Angus Deaton, Joseph Stiglitz, and a number of other prominent economists recently published an op-ed in the Guardian taking the aid industry to task for focusing on studies of aid effectiveness rather than “[tackling] the real root causes of poverty, inequality and climate change.” They offer the standard critiques of RCTs, including their cost and their micro-level focus, and call for systems-level thinking which evaluates public policies as a whole, rather than tinkering with them at the edges. They argue that this will help aid agencies to engage with the “broader macroeconomic, political and institutional drivers of impoverishment and underdevelopment.”
The op-ed manages a curious feat: it suggests a range of deeply political solutions to global poverty, all while using the words “government” and “political” each only once. The authors call for a number of commendable policies, such as public healthcare and education, robust social safety nets, and labor protection. But building the legislative frameworks and administrative systems that would achieve these goals is the role of national governments. Aid agencies can offer funding and technical assistance, but they ultimately have limited control over how national education or welfare systems are run. It’s a rhetorical sleight of hand to put the statements “the aid system is ineffective” and “public service delivery in poor places is ineffective” next to each other, and use this to imply that better aid would somehow improve public service delivery without any involvement from national governments.
In places where governments are failing to provide good public services, there are complex problems of political economy underpinning this. You’ve got countries with active conflicts, like South Sudan or the Central African Republic. You’ve got countries where the government has really low administrative capacity, like Liberia or the DRC. You’ve got countries with middling to high administrative capacity, like Kenya or South Africa or the US, where lots of poor citizens are seen as politically expendable and thus aren’t offered services. It’s difficult for aid agencies to build administrative capacity and solve problems of political exclusion from the outside, because these are fundamentally questions of domestic political settlements. The practice of thinking and working politically could result in some improvements to aid delivery, but it doesn’t solve the underlying political economy problems.
This point highlights that the “aid is ineffective because it’s overly focused on RCTs” argument is specious. Aid agencies haven’t ended global poverty because global poverty is a really complex problem — one whose solution involves contentious, long-term processes of political reform as well as short-term flows of funding and technical assistance. At a practical level, there’s lots of room in here for aid agencies to both engage with local political actors, and collect rigorous evidence on whether the programs that they’re promoting are working. At best, these activities can work together, producing strong evidence of program effectiveness which can be used to lobby domestic politicians. A great example of this comes from the spread of cash transfer programs in Latin America. Mexico was one of the first countries in the region to launch this type of social protection program, and hired researchers to run a rigorous evaluation as well. The results showed that program beneficiaries were healthier and more financially secure, and politicians throughout the region often relied on these findings to promote cash transfers in their own countries. The rich body of evidence created by subsequent evaluations of these other Latin American programs has become the cornerstone of efforts by DFID and other aid agencies to encourage African governments to adopt cash transfers. RCTs are one of many tools that can be used to do the political work of promoting policy change in low income countries; they’re not inherently a roadblock to it, or a substitute for it.