When randomistas rule the field: Randomized experiments, entanglement, and funding for the poor

Social scientists have long debated the merits of randomized controlled trials (RCTs), often asking whether their results are valid or the approach is appropriate. This article instead treats RCTs as interventions that become entangled with the projects they study. To achieve internally valid results, experimenters impose strict research protocols onto implementers, reshaping project implementation and, under certain conditions, constraining projects’ performance. Drawing on science and technology studies, this case study of five large-scale farmer-training projects across the Global South shows how random assignment displaced implementers’ purposive targeting and segregation of the treatment and control groups over space and time interfered with implementation. The result was the production of hybrid experiment–project assemblages whose findings were technically valid but distorted. Where experimenters’ rules ruled, implementers’ expertise was constrained and impacts failed to materialize. The article argues that such dynamics have implications for how development interventions are judged and financed, advancing evaluation risk as a political-economic concern distinct from reactivity or research ethics.