa- location of donors’ funding priorities to energy, support tion financing data in a timely and consistent way at a more for Ukraine, and healthcare in response to the COVID- disaggregated level. While about 7 in 10 countries publish 19 pandemic. By 2022, ODA accounted for 12.2 percent key education financing data, the absence of disaggregation of education funding in LICs (versus 13 percent in 2021) by type of expenditure or by level of education makes it diffi- and just 0.29 percent of total education funding globally. cult to monitor education financing allocations. However, EFW2024 has been able to access more household-level 4. In the past 10 years, interest payments on public debt have data than in previous years with five times more data points, increased faster than government education spending although there is still a lack of available post-pandemic data, in developing countries. The debt situation has become especially from poorer countries where households spend particularly worrisome for LICs and LMICs, some of much more out of pocket on education in relative terms. which are allocating nearly the same per capita resources Without sound and extensive data, forward thinking poli- to debt servicing as they do to education.4 Mounting cymaking is stymied. 3 IMF estimates (under preparation) using Carapella et al., 2023 suggest that LICs would need to invest between an additional 4.5 to 5.5 percent of their GDP to meet the education-related Sustainable Development Goals (SDGs). UNESCO (2023b) has estimated that LICs and LMICs would need to invest an additional 2.3 percent of their GDP to achieve their national SDG 4 benchmarks by 2030. 4 In EFW2024, “per capita” refers to the total population. In other words, a country’s per capita debt servicing burden is determined by dividing the total debt servicing burden by the coun- try’s total population. 3 1 Mobilizing Resources for Education How has global education spending changed over the last ten years? Globally, total education spending5 by governments, house- holds, and donors over the past decade has increased more © Dominic Chavez/World Bank increased by only 10 percent over the same period. Despite this slower growth, HICs accounted for 64 percent of the world’s total education expenditure in 2022, amounting to US$3.71 trillion, although this is a decrease from 72 percent in 2010. In the year from 2021 to 2022, total education spending grew slowly than economic growth. Total global education spend- in UMICs and HICs, remained static in LICs, and declined ing grew in real terms by an average of 1.8 percent per year in LMICs (Figure 1b). between 2010 and 2022. This rate of increase is slower than global economic growth rates6 and masks two diverging trends. In 2021, total education spending increased moderately from Since 2010, total education spending in LICs, LMICs, and US$5.7 trillion in 2021 to US$5.8 trillion in 2022 (in 2022 upper-middle-income countries (UMICs) nearly doubled in constant US dollars) (Figure 1a). This increase was driven real terms. In contrast, in high-income countries (HICs), it has by a slight decrease in total government expenditure in real Figure 1. Total education spending has increased by 60 percent in low-income and middle-income countries since 2010 a. Total education spending (government, aid, and household) in b. Growth in real education spending (all sources) by country constant 2022 US$, trillion, 2010–2022 income group, 2010-2022, with 2010 = 100 $501 bil 6 5.7 5.8 180 $507 bil 5.5 5.6 LIC, $16 bil 5.3 5.4 $26 bil 5.1 5.0 5.1 170 LMIC, $304 bil $26 bil 4.9 5.0 5 4.7 4.8 1.35 1.45 UMIC, $979 bil 160 0.01 0.02 HIC, $3.38 tri 0.99 $1.56 tri 4 0.01 150 140 $1.53 tri Trillion $ 3 4.39 4.33 130 3.68 2 120 $3.71 tri 110 1 100 $3.69 tri 0 90 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Government Development Assistance Household Total LIC LMIC UMIC HIC Source: Author estimates using the EFW2024 database. Notes: Interpolation was used to fill in missing data and ensure a comparable sample of countries in all periods. The variation in the numbers from EFW2023 can be attributed to: (i) recent updates, increased access to data on government education spending and aid, and greater data availability across different countries; and (ii) the availability of more data on household spend- ing in HICs. Spending patterns in households from HICs often differ from those in LICs, LMICs, and UMICs. The team has followed the World Bank’s country income classification published in 2023: LICs = low-income countries, LMICs = lower-middle-income countries, UMICs = upper-middle-income countries, and HICs = high-income countries. (https://data- helpdesk.worldbank.org/knowledgebase/articles/906519-world-bank-country-and-lending-groups, accessed in May 2024). 5 Global education spending refers to expenditure on education services by governments, households, and donors in accordance with UIS definitions. 6 During the same period, average GDP globally grew by 2.8 percent annually (WDI data). 4 Education Finance Watch 2024 terms (from US$4.39 trillion in 2021 to US$4.33 trillion in 2022) that was more than offset by global household spending (from US$1.35 trillion in 2021 to US$1.45 trillion in 2022) (Figure 1a). The contributions of governments, households, and donors to global education spending have remained relatively constant over time. As of 2022, governments contributed around three-quarters to the total (74.6 percent), while house- holds contributed one-quarter (25.1 percent), and donor’s ODA accounted for 0.29 percent in 2022 (Figure 1a). While governments are the largest funders of education in all country income groups, their contributions differ significantly among those groups, ranging from 80.4 percent in HICs to 61.9 percent in LICs in 2022 (Figure 2). In the same year, donor’s ODA represented 12.2 percent of total education spending in LICs, while in LMICs, it only accounted for 2.1 percent. Household spending on education in LMICs averaged 2.1 percent of GDP, while in HICs, it averaged 0.8 percent of © Dominic Chavez/World Bank GDP. Household spending on education varies significantly across countries within income groups. For example, in LICs, household contributions to education range from 0.1 percent countries. In 2022, household contributions accounted for to 7.6 percent of GDP. Household education expenditures as approximately one-quarter (25.8 percent) of education spend- a percentage of GDP also vary across country income catego- ing in LICs and over two-fifths (43.9 percent) in LMICs, ries. In poorer countries, the direct contribution of households whereas in HICs, they represented only 19.6 percent of total to education spending tends to be greater than in wealthier education spending. 7 Figure 2. Governments funded nearly three-quarters of all education expenditure in 2022 Distribution of total education spending by source, by income group, percentage, billion US$ 2015 12.0 2.3 5.7 2020 14.1 4.4 6.7 LIC 2021 15.7 3.6 6.6 2022 16.0 3.2 6.7 2015 227.2 6.8 164.9 2020 266.8 8.2 207.1 LMIC 2021 Billion US$ 289.9 8.1 208.8 2022 270.6 10.4 220.0 2015 947.8 2.0 375.7 2020 1,050.8 2.9 458.8 UMIC 2021 1,041.7 2.5 481.6 2022 1,053.2 2.9 499.1 2015 2,663.6 0.0 641.0 2020 2,947.4 0.0 648.8 HIC 2021 3,038.3 0.0 651.2 2022 2,985.3 0.0 728.1 0 10 20 30 40 50 60 70 80 90 100 Percent Government Development Assistance Household Source: Author estimates using the EFW2024 database. Note: Interpolation was used to fill in missing data and to ensure a comparable sample of countries in all periods. A total of 218 countries and territories were included in the EFW2024 data- base. To avoid double-counting, government expenditure nets out part of the ODA received by countries. The number changed compared to EFW2023 because of: (i) changes in govern- mental educational spending and aid due to data updates and changes in the countries with available data, and (ii) a significant increase in household spending in HICs due to data updates and to more household spending data becoming available. 7 The availability of new and additional household spending data from the UN and GEM Report database indicates that household spending accounts for a lower proportion of education expenditure in LICs compared to earlier EFWs. 5 Education Finance Watch 2024 How has government education over the past decade. The gap between LICs and LMICs was spending changed over the last ten 1.6 percentage points in 2010, but it decreased to 0.3 percentage years? points in 2022 (3.9 percent in LICs versus 4.2 percent in LMICs) (Figure 3a.). However, these averages conceal variations in trends Over the past decade, government funding for education as in individual countries. For example, in Burkina Faso (a LIC), a percentage of national income has increased in LICs and government spending as a share of GDP increased from 3.9 to 5.5 declined in all other country income groups.8 Between 2010 percent between 2014-15 and 2018-19, whereas it declined from and 2022, government education spending in LICs grew from 5.2 to 4.4 percent in Malawi (also a LIC) over the same period.9 2.9 to 3.9 percent of GDP, while it decreased by between 0.3 and 0.4 percentage points in all other country income groups. From Government education spending as a percentage of national 2021 to 2022, the share of government spending as a percentage income has also gradually converged across regions. In 2010, of GDP continued to decline in LMICs, UMICs, and HICs. For the largest gap observed was between East Asia and the Pacific the first time since 2016, education spending as a percentage of (at 5.5 percent in 2010) and South Asia (at 3.1 percent in GDP declined in LICs from 4 percent in 2022 to 3.9 percent in 2010), a difference of more than two percentage points. The 2021 (Figure 3a.). Globally, education spending as share of GDP difference between the highest and the lowest shares of educa- has decreased from 4.5 percent in 2010 to 4.3 percent in 2022. tion expenditures became narrower in 2022 between these two regions (5.1 percent versus 3.6 percent respectively, a difference The disparity in government education spending as a percentage of 1.5 percentage points). Regional rankings have remained of GDP across different country income groups has narrowed unchanged during the period of analysis (Figure 3b). Figure 3. Government spending on education as a share of GDP in low-income countries has converged with the shares in middle- and high-income countries a. Government education spending as a percentage of GDP by b. Government education spending as a percentage of GDP by income group, 2010-2022 region, 2010-2022 6.0 6.0 5.2% 5.2% 5.5 5.5 5.0% 4.9% 4.9% 5.1% 5.0 4.8% 5.0 4.8% 4.6% 4.6% 4.6% 4.5 4.6% 4.5 4.4% 4.4% 4.7% 4.6% 4.7% 4.4% 4.2% 4.0% % % 4.5% 4.2% 4.2% 4.0 4.0 3.9% 4.0% 3.8% 3.9% 3.7% 3.6% 4.0% 3.5 3.5 3.6% 3.1% 2.9% 3.0 3.0 2.5 2.5 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 LIC LMIC UMIC HIC East Asia & Pacific North America South Asia Middle East & North Africa Latin America & Caribbean Sub-Saharan Africa Europe & Central Asia Source: Author estimates using the EFW2024 database. Note: The number is the mean of government education spending as a percentage of GDP by income group. Estimates of spending as a percentage of GDP include interpolations to fill in missing data and ensure that there is a comparable sample of countries in all periods. The corresponding median of government education spending as a percentage of GDP in LICs, LMICs, UMICs, and HICs in 2022 are 3.8 percent, 3.6 percent, 4.0 percent, and 4.5 percent respectively. The corresponding median value by regions are 3.4 percent in Sub-Saharan Africa (SSA), 3.7 percent in South Asia (SAR), 4.0 percent in Latin America and the Caribbean (LAC), 4.1 percent in East Asia and the Pacific (EAP), 4.1 percent in North America, 4.6 percent in Europe and Central Asia (ECA), and 5.0 percent in Middle East and North Africa (MENA) in 2022. EFW2024 continues the use of mean values to remain consistent with earlier editions and to ensure that each country, including those exhibiting outlier behavior, is accorded equal weight. 