EDUCATION FINANCE WATCH 2022 Introduction The Education Finance Watch (EFW) is a collaborative effort between the World Bank (WB), the Global Educa- tion Monitoring (GEM) Report, and the UNESCO In- stitute of Statistics (UIS). The EFW aims to provide an annual analysis of trends, patterns, and issues in education financing around the world. The EFW uses various sources of data, including from UNESCO, the WB, the International Monetary Fund (IMF), and the Organisation for Economic Co-opera- tion and Development (OECD). Data quality and cover- age of education finance data has improved, but there are still large gaps in data availability and consistency across the major data sources. The EFW, where needed, uses a coherent methodological approach to deal with missing data. The specific methodology and rationale are described in detail in the accompanying Technical Note. The EFW 2021 reviewed trends and patterns of education spending for the past 10 years and captured preliminary ef- fects of COVID-19 on the education budgets of a glob- al sample of countries. The EFW 2021 documented that global education spending had increased continuously, in absolute terms, but there were indications that the pandem- ic would interrupt this trend. Furthermore, it highlighted that the policies that countries adopt to protect and increase impact on national economies globally, the EFW 2021 un- education spending differ, but a majority could make better derscored diverging trends in education financing, especial- use of the funds allocated. Lastly, in view of the pandemic’s ly between lower and higher income countries. EDUCATION FINANCE WATCH 2022 The EFW 2022 digs deeper into these topics, using the full breadth of available data. The EFW 2022 sheds fur- ther light on the impact of COVID-19 on global educa- tion financing in 2020, 2021, and 2022 and conducts a fo- cused analysis of recent trends in government education budgets using the latest data available for a subsample of high-income (HICs), low-income and lower-middle in- come countries (LICs and LMICs), as of May 2022. Key findings of EFW 2022: 1. In 2020, first year of the COVID-19 pandemic, half of the sample of countries analyzed reduced their annual real spending on education, compared to 28 percent of countries in 2019. On average, per capita spending in low-income countries grew by just 1 per- cent in 2020, fell by 4 percent in LMICs and by 6 per- cent in UMICs. Forty percent of LICs and LMICs reduced their spending on education after the onset of the COVID-19 pandemic in 2020, with an average decline in spending of 13.5 percent. 2. The available data on government budgetary com- mitments suggest that education spending lost space in national budgets of low- and lower-middle income countries in 2021 and 2022. The decline in the prioritization of education spending in govern- ment budgets in LICs and LMICs over the last two years is in contrast with trends observed in UMICs and HICs where education as a share of total govern- ment budgets was stable in 2019–2021 and in 2022 remained higher than in 2019. In LICs and LMICs, the share of education fell in 2020, rebounded slight- ly in 2021 but fell again in 2022 below 2019 levels. countries maintained higher levels of per capita pub- With overall total public spending being strained by lic education spending in 2019-2020 relative to 2014- increasing fiscal pressures, there is a risk that educa- 2015, with South Asia registering the largest increase tion spending in LICs and LMICs will not meet the (35 percent) in spending during this period. needs to implement urgent actions to recover learn- ing losses and address the already high learning pov- 4. Despite the increased efforts of low-income countries erty levels. to improve their public education spending, spending levels remain strikingly below those of higher income 3. The COVID-19 pandemic reversed a steady upward countries. As a share of GDP, government education trend in per capita real public spending on education spending in LICs rose steadily from the mid-2010s to in many middle income countries. One-third of low- reach an average of 3.6 percent in 2020, while fluc- er-middle income countries and half of upper-middle tuating around 4.8 percent for middle- and high-in- income countries spent less per capita on education in come countries. However, in 2020 government per 2019-2020 than they did in 2014-2015. Latin Ameri- capita spending was on average nearly 150 and 20 ca and the Caribbean is the one region where per cap- times higher in high- and upper middle-income coun- ita government spending on education fell below its tries (US$7,787 and US$1,079, respectively) than in mid-2010s levels, with an average decline of 3 percent low-income countries (US$53). In 2020, average gov- between 2014-2015 and 2019-2020. Low-income ernment per capita spending in sub-Saharan Africa 2 EDUCATION FINANCE WATCH 2022 (US$254) and South Asia (US$358) was less than education costs, accounting for 39 percent of the to- one-tenth of average per capita spending in Europe tal spending in education compared to 16 percent in and Central Asia (US$6,156). high-income countries. Moreover, within countries, the richest spend far more on education, further en- 5. Overall, direct aid to education was stagnant in 2020 trenching inequality: in 33 low- and middle-income and fell by US$359 million in the case of bilateral do- countries, households from the richest quintile spent nors. The decline in bilateral aid to education in 2020 4.2 percent of their budget on education compared has since been followed with cuts to aid earmarked to to just 2.4 percent among households in the poorest education by major donors, with the aftermath of the quintile. COVID-19 crisis and the war in Ukraine and shifts in some donor governments’ priorities. 7. Data availability has improved considerably. Never- theless, big gaps remain, with just 1 percent of coun- 6. Households in low- and lower-middle-income tries reporting 2020 spending according to level of countries continue to bear a significant portion of education. 3 1 Government education spending and external aid during COVID-19 The COVID-19 pandemic has worsened the global 1 uses only actual data on government education spend- learning crisis. The pandemic has led to significant learn- ing up to 2020 and budgetary allocations for 2021 and ing losses and increasing learning inequalities, between 2022 from the World Bank BOOST and UIS databas- and within countries. Schools around the world closed es, without any imputed data. This limits the analysis to a at a scale never seen, and at the height of the pandemic smaller sample of countries, but ensures time comparabil- learning was disrupted for more than 1.6 billion students ity while ensuring representation of all regions and coun- (World Bank, UNESCO, and UNICEF 2021). Most try income groups1. International aid was analyzed follow- countries tried to continue providing learning opportu- ing OECD’s definition of official development assistance nities by partially reopening schools or providing remote (ODA)2, and calculated based on data obtained from the or hybrid learning. Despite these efforts, it is estimated OECD Creditor Reporting System (CRS). No imputa- that globally learning losses from COVID-19 could cost tion was done on aid expenditures data, and imputation this generation of students close to US$21 trillion in life- was done when total aid to education was estimated using time earnings, which far exceeds the original estimate reported aid data in OECD CRS.3 of US$10 trillion made immediately after the pandemic outbreak and even the US$17 trillion estimated in 2021 (Azevedo et al. 2022). 1.1. Government education spending during the onset of COVID-19. The effects of COVID-19 on education finance are sig- nificant. In many countries, the onset of the COVID-19 More countries decreased the level and prioritization of pandemic resulted in significant mid-year budget revi- education spending with the onset of the COVID-19 sions. To alleviate the abrupt impact on economies, ad- pandemic. Half of the countries analyzed reduced their dress emergency needs, and provide fiscal stimulus, addi- annual spending on education in 2020, compared to 28 tional resources were mobilized through different means, percent in 2019, a trend observed in both higher and low- but education systems struggled to garner additional fi- er country income groups (Table 1). Forty-one percent of nancial support or to adapt to the crisis (World Bank, lower income countries reduced their spending on educa- UNESCO, and UNICEF 2021). tion after the onset of the pandemic in 2020, with an av- erage decline in spending of 13.5 percent. Moreover, the This section uses available data to analyze patterns in share of education spending in total government spending government spending on education in the aftermath of fell, albeit modestly, from an average of 15.2 to 14.3 per- COVID-19 and in the education allocations in nation- cent from 2019 to 2020 across the countries analyzed. The al public budgets in 2021 and 2022, as well as interna- decline in the prioritization of education in total public tional aid to education before and after COVID-19. For spending was more marked among low- and lower-middle the analysis of government education financing, Section income countries (Table 1). 