Radical Simplification: A Practical Way to Get More Out of Limited Foreign Assistance Budgets

With major cuts to foreign assistance announced or anticipated in the US, the UK, and across the EU, and in the World Bank’s subsidized lending program, we must urgently work to get the most out of shrinking aid budgets. Nothing can prevent the cuts from leading to deaths, but we can still do better with what we have left.

Aid agencies already try to cover too many countries and sectors, incurring high costs to set up small programs. Aid projects are far too complicated, resembling a Christmas tree weighed down with everyone's pet cause. With less money (and in the US, very few staff), now is the time to radically simplify. By choosing a few highly cost-effective interventions and doing them at large scale in multiple countries, we would ensure

• aid funds are spent on highly effective projects;
• we benefit from the substantial economies of scale seen in development;
• a much higher proportion of aid money goes to recipient countries, with less spent on consultants; and
• politicians and the public can more easily understand what aid is being spent on, helping build support for aid.

Radically simpler projects

When I served as chief economist of the UK’s Foreign, Commonwealth and Development Office (FCDO), my office reviewed all projects over £40 million ($50 million) and consistently found projects contained multiple, often unrelated components. An education project in Nigeria promoted evidence-based teaching approaches but also sought to advise on national data collection and private school regulation, and even covered village savings and loan associations. Each task demands different skills requiring contracts with different organizations, making it hard to effectively design, procure, implement, and evaluate the resulting complex beast.

Instead, projects should do one thing on a large scale. The cost per person of implementing a project falls sharply the more people it covers because of the high fixed costs of design, setup, and oversight. In the context of direct monetary transfers, the IRC note “the biggest single factor driving cost efficiency is the scale at which programs are run—reaching more households spreads the fixed costs of country support over a wider pool of beneficiaries, driving down per household costs dramatically.” The same declining costs per head is also found when distributing non-food items during emergencies, building latrines, and distributing hygiene kits. Simpler projects don’t have to be small; they could simply cover more people.

Fewer interventions

Lots of projects means small projects—the median ODA project is worth less than $100,000. It is extremely inefficient to start a new design process from scratch every time we set up a new project. It is not possible to do things the old way with the dramatically smaller staff now in charge of foreign assistance in the US. Others have suggested limiting the number of sectors an agency works in but we should be more radical and limit what we do in a sector: for example in health, one agency could specialize in just buying antiretroviral drugs or combatting malaria.

A selective portfolio of highly cost-effective interventions could absorb significant funding while maintaining high returns. For example:

• Fully funding the rollout of the new malaria vaccines could absorb around $4 billion to $5 billion. An initial $2 billion expansion of malaria vaccines costing just $4,200 per life saved would absorb about 7 percent of the budgetary resources (~US$27 billion) allocated to USAID in FY25.
• Antimalarial bednets can save lives for between $3,000 and $8,000 in priority locations and can be distributed through large-scale mass campaigns.
• Graduation programs targeting the ultra-poor, which have been proven (in multiple countries) to lift people permanently out of poverty, could absorb all the bilateral aid programs of all the G7 countries combined and still not reach all 700 million people in extreme poverty in the world.^[1]
• Rolling out a simplified combined protocol for treating malnourished children is highly effective (costing just $918 per child treated) and could be scaled with community health workers. A total of 45 million children under five could benefit.
• A simple, effective, and highly leveraged investment for bilateral donors is funding the World Bank‘s subsidized loans and grants to poor countries through its International Development Association (IDA). IDA can provide $3-$4 in highly subsidized loans or grants for every $1 it receives from donors by leveraging its high credit rating.

Fewer countries

Bilateral donors often support aid programs in many countries. More countries means smaller programs with higher overheads, which reduces cost-effectiveness. My colleague Charles Kenny finds only 7 percent of bilateral aid relationships reported to the OECD’s Development Assistance Committee (DAC) were greater than $100 million. One hundred million sounds like a lot, but it is similar to the cost of a high school renovation in Indiana. And these small country programs are far from simple. New Zealand gave Indonesia $16 million (about 0.001 percent of Indonesia’s GDP) in 2021 yet, as Charles points out, had a partnership agreement with Indonesia covering energy, agriculture, risk management, knowledge, and skills. Consultants would need to be hired to design strategies for each of these sectors, with little money remaining per sector to actually fund change. Working in fewer countries would allow for better engagement, with greater focus on fewer, larger interventions.

Being more selective in the countries we support would help us target aid, especially grants, to the poorest countries that need it most. Recent data suggests nearly 40 percent of DAC grants went to upper-middle-income countries. CGD analysis of ODA flows suggests whether a country receives grants or loans varies less by income than effective aid policy would imply.

Objections

Radical change will always face opposition. There are three main objections to simplification: (i) every country has different needs and requires individually tailored programs, (ii) poverty has multiple drivers, calling for complex responses, and (iii) we need some small programs to invest in learning what are even more effective interventions. The first two fail to recognize that we cannot fix everything. With limited funds we should identify the places with needs we know how to fix. It is true, for example, that malnutrition programs are not appropriate everywhere because not everywhere faces concentrated caseloads of acute childhood malnutrition. However, where the problem exists, applying a standardized well-tested feeding protocol is the right response, whatever the context. When implementing graduation programs aimed at the ultra-poor it makes sense to adjust the business asset given to beneficiaries in different countries, but the fundamentals of the program remain valid across countries.