8 Given data limitations, the EFW cannot analyze government spending by levels of education. 9 Author estimates using the EFW2024 database. 6 Education Finance Watch 2024 Figure 4. Public education spending per school-age individual has more than doubled in South Asia since 2010 a. Government education spending per school-age individual b. Government education spending per school-age individual (constant 2022 US dollars) by income group, 2010-2022 (constant 2022 US dollars) by region, 2010-2022 14,069 13,013 7,730 8,837 8,532 12,730 US$, 2022 constant US$ 2022 constant 6,851 1,273 6,024 6,570 958 4,397 1,271 4,151 3,502 3,838 2,331 3,351 309 2,290 316 2,478 278 2,041 458 57 55 220 515 42 288 218 283 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 LIC LMIC UMIC HIC East Asia & Pacific North America South Asia Middle East & North Africa Latin America & Caribbean Sub-Saharan Africa Europe & Central Asia Source: Author estimates using the EFW2024 database. Note: The number is the mean of government education spending per child by income group. Estimates of spending as a percentage of GDP include interpolated values to fill in missing data and ensure that there is a comparable sample of countries in all periods. The corresponding median of government education spending per child is US$51 for LICs, US$173 for LMICs, US$1,199 for UMICs, and US$7,317 for HICs. The median values of per child government education spending by regions are US$87 for SSA, US$146 for SAR, US$823 for ECA, US$889 for MENA, US$1,470 for LAC, US$4,851 for ECA, and US$11,248 for North America in 2022. Despite a slight convergence over the past decade, disparities to 2022, these per child government education expenditures by in annual government spending per child by country income country income level stayed mostly stable. However, in HICs, group remain significant. For every US$100 that HICs allo- there was a noticeable decrease from US$8,837 to US$8,532 cated on education per child in 2022, LICs allocated less than (adjusted for inflation) (Figure 4a).10 US$1. While these disparities are largely due to differences in countries’ economic development (and become less notable Although government spending per child in South Asia and when assessing the ratio of per child spending to per capita Sub-Saharan Africa is still low relative to other regions, it GDP), they indicate that even when countries reach recom- has increased significantly since 2010. Annual government mended investment levels according to international bench- spending per child is highest in North America and in Europe marks, they still face challenges in securing sufficient funding and Central Asia ranging from US$6,500 to US$13,000. In levels to produce adequate learning outcomes. Children in the Middle East and North Africa, Latin America and the HICs received an education worth 155 times more (US$8,532 Caribbean, and East Asia and the Pacific, annual govern- per child in 2022 constant US dollars) than the education ment spending per child ranges from US$2,400 to US$3,800 received by children in LICs (US$55 per child in LICs). About (Figure 4b.). In South Asia, although still very low, govern- four times as much was spent per child per year in UMICs ment spending per child has more than doubled over the past (US$1,273) than in LMICs (US$309). Even taking into decade, climbing from US$218 in 2010 to US$515 in 2022. account differences in purchasing power between countries, Because enrollment rates at private institutions have been the difference in per child government education spending is rising across South Asia, students in public schools should be huge: PPP$11,413 in HICs and PPP$172 in LICs. From 2021 reaping greater benefits from these government educational 10 Trends in spending have been tracked based on overall government education spending per child (defined as school-age individuals at the pre-primary, primary, secondary, and tertiary levels). This made it possible to compare levels of funding between countries or groups of countries. It also accounts for differences in the size and growth of child and youth popula- tions across countries and enables us to assess the availability and adequacy of funding for all children rather than only those who can attend schools, universities, and other educational institutions. 7 Education Finance Watch 2024 © Maria Fleischmann/World Bank investments.11 In Sub-Saharan Africa, education funding has shows government spending as a proportion of GDP and grown since 2010, reaching US$283 in 2022, but this is only the share of this spending devoted to education in LICs and a modest increase given the ongoing growth of the region’s LMICs. The dashed lines plot the combinations of the two population. (Figure 4b). spending indicators to mark out the zone between two bench- marks of 4 and 6 percent of GDP that were set in the Incheon In 2021-2022, regional government spending per child on Declaration in 2015.12 Many of the LICs and LMICs in the education declined in four out of seven regions. In the Middle EFW2024 analysis were spending less than 4 percent of GDP East and North Africa, it fell by 13 percent from US$4,397 and devoting less than 15 percent of their total public expen- to US$3,838. In North America, it fell by 2 percent from diture to education. Of 80 countries with available data in US$13,013 to US$12,730. In Europe and Central Asia, it 2022,13 41 countries met neither target, 25 countries met both, fell by 4 percent from US$6,851 to US$6,570, and in East and 14 countries achieved either one or the other (Figure 5). Asia and the Pacific, it also fell by 4 percent from US$3,502 Some countries such as Sri Lanka and Uganda fell far below to US$3,351, mainly due to a reduction in total government the average for their income group on both indicators, while expenditure. Government spending increased modestly in other countries such as Mozambique and Uzbekistan exceeded Latin America and the Caribbean, rising from US$2,290 to the average. US$2,478 (8 percent). To overcome their learning crises, LICs and LMICs would Many LICs and LMICs have not yet met international need to boost domestic resource mobilization and allocate a benchmarks for education spending allocations. Figure 5 higher percentage of their budgets to education. To achieve 11 UNESCO (2022). 12 Many countries have agreed to these international benchmarks, namely that they should spend 4 to 6 percent of their GDP and/or 15 to 20 percent of total government spending on educa- tion (according to the Education 2030 Incheon Declaration). For progress on these and other national SDG 4 benchmarks, see UIS and GEM Report (2024). 13 See the EFW2024 technical note for the complete list of country codes. 8 Education Finance Watch 2024 national SDG 4 benchmarks, it has been estimated that LICs and LMICs would need to increase total education spend- ing from governments, households, and donors between 4.2 and 6.5 percent of their GDP over the period from 2023 to 2030.14 For example, just to meet the public spending bench- mark of 4 percent of GDP (represented by the green dotted line in Figure 5), Madagascar would need to allocate nearly 30 percent of its total public expenditures to education Given the competing priorities for government spending in LICs and LMICs, it seems unlikely that significant increases in educa- tion funding will be realized simply by making education a higher priority within the government budget alone. In many countries, it will also be necessary to mobilize more domes- tic resources to increase government revenues (Figure 5). In fact, 83 percent of LICs and 43 percent of LMICs are below the international tax collection benchmark of 15 percent of GDP.15 Although it is challenging for developing countries to increase their domestic tax revenues because of their large informal sectors, widespread misreporting of income and asset ownership, and narrow tax bases, the IMF estimates that LICs and LMICs could boost their tax-to-GDP ratio by up to 9 percentage points by improving the design of their taxation systems (Gaspar et al., 2023). © Fernanda Reyes/Shuttersctock Figure 5: Fiscal space for mobilizing greater funding for education varies considerably across countries Education as a share of total government expenditure and as a share of GDP in LICs and LMICs (%), 2010-2022 30 LIC Education as a share of total government expenditure (%) 25 MAR LMIC ETH HND BOL IRN SENNIC BFA GNB UZB KGZ TJK BDI TUN 20 BTN COD MLI MOZ MWI WSM FSM TCD BEN KEN STP COG ZWE SWZ SLB 15 KHM VNM CIV PHL GMB CPV TGO MDG TZA HTI GHA DZA CMR DJI NER NPL IND LSO BGDGIN SDN LBRZMB EGY SSD CAF COM Combinations equal to 10 MRT AFG LBN LAO MMR JOR public education spending UGA of 6% of GDP PAK PNG AGO Combinations equal to LKA MNG TLS 5 VUT public education spending SOM of 4% of GDP 0 5.0 15.0 25.0 35.0 45.0 55.0 65.0 Total government expenditure as a share of GDP (%) Source: Author estimates using the EFW2024 database. See the EFW2024 technical note for the list of country codes. Note: N=80 for all countries (LICs = 26, LMICs = 54). 14 UNESCO (2023b). 15 Author’s calculations using the World Economic Outlook database. 9 Education Finance Watch 2024 What are the main trends in education aid? Total education aid (or ODA) reached a record high of US$16.6 billion in 2022, up from US$14.3 billion in 2021 (a growth of 16 percent) (Figure 6a). Between 2010 and 2022, ODA to educa- tion globally increased by 41 percent (or 2.9 percent per year – roughly at the same pace as global GDP). Basic education, which encompasses pre-primary and primary levels, usually receives the largest portion of ODA, but its share declined from 52 percent in 2010 to 46 percent in 2022. In contrast, the share allocated to secondary education rose from 20 to 26 percent, while the share for post-secondary education remained steady at approximately 28 percent (Figure 6b). Between 2021 and 2022, ODA for basic education increased by US$883 million (13 percent), for second- ary education by US$684 million (19 percent), and for post-sec- ondary education by US$723 million (18 percent).16 Education is getting lower on donors’ list of priori- ties. Although the overall volume of ODA to education has increased, the share of education in total ODA, which increased from 8.2 percent in 2013 to 9.3 percent in 2019, has fallen in recent years, down to 7.6 percent in 2022. In contrast, © Omotayo Kofoworola/Shutterstock Figure 6a. Education aid reached US$16.6 billion Figure 6b. The share of total aid allocated to globally in 2022 secondary education has been increasing Total aid to education by level of education, in 2022 constant US$, Distribution of total aid to education by level of education, in 2022 2010-2022 constant US$, 2010-2022 100 16 14 4.7 27.8 28.3 28.1 28.2 26.0 27.2 25.7 26.3 27.8 29.4 27.2 27.5 28.1 4.2 80 12 3.9 3.8 4.0 Constant US$ 2022 billion 3.3 3.3 Constant US$ 2022 billion 10 3.3 60 19.9 19.1 21.1 23.4 24.5 21.5 21.6 22.2 21.9 23.8 24.4 25.2 25.8 3.2 3.2 2.8 3.1 2.9 3.8 3.6 4.3 8 2.8 2.8 3.0 3.2 2.3 2.1 2.7 2.6 2.4 2.2 40 6 4 52.3 52.6 50.8 48.4 49.4 51.3 52.6 51.4 50.3 46.8 48.3 47.3 46.1 6.1 5.9 5.3 5.5 5.3 5.8 6.8 6.5 6.9 6.3 7.5 6.8 7.6 20 2 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Basic Secondary Post-secondary Basic Secondary Post-secondary Source: Author estimates based on data from the OECD’s Development Assistance Committee (DAC) Credit Reporting System (CRS). 