Global development assistance for early childhood care and education in 134 low- and middle-income countries, 2007–2021

Data sources

We extracted global development-assistance data from OECD’s Creditor Reporting System (CRS),24 where development assistance refers to aid intended to support the socioeconomic development and welfare of LMICs.25 The CRS database is a publicly accessible source that records information on aid projects. These projects are mandatorily reported by all OECD Development Assistance Committee (DAC) member countries and voluntarily reported by non-DAC states, multilateral institutions and private entities.26 27 We downloaded the 2007–2020 CRS data in May 2022 and the 2021 CRS data in January 2023. The CRS data included information on all aid projects implemented in 155 recipient countries or territories. We excluded 21 recipients classified as high-income or without income classifications according to the World Bank’s 2020 standard.28 This resulted in 134 LMIC recipients being included in our study (online supplemental table S1).

Donors disbursed aid to the 134 LMICs across various years. Online supplemental table S2a–d presents a breakdown of donors and the corresponding years in which each donor reported aid during our study period. We estimated ECCE aid using each year’s available aid data from all reporting donors in that year. Following previous studies and OECD’s recommendations, we excluded CRS data before 2007 because of the high missing data rate; for 2007 onwards, the completeness of the data reached nearly 100%.26 29–31 To assess ECCE aid per ECCE-age child, we calculated the total number of ECCE-age children for each country using data from UNESCO Institute for Statistics.32 This calculation involved aggregating each country’s eligible population for early childhood educational development programmes (typically children aged 0–2 years) and preprimary educational programmes (typically children aged 3 years to the primary school entry age).33 For cases with missing data in these eligible populations, we applied imputations as detailed in online supplemental text S1.

Defining and estimating aid for early childhood care and education

Defining and estimating ECCE aid to projects with COVID-19 activities

We paid special attention to aid for ECCE programmes with COVID-19 activities. The CRS database used a purpose ‘COVID-19 control’ to track projects with activities related to COVID-19 immunisation, testing, prevention, treatment and postrecovery therapies and a hashtag variable ‘#COVID-19/COVID-19’ to track projects aimed at mitigating socioeconomic impacts of the pandemic.27 39 However, our project review indicated that projects listed under the ‘COVID-19 control’ purpose or projects marked with the hashtag may not necessarily specify COVID-19 activities in their descriptions. Therefore, we compiled a list of COVID-19 keywords and searched for their presence in titles and descriptions among all the ECCE projects we identified for 2020 and 2021 (see keywords in online supplemental table S4c). We manually reviewed each project that fulfilled the search requirements and eliminated false positives. We compared the total ECCE aid in 2020 and 2021 with aid disbursements to ECCE projects with COVID-19 activities.

Defining and estimating ECCE aid to projects in conflict-affected countries

There is extensive evidence demonstrating the pervasive negative effect of armed conflicts on children’s development.40 41 This adverse impact on children is partially mediated by the lack of access to ECCE.13 We estimated ECCE aid invested in conflict-affected countries throughout the study period. Consistent with previous studies,20 we defined conflict-affected countries as those who were either in conflicts or in postconflict stages in the period between 2007 and 2021 and defined ‘conflict’ as major armed battles with at least 1000 battle-related deaths in 1 year.

We identified a country as being in its postconflict stage if its conflict occurred between 2004 and 2006 but ended before 2007. We used Uppsala University Conflict Database42 for battle-related deaths to identify conflict-affected countries. We also supplemented our list of conflict-affected countries with the World Bank’s classification of fragile and conflict-affected situations.43 We identified 28 conflict-affected countries within the study period (online supplemental table S6). We compared the average ECCE aid per child in conflict-affected countries with the corresponding amount in non-conflict-affected countries.

Analysing ECCE aid

Using both the lower-bound and upper-bound estimates, we tracked levels and trends of ECCE aid in terms of total amount and per ECCE-age child between 2007 and 2021 at global, regional and country levels. We also assessed levels and trends by recipient countries grouped into low-income, lower-middle-income and upper-middle-income categories.

We identified the top donors and recipient countries, tracked the CRS sectors to which ECCE aid was allocated, and identified the channel agencies that were responsible for implementing ECCE aid projects in the field. In cases where aid projects had missing information regarding channel agencies, we substituted their missing information with the most frequently employed agency of the same donor throughout the study period. For donors who did not report channel agencies for any ECCE projects, we assumed that the donors implemented the projects themselves.

