Are Aid Agencies Paying Attention to What Works in Education?

When aid budgets come under pressure, programmatic choices within sectors matter more. In education, a growing body of evidence shows that the best interventions to improve learning outcomes can be orders of magnitude more cost-effective than the average intervention. But how often do aid projects actually deliver the best interventions?

This blog focuses on “smart buys” in education: interventions that the Global Education Evidence Advisory Panel (GEEAP) has identified in its two reports published in 2020 and 2023 as especially effective, and in some cases especially cost-effective, for improving learning. We focus on two leading classroom-based approaches: targeted instruction and structured pedagogy.

We track the uptake of these approaches in education at two of the largest foreign aid providers, the World Bank and the UK’s Foreign, Commonwealth and Development Office (FCDO). Rather than measuring actual spending or implementation (which we intend to do in future installments of the series), we track whether official project documents (business cases in the case of FCDO and project appraisal documents [PADs] in the case of the World Bank) explicitly reference these interventions. That is an imperfect proxy for adoption, but it offers a useful first indication of whether evidence-informed ideas are beginning to shape mainstream programming.

A clear shift in project language

We see a clear increase in references to the two smart buys over time. Between 2015 and 2019, only 12 percent of World Bank education projects mentioned these smart buys. That nearly doubled between 2020 and 2024 to 23 percent and increased again to 68 percent in 2025. FCDO funds fewer education projects, but the direction is similar: before 2020, only three projects (8 percent of the total over that period) directly mentioned targeted instruction, and none mentioned structured pedagogy. From 2020 to 2025, 38 percent of projects mentioned one of the two approaches. In 2025, the two education projects approved by FCDO mentioned both.

Education aid projects increasingly mention smart buys

The World Bank trend is especially salient given its position as the largest external financier of education in the developing world. Explicit references to structured pedagogy or targeted instruction were relatively uncommon in 2015, became more visible from 2016 to 2024, and became much more common in 2025. The timing lines up with the renewed focus on learning outcomes as an objective of the World Bank’s education projects, especially after the 2018 World Development Report helped make the learning crisis a central organising idea.

FCDO shows the same broad direction, but with more year-to-year volatility because it approves fewer projects. Even so, the increase after 2020 is notable. Since 2020, 13 in 34 FCDO education projects have mentioned either structured pedagogy or targeted instruction. That share would be higher if one excluded categories of spending in which the UK finances education more indirectly. For example, some business cases related to core contributions to multilateral funds, such as International Finance Facility for Education (IFFEd) and Global Partnership for Education (GPE), referenced the GEEAP and evidence use but did not discuss specific interventions in detail. Several crisis-response programmes also focused more on broad service delivery than on classroom pedagogy. Excluding such cases, the post-2020 share of FCDO education business cases mentioning one of these smart buys would be close to 50 percent.

Reference to both leading smart buys have increased across the two organisations

World Bank

Image

FCDO

Image

Note: Y axis represents number of project appraisal documents (World Bank) or business case documents (FCDO)

One especially interesting pattern in the FCDO documents is that almost every business case that mentions structured pedagogy or targeted instruction also explicitly refers to smart buys or cites the GEEAP report. That suggests the report itself might be shaping how FCDO staff frame and justify education investments.

What are the education smart buys?

The GEEAP published a pair of reports in 2020 and 2023 on smart buys for improving learning in low- and middle-income countries, providing practical, evidence-informed recommendations for the prioritisation of education interventions. The interventions that deliver the biggest learning gains per dollar, especially for foundational literacy and numeracy, drew on rigorous evaluations and available cost-effectiveness evidence. Across the two reports, interventions are grouped into tiers: great buys, good buys, promising but limited evidence, effective but relatively expensive, and bad buys. The 2023 update expanded the evidence base and sharpened guidance on scalability and implementation.

The latest edition identifies three interventions as “great buys” based on evidence of both effectiveness and cost-effectiveness: (1) providing information on the benefits, costs, and quality of education; (2) supporting teachers through structured pedagogy; and (3) targeting instruction to students’ learning level rather than their grade. The first aims to increase effective demand for quality education, while the latter two operate on the supply side by improving the quality of teaching. In many contexts, limited provision of high-quality instruction is a binding constraint on foundational learning. Accordingly, our review focuses on the extent to which structured pedagogy and targeted instruction are incorporated into internationally financed education projects.

Why focus on the World Bank and FCDO?

We focus on the World Bank and FCDO for three practical reasons:

1. Both are closely connected to the smart buys agenda: as conveners of the GEEAP and its smart buys reports (along with UNICEF and, until its closure, USAID).
2. The World Bank is the largest provider of education financing after national governments, including through its concessional arm, the International Development Association, making it especially relevant for low-income settings where the marginal returns to well-chosen interventions can be large.
3. Both FCDO and the World Bank have relatively accessible and comparable documentation for nearly all education projects, making it feasible to track how programming language evolves over time and across contexts.

Together, these two organizations offer a useful window into whether ideas from the smart buy reports are translating into mainstream programming in both multilateral and bilateral contexts.

What we do and don’t measure

Our approach is deliberately simple. We identified references to targeted instruction and structured pedagogy in publicly available project documentation as a proxy for adoption. For FCDO, we use business cases for new education projects, excluding cost extensions. For the World Bank, we used project appraisal documents for projects forecast to spend at least half of their budgets on education, approved from 2015 onward. In total, the dataset includes 72 FCDO business cases (downloaded from the International Aid Transparency Initiative) and 237 World Bank PADs (drawn from its public repository).

This approach has obvious limitations. A project may mention a smart buy without implementing it seriously, at scale, or with fidelity. Equally, a project may adopt important elements of structured pedagogy or targeted instruction without using those exact labels. Some projects are also harder to classify because documentation varies in depth and precision. This exercise is the first snapshot of explicit uptake in official programming language, not a full accounting of education aid spending on smart buys. Later posts in this series will look at deeper patterns of adoption, including project components, implementation details, and the weight these interventions carry within larger programmes.

Smart buys are increasingly mainstream

The evidence from project documentation suggests that smart buys are becoming more visible within the international education aid architecture. This suggests notable progress in embedding evidence into large-scale and ambitious programme designs in education.

Despite the increasing mention of smart buys, substantial amounts of money are still not spent on them. In some cases, that will be for very good reasons, but for others not. That implies a margin for further improvement.

A final lesson we learnt was about just how hard it was to categorise the projects we read. Transparency in documentation is an important step, but it’s only part of the battle; we need to improve its quality, too.