16 For better comparability across donor countries, the amount of ODA in EFW2024 excludes imputed student costs (in other words, the cost of tuition in donor countries for nationals of ODA recipient countries in countries in cases where education systems are tuition-free or where fees do not cover the full cost of tuition). Only some European countries have been includ- ing imputed student costs in their ODA definition, which has distorted comparisons. Additionally, some countries, like Belgium did in 2022, have recently stopped reporting imputed costs. Excluding imputed student costs also aligns with the introduction of OECD’s Total Official Support for Sustainable Development (TOSSD) framework, which reclassified imputed student costs as global public goods. If imputed student costs had been included, they would have accounted for 15 percent of the total ODA. 10 Education Finance Watch 2024 Figure 7. Between 2019 and 2022, there was a marked Figure 8. Donors prefer different aid modalities decline in the proportion of total aid allocated to education Share of education, population and health, and energy in sector Aid to education by modality, top six donors, 2022 allocable aid, 2010-2022 100 25 24.4 23.8 80 20 19.3 60 15 % % 10 9.9 40 8.0 7.9 6.5 7.6 7.0 5 20 0 0 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 World Bank United European Germany United France (IDA) States Union Kingdom Health and population Education Scholarships Budget support Project interventions Energy Core contributions/Pooled funds Expert cost/Technical assistance Source: Author estimates based on data from the OECD’s DAC CRS data. Source: GEM Report team analysis based on data from the OECD’s DAC CRS. the share of the health sector in total ODA increased from 17.5 increased, whereas those from bilateral donors like the United percent in 2019 to 23.8 percent in 2022, likely because of the Kingdom and the United States have decreased. The Global increased need during the pandemic (Figure 7). However, this Partnership for Education (GPE), which disbursed US$454 decline in the proportion of ODA going to education appears million annually on average between 2021 and 2022, increased its to be due not just to temporary shocks but also to a more struc- disbursements to US$521 million in 2023 (GPE, 2024). These tural shift in global priorities towards health and energy. figures are included in the funding reported to the OECD by its donors, such as the European Union and the United Kingdom. Disbursements of ODA for education to LICs and LMICs have followed different trajectories in recent years. The volume Donors’ education sector priorities vary. For instance, 75 of aid to LICs increased gradually throughout the 2010s and percent of the United States’ total ODA allocations and 66 then declined by 4 percentage points between 2019 and 2022 percent of Norway’s ODA allocations were earmarked for due to the COVID pandemic and the Ukraine war. During basic education. In contrast, France and Japan allocated 60 this four-year period, the share of aid channeled to LMICs percent and 53 percent of their ODA respectively to post-sec- increased, with a particularly sharp increase in 2022 (6 percent- ondary education. age points higher than in 2021). The increase was predom- inantly driven by the surge in aid to Ukraine, from US$187 Donors also vary in their preferred aid modality. Project-based million in 2021 to US$2.1 billion in 2022. Other than in Europe funding is the main modality for providing this aid, accounting (driven by Ukraine) and Sub-Saharan Africa, ODA for educa- for 64 percent of all aid allocations in 2022. Bilateral donors tion declined in 2022. Between 2020 and 2022, the World differ in their strategies. Some, such as Germany and Norway, Bank’s International Development Association (IDA) was the have increased their funding to multilateral financing orga- leading provider of ODA globally for education, disbursing an nizations to enhance their aid effectiveness (OECD, 2021 average of US$2 billion per year. It was followed by Germany and OECD, 2024). In 2022, Germany allocated 30 percent (US$1.4 billion), the United States (US$1.3 billion), and the of its aid through budget support and pooled funds, while the European Union (US$1.2 billion). The next three largest United States allocated only 2 percent of through this modal- donors by volume—France, Japan, and the United Kingdom— ity. The European Union, France, and the United Kingdom contributed a combined total of less than US$1 billion annually. allocated one-fifth of their education aid through scholarships Collectively, these donors provided an average of US$9 billion (Figure 8). Across all donors, core funding, technical assistance, per year, accounting for nearly 60 percent of total aid for educa- and budget support decreased from 36 percent in 2010 to 23 tion. Since 2018, the World Bank’s ODA disbursements have percent in 2022, suggesting a shift towards project funding. 11 Education Finance Watch 2024 What do we know about household education spending? As already mentioned, EFW2024 estimates that households contribute one-quarter of all global education expenditures. This result is consistent with the findings of the last two EFW editions even though EFW2024 has used a new data source. EFW2022 and EFW2023 relied primarily on the final reports of national household income and expenditure surveys. Most of these reports included a table that classified individual consumption by purpose, one of which was education, either on its own or combined with culture and/or recreation. The share of education in total household consumption was then multiplied by the share of household consumption in GDP17 to estimate household education consumption expenditure as a share of GDP. EFW2024 has the benefit of new and better data to assess household education spending. It uses the United Nations’ official country data reported from national accounts, which includes education and total household consumption expen- diture figures from which the share of education in house- hold budgets has been calculated. This database contains five times more data points as those used for the estimates in the last two EFW editions. In the absence of such information in © John Hogg/World Bank past EFW editions, the assumption had been that the share of household spending in total education spending was constant. These new data points provide time-series information that household education spending data prior to 2018. In contrast, make it possible to test whether the assumption of a constant in LMICs, the priority that households assign to education is share was valid or not. The disadvantage of the new source is higher, reaching a median of 2.9 percent and an unweighted that most of these data points come from UMICs, and partic- mean of 3.6 percent (based on data from the latest available ularly HICs. While these countries account for the bulk of years 2018 to 2022), though these percentages vary widely global education spending, it was necessary to complement among LMICs. Household spending on education in any the UN’s database with original data from national reports country depends on characteristics such as economic status, to increase the number of observations from LICs and, espe- household location, and the type of school children attend cially, LMICs where some of the highest household out-of- (World Bank and UNESCO, 2023). For instance, in African pocket rates have been recorded (Figure 9). For instance, the countries, it costs families between 1.5 and 5 times more to UN National Accounts contain data on household spending enroll a child in a private school than in a public school. The on education for 23.1 and 31.5 percent of LICs and LMICs share of household spending on education is inversely related respectively. When these data are supplemented by data from to the volume of public spending on education in a given coun- national household expenditure reports, the EFW2024 sample try (UNESCO, 2022). The median share of household educa- of LICs and LMICs increased to 53.8 and 85.2 respectively. tion spending drops to 1.5 percent in UMICs and 1.3 percent in HICs (Figure 9). Household education spending varies across regions, with the highest allocations being in LMICs. Households in the The proportion of GDP that households allocate to educa- world’s poorest countries generally cannot afford to spend tion has generally remained consistent over time. When the more than 1 percent of their budget on education, and often share of household spending devoted to education is multi- much less than that. This is evident from UN National plied by household consumption expenditure as a share of GDP, Accounts Data for the four LICs included in the sample with it is possible to form a picture of whether household educa- available recent data, as well as in 14 additional LICs with tion spending as a share of GDP has changed. Analyzing 17 This includes the market value of the final consumption expenditure of households and of non-profit institutions serving households (NPISH) as reported in countries’ National Accounts. 