Additionally, we examined the flow types of ECCE aid. The CRS database categorises aid into official development assistance (ODA) grants, ODA loans, other official flows and private development finance. ODA grants are non-repayable funds to recipient countries, while ODA loans require repayment that offer favourable terms such as lower interest rates or extended repayment periods. Other official flows are aid transactions with developmental goals but do not meet the criteria of ODA grants or loans. We included ‘private development finance’ in our analysis due to its primary development-focused nature, as compared with commercial revenues.26 44

Furthermore, we calculated each donor’s ECCE aid as a percentage of its development assistance to education (DAE) disbursed during the SDG study years (2016–2021). We compared these percentages to UNICEF’s recommendation, which suggests that donors allocate a minimum of 10% of their DAE to the ECCE sector.17 For recipient countries with available data on ECCE gross enrolment rates in 2020,45 we matched each country’s rank in enrolment with that of country’s rank in cumulative ECCE aid per child received throughout the study period.

We also examine countries’ cumulative ECCE aid per child against their most recent available data on the percentages of young children (36–59 months) developmentally on track as measured by the Early Childhood Development Index (ECDI).46 The ECDI is a 10-item index summarising caregiver-reported information on children’s achievement of some universal developmental milestones across countries. A child is deemed ‘developmentally on track’ if the child is on track in at least three of the four ECDI domains, including literacy-numeracy, physical development, social-emotional development and learning. Within each domain, a child is considered to be on track if the child has at least half of the relevant skills.47 Seventy-nine LMICs have available data on the percentages of children developmentally on track.

This article presents the lower-bound estimates as the main findings in the text, while the online supplemental file 1 document presents the upper-bound results. All disbursements were expressed in constant 2020 US dollars. We used STATA V.18.0 for analyses.

General levels and trends of ECCE aid

Between 2007 and 2021, a total of US$3646 million in ECCE aid was allocated to 134 LMICs through 7977 projects. Figure 1 illustrates that the annual ECCE aid increased more than fivefold from US$71 million in 2007 to US$360 million in 2010. It then experienced a slight decrease to US$297 million in 2011, followed by a sharp decline to US$115 million in 2012. However, since 2013 it has been on an upward trajectory, reaching US$562 million in 2021. The peak in 2010 and 2011 was primarily due to large disbursements from the World Bank to Brazil and Turkey, amounting to US$243 million and US$181 million, respectively (figure 2). On average, ECCE aid accounted for 1.7% of the total DAE (US$2 13 279 million), with yearly percentages ranging from 0.6% in 2007 to 3.2% in 2021 (figure 1). Globally, ECCE aid per ECCE-age child increased substantially from US$0.1 in 2007 to US$0.8 in 2021, with an annual growth rate of 15.3% (table 1). For children enrolled in ECCE programmes, the aid amount per child rose from US$1.5 to US$3.4, implying an annual growth rate of 6%.

Request permission

Request permission

ECCE aid by recipient countries’ income group

Across all country-income groups, there was a noticeable increase in total ECCE aid during the study period. However, the LIC group received less aid than the middle-income groups for most of the years, reaching a share of only 12.3% (US$449 million) of total ECCE aid. Of concern, between 2020 and 2021, LICs saw a reduction in their funding from US$103 million to US$77 million, while middle-income countries continued to see increasing amounts (online supplemental figure S2). When considering population size, LICs consistently received less ECCE aid per ECCE-age child before 2016 but started to catch up afterwards. However, from 2020 to 2021, LICs then saw a decrease from US$0.8 to US$0.6 per child, while lower-middle-income and upper-middle-income countries increased from US$0.4 to US$0.6 and from US$1.0 to US$1.3, respectively (table 1).

ECCE aid by region of recipient countries

Over the span of 15 years, Latin America and the Caribbean region received the largest share of total ECCE aid (30.7%, US$1118 million), followed by sub-Saharan Africa (21.8%, US$794 million), East Asia and Pacific (14.0%, US$511 million), Middle East and North Africa (12.7%, US$461 million), South Asia (10.6%, US$386 million) and Europe and Central Asia (10.3%, US$377 million) (online supplemental figure S3). However, when considering population size, the South Asia, East Asia and Pacific and sub-Saharan Africa regions received a smaller amount of aid per child compared with the other regions. Notably, sub-Saharan Africa was the only region that experienced a decline in its aid per child from 2020 to 2021, dropping from US$0.8 to US$0.7 (table 1).

