Needed: A Modernized Multilateral Global Education Architecture

In this blogpost, which is part of NORRAG’s blog series “International Organisations and the Global Governance of Education”, Nicholas Burnett advocates for the urgent streamlining and modernization of the multilateral global education architecture to meet ten priorities, six for country support and four at the global level. A longer unpublished version of this blog is available here.

The Multilateral Global Education Architecture is in deep trouble, as I have noted since 2019. Failure to meet international targets has led to a continual muddling along, as Jenkins and I argued recently. Indeed, failure is now seen as inevitable, undermining the bold action needed to serve the world’s children and to fix international education aid financing and its governing global architecture.

This architecture was initially limited to UNESCO and UNICEF, joined in 1962  by the World Bank and later other regional multilateral development banks (MDBs). The 20^th century architecture was characterized by alternating rivalry and collaboration. If the architecture was limited in the 20^th century, however, it exploded in the 21^st. The Global Partnership for Education (GPE) was set up in 2002 to help finance the international goals, implementing its grants mainly through the World Bank and UNICEF; Education Cannot Wait (ECW), housed in UNICEF, was established in 2016 to finance education during humanitarian crises; the Education Outcomes Fund (EOF), also housed in UNICEF, was created in 2018 to promote impact investing in non-state education; and the International Finance Facility for Education (IFFEd) was founded in 2023 to help Lower Middle-Income Countries (LMICs) increase their access to concessional financing. Other than GPE, however, none has yet provided any major new external financing. Indeed, there are now too many small agencies, and also insufficient coordination among them and the few larger ones. This has resulted in much duplication, placing a heavy administrative burden on developing countries.

All agencies are now in increasingly deep trouble, following reduced bilateral interest in development aid, both general and for education. This became starkly apparent in early 2025, with the end of support from USAID, the largest bilateral donor. In fact, the trend was already visible in 2024, with most bilateral donors cutting their aid. Moreover, aid to education will further decline by a quarter by 2027, forecasts the Global Education Monitoring Report (GEMR). Though a few bilateral donors like Japan and philanthropies like the Gates Foundation have announced offsetting increases, these are in the tens of millions of dollars, relatively small compared to the likely overall cut of over $3 billion by 2027.

GPE and ECW are both seeking replenishments in 2027. Astonishingly, however, planning for replenishing GPE and ECW is happening largely as if the world has not changed and as if their missions should also continue unchanged.

There is one exception. There is some current behind-the-scenes discussion about possibly merging ECW with GPE to simplify their replenishment. This seems a highly inappropriate solution. ECW’s function is very different from the other agencies, and deliberately mixing humanitarian and development aid would only compound the already seriously inadequate humanitarian aid for education, with less than one third of 2024 funding requested for education materializing.

Glennerster and Crawfurd from the Center for Global Development (CGD) suggest that donors use four criteria to prioritize replenishment decisions for education agencies: track records of funding cost-effective interventions; Foundational Learning and Numeracy (FLN) focus; poverty focus; and maximization of financial and policy leverage. They hold the first to be the most important. These suggestions are generally sensible but much too limited, framed in terms of existing agencies and modes of operation.

A bolder approach would be to reduce the number of agencies, focusing the resulting new architecture on ten priorities, six country-based and four global:

Country-based Priorities

1. Massively increasing grant funding for LICs, the top priority for the architecture. LICs currently receive only one fifth of all education aid; only one third of all IDA education funding; and only one half of GPE grants. IDA should earmark about half of its education funding for LICs, and GPE should concentrate almost entirely on them. The two should also work more closely together; the simplest way to achieve this would be for GPE to revert to being a trust fund managed by the World Bank with a specific mandate to support LICs.
2. Subsidizing prudent borrowing by LMICs. The corollary of increasing grant funding for LICs is increased borrowing by LMICs. Subsidized borrowing can come from IDA, from co-financing education loans and grants (as the GPE Multiplier achieves), and from expanding IFFEd, even though it has yet to commence operations. Indeed, IFFEd represents the only currentnew mechanism to generate a real increase in external support to LMICs on attractive terms. To improve visibility and coordination and, above all, to speed things up, IFFEd could be housed within the World Bank while remaining an independent foundation.
3. Prioritizing External Support for Domestic Resource Mobilization (DRM). Domestic financing will always be the principal source of public spending on education. Yet the architecture has done little to support DRM. It is a high priority to correct this, for example by establishing a knowledge program on DRM in education at either UNESCO or the World Bank in collaboration with the effective tax support programs of the International Monetary Fund and the Organization for Economic Cooperation and Development.
4. Focusing support on programs to achieve FLN, especially in LICs. FLN and primary education should be the top priority for all education aid, and especially the multilateral priority, since higher and vocational education are more appropriate for bilateral donors, using mechanisms like institutional twinning. Yet basic education’s share of education aid is falling, down from 40 percent in 2016 to 30 percent in 2023.
5. Delivering country support via pooled funding to peer groups of developing countries. Budget support and pooled funding represent only 7 and 14 percent, respectively, of education aid. As Jenkins and I suggest, In general, country aid should be allocated by peer groups of countries based on performance, linking it to DRM, and rebalancing the donor-dominated architecture.
6. Supporting non-state education where it can complement or directly improve public education. Non-state education, encompassing everything from private schools for the elite through NGO and religious schools and for-profit Low-Fee Private Schools for the poor, is viewed with suspicion by many in the education sector. The overall shortage of finance for public schools, however, means that it makes sense for donors to support non-state education in countries whose governments also do so. EOF should therefore continue to exist but should shift its focus more toward primary level FLN, away from its current limited emphasis on early childhood, vocational education and school management support.

Global Priorities

7. Adequately financing global public goods in education. UNESCO should ideally be the agency for this, focused on data, analysis of what works, and the promotion of educational innovations. UNESCO’s ability to conduct and even synthesize research is somewhat limited but this could be offset by an agreement among all agencies to coordinate their research and to establish a database and an innovation fund within UNESCO, a potentially important first step to restoring UNESCO’s knowledge leadership role.
8. Adequately financing humanitarian aid for education. ECW is already housed at UNICEF, the only agency with humanitarian experience. ECW need not remain a separate agency but could itself be integrated into UNICEF.
9. Making the global architecture more inclusive. The current architecture is not just ineffective; it is also not inclusive, consisting only of the main global multilateral agencies and the main bilateral donors but excluding the regional MDBs, philanthropies and, above all, developing countries. Mechanisms are needed to include regional MDBs and philanthropies and to increase the voice of developing countries in the governance of both individual agencies and the architecture as a whole. Allocating aid via peer groups of countries could be an important first step.
10. Promoting education’s importance. Education lacks leadership in the overall contests for resources, both domestic and international. The World Bank and UNESCO could find a way to work together and assume shared global leadership; and could jointly establish an independent group to advocate for education and to fix the global architecture.

The need is urgent. Not only has the overall development architecture and prospects taken a turn for the worse, so has the condition of education in the developing world, especially the LICs. The latest UIS/GEMR estimate of the number of children out of school has risen to 272 million for 2023, up 21 million from the previous estimate, though only about 40 percent of this increase is due to an absolute increase in out of school numbers, the remainder reflecting revised population estimates. The number of displaced children has almost tripled from 17 million in 2010 to 49 million in 2024.  COVID-19 sharply increased the rate of learning poverty, defined by the World Bank as the proportion of 10-year-old children unable to read a simple text with comprehension in their own language, from 57 percent in 2019 to a staggering 70 percent by 2022, with the latest 2024 data indicating only very modest improvements.

The Multilateral Global Education Architecture has to address these issues; it cannot do so efficiently and effectively in its present form. Simply replenishing funds like GPE and ECW, even at what are at best likely to be reduced levels, will not suffice – the whole architecture needs to be revamped along the lines of the ten priorities advocated above. If not, despite demographic change, the proportion of uneducated and poorly educated children will continue to rise, especially in LICs, a huge stain on our global conscience and on the ability of multilateral education agencies, and their member countries, to collectively reform themselves.

The Author:

Nicholas Burnett is a Senior Fellow at Results for Development Institute whose Global Education Practice he previously founded, having previously held senior education management positions at the World Bank, the Global Education Monitoring Report and UNESCO.