12 Education Finance Watch 2024 data from selected countries from five regions reveals that, in Figure 9. Households in LMICs spend at least twice as general, household education spending as a share of GDP tends much on education as households in HICs to remain stable over time. Nevertheless, in some countries, Distribution of the share of education in total household significant changes occur within a relatively short period. For consumption spending by country income group, 2018-22 instance, between 2010-2011 and 2020-2021, household educa- 6 tion spending as a share of GDP rose from 0.9 to 1.5 percent in the United Kingdom, from 1.7 to 2.4 percent in South Africa, 5 from 2.1 to 2.8 percent in India, and from 2.2 to 3.5 percent in Mongolia. In other countries, there have been considerable 4 3.6 ∆ declines during the same decade, for example, from 4.3 to 3.0 percent in Kenya and from 3.6 to 2.3 percent in Republic of 3 2.9 % 2.3 Korea. In several countries with available data from the 2020- ∆ 2022 period, including Colombia, Montenegro, Namibia, and 2 1.7 Singapore, it is evident that household education spending has 1.5 ∆ 1.3 0.9 fallen. This trend is likely a consequence of the economic shocks 1 ∆ experienced by many households due to COVID-19. Finally, 0.5 there are considerable differences in household education spend- 0 ing within regions. For example, in Latin America, households LIC LMIC UMIC HIC in Costa Rica spent three times as much as those in Mexico, Source: Author analysis based on the United Nations National Accounts Official Country while in Sub-Saharan Africa, households in Namibia spent six Database using the latest data available between years 2018 and 2022. Notes: The box plots show the distribution of the share of education in total household times as much as those in Senegal. Also, in Europe, households consumption expenditure by country income group. The box limits show the range within which the central 50 percent of the data is found. The central line (and label) indicates the in Greece spent six times as much as those in France. In South median value, and the ∆ sign (and label) indicates the unweighted mean for each country Asia, households in India spent three times as much as those in income group. Lines extending from each box illustrate the range of the remaining data, with dots placed past the line edges indicating outliers. The graph includes observations Sri Lanka, and in East Asia, households in the Philippines spent from 4 LICs, 12 LMICs, 17 UMICs, and 42 HICs. five times as much as those in Thailand (Figure 10). Figure 10. While household education spending tends to remain stable over time, significant differences exist within regions Household education consumption expenditure as a share of GDP, selected countries, by region 2010–22 5 Latin America and Sub-Saharan Africa Europe and Central Asia South Asia East Asia and Pacific the Caribbean 4 Namibia Philippines Mongolia Colombia 3 Kenya India Korea, Rep. Costa Rica % South Honduras Africa Hong Kong SAR, China 2 Guatemala Togo Greece Australia Brazil Nicaragua United Kingdom Singapore Montenegro Sri Lanka Malaysia 1 Japan Mexico New Zeland Türkiye Senegal Ukraine Bhutan Thailand Burkina Faso France Finland 0 2010 2014 2018 2022 2010 2014 2018 2022 2010 2014 2018 2022 2010 2014 2018 2022 2010 2014 2018 2022 Source: Author analysis based on the United Nations National Accounts Official Country Database (on share of education in total household consumption expenditure) and World Development Indicators (on consumption as share of GDP). 13 2 Using Funds Equitably and Efficiently Public spending on education can be highly unequal within countries, with wealthier groups often capturing a dispropor- tionate share of available resources. Inequalities tend to be highest in poorer countries, where differences in enrollment patterns by income quintile tend to be most pronounced and © Salahaldeen Nadir/World Bank and data published in government budget and expenditure reports. The report revealed that government spending on primary education favored poorer households in both coun- tries, whereas spending on secondary and tertiary education disproportionately benefited wealthier households, reflect- can result in significant inequalities in public education spend- ing their higher enrollment rates at those levels, especially in ing across the income distribution (Figure 11). These inequali- tertiary institutions. ties can be exacerbated by subnational differences in education spending. It is very common for a child living in one part of This year, EFW2024 expands the analysis by examining the a country to attend a school that receives much more fund- distribution of public spending on education across household ing than a comparable school in another region. For example, income quintiles at each education level18 in twelve countries in EFW2021, it was noted that spending per child in Sudan (six LICs, three LMICs, and three UMICs). EFW2024 also was approximately six times higher in the highest spending uses microdata from household surveys and data published region than in the lowest (World Bank and UNESCO, 2021). in government budget and expenditure reports (Figure 11). Subnational public spending differences can reinforce exist- ing patterns of poverty and disadvantage. In many countries, Results are mixed in the case of pre-primary education. spending per enrolled child is significantly lower in poorer Recognizing the significant impact of early childhood devel- regions than in wealthier regions. For example, in Uganda, the opment on reducing delayed school entry and enhancing life- relationship between per enrolled child spending on education long learning, countries are increasingly prioritizing and funding and levels of poverty by districts has been found to be negative these initiatives due to their high return on investment and posi- and statistically significant (World Bank, 2023). tive effects on human capital development (World Bank and UNESCO, 2023). Yet available data on public spending on pre-primary education shows mixed results. Some countries, Equity like Thailand, fund pre-schools in a way that enable greater access for lower-income families, while in others, like Niger, public Education is critical for equalizing opportunities and provid- spending for pre-primary education disproportionately bene- ing each child with the skills to achieve their full potential. fits higher-income families. This likely unintended consequence Children in disadvantaged and vulnerable situations face (negative externality) occurs because even subsidized pre-primary barriers to accessing school and learning related to household education can still impose a financial burden on low-income income and location, gender, ethnicity, and disability, among families, who may need to cover ancillary costs. Additionally, others (UNICEF, 2023). EFW2023 analyzed the distribu- the limited availability of infrastructure and teachers, especially tion of public spending on education across household income in low-income countries, can lead to quality gaps between urban quintiles at each education level (pre-primary, primary, second- and rural areas (Agyekum et al., 2023). Three of the twelve coun- ary, and tertiary) in two countries (Côte d’Ivoire in 2015 and tries (Armenia, Pakistan and Togo) have not reported data on 2019 and Ghana in 2013 and 2017), using household surveys government spending on pre-primary education. 18 Pre-primary, primary, secondary, and tertiary. 14 Education Finance Watch 2024 Figure 11. Public spending on education can be highly unequal, with wealthier groups often capturing a greater share of available resources Distribution of total public education funding by income quintile and by education level, selected LICs and LMICs, 2013-2021 Burkina Faso 2013 LIC Ethiopia 2013 LIC Niger 2014 LIC Malawi 2016 LIC 1.0 1.0 1.0 1.0 0.9 0.9 0.9 0.9 0.8 0.8 0.8 0.8 0.7 0.7 0.7 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.4 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Rwanda 2016 LIC Togo 2017 LIC Pakistan 2015 LMIC Bangladesh 2016 LMIC 1.0 1.0 1.0 1.0 0.9 0.9 0.9 0.9 0.8 0.8 0.8 0.8 0.7 0.7 0.7 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.4 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Bhutan 2017 LMIC Thailand 2020 UMIC Mexico 2020 UMIC Armenia 2021 UMIC 1.0 1.0 1.0 1.0 0.9 0.9 0.9 0.9 0.8 0.8 0.8 0.8 0.7 0.7 0.7 0.7 0.6 0.6 0.6 0.6 0.5 0.5 0.5 0.5 0.4 0.4 0.4 0.4 0.3 0.3 0.3 0.3 0.2 0.2 0.2 0.2 0.1 0.1 0.1 0.1 0.0 0.0 0.0 0.0 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Q1 Q2 Q3 Q4 Q5 Pre-primary Primary Secondary Tertiary Line of perfect equality Source: Author estimates using household survey micro data and EFW2024 database. Note: Some countries do not include any estimates for the pre-primary or tertiary education levels due to a lack of available data. Government spending on primary education tends to bene- al., 2021). Additionally, wealthier households typically have fit lower-income households, which typically have a higher greater access to private education opportunities.19 These find- number of children attending government schools. Spending ings were consistent across the country income groups studied: on primary education was found to be pro-poor. In Figure 11, LICs, LMICs, and UMICs. the Lorenz curves show the distribution of expenditure by consumption quintile above the diagonal line (the 45-degree Funding for secondary and tertiary education becomes more line), which means that poorer households in 11 of the coun- equitable as enrollment increases among the poor. Typically, tries received a higher share of government spending in expenditure on secondary and tertiary education is skewed primary education (yellow curve) than wealthier households. toward the wealthiest as those with access to these educa- This proportion is affected by the fact that lower-income fami- tion levels mainly come from the richest households. This lies often have larger numbers of children (Munoz Boudet et was confirmed by our findings. Where data were available 19 Bangladesh was the outlier, with wealthy families receiving more public education funding than did poorer families. 15 Education Finance Watch 2024 for the countries in our sample, public funding for tertiary the percentage of 10-year-olds who achieve minimum profi- education for all countries, except for Bhutan and Mexico20, ciency level in reading with per child education expendi- was skewed toward richer households. While it is essential tures shows that learning poverty is highest in countries to recognize the challenges involved in making post-second- that spend the lowest amounts per school-age individual. ary education accessible to the poor, it is equally important to However, some countries with lower levels of per child avoid the misconception that governments should not invest spending have achieved similar minimum proficiency levels in post-secondary education on the grounds of it not being in reading among 10-year-olds as those that spend more. pro-poor. Effective and functional post-secondary education For instance, Armenia spends less than one-third as much systems have the potential to transform the lives of individu- per child as Chile, yet a similar proportion of each coun- als from impoverished backgrounds by boosting their future try’s ten-year-olds – nearly two-thirds (63 percent) – can earnings (Shimeles, 2016). Indeed, progressive investments in read and understand a paragraph of age-appropriate text. post-primary education have proven to be effective in reduc- Türkiye spends less (US$2,630) to educate a child than the ing inequality in Africa, as individuals from poorer households Dominican Republic (US$3,173), but fewer than 20 percent who attain secondary and tertiary education have been found of its ten-year-olds are in learning poverty, compared to 80 to receive higher than average increases in earnings (Abdullah percent in the Dominican Republic (Figure 12). Some of et al., 2013). the factors that influence these outcomes relate to charac- teristics of service delivery that are difficult to change. It is generally cheaper, for instance, to provide education services Efficiency in densely populated and more urbanized countries than in more sparsely populated ones. However, many educa- Countries vary greatly in their effectiveness at converting tion systems also suffer from spending inefficiencies due to government education spending into better educational suboptimal spending decisions, limited accountability, and outcomes. On average, richer countries tend to have both the diversion of education funds to other uses (World Bank higher spending and better learning outcomes. Comparing and UNESCO, 2023). Figure 12. Countries differ in how effectively they translate funding into improved learning outcomes Education expenditure per primary-school-age child and learning poverty, latest year since 2015 100 ZMR LAO LIC TCD 90 BND PHL DRB HMR LMIC UMIC GHL TGO ZAF HIC 80 PAK GAN GTM PRY DOM BFA CMR 70 SEN SLV ECU JOR Learning poverty, % 60 ARG BEN NEG EGY IDN COL KWT 50 MEX BRA PER URY IRN MYS OMN 40 AZE CRI UZB 30 ARM CHL ARE 20 SVK TUR BGR CZE ICR DEU AUT 10 ALB NZL MITCYP USA HUNLVA PRT AUS BEL ESP SVN BA POL IPN HKA FIN SWE NOR LTU IRL KOR HRV SGP 0 0 5,000 10,000 15,000 20,000 Government spending on primary education per child of primary school age (constant PPP$, 5 year average) Source: Author estimates using the EFW2024 database (education spending), the Learning Poverty database, and the UN Population Database. 