ECCE aid by CRS sector

The educational sector received the largest share of ECCE aid, totaling US$3454 million, accounting for 94.7% of the total ECCE aid. It was followed by the social and economic infrastructure sector, which received US$149 million (4.1%), the emergency and humanitarian sector with US$18 million (0.5%), the health sector with US$9 million (0.3%) and other sectors with US$17 million (0.5%). Notably, the social-infrastructure and economic-infrastructure sector gradually increased its annual share from 2.3% in 2007 (US$2 million) to 10.1% in 2021 (US$57 million) (online supplemental figure S4), with the Inter-American Development Bank and the World Bank each contributing US$60 million during the study period.

Within the CRS emergency and humanitarian sector, a total of 149 ECCE projects were funded over the study period. The EU Institutions, as a single donor category in CRS, was the most prominent donor, contributing US$5 million, followed by UNICEF with US$4 million, and the USA with US$2 million. These projects were primarily focused on material-relief assistance and emergency-support services, including the provision of food assistance in ECCE settings, establishing ECCE classes, training teaching personnel for young refugee children and offering psychosocial support to preschool-aged children in affected areas. However, when comparing the total aid disbursements to the CRS emergency and humanitarian sector across the study period (US$477 047 million), the ECCE projects with humanitarian characteristics (US$18 million) appeared almost negligible.

ECCE aid by implementation channel

Of the 7977 ECCE projects, 1976 (24.8%) lacked information on implementation channels, amounting to US$714 million (19.6%) over the study period. Among projects with missing channels, UNICEF contributed 1851 projects. We replaced missing channel data using the method described above. For UNICEF, we treated the missing data as self-implemented projects. After replacing the missing values, it was found that the governments of recipient countries played the most significant role in implementing ECCE aid-funded projects on the ground, accounting for 60% (US$2185 million) of total ECCE aid. Non-governmental organisations (NGOs) and multilateral organisations were the second and third largest channels, implementing 20.8% (US$757 million) and 11.8% (US$429 million) of total ECCE aid, respectively (online supplemental figure S5). The relatively large proportion (60%) of ECCE aid implemented by governments is unsurprising, given that the World Bank contributed 53.3% of the total ECCE aid over the study period, with nearly all its aid directed to recipient governments.

ECCE aid by flow type

Throughout the study period, the other official flows, defined as aid transactions with developmental goals but that do not meet the criteria of ODA grants or loans, constituted 41.6% (US$1516 million) of total ECCE aid. They were followed by ODA grants at 31.7% (US$1156 million), ODA loans at 14.2% (US$518 million) and private development finance at 12.5% (US$456 million). By country-income groups, ODA grants and loans accounted for the largest proportion of funds allocated to low-income and lower-middle-income countries, representing 81.5% and 74.5%, respectively. In contrast, 76.9% of the total ECCE aid for upper-middle-income countries was in the form of other official flows. It is noteworthy that private developmental finance steadily increased since 2013, accounting for 25.4% (US$143 million) of the ECCE aid in 2021 (online supplemental figure S6).

ECCE aid by donor

In the CRS database, the number of donors reporting any type of aid to the 134 LMICs increased from 42 in 2007 to 138 in 2021, while the number of donors reporting ECCE aid rose from 23 to 53 over the same period, partially contributing to the upward trend in ECCE aid. The World Bank was the leading donor with a cumulative aid disbursement of US$1944 million, accounting for 53.3% of total ECCE aid disbursed by all donors (figure 2). Moreover, the World Bank financed nine out of the 10 projects with the largest project values, including a cumulative total of US$215 million to three lower-middle-income countries (Vietnam, Indonesia and Morocco) and US$603 million to three upper-middle-income countries (Brazil, Turkey and Mexico) (table 2). Together with the World Bank, the top 10 donors contributed 80% (US$2922 million) of the total ECCE aid. Among them, eight donors have available aid data between 2007 and 2021, including the World Bank, UNICEF, Inter-American Development Bank, EU Institutions, The Netherlands, Canada, Italy and South Korea, contributing 74% (US$2693 million) of total ECCE aid. The other two donors—LEGO Foundation and John D. & Catherine T. MacArthur Foundation—started reporting to CRS in 2018 and 2017, respectively. These two donors cumulatively contributed 6% (US$229 million) of total ECCE aid.

