Smart Buys Alliance: An Initiative for the Effectiveness of International Aid

Romain Barbe

Founder · Mieux Donner · Reading time: 8 min

An opportunity to share knowledge and better direct financing

Imagine a ministry that must decide how to allocate additional millions of euros to its existing cooperation budget. Several options are available: funding a vaccination campaign, a nutrition programme, a water access initiative, or training for local healthcare staff. How do you choose among interventions that all seem legitimate?

This question deserves a rigorous answer. Yet producing a high-quality synthesis of the most effective interventions in a given sector is a considerable undertaking: it requires identifying relevant studies, assessing their methodological quality, harmonising heterogeneous units of measurement, and translating the results into a form usable by operational teams who are not necessarily researchers. This work takes time, requires specialised expertise, and demands resources that not all agencies invest sufficiently in alongside their day-to-day activities.

Some of these syntheses already exist, in the form of internal lists developed by certain agencies for their own needs, or reports produced by research institutions on specific topics. But these resources often remain siloed, not harmonised with each other, and unavailable across the full range of sectors where they would be useful.

What is a smart buy?

A smart buy is a development intervention whose cost-to-impact ratio is validated by rigorous evaluations, typically randomised controlled trials or quasi-experimental studies with robust comparison groups. Examples include structured pedagogy programmes in education, simplified malnutrition treatment protocols, vaccination programmes, and improved weather forecasting for farmers.[1]

Two important clarifications about what this concept is not.

The central idea is comparative. When seeking to improve learning outcomes in low-income countries, not all interventions are equal. Identifying these effectiveness gaps is the starting point. We have documented this logic of large effectiveness differences between interventions separately.

How is a smart buys list built? The education example

To understand concretely what smart buys lists are and how they are constructed, the education sector provides the best-documented example to date.

Measuring impact: what does a unit look like in education?

Before comparing interventions with each other, a common unit of measurement is needed. In health, DALYs (disability-adjusted life years) are commonly used: how many healthy life years does an intervention generate, and at what cost? In education, the equivalent is the LAYS (Learning-Adjusted Years of Schooling).

The LAYS corrects an important bias: a year of schooling does not produce the same amount of real learning across countries, education systems, or teaching quality. In some contexts, a child can spend five years in school without acquiring basic reading and numeracy skills. The LAYS accounts for this: it measures not time spent in the classroom, but learning actually produced.[2]

What the GEEAP list reveals: considerable gaps between interventions

The Global Education Evidence Advisory Panel (GEEAP), an independent expert group co-led by the World Bank, FCDO, UNICEF, and USAID, published an update of its report in 2023 on the most cost-effective approaches to improving learning in low- and middle-income countries.[3] For this report, the team conducted a systematic review of over 13,000 additional studies, resulting in the analysis of more than 400 high-quality evaluations.[4]

The report distinguishes four categories:

The "Bad Buys" category deserves attention. Equipping schools with computers is an intuitively appealing, visible, and politically easy intervention to defend. Yet the evidence indicates it produces little or no improvement in learning outcomes when not accompanied by deep pedagogical change. Significant resources can thus be invested in an intervention with very low educational returns, while far more cost-effective interventions remain underfunded.

Teaching at the Right Level: from rigorous evaluation to 80 million children

The Teaching at the Right Level (TaRL) approach, developed by the Indian NGO Pratham, concretely illustrates what a smart buy can produce when properly identified, evaluated, and scaled.

The pedagogical problem TaRL solves

In many low-income countries, millions of children are enrolled in school but fail to acquire foundational reading and numeracy skills during primary education. This is not primarily a problem of school access: enrolment rates have risen sharply in these countries since the 2000s. The problem is pedagogical.

Most education systems organise teaching by grade level, following an official curriculum that assumes all children of the same age share the same prior knowledge. In reality, this is not the case. A child in Grade 5 who has not yet grasped the basics of reading finds themselves in grammar class, without ever getting the chance to consolidate the foundational skills they need.

What evaluations show

Since 2001, Pratham has evaluated TaRL in partnership with J-PAL through six randomised evaluations across seven Indian states, producing some of the largest learning gains ever measured in education research.[5] In the state of Uttar Pradesh, the number of children able to read a paragraph or story doubled following the intensive learning programmes.

Multiple independent evaluations have shown that the TaRL approach improves learning outcomes in most tested contexts, with gains ranging from 0.1 to 0.3 standard deviations per student, and estimates reaching 0.7 standard deviations in some evaluations.[6] As a benchmark: a gain of 0.2 standard deviations in reading corresponds roughly to several months of additional learning for a child, at generally very low cost per student. The TaRL approach was classified as a "Great Buy" by the 2023 GEEAP report.[7]

Deployment conditions and known limitations

Several African governments have adapted and deployed TaRL with technical support from Pratham, J-PAL, and TaRL Africa: Zambia extended its "Catch Up" programme nationwide, Nigeria implemented it across seven states, and Côte d'Ivoire developed the Targeted Teaching Programme. TaRL Africa reached more than 7 million students on the continent, and overall the approach had reached more than 80 million children in India and Africa according to data available for 2025.[8]

These results are not uniform, and evaluators document this honestly. Learning gains tend to diminish when programmes are deployed at scale through government systems, where ongoing teacher training and quality monitoring are harder to sustain.[9] Significant variation has been observed across countries, and sometimes across states within the same country, depending on implementation conditions.

What the Smart Buys Alliance does

Faced with the proliferation of sector-specific lists produced by different institutions and the risk of costly duplication among funders, the Smart Buys Alliance proposes a coordinated response built around three axes.

Producing smart buys lists as a public good

One of the current structural problems is that producing high-quality syntheses of the most effective interventions is costly in time and resources. When each agency funds its own literature reviews, the same studies are analysed in parallel by different teams, with results that can diverge depending on methodological choices. The resources invested in this duplication could be better used.

The Smart Buys Alliance aims to pool this work and generate new evidence: either by harmonising and updating existing but unpublished lists, or by developing entirely new lists on topics of common interest that are still underserved. These reference documents are intended to become resources accessible to all funders, with shared vocabulary and implementation conditions spelled out as precisely as possible to facilitate their use. This is a complementary approach to the Innovation for Development Fund, which also finances the production of evidence in international aid, and whose findings will contribute to the determination of future smart buys.

Facilitating co-financing of validated interventions

Beyond knowledge production, the Alliance plays a matchmaking role. When an agency identifies a smart buy intervention it wishes to fund, it may lack sufficient resources to deploy it at the relevant scale. Other funders may share the same thematic priorities without having access to an evidence synthesis to guide their investment.

The Alliance facilitates these connections to organise targeted co-financing of validated interventions. An initial investment from one agency can thus be doubled when a co-funder is mobilised, increasing the potential impact without raising the cost per funder.

Supporting implementation

The third component responds directly to the limitation documented in the TaRL example: a smart buys list is only useful if the organisations seeking to apply it have support to adapt these interventions to their specific context.

The Alliance offers implementation consultancies, coordinated by the co-hosting organisations (J-PAL, CEGA, Center for Global Development, Global Poverty Research Lab). These can take the form of rapid evidence reviews on a specific topic (two to three weeks), in-depth cost-effectiveness analyses tailored to a given context (approximately six months), or support for integrating smart buys criteria into internal procurement processes, RFP design, and proposal evaluation.

Alliance members

The Smart Buys Alliance is co-hosted by a consortium of five research and public policy organisations, with two founding public agency members.

Frequently asked questions