20 The more young people from the richest households enroll in private higher education institutions, as in the case of Bhutan and Mexico, the less pressure they place on the public educa- tion system and, hence, the more progressive investments in post-secondary education become. 16 3 Spotlight on Public Debt External borrowing plays a critical role in education financ- ing, particularly in developing countries whose domestic resources may not be sufficient to meet the educational needs of a growing population. By accessing funds from international financial institutions, governments can invest in expanding programs (World Bank, 2023a). Additionally, even when funds © Maria Fleischmann/World Bank are used efficiently, the returns on education investments typi- cally take at least 15 to 20 years to materialize. In this section, we evaluate the impact of debt on education spending in terms of both magnitude and management. and improving their educational infrastructure, teacher train- ing, and learning materials. This influx of capital can help to Over the past decade, debt relative to gross national income bridge the gap between limited domestic budgets and the (GNI) has increased in LICs and LMICs.21 Between 2012 financial resources needed to achieve educational goals. For and 2022, the accumulation of external debt outpaced growth countries facing tight fiscal constraints, limited fiscal space, in the GNIs of LICs and LMICs, as well as global trade. low household contribution capacity, and sluggish economic During this period the GNI of LICs and LMICs rose on aver- growth, international financing may be the only viable option age by 33 and 21 percent, while their combined external debt for enabling their governments to make adequate public stock rose by 109 and 46 percent, respectively (World Bank, investments in education. Borrowing for education can be a 2023a).22 This decade-long asymmetry between economic sound investment if it leads to improved economic outcomes growth and debt accumulation has created or exacerbated debt by producing a more educated population, which increases vulnerabilities in many LICs and LMICs, making it a matter productivity and earnings, which in turn can help the country of urgency to address these vulnerabilities.23 to repay its debt (World Bank and UNESCO, 2023). As a result, the percentage of LICs in debt distress or at high However, borrowing must be managed prudently to ensure risk of falling into it increased from 21 percent to 58 percent that it does not lead to unsustainable debt levels. Although between 2013 and 2022. Moreover, about 60 percent of the external financing can help countries increase their levels of countries eligible to receive IDA resources have been assessed investment in education, they must also allocate sufficient as being at high risk of, or already in, debt distress (World domestic resources to education. Over-reliance on exter- Bank, 2023a). In some countries in Sub-Saharan Africa, such nal financing can result in public finance volatility, depen- as Rwanda, Zambia, and Sudan, the ratio of debt service to dency, and unsustainable debt. Therefore, mobilizing greater GNI has increased by 7, 8, and 9 percent, respectively (IMF, domestic resources is essential to ensure long-term educational 2022). and economic stability. Furthermore, higher levels of debt to finance education spending do not necessarily lead to better Surging interest rates have intensified debt vulnerabilities educational outcomes if institutions are weak and/or if invest- in developing countries. In 2022, LICs and LMICs incurred ments are not efficiently directed towards effective policies and a historic high of US$443.5 billion in public and publicly 21 Gross domestic product (GDP) and gross national income (GNI) both measure a country’s income, but GDP counts only income received from domestic sources, whereas GNI includes net income received from abroad. The World Bank favors the use of GNI in public debt analysis for operational purposes. The International Debt Statistics database follows this conven- tion and provides users with GNI data for each reporting country and the relevant external ratios of debt stock and debt service to GNI ratios (World Bank, 2023a). 22 In US dollars. 23 The International Debt Report (IDR) is a longstanding annual publication of the World Bank featuring external debt statistics and analysis for the 122 countries that report to the World Bank Debtor Reporting System. IDR 2023 is the 50th annual edition. 17 Education Finance Watch 2024 © Bisual Photo/Shutterscotk guaranteed (PPG) debt service payments due to elevated The debt situation has become particularly worrisome for borrowing, and it is projected that this rising trend will persist.24 LICs, some of which are allocating nearly as much per capita Debt service on PPG external debt alone is expected to rise to debt servicing as to education. Over the past 10 years, inter- by 10 percent in 2023-24 compared to the previous two years. est payments have increased at a pace exceeding the growth This huge increase in borrowing has occurred during a time of education spending in developing countries (UNCTAD, of rising interest rates and largely unfavorable exchange rate 2024), although this trend varies by region. In 2022, develop- movements, which has exacerbated the fiscal burden of exter- ing countries in Africa invested nearly as much per capita in nal debt service payments. As a result, servicing external debt education (US$65 per capita25) as they paid in interest (US$50 could become increasingly burdensome for many LICs and per capita), while those in South Asia spent more on inter- LMICs, potentially crowding out spending on other priorities est (US$103 per capita) than on education (US$81 per capi- such as education (World Bank, 2023a). Over the past decade, ta).26 For instance, the median level of government spending interest payments on debt by IDA countries have quadrupled, on education as a share of total public expenditure in Africa reaching an all-time high of US$23.6 billion in 2022. More was 15.5 percent in 2021, down by 0.9 percentage points 2012 than one-third of their external debt involves variable inter- (UNESCO, 2024), largely due to increased debt servicing est rates that could rise suddenly, further worsening the fiscal costs and the shift of resources to address the health needs situation in these countries (World Bank, 2023a). In the past resulting from the COVID-19 pandemic (AUC et al., 2024). three years alone, there have been 18 sovereign defaults across In recent years, rising interest payments have coincided with 10 developing countries- more than at any time over the past a decline in the share of government budgets allocated to two decades. education in countries like Ghana, the Republic of the Congo 24 Publicly guaranteed debt (PPG) involves loans or credits that are guaranteed by the government, ensuring that the debt will be repaid even if the original borrower defaults. PPG debt service payments refer to the payments made by governments to service the part of their debt that is publicly guaranteed. This includes the aggregate number of repayments, interest, and other charges on the debt. 25 In this section, we use “education spending per capita” to compare with debt servicing per capita. This figure is derived by dividing total education spending by the overall population. This differs from metrics such as education spending per child or per school-age individual, which we use in other sections of EFW2024. 26 More developed regions typically spend significantly more per capita on education than they do on annual interest payments (1.5 times more in North America, 5.2 times more in the Middle East and North Africa, 3 times more in Europe and Central Asia, and 6 times more in East Asia and the Pacific) according to UNCTAD data (Figure 11a.) 18 Education Finance Watch 2024 Figure 13. Poorer countries’ interest payments have been growing faster than their education spending a. Public expenditure per capita on interest and education in 2022 b. Nominal change in public expenditure per capita (US$) between (US$) in selected countries 2010-2012 and 2020-2022 in selected countries 67 39 Republic Republic of Congo 78 of Congo -32 166 127 Ghana Ghana 64 -31 30 23 Lao PDR Lao PDR 29 18 91 53 Zambia Zambia 52 -15 0 50 100 150 200 -60 -30 0 30 60 90 120 150 US$ US$ Interest Education Interest Education Source: Author estimates using EFW2024 database and UNCTAD data. and Zambia (Figures 13a and 13b). Although reducing debt economic stability. Debt restructuring initiatives, such as debt service payments could free up domestic resources for educa- relief and debt swaps, offer significant benefits and challenges tion, it is important to recognize that this does not guarantee for countries with high debt burdens (World Bank, 2024b). these resources will be allocated to education, as competing The main benefit of debt restructuring is that it alleviates a fiscal priorities may necessitate funding for other critical needs. country’s immediate financial pressure, allowing it to redi- rect resources toward critical social investments, including As the cost of debt servicing in developing countries has education. However, the process of debt restructuring can be climbed, opportunities to borrow have dwindled. In 2022, complex and time-consuming, often requiring coordination new external loan commitments to PPG entities in developing among multiple creditors, and may still not lead to long-term countries dropped by 23 percent to US$371 billion—the lowest fiscal stability. Moreover, these restructuring efforts along with level in a decade. When a country has a high debt burden, its the associated fiscal consolidation policies adopted by coun- ability to borrow additional funds is often severely constrained tries reduce debt, often result in expenditure reductions that due to a loss of confidence among lenders and investors. This can actually reduce critical human capital and social invest- can lead to higher borrowing costs, making new loans more ments (Miningou, 2023). expensive and less accessible (World Bank, 2023). This is not the first time a debt crisis has had a negative impact on devel- Debt swaps