Which Devil in Development? A Large-N Survey and Randomized Field Experiment Comparing Bilateral Aid, Multilateral Assistance, and Government Action in Uganda

Abstract

Debate over the effectiveness of foreign aid has raged on despite a paucity of information about recipients’ actual views of development assistance, especially as citizens compare aid to domestic government programs. We argue that citizens may see foreign aid as an escape from clientelism because aid is less politicized than government programs, citizens trust donors more, and they support aid conditionality. They may also favor multilateral donors over bilateral donors for similar reasons. We test the argument with a randomized experiment on a subject pool of roughly 3,600 Ugandan citizens – to our knowledge the first nationally representative, large-n study of aid recipients. We randomly assigned the project funders – multilateral banks, bilateral donors, and a control implying the domestic government – for actual co-financed “pipeline” projects and invited citizens to sign a petition and send a text message in support. We find that citizens are significantly more willing to sign a petition or send a text message in favor of foreign aid projects compared to government programs. A companion survey to the experiment reveals evidence that citizens perceive aid as less prone to politicization. Some evidence suggests that Ugandans also see multilateral donors as superior to bilaterals. The findings suggest that recipients view foreign aid as relatively effective compared to domestic government programs.