1 Income level grouping follows the 2020 World Bank country income classification. In this section, low- and lower-middle-income countries are grouped as “lower income” countries, and upper-middle-income and high-income countries are grouped as “higher income” countries. 2 https://www.oecd.org/dac/financing-sustainable-development/development-finance-standards/officialdevelopmentassistancedefinitionandcoverage. htm (accessed on June 12, 2022). 3 Non-concessional loans to several middle- and high-income countries, by development banks, are not counted as part of ODA. 4 EDUCATION FINANCE WATCH 2022 Table 1. Changes in real education spending between pre- (2019) and post-pandemic onset (2020), by income group Global HIC and UMIC LMIC and LIC 2019 2020 2019 2020 2019 2020 Total education spending Change in education spending (%) 6.7 1.9 2.9 3.5 11.5 0.0 Share of countries decreasing education 27.7 50.8 22.2 58.3 34.5 41.4 spending from previous year (%) Average decrease in countries that decreased -7.9 -8.2 -8.7 -5.1 -7.2 -13.5 their spending from previous year (%) Average increase in countries that increased 12.4 12.4 6.2 15.6 21.4 9.6 their spending from previous year (%) Education’s share of the total spending Average change in percentage points from previous year 0.3 -1.0 -0.1 -0.8 0.8 -1.2 Average share (%) 15.2 14.3 13.8 13.0 17.0 15.9 Total number of countries (N) = 65 (LIC & LMIC = 29, UMIC & HIC = 36). Note: Changes are expressed in real terms (see technical note for details). Source: Own estimates using the World Bank BOOST database, as of May 11, 2022. The disruptions caused by the pandemic also resulted Lower income countries, on average, increased the utili- in a decline in the execution rates (actual utilization) of zation of their capital budget to 80 percent in 2020, com- planned education budgets in 2020.4 While the overall pared to 68 percent in 2019.5,6 education budget execution rates fell very little, there are significant differences by types of expenditures (Table 2). The COVID-19 pandemic impacts on education financ- While wage expenditures were largely executed accord- ing may amplify the gap in learning outcomes between ing to plans and similarly to 2019, the budget for goods lower and higher income countries. Low- and lower-mid- and services saw a significant fall in execution. On average, dle income countries with low education and learning across the sample, 76 percent of the budget for goods and outcomes largely registered a more marked decline in the services was executed, a 10-percentage point decline from share of education spending in total government spending the pre-pandemic execution rate of 86 percent. Under-ex- (Figure 1). The fact that lower-income countries with be- ecution on goods and services was more pronounced in low-average learning-adjusted years of schooling (LAYS)7 low and lower-middle income countries, which had an av- devote fewer of their resources to education is likely to in- erage execution rate for goods and services of 70 percent crease learning inequalities since public education spend- in 2020 compared to 79 percent in 2019. Meanwhile, ex- ing tends to benefit relatively more disadvantaged learners. ecution rates for capital expenditures increased from 79 to 86 percent from 2020 to 2019, likely reflecting shifts If not redressed, the diminished prioritization of education in expenditures to provide remote learning opportunities. could undermine human capital development, slowdown 4 The execution rate is the proportion of the originally approved budget that was spent in the fiscal year. 5 Countries potentially driving this increase include Mozambique (150 percent), Cabo Verde (120 percent), Ukraine (115 percent), El Salvador (113 per- cent) and Lesotho (112 percent) (BOOST database). 6 Research has found that lower income countries tend to over-execute remuneration budgets and under-execute “everything else” (Carvalho, Crawfurd, and Minardi 2020; Crawfurd and Pugatch 2020) and tend to underspend on infrastructure because of limited absorptive capacity (Presbitero 2016). An analysis of the impacts of the pandemic on expenditures composition and budget execution is beyond the scope of EFW 2022. 7 Median LAYS is 7 based on the sample of 91 countries in the BOOST database. The LAYS metric combines quantity (expected years of schooling) and quality (harmonized learning outcomes). The expected years of schooling measures the number of years of school a child born today can expect to obtain by age 18. It is based on age-specific enrollment rates between ages 4 and 17 and has a maximum value of 14. Meanwhile, harmonized learning outcomes are calculated using a conversion factor. For more details on the methodology, see Filmer et al. (2018); Kraay (2018); Patrinos and Angrist (2018). 5 EDUCATION FINANCE WATCH 2022 Table 2: Budget execution rates before and during COVID, overall, by type of expenditure and by country income group Global HIC and UMIC LMIC and LIC 2019 2020 2019 2020 2019 2020 Total education budget 94.7 93.4 95.5 93.4 93.9 93.4 Wages 101.5 99.5 100.4 98.1 102.7 101.0 Goods and services 86.0 76.0 92.5 81.5 79.2 70.2 Capital expenditures 78.6 85.5 88.9 91.2 67.8 79.6 N = 43 (LIC & LMIC = 21, UMIC & HIC = 22) Note: Presented here are the execution rates during pre- and post-COVID onset periods. Sample is limited to those with complete execution data on all type of expenditures. Source: BOOST database, as of May 11, 2022. economic growth, and in turn further lower public resourc- Figure 1: Lower income countries with poor learning es available for education financing. Research indicates a outcomes are committing less to education positive relationship between education quality and eco- Change in the share of education in total government spending vs learning-adjusted years of schooling (LAYS) nomic growth (Hanushek and Woessmann 2013). In turn, economic growth enables higher spending on education, 5 initiating a virtuous cycle (Al-Samarrai et al. 2019). Never- theless, as analyzed in the EFW 2021, increases in education Change in share of education in total spending 2019-2020 spending do not automatically result in improved learning outcomes; improving spending efficiency is also paramount (World Bank and UNESCO 2021). 0 The impacts and aftermath of the pandemic makes it more challenging for many countries to achieve recommend- ed international benchmarks on government education -5 spending. The commonly used benchmarks for education spending are to reach 4 to 6 percent of GDP and/or 15 to 20 percent of total government expenditure.8 Of 33 low- and lower-middle-income countries in the EFW 2022 da- -10 tabase, we can distinguish 3 groups of countries: 15 coun- tries that met both targets, 4 countries that achieved one Mediam LAYS = 7 or the other, and 14 countries that meet neither (Figure 2 2 4 6 8 LAYS, pre-COVID 10 12 and Appendix C). Examples of the first group of countries UMIC & HIC LIC & LMIC are Burundi, Cabo Verde, Honduras and Moldova, where N = 67 countries. education spending reaches around 6 percent of GDP and Source: World Bank BOOST and Harmonized Learning Outcomes 20 percent of total government expenditure. Countries in database. the latter group would need some combination of expand- ing the share of government spending in the economy and increasing the share of education in total public spending. while maintaining its current government spending rel- For instance, for Uganda to reach 4 percent of GDP, ei- ative to GDP (Figure 2). Nepal, a country close to the 4 ther government spending as a share of GDP would need percent of GDP benchmark, would need to either increase to significantly increase (from the current 19.6 percent to the share of government spending in GDP from 27.4 per- approximately 35 percent) or the country would need to cent to approximately 35 percent or increase the share of almost double their share of education spending in total education in total government expenditures from 11.4 to public spending (from 11.4 to approximately 22 percent) at least 14 percent, or some mix of the two. 8 See the recommendations of the Education 2030 Framework for Action. 