are increasingly being used to alleviate countries’ opment funding. During the peak of the previous debt crisis in debt servicing burdens. Debt swaps are financial arrange- 1994, the median country’s debt-to-GDP ratio was 72 percent, ments where a portion of a country’s external debt is forgiven whereas it was only 33 percent at the end of 2021. However, if or reduced by the donor in exchange for the debtor country current trends continue, there may be a return to the debt ratios committing to investing the equivalent amount in local devel- of the 1990s within the next seven years (Kose et al., 2021). opment projects. These projects often focus on areas such as education, healthcare, environmental conservation, and infra- While not a panacea, debt restructuring, when combined structure. Before a debt swap is implemented, countries need with other funding sources, can sometimes help increase to undertake a comprehensive evaluation to assess whether the education funding in highly indebted poor countries. As debt instrument is viable and beneficial. From a debt and financial levels soar, countries are increasingly considering restructuring perspective, the most important criteria are: (i) the country’s their debt to better manage their financial burdens and foster initial debt position and the swap’s likely effects on its debt 19 Education Finance Watch 2024 Box 1. Somalia - Less debt and more investment in education Before the World Bank and the IMF launched the Heavily Indebted Poor Countries (HIPC) Initiative in 1996, many coun- tries were spending more on debt service than on health and education combined. As a result of the initiative, their spending on social services, including health and education, has increased to five times the amount of their debt payments. However, not all indebted countries qualify for HIPC. To be considered for HIPC Initiative assistance, a country must: • Be eligible to borrow from the World Bank’s International Development Association (IDA), which provides interest-free loans and grants to the world’s poorest countries, and from the IMF’s Poverty Reduction and Growth Trust, which provides loans to low-income countries at concessionary rates. • Be carrying an unsustainable debt burden that cannot be addressed through traditional debt relief mechanisms. • Have a track record of reform and sound policies through IMF- and World Bank-supported programs. • Develop a Poverty Reduction Strategy Paper (PRSP) through a broad-based participatory process. The Executive Boards of the IMF and the World Bank formally decide on a country’s eligibility for debt relief and the inter-  onor community commits to reducing debt to a sustainable level. This stage is referred to as the decision point. national d Once a country reaches this point, it can immediately obtain interim debt relief. To receive a full reduction of its debt under the HIPC Initiative, a country must: • Establish a further track record of good performance under programs supported by loans from the IMF and the World Bank. • Successfully implement key reforms to which it agreed at the decision point. • Adopt and implement its PRSP for at least one year. A country that has met these criteria has reached its completion point, which allows it to receive the full debt relief agreed upon at the decision point. O  f the 39 countries that are eligible or are potentially eligible for HIPC Initiative assistance, 37 had reached their comple- tion points as of June 2024 and are receiving full debt relief from the IMF and other creditors. In December 2023, Somalia became the latest to reach its completion point. Somalia has increased government education spending in conjunction with obtaining debt relief under the HIPC Initiative (IMF, 2023). Debt restructuring irrevocably reduced Somalia’s debt from 64 percent of GDP (US$5.2 billion) in 2018 to less than 6 percent of GDP (US$557 million) in 2023 in net present value terms. In March 2020, the World Bank announced that the IMF and IDA had determined that Somalia had taken the necessary steps to begin receiving debt relief under the enhanced HIPC Initiative (World Bank, 2020a). Upon reaching the HIPC decision point, Somalia regained access to inter- national financial markets. Somalia is increasing its investment in education focusing on teachers, a key driver of student learning. In January 2023, President Hassan Sheikh Mohamud announced the country would hire a record 3,000 new teachers following a four-fold increase in the country’s education budget (to US$34 million) for the year. He also reported that, for the previous five years, only 1,000 teachers had been on the public payroll. This move was intended to address the substantial teacher shortage, a key contributing factor to Somalia’s high out-of-school rate. It will be crucial for Somalia to invest these resources effectively by hiring the most qualified teachers and by funding the provision of supporting services, particularly by financing schools. Currently, fewer than one in five children complete primary school in Somalia (UNESCO, 2024, pp.33-34). As Somalia moves towards stability and development after 30 years outside the international financial system, the immediate normalization of its relations with the international community as a result of reaching the HIPC decision point will re-open its access to critical additional financial resources that can gener- ate growth and sustainable employment for Somalis (World Bank, 2020a and 2020b). 20 Education Finance Watch 2024 © Sarah Farhat/World Bank sustainability;27 (ii) the net financial gains for the debtor; (iii) Debt-for-development swaps can be used across a wide range the country’s debt management capacity and commitment of public expenditure programs, but it is important to ensure to transparency; and (iv) the opportunity costs for both the that the country’s spending commitments are fully aligned borrower and the donors. Countries that are potentially good with its development goals and strategies. Swaps are intrin- candidates for swaps are those at “moderate” or “high” risk of sically complex, and it is crucial that these new spending debt distress with a sustainable long-term debt outlook, but commitments maintain or increase the overall efficiency of experiencing temporary liquidity pressures. These are usually the budget (World Bank, 2024b). smaller economies where the transaction can provide criti- cal short-term relief and increase the country’s debt sustain- Since the 1980s, several nations have used debt swaps to fund ability prospects. In such countries, debt swaps can smooth education projects,28 although their impact has not been debt amortization profiles and improve liability manage- widely assessed.29 The GPE’s Debt2Ed program30 can be used ment, while supporting high-impact development projects. for debt swaps or loan buydowns, helping partner countries 27 Countries with unsustainable debt levels or those requiring (or already undergoing) comprehensive debt restructuring are not suitable candidates for debt swaps, which are not appropriate tools for restoring debt sustainability. These countries will need to negotiate substantial reductions in their debts from all creditors and implement a fully funded macroeconomic adjust- ment program. 28 Debt-for-education swaps are broadly defined as the cancellation of external debt in exchange for the debtor government’s commitment to mobilize domestic resources to spend on educa- tion (UNESCO, 2009). 29 In addition to Côte d’Ivoire’s recent agreement, other debt swap examples include: (i) €26 million in 2002 and €23 million in 2004 between Germany and Indonesia; (ii) US$10 million in 2005 between Spain and El Salvador; and (iii) €1.2 billion in 2006 between France and Cameroon (Cassimon et al., 2011; UNESCO, 2011; Ito et al., 2018; GPE, 2023a; and World Bank, 2024). 30 GPE’s Debt2Ed is an innovative financing instrument to transform repayments on national borrowing into investments in education while securing significant grant financing from the GPE through the Multiplier. The Multiplier unlocks grant financing for a partner country’s education system when external partners mobilize new and additional financing (https://www. globalpartnership.org/funding/gpe-multiplier). 21 Education Finance Watch 2024 © Dominic Chavez/World Bank improve their debt metrics (by lowering debt stock and shown that each additional year of schooling that a person debt service payments), lower borrowing costs, and increase completes yields 10 percent more income on average, which resources for education to promote economic growth (GPE, is higher than the average annual returns on the US stock 2023a). While debt swaps have the potential to be used as a market. However, the rising tide of debt service obligations debt restructuring mechanism, they have not yet been widely in developing countries, particularly in LICs, threatens to implemented. Available evidence suggests that debt-for-ed- undermine critical investments in human capital. As debt ucation swaps could be a viable instrument for mobilizing service costs soar, they risk overshadowing and potentially higher levels of education funding under certain conditions. crowding out funding for education, which can have detri- However, there is still no evidence demonstrating a direct mental effects on a country’s human and economic devel- correlation between debt swaps and improved student learn- opment trajectory. Long-term, sustainable increases in ing outcomes (Ito et al., 2018). government education spending per student, such as those made by Republic of Korea during the 1990s, have been Amid all the uncertainty, one thing is clear: investing in proven to increase student learning when phased appro- learning is one of the most cost-effective, forward-look- priately, aligned with population growth, and coupled with ing strategies a country can adopt. Global evidence has strong economic growth.31 31 Republic of Korea has higher than average expenditures on education. Its expenditure on education increased dramatically in the 1990s, both as a percentage of GDP (rising from 4.9 percent in 1990 to 6.8 percent in 1998) and on a per student basis, driven by the country’s high rate of economic growth (averaging 4 percent per year). In 2023, Korea spent US$14,113 annually per full-time equivalent student (adjusted for purchasing power and including expenditure on research and development), compared to the OECD average of US$12,647 (OECD, 2023). 