In response to the second objection, the evidence shows that fixing one part of the problem at a time can work. The interventions we mention above, and many others, have been found to be effective in isolation—i.e., within poorly functioning systems and without fixing anything else. The assumption that adding more, often untested, interventions will improve outcomes is just that, an assumption, and it will almost certainly reduce cost-effectiveness.

The third objection has more merit. If we only implement a few of the currently most cost-effective interventions, we will never discover new, potentially even more cost-effective approaches. Aid donors including USAID, FCDO (see DFID in chart here) and the World Bank have played an important role in generating evidence on the most cost-effective interventions, which governments and donors have scaled up, generating high returns.^[2] Foreign assistance funders should therefore continue to support evidence gathering.

The incentives that drive complexity

There are reasons why complexity exists: understanding and addressing these drivers will help us achieve radical simplification. We identify several forces that encourage complexity.

Donor optics bias officials away from choosing simple programs. Junior bureaucrats want to showcase visible impacts to their seniors, journalists, or visiting politicians—perfect-looking farms or schools. It is often hard to see the difference caused by cost-effective interventions with modest impacts at very low cost (e.g., digital extension for farmers).

“Christmas tree” projects reflect review processes where experts from multiple sectors are asked to comment on a project and add their favored interventions, and proposal are mandated to address specific subjects. FCDO business cases typically require text on equality, climate and environment, and counterterrorism regardless of the intervention. Projects supported through World Bank Investment Project Financing are required to follow 10 environmental and social standards, including biodiversity conservation, cultural heritage, and community health. The laudable intention of these requirements is to make sure teams do not overlook unintended consequences but it is easier to simply add an environmental element (solar panels on schools, as in a recent World Bank education project appraisal document for Nigeria) or gender element (add gender awareness to the curriculum) than adjust the main program to ensure it does not cause unintended consequences. CGD analysis finds $15 billion worth of the World Bank’s climate finance portfolio is attributed to projects where climate accounts for less than 20 percent of the project’s value, suggesting officials face incentives to add climate elements to projects that are not primarily about climate. The length of the review process also encourages senior staff to bundle projects that otherwise might not fit together.

These incentives are compounded in organizations such as the World Bank where staff have to bill out their time: if the environmental team raises environmental issues, they can bill their time to address them.

Simplification requires closing offices and letting staff go. One of the benefits is that more of the money gets spent on beneficiaries but that means less on staff. There are always advocates for a particular program or sector, and they will be louder than those who think it is not a priority, making it harder to achieve simplification.

How to address incentives towards complexity

The case for focusing limited resources on the most cost-effective interventions is always strong but politically challenging because it means disappointing some advocates. The present moment—where staffing and funding cuts are inevitable—provides an opportunity to make hard decisions. It takes much fewer staff to commission a simple type of program in multiple countries—a consideration with particular relevance today.

An era of reduced aid budgets also calls for greater emphasis on outcomes, not inputs. Focusing on the amount and cost of averted greenhouse gas emissions would reduce the incentive to add climate components to every program. During my time at FCDO, ministers wanted easy metrics to judge the likely cost-efficiency of a project: e.g., the number of people reached versus the cost of the program. Officials worried these simple metrics would be misleading but the lack of transparency undermined faith in the system. We should prominently display and track indicators of scale—numbers reached, CO₂ mitigated—on project documents. This would provide a nudge towards implementing projects that reach large numbers of people with simple interventions.^[3]

The review process should be reformed, with the extent of review reflecting project complexity, not spend. I saw long, detailed review documents for support to the World Bank or GAVI because the projects had high price tags even though the case was simple. Highly cost-effective interventions that have been thoroughly vetted and frequently implemented by the agency should only be subject to a short and simple review process. In these cases, reviews should simply ask whether this is an appropriate context for this intervention (e.g., is there a significant acute malnutrition caseload?) and whether the implementing team is following standard guidelines. Central review teams, such as FCDO’s Quality Assurance Unit, should also be empowered to reject requests by sector experts for add-on components that add complexity.

Given the reality of smaller aid budgets, it is imperative we use them to best effect. We need to focus on a select number of simple, cost-effective, large-scale interventions. This will require reforming the review process and making better use of fewer staff, but it will make sure more of our aid gets to those who need it and make it easier for politicians and the public to understand what they are getting from aid. Reform is hard but the aid industry is in a period of change, so now is the time to act.

We thank Charles Kenny, Justin Sandefur, Ian Mitchell, Janet Hodur, Erin Collinson, Emily Schabacker, and Clemence Landers for thoughtful comments on this post.

^[1] Banerjee et al. (2015) Table S7 (supplementary material) report program costs in six countries in 2014 USD exchange rate terms. The average program cost per household is ~$1,700 (2014 USD). We divide that by household size of ~6 to get a cost per person of ~$288. Re‑expressing that in 2024 dollars with (i) the U.S. CPI‑U (+ ~33 %) and (ii) the cumulative CPI inflation for the IMF’s ‘emerging‑market & developing‑economies’ aggregate (+ ~87 %) gives a range of ~$380–$540 per person. Applied to the World Bank’s ~700 million people in extreme poverty, the total graduation program bill is ~$270–380 billion. For comparison, G7 official development assistance in 2024 is preliminarily estimated at ~$160 billion in total and ~$127 billion in bilateral flows.

^[2] For full disclosure, Rachel Glennerster has received funding from all three units that fund evidence collection referenced in the links. Michael Kremer was scientific director to the USAID unit "Development Innovation Ventures" and is married to Rachel Glennerster.

^[3] There is a risk that this encourages interventions such as national regulation which can be said to apply to large numbers, but this could be mitigated with appropriate guidance.