6 EDUCATION FINANCE WATCH 2022 Figure 2: About 40 percent of low- and lower- Figure 3. The prioritization of education in middle-income countries spend below international government budgets recovered from its pandemic benchmarks for public education spending dip in high-income countries but not in low and Education as a share of total government expenditure, lower-middle income countries and government spending as a share of GDP in low- Share of education in total government budget, 2019–2022 and lower-middle-income countries (%), 2020 25 15.0 HND 14.5 SEN Combinations equal Low and lower middle to public ed. spend income countries BDI of 6% of GDP 14.0 20 % Education as share of budget (%) COG 13.5 TJK MOZMDA CPV STP 13.0 BTN SWZ MDG Upper middle and high 15 CMR MLI BOL income countries GIN 12.5 NER SLV ZMB TCD MWI NPL 12.0 UGA RWA 2019 2020 2021 2022 10 AFG MRT CAF Combinations equal to 4% of GDP Note: 54 countries with information for each year were used (LIC & LMIC = 28, UMIC & HIC = 26). AGO Low & lower middle Source: UIS database. income average 5 VUT 5 10 15 20 25 30 35 40 45 Govt. spending as a share of GDP (%) space in national budgets. The share of education in total N=33 LIC & LMIC. Not all countries are presented in Figure 2. See Appendix C for a full list of countries names and ISO codes in government budgets in higher income countries remained EFW2022 database. constant between 2019 and 2021 but increased in 2022. In Note: Data presented here do not include interpolated values. contrast, in low- and lower-middle income countries this Source: UIS database, accessed March 2022. share decreased in 2020, rebounded in 2021 but fell again in 2022, remaining below 2019 levels (Figure 3). With over- all total public spending being strained by increasing fis- cal pressures, there is a risk that education spending in low 1.2. Government education budgets after and lower-middle income countries is not adequate to meet the onset of COVID-19. the needs to implement urgent actions to recover learning losses and address the pre-pandemic already poor learning Unfortunately, there is not readily available data across outcomes. countries to track actual education spending after the on- set of the pandemic. In order to gauge at likely trends, the EFW 2022 extends the analysis carried out in the EFW 1.3. Aid to education during the onset of 2021 of the share of education in national public bud- COVID-19. gets (planned spending) using 2021 and 2022 data ob- tained from UIS for a sub-set of low, middle-income and Total aid to education, including imputations from aid high-income countries. devoted to overall budget support, reached US$18.1 bil- lion in 2020.9,10 In real terms, between 2019 and 2020, to- While higher income countries increased the prioritiza- tal education ODA increased by 15 percent overall, 21 tion of education in their government budgets in 2022, in percent for basic education,11 19 percent for secondary ed- low- and lower-middle income countries education lost ucation, and 8 percent for post-secondary education. 9 The most recent year with sector and recipient level data. 10 Total aid to education includes direct aid to education plus 20 percent of general budget support (aid provided to governments without being earmarked for specific projects or sectors). 11 Basic education covers primary education, basic life skills for youth and adults, and early childhood education. 7 EDUCATION FINANCE WATCH 2022 When aid derived from direct budget support flows are crises, which may have longer-term impacts on aid to ed- excluded, direct (earmarked) aid to education was stag- ucation. The COVID-19 crisis and its aftermath, the war nant in 2020 and dropped by US$359 million among bi- in Ukraine, and shifts in some governments’ priorities have lateral donors (Figure 4.b). Among the few major donors led to three major donors in education announcing severe in education to increase their direct aid to education, Ger- cuts to their overall aid budget, also negatively impacting many and the World Bank increased their aid to education aid to education. Following the announcement in No- to sub-Saharan Africa by 25 percent in 2020. Meanwhile, vember 2020 that it would reduce its ODA target from even including imputations from overall budget support, 0.7 percent to 0.5 percent of GNI, the United Kingdom 43 bilateral donors collectively decreased their aid to ed- DA fell by GBP 4.5 billion in 2021 (ICAI 2021). Nor- ucation by US$153 million from 2019 to 2020. The de- way reduced its support to education by US$19 million crease was concentrated among four donors (Saudi Arabia, in 2020 and postponed its payments to the Global Part- Turkey, United Kingdom, and United States) who collec- nership for Education until 2021 (Donor Tracker 2022). tively decreased aid by US$747 million while France, Ger- Germany has signaled a repurposing of their aid budgets many, and Japan increased their education aid by US$597 to finance spending derived from the influx of refugees million. from Ukraine, reducing ODA spent abroad by 12 percent (DW 2022). Sweden announced a 20 percent cut in its aid The increase in total aid to education was largely driven budget, but an amendment was recently rejected by Par- by imputations from aid devoted to direct overall budget liament (United Nations 2022; Riksdagen 2022). Cana- support to countries to help manage the impacts of the da and France look set to maintain aid as a priority, with pandemic. The imputations derived from direct aid trans- France committed to hit the target of allocating 0.7 per- fers to support the overall government budget as a per- cent of GNI to ODA by 2025 (Devex 2021; Donor Track- centage of total aid to education increased from a 10-year er 2022). The United States, the largest donor to educa- average of 9 percent to 19 percent in 2020. Almost the en- tion, increased total ODA by 14 percent in 2021; much tire increase in total aid to education came from the Euro- of which was emergency funding for COVID-19 vaccines pean Union (US$0.9 billion) and the IMF’s concessional (OECD 2021). trust funds (US$1.6 billion) (Figure 4.a). Education aid lost ground to other sectors in 2020. It is The decline in direct bilateral aid to education is worri- estimated that OECD DAC countries spent 7 percent some considering the potential effects of several ongoing of total aid, or US$12 billion, on interventions related to Figure 4. Increases in aid to education in 2020 were mainly driven by direct budget support Changes in aid to education, 2019–2020 a. Total aid to education, US$ million b. Direct (ear-marked) aid to education, US$ million 4,000 4,000 3,000 3,000 IMF 2,000 2,000 European Union European 2,369 Germany 1,000 Union 1,000 France Germany World Bank (IDA) Japan Japan Saudi Arabia France Other Qatar Other United States 0 0 ADB United States -38 ADB Turkey Turkey United Kingdom -1,000 -1,000 United Kingdom Decreased Increased Total Decreased Increased Total Note: Total aid to education includes direct aid to education plus an imputation of 20 percent of aid devoted to general budget support (aid provided to governments without being earmarked for specific projects or sectors). Direct aid to education is specifically earmarked to fund education spending. Source: Own estimates based on OECD CRS database. 8 EDUCATION FINANCE WATCH 2022 COVID-19 in 2020. The share of the health sector in to- Figure 5: Education was not prioritized in total aid tal ODA increased, as expected, from 16.5 percent in 2019 flows in 2020 to 19.3 percent in 2020. The share of government and civ- Education, government and civil society, and health, population policies/programs and reproductive health il society support12 in total ODA rose to 16.6 percent, as a share of total sector allocable aid, 2019–2020 whereas the share of education fell from 11.0 percent to 9.7 percent, the lowest in five years (Figure 5). 25.0 It should be noted that while overall external aid in- creased in response to COVID-19, most donor coun- 20.0 19.3 tries remain below international aid benchmarks.  