22 4 Monitoring and Reporting on Education Spending Delays and gaps in country-level reporting of education spending data undermine accountability and monitoring, thereby restricting policy and technical discussions on educa- tion financing. Effective monitoring of educational outcomes requires detailed analysis of funding amounts and the specific had not reported data for 2022 (the last year for which data were used in EFW2024) on education expenditure as a share of GDP, while 58 percent had not reported their education spending as a share of government spending (Figure 14c). Furthermore, most of the reported data were not broken down © studiolaska/Shutterstock use of these financial resources. This information is also vital by type of expenditure or level of education, making it difficult for accountability. The share of education spending in total to effectively monitor the adequacy and efficiency of education government spending is a global Sustainable Development spending allocations. Without this information, it is impos- Goal monitoring indicator under SDG 1 on poverty reduc- sible to track trends and analyze deeper the inequalities and tion. The EFW uses the UIS database as its main source of inefficiencies in education spending outlined in the EFW2024. data on government spending on education, while also making Reporting these data in a timely manner is essential to ensur- use of data from the IMF and from the World Bank’s BOOST ing effective policymaking and making any necessary course Open Budgets Portal32 to impute any missing data. Seventy- corrections. three percent of the countries studied for the EFW2024 had reported 2020 data on government expenditure as a share Despite the importance of data for monitoring, only a of GDP in the UIS, while 70 percent had done so for 2021. minority of countries have reported comparable data on core Similarly, 72 percent of countries had reported 2020 data on education finance indicators to the UIS. Data on spending education spending as a percentage of total government expen- on different education levels was available for only 47 and diture, and 68 percent had done so for 2021 (Figure 14a and 22 percent of countries in 2018 and 2022, respectively, and Figure 14b). much of that data was incomplete or inconsistent. More importantly, there is a significant lack of household-level Many countries do not report data in a timely way. As of data on education spending, especially in the post-COVID February 2024, 51 percent of countries in the UIS dataset period. 32 The BOOST Open Budgets Portal is an effort to create a one-stop shop for budget data worldwide, with the hope of highlighting countries’ efforts to report their budget data, facilitating access and promoting use of spending data, and motivating other countries to report their own data https://www.worldbank.org/en/programs/boost-portal/about-the-portal. 23 Education Finance Watch 2024 Figure 14. Although the country database that informed this report is adequate, there is scope for further refinement a. Availability of data on government b. Availability of data on government c. Availability of data for the latest year for which education spending as a share of GDP in education spending as a share of total data are used in EFW reports the EFW2024 database government spending in the EFW2024 database 100 100 100 19% 19% 20% 22% 40% 20% 21% 20% 24% 41% 41% 80 6% 80 80 6% 7% 6% 39% 35% 40% 36% 50% 7% 8% 8% 8% 60 60 60 11% % % % 17% 40 40 40 20 76% 74% 73% 70% 48% 20 74% 72% 72% 68% 42% 20 61% 65% 60% 64% 50% 59% 0 0 0 2018 2019 2020 2021 2022 2018 2019 2020 2021 2022 2020 2021 2022 2020 2021 2022 Data Data Data Data Data Data Data Data Data Data data in data in data in data in data in data in EFW2022 EFW2023 EFW2024 EFW2022 EFW2023 EFW2024 UIS Others UIS Others Education spending Education spending as % of as % of GDP total Governmental expenditure Not reporting Not reporting With data Not reporting Source: Author estimates using the EFW2024 database and the EFW2022 and EFW2023 reports. Note: The percentage is calculated for all of the 218 countries and territories covered by EFW2024. © Jonathan Ernst/World Bank 24 5 Summary To achieve national and international education goals, many countries will need to invest more and better in their education systems. Over the last decade, government education spend- ing has increased steadily, but this has not led to any signifi- cant increases in per child allocations, particularly in LICs, due © Liang Qiang/World Bank teacher performance, strengthening governance, and chan- neling resources to cost-effective policies and programs. • The amount of aid provided for education in LICs is high, but the proportion of aid allocated to education has to rapid population growth. Learning outcomes remain poor, declined in recent years. Globally, the volume of educa- especially in countries that spend the least per school-age child. tion aid in 2022 reached a record high of US$16.6 billion, The World Bank estimates that learning poverty in low- and growing by 16 percent in real terms. However, education’s middle-income countries increased after the pandemic from 57 share of total ODA declined from 9.3 percent in 2019 to percent in 2019 to 70 percent in 2022 (Azevedo et al., 2022). 7.6 percent in 2022. In 2022, ODA accounted for 12.2 The COVID-19 pandemic also strained public finances, reduc- percent of education spending in LICs and only 0.29 ing the prospects for maintaining pre-pandemic investment percent globally.   levels in education. Additionally, growing student populations, particularly in LICs and the increasing negative impacts of • Over the past 10 years, interest payments on public debt climate-related events on education systems are putting further have increased faster than government education spend- pressure on limited resources in countries that invest the least ing in developing countries. On a per capita basis, some in education. In examining trends and patterns in education LICs are allocating nearly as much to debt servicing as financing from 2010 to 2022, the report concluded following: they are to education. Mounting fiscal challenges are significantly reducing the ability of LICs to fund their • Although total education spending has increased over the education systems. Some countries are exploring inno- past decade, the amount of funding available is still insuf- vative financing mechanisms for short-term relief, such ficient to address the learning crisis, especially in LICs. as debt restructuring, debt swaps, debt-for-development LICs and LMICs have experienced a more rapid annual agreements. However, these measures must be comple- increase in education spending than wealthier nations, mented by sustained domestic resource mobilization, effi- but even those that have reached international education cient spending, effective public financial management, and spending benchmarks are struggling to secure enough robust economic growth to ensure that their populations funding to enable all students to learn effectively. can receive quality education. • LICs and LMICs face a double jeopardy: not only are their • To maintain a clear global picture of education financing financial resources scarce but also often inefficiently used, trends, it is imperative for countries to report their educa- a persistent issue highlighted in previous EFW editions. tion funding data in a timely and consistent manner, with Increased investment could enhance students’ educa- more disaggregation. When data are reported late or not at tional performance, particularly in LICs, but inefficient all, it becomes difficult to conduct an analysis of education use of resources limits its impact. To improve educational spending to inform forward-thinking policymaking. This outcomes, governments should prioritize enhancing the lack of timely and accurate data is a disservice to education, as efficiency of their current spending by optimizing public it hinders the ability to make informed decisions that could financial management, improving school management and improve resource allocations and educational outcomes. 25 Education Finance Watch 2024 References Abdullah, Abdul, Hristos Doucouliagos, and Elizabeth Manning Gasper, V., M. Mansour, C., Vellutini (2023) “Countries Can Tap Tax (2015). “Does Education Reduce Income Inequality? A Meta-Re- Potential to Finance Development Goals: Emerging market and gression Analysis.” Journal of Economic Surveys, Wiley Blackwell, low-income economies have a significant untapped tax poten- vol. 29(2), 301-316, April. tial of 8 percent to 9 percent of GDP”. IMF Blog. International Azevedo, J.P., M. Akmal, M-H. Cloutier, H. Rogers, and Y.N. Wong, Monetary Fund, Washington, D.C. https://www.imf.org/en/ 2022. “The State of Global Learning Poverty: 2022 Update.” Blogs/Articles/2023/09/19/countries-can-tap-tax-potential-to- World Bank, Washington, D.C. finance-development-goals https://thedocs.worldbank.org/en/doc/e52f55322528903b27f- GPE (2023a). “Factsheet: Debt2Ed.” Global Partnership for Educa- 1b7e61238e416-0200022022/original/Learning-poverty-re- tion, Washington, D.C. port-2022-06-21-final-V7-0-conferenceEdition.pdf GPE (2023b). “Turning debt payments into financing for education: Agyekum M.W., S.B. Yeboah, C. Dzradosi, K. Ofosu-Ampong, M.O. Leveraging GPE’s Debt2Ed in Côte d’Ivoire.” Global Partnership Quaye, C. Donkoh, Andrews Acquah, Cosmos Kwame Dzikunu, for Education, Washington, D.C. https://www.globalpartnership. Edison Pajibo, Daniel Yelkpieri, Emmanuel M.J. Tamanja, and org/blog/turning-debt-payments-financing-education-leverag- Ephraim Avea Nsoh (2023). “Rural-urban differentials in early ing-gpes-debt2ed-cote-divoire childhood education and child development: Evidence from Multi- GPE (2024). “Results Report 2023.” Global Partnership for Educa- ple Indicator Cluster Survey (MICS) in Ghana.” PLOS Global tion, Washington, D.C. Public Health 3(8): e0002171. https://doi.org/10.1371/journal. Hanyang, Y.L. (2021). “Education Finance in Korea: Achievements pgph.0002171 and New Challenges.” Hanyang University, Seoul, Korea. https:// Arias, O. and I. Kheyfets (2023). “The Adequacy of Public Expendi- www.kdi.re.kr/upload/7322/3-2-2.pdf ture on Education and the Needs Post-COVID-19.” World Bank, IMF (2022). “Crisis upon Crisis: IMF Annual Report 2022.” Interna- Washington, D.C. https://thedocs.worldbank.org/en/doc/9b9ec- tional Monetary Fund, Washington, D.C. b979e36e80ed50b1f110565f06b-0200022023/original/Adequa- https://www.imf.org/external/pubs/ft/ar/2022/downloads/imf-annu- cy-Paper-Final.pdf al-report-2022-english.pdf AUC, ITU, UNECA, OSAA and World Bank (2024). “Transform- IMF (2024). “Debt relief under the heavily indebted poor countries ing Education in Africa by Leveraging Innovative Finance and initiative - HIPC. 2023 Factsheet.” International Monetary Fund, the Digital Revolution.” Africa Dialogue Series (ADS) 2024. Washington, D.C. Accessed on September 9, 2024. https://www. African Union Commission, Addis Ababa, the International imf.org/en/About/Factsheets/Sheets/2023/Debt-relief-un- Telecommunication Union, Geneva, United Nations Economic der-the-heavily-indebted-poor-countries-initiative-HIPC Commission for Africa, Addis Ababa, the UN Office of the Ito, H., R. Sekiguchi, T. Yamawake, and J. Sapena (2018). “Debt swaps Special Adviser on Africa, New York, and World Bank, Wash- for financing education: Exploration of new funding resources.” ington, D.C. Cogent Economics & Finance, 6(1). https://doi.org/10.1080/2332 Carapella, Piergiorgio M., Tewodaj Mogues, Julieth C. Pico-Me- 2039.2018.1563025 jia, and Mauricio Soto (2023). “How to Assess Spending Needs Kose, Ayhan, Franziska Lieselotte Ohnsorge, and Naotaka Suga- of the Sustainable Development Goals: The Third Edition of wara (2021). “A Mountain of Debt: Navigating the Legacy of the the IMF SDG Costing Tool.” IMF Fiscal Affairs Department Pandemic.” World Bank, Washington, D.C. https://documentsin- 2023/005, International Monetary Fund. https://www.imf.org/en/ ternal.worldbank.org/search/33476302 Publications/imf-how-to-notes/Issues/2023/12/14/How-To-As- Miningou, Elise