However, when considering each donor’s ECCE aid as a percentage of its total educational aid (DAE) disbursement across the SDG study years (2016–2021), the DAC member countries, on average, allocated only 1.9% of DAE to the ECCE sector. Greece (18.5%) was the sole DAC member country whose allocation exceeded the UNICEF-recommended 10% threshold. In contrast to the DAC member countries, multilateral institutions allocated an average of 9.3%, with UNICEF leading at 25.6% (online supplemental figure S7).

ECCE aid by recipient country

Out of the 134 LMICs, 129 received ECCE aid directly from donors, while Grenada, Libya, Saint Lucia, Saint Vincent and the Grenadines and Suriname obtained ECCE aid through allocations of related regional or bilateral funds. Aggregated across the 15 study years, the top 10 recipients were middle-income countries, with Brazil leading with a total receipt of US$394 million. When considering the progress in ECCE aid per child, before the SDG study years (2007–2015), 52 recipient countries, half of which were in sub-Saharan Africa, received an average of <US$0.1. However, the number decreased to 23 during the SDG study period (2016–2021), with nine sub-Saharan Africa countries receiving <US$0.1 (figure 3). See online supplemental table S7 for each country’s yearly ECCE aid per child.

Request permission

In figure 4 and online supplemental figures S8 and S9, we ranked 128 countries with available data on ECCE enrolment rates in 2020 in decreasing order on the right and ranked countries’ cumulative ECCE aid per child throughout the study period in a decreasing order on the left. Among the 25 low-income recipient countries in the lists, only six had a higher rank on the enrolment list compared with their ranking on the aid list. Notably, Sudan had the largest ranking gap among the six countries, ranking 63rd in terms of enrolment but 118th in aid. By contrast, Syria represented the other end of the spectrum, ranking 122nd in terms of enrolment rate but 47th in aid received.

Request permission

When examining countries’ cumulative ECCE aid per child against their percentages of children developmentally on track, most LICs with available data exhibit both low levels of aid and small percentages, forming a cluster in the bottom left corner of figure 5.

Request permission

ECCE aid with COVID-19 activities

ECCE aid allocated to projects with COVID-19 activities totaled US$87 million, with a decrease from US$50 million in 2020 to US$37 million in 2021, representing 11.4% and 6.6% of each year’s ECCE aid, respectively. The World Bank contributed the largest amount of ECCE aid to projects encompassing COVID-19 activities, constituting 9.3% (US$46 million) of its total ECCE aid over 2020 and 2021. Notably, the LEGO Foundation allocated US$36.4 million to ECCE projects including COVID-19 activities, accounting for 36.8% of its total ECCE aid over the 2 years. For recipient countries, Mexico received the largest amount of COVID-19-related ECCE aid, totaling US$40 million, which accounted for 94.1% of its total ECCE aid received during the 2 years. Panama and Kenya followed, receiving US$6 million and US$4 million, respectively, which represented 89.8% and 37.9% of their respective ECCE aid received during the same period (online supplemental tables S8 and S9).

ECCE aid to conflict-affected countries

During the 15-year study period, the average ECCE aid per child in conflict-affected countries remained strikingly lower compared with non-conflict-affected countries. Conflict-affected countries received an average of US$0.3 per child, which was only a quarter of what non-conflict-affected countries received (US$1.2 per child). While non-affected countries saw a consistent increase in aid per child, conflict-affected countries experienced a slower and fluctuating growth (online supplemental figure S10).

General levels and trends of upper-bound ECCE aid

The upper-bound ECCE aid showed a similar trend pattern to the lower-bound estimates. It totaled US$18 180 million across the 15-year period, with the annual aid spanning from US$664 million (2007) to US$1990 million (2021). When considering population size, the aid per ECCE-age child increased approximately threefold from US$1.0 (2007) to US$2.8 (2021). See online supplemental tables S10–S13 and online supplemental figure S11–S18 for more information on the upper-bound estimates.