To- tal ODA increased in 2020 and further in 2021, reaching 16.6 16.5 US$179 billion, up by 4.4 percent in real terms relative to 15.0 15.7 2020 according to preliminary OECD data. However, in 2021, only five of the 30 OECD Development Assistance % Committee (DAC) countries (Denmark, Germany, Lux- 11.0 10.0 embourg, Norway, and Sweden) achieved the internation- 9.7 al target of allocating 0.7 percent of gross national income (GNI) to ODA. Germany began meeting this target in 2019. Conversely, the United Kingdom decreased its aid 5.0 from 0.7 percent in 2019 to 0.5 percent in 2020. The lev- el of ODA as a share of DAC members’ GNI increased slightly, from 0.30 percent in 2019 to 0.33 percent in 2020 0.0 2019 2020 and 2021, although this maintains the long-term stag- Education Government Health nant trend since 2005. Of non-DAC members, the United Source: Own estimates based on the OECD CRS database (2022). Arab Emirates spent nearly 1 percent of GNI on ODA. 12 “Government and civil society” include public sector policy and administrative management; public finance management; decentralization and support to subnational government; procurement; domestic resource mobilization; law and justice; democratic participation and civil society etc. 9 2 Education spending - Longer term trends and patterns This section puts the observed trends in education spend- of total real spending in education, considering spending ing during the onset of COVID-19 in a longer term per- by governments (net of ODA), international education spective, and also analyzes changes in the patterns of the aid, and expenditures made by households. All country in- sources of spending. come groups increased total real spending on education over the period up to the onset of COVID-19. The relative increase was greater in low- and middle-income countries 2.1. How has total education spending than in high-income countries.13 Only low-income coun- changed over the last ten years? tries continued to increase total real education spending in 2020, reaching a level that was 65 percent higher in 2020 Total global education spending over the last 10 years be- than in 2010 (Figure 6.b). Meanwhile, the dip in total ed- fore COVID-19 increased steadily, from US$4 trillion in ucation spending in 2020 was largest among middle-in- 2010 to US$4.9 trillion in 2018, and then stagnated with come countries, falling from its peak of US$1.81 trillion the onset of the pandemic. Figure 6.a depicts the evolution in 2019 to US$1.78 trillion in 2020. Figure 6: Total real spending on education has remained constant in 2020 a. Government, household, and official development assistance b. Evolution of total real education spending (all sources), spending on education, constant 2020 US$, trillion, 2010 – 2020 by country income group 2010 - 2020 (2010 = 100) 5 4.8 4.9 4.9 4.9 180 4.5 4.6 4.3 4.4 170 LIC, $29 bil. 4.1 4.2 4.0 4 160 LMIC, $391 bil. 150 UMIC, $1.4 tri. 3 140 LIC, $17.5 bil. Trillion $ 130 LMIC, $241 bil. 2 UMIC, $928 bil. 120 HIC, $2.8 tri. 1 110 HIC, $3.08 tri. 100 0 90 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 80 Government Development Assistance Household Total 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Note: Income grouping follows the 2020 World Bank country income classification. Interpolation was done to fill missing data and ensure comparable sample of countries in all periods. Source: Own estimates using EFW 2022 database14. 13 High-income countries account for more than 60 percent of global education expenditures. 14 To avoid double-counting, government expenditure nets out actual ODA received by countries. The EFW 2022 estimates are different from EFW 2021 estimates for three reasons: (1) The difference in the methodology in estimating the imputed values for government and household spending; (2) avail- ability of more country data, supplemented by the data mapping exercise conducted by UIS, as well as the World Bank BOOST database; and (3) the downward adjustments in GDP estimates from last year’s data, especially for upper-middle- and high-income countries. Accordingly, the estimate was updated, for instance, from US$4.7 trillion to US$4.9 trillion for 2018 and 2019. 10 EDUCATION FINANCE WATCH 2022 Governments’ own resource mobilization remains the Figure 7: Governments account for less than half of main funding source for education, although households total education spending in low-income countries account for a sizable share of total spending in low and Distribution of total education spending by source, year, and country income group, percentage and billions US$ lower-middle income countries. More than three quar- ters (76 percent) of global education spending stems from 2010 7 3 7 government spending net of ODA contributions to edu- low income 2015 12 2 8 cation. Globally, households contributed a little less than 2019 14 4 10 one-quarter of global education spending in 2020, but in 2020 14 5 10 low-income countries that share was 35 percent. In com- 2010 134 6 101 parison, households in high-income countries contribut- lower-middle 2015 189 7 130 income ed 16 percent of total education spending. The continued 2019 239 8 162 increase in total education spending in low-income coun- Billion US$ 2020 227 9 155 tries was sustained by external aid to education, which in- 2010 627 4 297 creased by 15 percent from 2019 to 2020 (Figure 7). upper-middle 2015 839 3 389 income 2019 931 4 468 2020 915 4 466 2.2. How has government education 2010 2,373 0 443 spending changed over the last five to ten high income 2015 2,467 0 486 years? 2019 2,505 0 528 2020 2,578 0 503 Government education spending, as a percentage of 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% GDP, has remained largely unchanged in all country in- Government Development Assistance Household come groups since 2010 – except for low-income coun- Note: Interpolation was done to fill missing data and ensure compara- tries. Across the 10-year period, the share of government ble sample of countries in all periods. education spending in GDP has fluctuated marginal- Source: Own estimates using EFW 2022 database. ly between 4.7 percent and 4.9 percent for middle- and high-income countries, and in 2020 is on par with its level in 2010 (Figure 8.a). Meanwhile, the share of government education spending in GDP increased steadily in low-in- of per capita spending in upper-middle-income countries come countries since 2016, reaching an average of 3.6 per- (US$1,079). Government per capita spending was on av- cent in 2020, still below international benchmarks. erage nearly 150 times higher in high-income countries (US$7,787) than in low-income countries. In 2019-2020, The COVID-19 pandemic reversed a gradual upward average government per capita spending in sub-Saharan trend in per capita public spending on education in low- Africa (US$254) and South Asia (US$358) was less than and lower-middle-income countries. Per capita govern- one-tenth that of Europe and Central Asia (US$6,156) ment spending grew on average between 2010 and 2019 and less than 5 percent of North America’s (US$11,956). by 2.7 percent in low-income countries, 2.3 percent in low- Even within country income groups, per capita education er-middle-income countries, 2.9 percent in upper-mid- spending varies significantly, from US$17 in the Central dle-income countries, and by 1.1 percent in high-income African Republic to US$119 in Tajikistan, and in low- countries. In 2020, per capita public spending grew by just er-middle income countries from US$75 in Pakistan to 1 percent in low-income countries and fell by 4 percent in US$1,247 in Moldova. lower-middle countries and by 6 percent in upper-mid- dle-income countries. It should be noted that over 75 per- Zooming into the more recent evolution of public educa- cent of the world’s school-age population reside in mid- tion spending across regions reveals distinctive patterns in dle-income countries. the levels and prioritization of education spending with the onset of the pandemic (Figure 9). Despite its dip with There remain huge disparities in the levels of per capita the onset of the pandemic, average per capita government public spending across countries (Figure 8.b). In 2020, spending on education remained higher in 2019-2020 in all on average, per capita education spending in low-income regions except in Latin America and the Caribbean. Aver- countries was only 17 percent (US$52) of the spending in age per capita government education spending fell on aver- lower-middle-income countries (US$318) and 5 percent age by 3 percent in Latin America and the Caribbean, with 11 EDUCATION FINANCE WATCH 2022 Figure 8: In low-income countries, public spending has grown as a share of GDP but remains very low in per capita terms a. Government education spending as percentage b. Government education spending per capita (constant of GDP by income group, 2010-2020 2020 US dollars) by income group, 2010-2020 5.5 7,787 HIC LMIC, 4.9% UMIC, 4.9% HIC, 4.8% 6,878 LMIC, 4.7% 4.5 UMIC, 4.8% HIC, 4.8% LIC, 3.6% 1,079 UMIC 3.5 861 LIC, 3.0% 318 LMIC 263 2.5 40 53 LIC 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 Note: Estimates on spending as percentage of GDP include interpolated Note: To calculate, total public education spending is divided by total values. Interpolation was done to fill missing data and ensure comparable pre-primary to tertiary school-age population. Estimates (interpolated val- sample of