Wendlassida (2023). “External Debt, Fiscal Consol- sess-Spending-Needs-of-the-Sustainable-Development-Goals- idation, and Government Expenditure on Education.” World The-Third-Edition-of-the-541463 Bank, Washington, D.C. https://documentsinternal.worldbank. Cassimon, D., D. Essers, and R. Renard (2011). “An assessment of org/search/34076134 debt-for-education swaps. Case studies on swap initiatives between Munoz Boudet, A.M., A. Bhatt, G. Azcona, J.J. Yoo, and K. Beegle Germany and Indonesia and between Spain and El Salvador.” (2021). “A Global View of Poverty, Gender, and Household Comparative Education 47:2, 139-156. Composition.” Policy Research Working Paper Series 9553, World Fery, Marin, Marine de Talancé, and Miguel Niño-Zarazúa (2022). Bank, Washington, D.C. “Less Debt, More Schooling? Evidence from Cross-country OECD (2021). “OECD Development Co-operation Peer Reviews: Micro Data.” Journal of Comparative Economics, Volume 50, Issue Germany 2021,” OECD Publishing, Paris, https://doi. 1, 153-173. org/10.1787/bb32a97d-en. 26 Education Finance Watch 2024 OECD (2023). PISA 2022 Results (Volume I): The State of Learning VOA News (2023). “Somalia Hiring 3,000 Teachers after quadru- and Equity in Education, PISA, OECD Publishing, Paris, https:// pling education budget.” January 3, 2023. https://www.voanews. doi.org/10.1787/53f23881-en. com/a/somalia-hiring-3-000-teachers-after-quadrupling-edu- OECD (2023a). Education at a Glance 2023: OECD Indicators, OECD cation-budget/6902472.html (accessed on 8/11/2024) Publishing, Paris, https://doi.org/10.1787/e13bef63-en. World Bank (2020a). “Somalia to Receive Debt Relief under the OECD (2024). “Peer Review Mid-term Review of Norway.” March Enhanced HIPC Initiative.” Press Release, March 25, 2020. World 4, 2024. OECD, Paris. Bank, Washington, D.C. Patrinos, Harry, and Nobuyuki Tanaka (2024). “Education: Innovative World Bank (2020b). “Somalia To Reestablish Financial Relations Financing in Developing Countries.” Education Working Paper; with The World Bank Group After Thirty Years.” Press Release No. 1, April 2024. World Bank, Washington, D.C. http://hdl. No: 2020/067/AFR, February 27, 2020. World Bank, Washing- handle.net/10986/41476 License: CC BY-NC 3.0 IGO ton, D.C. Pholphirul, P., P. Rukumnuaykit, and S. Teimtad (2023). “Teacher World Bank (2023). “Uganda Economic Update.” World Bank, shortages and educational outcomes in developing countries: Washington, D.C. https://documents1.worldbank.org/curated/ Empirical evidence from PISA-Thailand.” Cogent Education, 10(2). en/099020224131540261/pdf/P1798401a450b40361963b1263 https://doi.org/10.1080/2331186X.2023.2243126 4ab074169.pdf Shimeles, Abebe (2016). “Can higher education reduce inequality World Bank (2023a). “International Debt Report 2023.” World Bank, in developing countries?” IZA World of Labor, Institute of Labor Washington, D.C. http://hdl.handle.net/10986/40670 Economics (IZA), pages 273-273, July. World Bank, UNESCO, UNICEF, USAID, FCDO, and Bill and UIS and GEM Report (2024). “SDG 4 Scorecard: Progress Report on Melinda Gates Foundation (2023). “The State of Global Learning National Benchmarks – Focus on Teachers.” UNESCO Institute Poverty: 2022 Update.” World Bank, Washington, D.C. of Statistics, Montreal and Global Education Monitoring, Paris. https://thedocs.worldbank.org/en/doc/34035a49acb5700ce- https://unesdoc.unesco.org/ark:/48223/pf0000388411 8b118aeda81a5cb-0510022023/original/TheStateOfLearning- UNCTAD (2024). “A World of Debt Report 2024: A Growing Burden Poverty-Feb2023Update-03-08-23.pdf to Global Prosperity.” UNCTAD, Geneva. World Bank (2024a). “Education: Innovative Financing in Developing UNESCO (2009). “Debt swaps for education.” UNESCO, Paris. Countries.” Education Working Paper No. 1. April 2024. World UNESCO (2011). “Debt swaps and debt conversion development Bank, Washington, D.C. bonds for education: Final report for UNESCO advisory panel World Bank (2024b). “Debt-for-Development Swaps and the Potential of experts on debt swaps and innovative approaches to education Role of the World Bank: Technical Note (English).” World Bank, financing.” UNESCO, Paris. Washington, D.C. http://documents.worldbank.org/curated/ UNESCO (2022).“Global Education Monitoring Report 2021/2: Non-state en/099080524122527783/BOSIB1f57baa3f0971916811e7b- Actors in Education – Who Chooses? Who Loses?” UNESCO, Paris. da53f7d5 https://unesdoc.unesco.org/ark:/48223/pf0000383550 World Bank and UNESCO (2021). “Education Finance Watch 2021.” UNESCO (2023a). “Global Education Monitoring Report 2023: World Bank and UNESCO, Washington, D.C. and Paris. Technology in Education – A Tool on Whose Terms?” UNESCO, World Bank and UNESCO (2022). “Education Finance Watch 2022.” Paris. https://unesdoc.unesco.org/ark:/48223/pf0000385723 World Bank and UNESCO, Washington, D.C. and Paris. UNESCO (2023b). “Can countries afford their national SDG 4 World Bank and UNESCO (2023). “Education Finance Watch 2023.” benchmarks?” Global Education Monitoring Report Policy Paper World Bank and UNESCO, Washington, D.C. and Paris. 49. UNESCO, Paris. https://unesdoc.unesco.org/ark:/48223/ World Bank and UNESCO (2024). “Education Finance Watch 2023. pf0000385004 Special Edition for the African Union Year of Education 2024.” UNESCO (2024). “Spotlight Report on Basic Education Completion World Bank and UNESCO, Washington, D.C. and Paris. and Foundational Learning in Africa.” UNESCO, Paris. https:// World Bank, UNESCO, UNICEF, USAID, FCDO, and Bill and unesdoc.unesco.org/ark:/48223/pf0000389481 Melinda Gates Foundation (2022). “The State of Global Learn- UNICEF (2023). “Transforming Education with Equitable Financ- ing Poverty: 2022 Update.” Conference Education, June 23, ing”. UNICEF, New York. https://www.unicef.org/reports/trans- 2022. forming-education-equitable-financing 27 Education Finance Watch 2024 Acknowledgements EFW2024 was prepared by Nobuyuki Tanaka, Diego Angel-Urdinola, May Bend, Yitong Hu, Angelica Rivera-Olvera (World Bank), Manos Antoninis, Yuki Murakami (GEM Report), and Silvia Montoya (UIS). It was prepared under the overall guid- ance of Luis Benveniste and Halil Dundar (World Bank). It benefited from comments from Albert Tumino, Fadila Caillaud, Fernando Blanco, Ivailo V. Izvorski, Fatine Guedira, Elisabeth Sedmik, Igor Kheyfets, Samer Al-Samarrai (World Bank). The team also received essential assistance from Dani Clark, Fiona Mackintosh, Stefano De Cupis (World Bank), Katherine Redman, Madeleine Barry (GEM Report). The team benefited from valuable support from education regional teams, poverty economists, and the BOOST team (World Bank). © Simone D. McCourtie/World Bank © 2024 International Bank for Reconstruction and Development / The World Bank and UNESCO Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank and UNESCO. This work is a co-publication of The World Bank and UNESCO. The findings, interpretations, and Views and opinions expressed in the adaptation are the sole responsibility of the author or conclusions expressed in this work do not necessarily reflect the views of The World Bank, its authors of the adaptation and are not endorsed by The World Bank and UNESCO. Board of Executive Directors, or the governments they represent, or those of UNESCO. Third-party content—The World Bank and UNESCO do not necessarily own each component of The World Bank and UNESCO do not guarantee the accuracy, completeness, or currency of the content contained within the work. The World Bank and UNESCO therefore do not warrant the data included in this work and do not assume responsibility for any errors, omissions, that the use of any third party-owned individual component or part contained in the work will or discrepancies in the information, or liability with respect to the use of or failure to use not infringe on the rights of those third parties. The risk of claims resulting from such infringement the information, methods, processes, or conclusions set forth. The boundaries, colors, rests solely with you. If you wish to re-use a component of the work, it is your responsibility to denominations, and other information shown on any map in this work do not imply any determine whether permission is needed for that re-use and to obtain permission from the judgment on the part of The World Bank and UNESCO concerning the legal status of any copyright owner. Examples of components can include, but are not limited to, tables, figures, territory or the endorsement or acceptance of such boundaries. or images. Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of All queries on rights and licenses, including subsidiary rights, should be addressed to World Bank the privileges and immunities of The World Bank, all of which are specifically reserved. Publications, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. DOI: https://doi.org/10.54676/NSCT2115 Rights and Permissions Designed by alejandro espinosa/sonideas.com. Photography credits: Photography credits: 1: Bart Verweij | World Bank, further permission This publication is available in Open Access under the Creative Commons Attribution-ShareAlike required for reuse. 2: Dominic Chavez | World Bank | 2013, further permission required for reuse. 3.0 IGO (CC-BY-SA 3.0 IGO) license (http://creativecommons.org/licenses/by-sa/3.0/igo/). 3: Dominic Chavez | World Bank | 2013, further permission required for reuse. 4: Steve Harris Under this license, you are free to copy, distribute, transmit, and adapt this work, including for | World Bank, further permission required for reuse. 5: Dominic Chavez | World Bank | 2013, commercial purposes, under the following conditions: further permission required for reuse. 8: Maria Fleischmann | World Bank, further permission required for reuse. 9: Fernanda Reyes | 2022 | Shutterstock, further permission required for reuse. Attribution—Please cite this publication as follows: The World Bank and UNESCO (2024). 10:  Omotayo Kofoworola | 2024 | Shutterstock, further permission required for reuse. 12: John Education Finance Watch 2024. Washington D.C., Paris: The World Bank and UNESCO. Hogg | World Bank, further permission required for reuse. 14: Salahaldeen Nadir | World Bank, Share Alike — If you remix, transform, or build upon the material, you must distribute your further permission required for reuse. 17: Maria Fleischmann | World Bank, further permission contributions under the same license as the original. required for reuse. 18: Bisual Photo | 2019 | Shutterstock, further permission required for reuse. 21: Sarah Farhat | 2016 | World Bank, further permission required for reuse. 2: Dominic Chavez | 2016 Translations—If you create a translation of this work, please add the following disclaimer along | World Bank, further permission required for reuse. 23: studiolaska | 2014 | Shutterstock, further with the attribution: This translation was not created by The World Bank or UNESCO and should permission required for reuse. 24: Jonathan Ernst | 2006 | World Bank, further permission not be considered an official World Bank or UNESCO translation. The World Bank and UNESCO required for reuse. 25: Liang Qiang | 2005 | World Bank, further permission required for reuse. shall not be liable for any content or error in this translation. 28: Simone D. McCourtie | World Bank, further permission required for reuse.