countries in all periods. ues) are used for missing country level data. Source: Own estimates using EFW database. Source: Own estimates using EFW database. Figure 9: Government education spending by regions, 2014-15 vs 2019-2020 a. Changes in education spending as percentage of total b. Changes in government education spending per government expenditure, by region (2014–15 vs 2019–20) capita by region (2014–15 vs 2019–20) Sub-Saharan Africa 15.2 15.5 Europe & Central Asia 5,931 6,156 Latin America & Caribbean 15.4 16.3 East Asia & Pacific 2,658 3,183 East Asia & Pacific 15.1 16.4 Middle East & North Africa 3,053 3,120 Middle East & North Africa 13.4 14.2 Latin America & Caribbean 1,772 1,819 South Asia 12.5 14.5 266 South Asia 358 Europe & Central Asia 12.2 12.3 239 Sub-Saharan Africa 254 10 11 12 13 14 15 16 17 0 1,000 2,000 3,000 4,000 5,000 6,000 7,000 Share of education in total government expenditure (%) Per capita education spending, 2020 US$ Note: Estimates are used in place of missing country-level data. Composi- Note: Estimates are used in place of missing country-level data. Composi- tion of countries is the same in the two time periods (2014-2015 and 2019- tion of countries is the same in the two time periods (2014-2015 and 2019- 2020). The blue arrows show a rise, while the orange arrows show a fall. 2020). Per-capita figures use total pre-primary to tertiary school-age pop- Source: Own estimates using EFW 2022 database. ulation. The blue arrows show a rise, while the orange arrows show a fall. Source: Own estimate using EFW 2022 database. 12 EDUCATION FINANCE WATCH 2022 significant variation across countries.15 East Asia and the Pa- Figure 10: The share of earmarked aid to education cific registered the largest absolute increase in the level of per has declined over the last 10 years capita government spending on education (US$525) from Education, government and civil society, and health, population policies/programs and reproductive health 2014-2015 to 2019-2020. This increase is driven by high-in- as a share of total sector allocable aid, 2010–2020 come countries in the region, with half of lower-middle-in- come countries and 60 percent of upper-middle-income 20 19.3 countries registered a reduction in their per capita education 18.3 Health spending in this period. In relative terms, South Asia showed the largest regional increase in per capita government spend- 16.6 16.3 ing by 34 percent between 2014-2015 and 2019-2020. 15 Meanwhile, the share of education spending in total gov- Government ernment expenditure fell in most regions with the onset of the pandemic. The exceptions were sub-Saharan Africa and the Middle East and North Africa, where this share % 11.7 remained marginally higher (by 0.2 and 0.8 percentage points, respectively) in 2019-2020 relative to 2014-2015. 10 Education 9.7 The prioritization of education in total government spend- ing as well varies significantly within regions. In sub-Sa- haran Africa and East Asia and the Pacific, the share of education spending in total government spending ranges from 4.3 percent in Somalia to 34.6 percent in Sierra Le- one and from 5.3 percent in Vanuatu to 31 percent in the 5 2010 2012 2014 2016 2018 2020 Solomon Islands (See Appendix A). Note: Calculated using sector specific allocation only. It does not in- clude imputed aid amount from general budget support to ensure com- parability across three sectors presented in the figure. 2.3. How has earmarked aid to education Source: Own estimates based on the OECD CRS database (2022). changed over the last ten years ? The share of earmarked aid to education has declined over the past 10 years. The share of education in total aid (not 2018 and 2020, low-income countries received US$1.6 including aid devoted to overall budget support) fell from billion per year in aid to basic education while lower-mid- 11.7 percent in 2010 to 9.7 percent in 2020. This is in con- dle income countries received US$2.6 billion per year. In trast to aid to health and other government and civil soci- effect, this amounted to US$16 per primary school–age ety spending16, which have increased their share. Aid to ed- child in low-income countries, and US$7 in lower-middle ucation is approximately half of aid allocated to these two income countries. sectors. While the health sector ODA increased its share in total aid significantly in 2020, as a natural response to COVID-19, education’s share dipped (Figure 10). 2.4. How important is household education spending across income groups? The share of aid allocated to each level of education has remained relatively stable over the past 10 years. Aid to The importance of household spending on education is basic and post-secondary education account for around 40 underappreciated. Out-of-pocket spending on education percent of total aid to education each, and secondary edu- is largely a result of low government spending, which forc- cation for 20 percent (Figure 11). es parents to pay for items that otherwise would be avail- able free of cost, but also varies with household wealth, am- As a result, the effective level of aid going to basic educa- bition, or peer pressure to ensure that their children have tion remains very low. Taking a 3-year average between access to education of the highest possible quality. The 15 Country figures are presented in Appendixes A and B. 16 “Government and civil society” include public sector policy and administrative management; public finance management; decentralization and support to subnational government; procurement; domestic resource mobilization; law and justice; democratic participation and civil society etc. 13 EDUCATION FINANCE WATCH 2022 Figure 11: The share of aid by level of education has remained relatively stable over the past 10 years Aid to education, by level, 2002-2020 18.1 18 Post-secondary Secondary 16 Basic 15.6 15.7 14.5 14.4 7.0 14 13.5 13.2 12.9 13.0 12.8 Constant US$ 2020 billion 12.2 12.3 6.1 12 6.5 11.4 11.2 5.3 5.5 10.4 5.2 5.3 10 5.2 5.0 5.0 9.3 4.9 4.7 8.4 4.8 4.6 3.8 8.1 8 4.6 3.0 2.8 2.8 3.2 4.3 2.3 6.0 2.4 2.1 2.6 2.4 6 3.8 2.2 2.6 4.0 1.8 1.9 2.4 1.5 4 1.2 1.2 1.0 1.0 2 2.7 3.0 3.3 3.9 4.2 4.8 4.8 5.6 5.9 5.6 5.1 5.3 5.1 5.5 6.5 6.1 6.5 6.0 7.3 0 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Source: Own estimates based on the OECD CRS database (2022). level of household spending on education may be related including Bosnia and Herzegovina, North Macedo- not only to the level and equity of government spending, nia, and Romania, and in several sub-Saharan African but also to the overall redistributive effect of taxation. The countries, including Burundi, Ethiopia, and Lesotho, to higher the share of households in total education spend- more than 6 percent in countries with a high percentage ing, the larger the risk of inequality in learning. of private schools, such as Haiti and Lebanon, and in other sub-Saharan African countries, including Rwan- A considerable portion of household income in low- and da, Uganda, and Zambia. middle-income countries is devoted to spending on edu- cation. Analysis of reports from approximately 100 low- Richer households tend to spend a higher proportion of and middle-income countries between 2009 and 2020 their income to education. Poorer households are more found that, on average, households allocated 3.2 percent likely to spend more of their income on food and other of their total expenditure to education. Evidence suggests necessities and are less likely to be able to afford to pay that this share remains relatively stable over time. For in- for education services. In 30 of 33 countries with data, stance, the share of total household spending on education households from the richest quintile spent a higher share was 3.8 percent in 2007 and 4.1 percent in 2017 in Peru, of their total consumption on education than those from 2 percent in 2010 and 2.3 percent in 2016 in Turkey, and the poorest quintile. The average shares were 2.4 percent 6.3 percent in 2002 and 6.1 percent in 2018 in Vietnam. for the poorest households and 4.2 percent for the richest However, there are a few cases where the share has been (Figure 12). rising, such as in some sub-Saharan African countries. In Uganda, the share increased from 5 percent in 2012 to 7.8 Households account for a major share of total education percent in 2016. In Ghana, three successive rounds of the spending in many countries. Household education ex- Living Standards Survey show that the share of education penditures as a share of GDP can be estimated by multi- spending was not only the world’s largest, but also had in- plying the share of education in household consumption creased from 8.9 percent in 2006 to 10.6 percent in 2013 expenditure with household consumption expenditure and 13.1 percent in 2017 (Ghana Statistical Service 2008; as a share of GDP, as published in the World Develop- 2016; 2019). ment Indicators. The latter varies by country because of considerable differences in economic, social, and political The share of education in total household spend- conditions. Globally, household consumption as a share ing varies across countries. It ranges from less than of GDP was 59 percent in 2020, ranging from less than 1 percent in some southeastern European countries, 40 percent in rich oil-producing countries, including 14 EDUCATION FINANCE WATCH 2022 Figure 12: Richer households spent a larger share of their budget on education than poorer households Share of education in total household spending, by income quintile, 2010s 16 Share of education in household spending (%) Poorest 20% Richest 20% 14 12 10 8 6 4 2 0 North Macedonia, 2017 Ethiopia, 2015 Mozambique, 2014 Micronesia, 2013 Russian Fed., 2019 Albania, 2019 Turkey, 2016 South Africa, 2011 Tuvalu, 2016 Bolivia, 2016 Burundi, 2013 Ukraine, 2019 Guinea, 2018 Ecuador, 2011 Kyrgyz Republic, 2016 Cabo Verde, 2015 Costa Rica, 2018 Azerbaijan, 2019 Malawi, 2010 Argentina, 2017 Nepal, 2015 Tunisia, 2015 Algeria, 2011 Namibia, 2015 Gambia, 2015 Bhutan, 2017 Rwanda, 2011 Nicaragua, 2016 Vietnam, 2018 Dominican Rep., 2018 Haiti, 2012 Ghana, 2016 Liberia, 2016 Source: Own estimates based on national household budget survey reports. Oman, Saudi Arabia, and the United Arab Emirates, to the UIS and OECD, generating a data set of almost 140 close to or even exceeding 100 percent in low-income countries. On average, households spend 1.9 percent of countries that rely on remittances, such as Haiti, Liberia, GDP on education. Given that governments in this sam- and Somalia (World Bank 2021). This indirect estimate ple of countries spend on average 4.5 percent of GDP of household education expenditure as a share of GDP on education (Figure 13.a), households account for 30 can be complemented with direct estimates provided by percent of total education spending.17 The share ranges Figure 13: Households account for almost 40 percent of total education spending in low- and lower-middle-income countries a: Education expenditure as share of GDP, by b: Share of education expenditure, by source, source, region, and income group, 2010s region, and income group, 2010s 8 100 7 80 6 5 60 GDP (%) 4 % 3 40 2 20 1 0 0 Low Lower middle Upper middle High Europe and Central Asia South Asia East Asia and Pacific Sub-Saharan Africa Middle East and North Africa North America World Low Lower middle South Asia Latin America and the Caribbean Upper middle High Europe and Central Asia East Asia and Pacific Sub-Saharan Africa Latin America and the Caribbean Middle East and North Africa North America World Household Public Source: Own estimates based on national household budget survey reports and UIS and OECD data. 17 This is an unweighted average and covers only the countries with available information. 15 EDUCATION FINANCE WATCH 2022 Figure 14: There is large variation in the share of households in total education spending among countries Share of education expenditure, by source and income group, 2010s 100 75 50 % 25 0 Liberia Uganda Niger Rwanda Madagascar D. R. Congo Gambia Benin Sudan Mali Togo Guinea U. R. Tanzania Chad Sierra Leone Cameroon Burundi Mozambique Kenya Ethiopia Haiti Nigeria Bangladesh Ghana Mauritania Myanmar Pakistan El Salvador Lao PDR Nicaragua Honduras Sri Lanka Palestine Comoros Egypt Nepal Zambia Senegal Indonesia Mongolia Cambodia Zimbabwe India Cote d'Ivoire Morocco Vietnam Samoa Cabo Verde Lesotho Angola Djibouti Vanuatu Timor-Leste Kyrgyz Republic Congo Ukraine Bhutan Algeria S. Tome/Principe Philippines Tunisia Low income Lower middle income 100 75 50 % 25 0 Russian Fed. U. A. Emirates Greece Rep. of Korea Barbados Czechia Croatia Slovenia Denmark Iceland Sweden Costa Rica Tonga Botswana Mexico Kazakhstan Rep. Moldova Romania Uruguay Chile Bahrain United Kingdom Australia United States Oman Canada Saudi Arabia New Zealand Macao, China Israel Netherlands Trinidad/Tobago Ireland Slovakia France Malta Estonia Austria Switzerland Poland Hungary Germany Lithuania Belgium Luxembourg Norway Finland Lebanon Jordan Dominican Rep. Fiji Jamaica Maldives Colombia Georgia Paraguay Brazil Azerbaijan Argentina Turkey South Africa Serbia Thailand Bulgaria Malaysia Namibia Belarus Cyprus Spain Portugal Italy Latvia Mauritius Armenia Albania Peru Panama Ecuador Upper middle income High income Private Public Source: Own estimates based on national household budget survey reports and UIS and OECD data. from 15 percent in high-income to 39 percent in low- spending. For instance, in 39 percent of low-income and lower-middle-income countries and from 14 per- and 26 percent of lower-middle-income countries but cent in Europe and Central Asia to 42 percent in South just 6 percent of upper-middle-income and 2 percent Asia, with North America (23 percent), East Asia and of high-income countries, households account for more the Pacific (31 percent), the Middle East and North Af- than half of total education spending. There is also signif- rica (36 percent), sub-Saharan Africa (37 percent), and icant variation within each income group. For instance, Latin America and the Caribbean (38 percent) falling in among low-income countries, households account for 5 between (Figure 13.b). percent of total education spending in Ethiopia and 10 percent in Mozambique but 59 percent in Uganda and 73 There is significant variation between countries in the percent in Liberia. Among lower-middle-income coun- share of household education spending in total education tries, the shares are 5 percent in Lesotho and Sao Tome 16 EDUCATION FINANCE WATCH 2022 and Principe but 71 percent in Bangladesh and 72 per- children attend public schools, including Ghana, Kenya, cent in Nigeria. Among upper-middle-income coun- and Zambia. tries, households account for 7 percent of total education spending in Romania and 9 percent in the Russian Fed- High education spending levels mean that households eration, 55 percent in Jordan, and 74 percent in Lebanon must save and borrow for education. About one in six (Figure 14). families in low- and middle-income countries saves to pay school fees. Analysis of the 2014 Global Findex Database Overall, most spending comes from households whose has found that about 8 percent of households also bor- children attend private schools. In countries in South row, with shares ranging from 6 percent in high-income Asia such as India and Pakistan and in countries in Lat- countries to 12 percent in low-income countries. Similar in America and the Caribbean such as Colombia and El proportions of the poorest 40 percent and richest 60 per- Salvador, most spending on primary and secondary ed- cent of households borrow to pay fees in low- and low- ucation in absolute terms comes from families whose er-middle-income countries, although slightly more of the children attend private schools – and goes mainly to pay richest manage to save. In Haiti, Kenya, the Philippines, for fees. But there are some cases in sub-Saharan Africa and Uganda, 30 percent or more of households borrow for where household spending takes places in families whose school fees (UNESCO 2021). 17 3 Data Spotlight: Monitoring education spending The availability of education finance data has improved. available data on public education spending for 2021. Data When the first EFW was published in 2021, data on the availability is better for the indicator on education spend- 2019 share of education in total government spending was ing as a share of total government spending, but more than available for just 16 percent of countries. A year later, this one-fourth of countries have not reported 2020 data (Fig- indicator is reported in the UIS database for 64 percent of ure 15.b). Most countries do not report data by level of countries (with data for an additional 9 percent of coun- education, and national reporting practices vary. Only 1 tries available from other sources). percent of countries and territories (3/218 countries and territories) reported education spending by level of educa- Nevertheless, significant gaps in education spending data tion in 2020 (Figure 15.c). remain. As of April 2022, nearly 20 percent of countries had not reported 2016 data on education spending as a The sources for education spending data are expand- share of GDP, and more than one-third had not report- ing. The EFW 2022 benefitted from additional sources ed 2020 data (Figure 15.a). Very few countries have made of data not included in the EFW 2021. Historically, the Figure 15: Despite improved overall data availability, the gap in disaggregated data is still severe Percentage of countries reporting on public expenditure indicators, by year a. As a share of GDP b. As a share of government spending c. By level of education 100% 100% 100% 19% 17% 17% 16% 19% 90% 20% 90% 22% 90% 24% 27% 5% 38% 4% 80% 5% 80% 7% 5% 80% 6% 9% 10% 55% 70% 70% 9% 70% 59% 63% 60% 60% 60% 81% 11% 91% 50% 50% 50% 40% 78% 40% 80% 40% 76% 74% 77% 77% 70% 21% 14% 66% 64% 30% 30% 30% 11% 51% 20% 20% 20% 27% 26% 10% 10% 10% 10% 25% 9% 8% 0% 0% 0% 1% 2016 2017 2018 2019 2020 2016 2017 2018 2019 2020 2016 2017 2018 2019 2020 UIS Others Not reporting Note: The ‘Others’ data represent all other data points not included in the UIS dataset but nonetheless reported in other data sources (WB BOOST, WB WDI, WB PER, and IMF GFS). Source: Own estimates using EFW database. 18 EDUCATION FINANCE WATCH 2022 Figure 16: Data availability improved significantly a. 2019 data in EFW 2021 vs EFW 2022 b. 2019 data in EFW 2021 vs 2020 data in EFW 2022 100% 100% 24% 27% 80% 80% 38% 60% 80% 60% 80% 83% 92% 91% 40% 40% 76% 73% 62% 20% 20% 20% 20% 17% 8% 9% 0% 0% As of Apr 2021 As of Apr 2022 2019 data in 2020 data in 2019 data in 2020 data in 2019 data in 2020 data in EFW2021 EFW2022 EFW2021 EFW2022 EFW2021 EFW2022 Education as % of GDP Education as % of GDP 2019 Share of Govt spending Ed. Spend by level of education With data Not reporting Source: Own estimates using EFW 2021 and EFW 2022 database. UIS has reported on total government spending on ed- However, disaggregated data by level remains a challenge ucation (as a percentage of GDP or of total government with just 9 percent of countries reporting in 2020 (Fig- spending) from Questionnaire B of the Education For- ure 16.b). mal Survey and the UNESCO OECD Eurostat finance questionnaire. In 2021, the UIS expanded the inclusion The EFW 2022 has adopted a different methodology of of alternative official sources to report. Data from inter- imputing missing values than the EFW 2021. The up- national institutions, such as the World Bank’s BOOST dated methodology, applied in Section 1, did not use initiative and Public Expenditure Reviews and the IMF’s statistical modeling to impute missing values. Instead, Government Finance Statistics, as well as national budget the EFW 2022 used available information from various and expenditure estimates published in government eco- sources, in particular trends and growth rates, to inter- nomic reports have been integrated in the database fol- polate or extrapolate missing information. Specifically, lowing a process of standardization to establish compara- the data that the UIS shared—which is already a com- bility. Integrating various data sources has helped develop bination of various data sources—served as the base or more complete and up-to-date series, both for recent and primary series, and where missing, interpolation or ex- earlier years.18 By contrast, data for the EFW 2021 (as of trapolation was done using information from IMF or April 2021) had included data that UIS collected using its BOOST, whichever was closer in terms of average val- own survey only. As such, with respect to 2019 data, cov- ues. With the new methodology, the number of coun- erage increased from 20 percent to 76 percent for data on tries covered in the analysis has increased by 21 percent the share of education expenditure in GDP (Figure 16.a). over the previous methodology, making the analysis The improvement in reporting of recent-year data (2019 more robust.19 An accompanying technical note to this data in the EFW 2021, 2020 data in the EFW 2022) is report provides a detailed presentation of the method- pronounced for education spending as both a percentage ology, as well as a comparison between the old and the of GDP and a percentage of total government spending. new database. 18 Global efforts on strengthening the monitoring mechanism have continued. As part of the Technical Cooperation Group on Sustainable Development Goal 4 Indicators, the Finance Working Group conducted a mapping exercise regarding education expenditure indicators, which resulted in the higher reporting rates. 19 The comparison was made using the period of 2009 to 2019, which was the period covered in the EFW 2021. 19 EDUCATION FINANCE WATCH 2022 Household income and expenditure surveys are un- An alternative is to rely on governments’ own reporting. derused as sources of household education spending. Although household budget survey data sets may be dif- Almost every country in the world conducts household ficult to access and process, household budget survey re- surveys, but their potential for analyzing education expen- ports are publicly available. In most cases, these reports diture has not been used, and these datasets vary in terms include a summary table of how households allocate ex- of accessibility and quality. There can be considerable vari- penditures to various purposes. Almost all countries sepa- ations in the questions, as well as the coverage in relation rate out the share of education in household consumption to education expenditures. For instance, some countries expenditures. Despite differences in formulations, averag- collect aggregated education expenditures by household, ing over thousands of households within each country and whereas others collect information on individual members. over many countries can help cancel out some of these po- The list of items included, how they are grouped, and recall tential biases and errors. Nevertheless, this offers a tem- periods also differ (Oseni et al. 2020). Significant resources porary solution to a long-standing problem. A concerted are required to standardize them so that comparable data effort must be made to directly analyze household budget on issues such as education expenditures can be extracted. survey micro data. 20 EDUCATION FINANCE WATCH 2022 Appendixes  Appendix A. Education spending as Appendix B. Per capita public education percentage of total government expenditures, spending, changes from 2014-2015 to changes from 2014-2015 to 2019-2020  2019-2020 Solomon Islands Moldova Nicaragua Ukraine Uzbekistan Morocco Honduras Solomon Islands Senegal Moldova Mongolia Congo, Rep. Bhutan Micronesia, Fed. Sts. Kiribati Bhutan Cabo Verde Kenya Eswatini West Bank and Gaza Vietnam Lower-middle-income countries Lower-middle-income countries Eswatini São Tomé and Principe El Salvador India Honduras Mongolia Philippines Kyrgyz Republic Uzbekistan Kiribati São Tomé and Principe Côte d'Ivoire India Cabo Verde Philippines Nicaragua Cameroon Kenya Vietnam Lesotho Bolivia Kyrgyz Republic Zambia Côte d'Ivoire Lesotho Congo, Rep. Ukraine El Salvador Senegal Pakistan Vanuatu Nepal Cameroon Myanmar Zambia Bangladesh Myanmar Sri Lanka Benin Angola Egypt, Arab Rep. Tanzania Nigeria Angola Vanuatu Pakistan 0 5 10 15 20 25 30 35 $0. $200. $400. $600. $800. $1,000. $1,200. $1,400. Share of education in total government expenditures (%) Spending per capita, 2020 US$ Sierra Leone Tajikistan Burundi Burkina Faso Mozambique Sierra Leone Togo Tajikistan Mali Madagascar Rwanda Low-income countries Low-income countries Mali Mozambique Guinea Malawi Guinea-Bissau Niger Gambia, The Malawi Guinea Chad Niger Uganda Afghanistan Rwanda Liberia Chad Afghanistan Madagascar Central African Republic Burundi Somalia Central African Republic 0 5 10 15 20 25 30 35 40 $0 $20 $40 $60 $80 $100 $120 $140 Share of education in total government expenditures (%) Spending per capita, 2020 US$ Source: UIS (education spending as percentage of total government ex- Source: UIS (education spending as percentage of GDP), and WDI penditures), Unit: US$ (2020=100) and IMF (GDP data in US$, rebased to 2020)  21 EDUCATION FINANCE WATCH 2022 Appendix C. Status of selected low-income and lower-middle-income countries in achieving international benchmark targets on education spending 2020, % Country name  Country code Income group % TGE  % GDP  Burundi BDI LIC 20.7 5.0 Central African Republic CAF LIC 9.8 2.2 Chad TCD LIC 11.7 2.9 Guinea GIN LIC 14.3 2.2 Madagascar MDG LIC 15.3 3.1 Malawi MWI LIC 11.5 2.9 Mali MLI LIC 14.5 3.8 Mozambique MOZ LIC 17.9 6.3 Niger NER LIC 13.3 3.8 Rwanda RWA LIC 10.8 3.3 Sierra Leone SLE LIC 34.2 8.8 Uganda UGA LIC 11.2 2.6 Angola AGO LMIC 6.5 2.4 Cabo Verde CPV LMIC 17.1 7.6 Cameroon CMR LMIC 14.4 3.2 Congo, Rep. COG LMIC 18.3 4.5 Côte d'Ivoire CIV LMIC 15.1 3.4 Eswatini SWZ LMIC 15.9 5.3 Honduras HND LMIC 24.3 6.4 India IND LMIC 16.5 4.5 Kenya KEN LMIC 17.9 5.1 Lesotho LSO LMIC 13.8 7.7 Mauritania MRT LMIC 9.7 1.9 Moldova MDA LMIC 18.0 6.4 Morocco MAR LMIC 14.8 6.8 Nicaragua NIC LMIC 22.8 4.6 Philippines PHL LMIC 14.2 3.9 São Tomé and Principe STP LMIC 16.1 5.0 Senegal SEN LMIC 22.1 5.5 Solomon Islands SLB LMIC 31.9 12.8 Uzbekistan UZB LMIC 20.5 4.9 Vanuatu VUT LMIC 5.1 2.2 Zambia ZMB LMIC 12.4 3.7 Note: Data presented here do not include interpolated values. Low- and lower-middle-income countries with information on both indicators are included in this list. %TGE: education spending as a share of total government expenditure, %GDP: education spending as a percentage of GDP, LIC: low-income country, LMIC: lower-middle-income country. Source: UIS database, accessed March 2022. 22 References Al-Samarrai, Samer et al. 2019. “Mobilizing Resources for Edu- Independent Commission for Aid Impact (ICAI). 2021. Man- cation and Improving Spending Effectiveness: Establishing agement of the 0.7% ODA spending target in 2020. A rapid re- Realistic Benchmarks Based on Past Trends.” Policy Research view. ICAI. London. Working Paper 8773. World Bank Group, Washington, DC. Kraay, Aart C. 2018. “Methodology for a World Bank Human Azevedo, João Pedro, Maryam Akmal, Marie-Helene Cloutier, Capital Index.” World Bank, Washington, DC. and Yi Ning Wong. 2022. “Covid-19 Learning Loss Sim- ulations: Global Update.” Policy Research Working Paper. OECD. 2022. ODA Levels in 2021 – Preliminary Data Detailed World Bank, Washington, DC. Summary Note. April 12, 2022. Carvalho, S, L. Crawfurd, and L. Minardi. 2020. “What Hap- Oseni, Gbemisola et al. 2020. “Measuring Household Expenditure pens to Teacher Labor Markets Following Shocks.” Cen- on Education: A Guidebook for Designing Household Survey ter for Global Development. https://www.cgdev.org/blog/ Questionnaires.” World Bank Group, Washington, DC. what-happens-teacher-labor-markets-following-shocks. Patrinos, Harry Anthony, and Noam Angrist. 2018. “Glob- Crawfurd, Lee, and Pugatch, Todd. 2020. “Teacher Labor Mar- al Dataset on Education Quality: A Review and Update kets in Developing Countries,” IZA Discussion Papers (2000–2017).” World Bank, Washington, DC. 12985. Institute of Labor Economics, Bonn, Germany. Presbitero, Andrea F. 2016. “Too Much and Too Fast? Public Devex. 2021. France striving to hit 0.7% aid target by 2025. Investment Scaling-up and Absorptive Capacity.” Journal of February 26, 2021. https://www.devex.com/news/ Development Economics 120: 17–31. france-striving-to-hit-0-7-aid-target-by-2025-99198 Riksdaden. 2022. No to the proposal from the Committee on Donor Tracker. 2022a. Norway – Education. https://donortracker. Finance regarding the spring amending budget and an extra org/norway/education#:~:text=Norway’s%20budget%20for%20 amending budget (FiU21). June 17, 2022. https://www.riks- education%2C%20research,of%20the%20COVID-19%20crisis. dagen.se/en/news/2022/jun/17/no-to-the-proposal-from- the-committee-on-finance-regarding-the-spring-amend- Donor Tracker. 2022b. Canada. https://donortracker.org/ ing-budget-and-an-extra-amending-budget-fiu21/ country/canada UN. 2022. Swedish UNA: Budget cuts hurt the world’s DW. 2022. Germany looks set to cut development aid. March 25. poorest people. April 20, 2022. https://unric.org/en/ 2022. https://p.dw.com/p/492XT. swedish-una-budget-cuts-hurt-the-worlds-poorest-people/ Filmer, Deon et al. 2018. “Learning-Adjusted Years of School- UNESCO. 2021. Global Education Monitoring Report 2021/2: Non- ing (LAYS): Defining a New Macro Measure of Education.” State Actors in Education: Who Chooses? Who Loses? Paris: UNES- World Bank, Washington, DC. CO. https://www.unesco.org/gem-report/en/non-state-actors Ghana Statistical Service. 2008. Ghana Living Standards Survey World Bank. 2021. “World Development Indicator 2021.” Round 5 (GLSS 5). World Bank, Washington, DC. . 2016. Ghana Living Standards Survey Round 6 (GLSS 6). World Bank and UNESCO. 2021. “Education Finance Watch 2021.” World Bank/Global Education Monitoring Report, . 2019. Ghana Living Standards Survey Round 7 (GLSS 7). Washington DC/Paris. Hanushek, Eric A., and Ludger Woessmann. 2015. The Knowl- World Bank, UNESCO, and UNICEF. 2021. “The State of the edge Capital of Nations: Education and the Economics of Global Education Crisis: A Path to Recovery.” World Bank, Growth. Cambridge, MA: MIT Press. UNESCO, and UNICEF, Washington DC, Paris, New York. 23 Acknowledgements The EFW 2022 was prepared by Nobuyuki Tanaka, Thomas Poulsen, Pedro Cerdan-Infantes, Marianne Joy Anacleto Vi- tal (World Bank) Manos Antoninis, Yuki Murakami and Marie Renée Andreescu (Global Education Monitoring Report). It was prepared under the guidance of Omar Arias and Jaime Saavedra and benefitted from data access to BOOST dataset from Massimo Mastruzzi (World Bank) and from UIS data access and comments from Silvia Montoya (UIS), and com- ments from Anna Olefir, Kebede Feda, Lars M. Sondergaard, Samer Al-Samarrai, Timothy Stephen Williamson, and Inga Afanasieva. The team was also supported by Juanita Bodmer. Administrative support was provided by Restituto Jr. Mijares Cardenas. © 2022 International Bank for Reconstruction and Development / The World Bank and UNESCO Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank or UNESCO and should not be This work is a co-publication of The World Bank and UNESCO. considered an official World Bank or UNESCO translation. The World Bank and UNESCO shall not be liable for any content or error in this translation. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent, or those of Adaptations—If you create an adaptation of this work, please add the following disclaimer along UNESCO. with the attribution: This is an adaptation of an original work by The World Bank and UNESCO. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the The World Bank and UNESCO do not guarantee the accuracy, completeness, or currency of the data adaptation and are not endorsed by The World Bank and UNESCO. included in this work and do not assume responsibility for any errors, omissions, or discrepancies in the information, or liability with respect to the use of or failure to use the information, methods, Third-party content—The World Bank and UNESCO do not necessarily own each component of the processes, or conclusions set forth. The boundaries, colors, denominations, and other information content contained within the work. The World Bank and UNESCO therefore do not warrant that the use shown on any map in this work do not imply any judgment on the part of The World Bank and of any third-party-owned individual component or part contained in the work will not infringe on the UNESCO concerning the legal status of any territory or the endorsement or acceptance of such rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If boundaries. you wish to re-use a component of the work, it is your responsibility to determine whether permission is needed for that re-use and to obtain permission from the copyright owner. Examples of components can Nothing herein shall constitute or be construed or considered to be a limitation upon or waiver of the include, but are not limited to, tables, figures, or images. privileges and immunities of The World Bank, all of which are specifically reserved. All queries on rights and licenses, including subsidiary rights, should be addressed to World Bank Rights and Permissions Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2625; e-mail: pubrights@worldbank.org. Designed by alejandro espinosa/sonideas.com. This publication is available in Open Access under the Creative Commons Attribution-ShareAlike Photography credits— 1: © Talukdar David/Shutterstock. Used with the permission of Talukdar David/ 3.0 IGO (CC-BY-SA 3.0 IGO) license (http://creativecommons.org/licenses/by-sa/3.0/igo/). Under Shutterstock. Further permission required for reuse; 2: © JLwarehouse/Shutterstock. Used with the this license, you are free to copy, distribute, transmit, and adapt this work, including for commercial permission of JLwarehouse/Shutterstock. Further permission required for reuse; 3: © Richard Juilliart/ purposes, under the following conditions: Shutterstock. Used with the permission of Richard Juilliart/Shutterstock. Further permission required Attribution—Please cite this publication as follows: The World Bank and UNESCO (2022). Education for reuse; 4: © Bill Lyons/World Bank. Further permission required for reuse; 10: © Manoej Paateel/ Finance Watch 2022. Washington D.C., Paris: The World Bank and UNESCO. Shutterstock. Used with the permission of Manoej Paateel/Shutterstock. Further permission required for reuse; 17: © Emmage/Shutterstock. Used with the permission of Emmage/Shutterstock. Further Share Alike — If you remix, transform, or build upon the material, you must distribute your contributions permission required for reuse; 18: © cfalvarez/Shutterstock. Used with the permission of cfalvarez/ under the same license as the original. Shutterstock. Further permission required for reuse; 20: © Talukdar David/Shutterstock. Used with the permission of Talukdar David/Shutterstock. Further permission required for reuse.