There has been a debate in development economics over the last 20 years as some claimed the use of RCTs as a tool for independent impact evaluation would significantly improve development practice and hence development. While right about the methodological claims about the superiority of randomization to produce cleaner estimates of the LATE (local average treatment effect) of projects and programs, this, in and of itself, does not change development practice. All of the five claims needed to sustain a positive model in which RCT/IIE has a major positive impact are demonstrably false. The proponents of RCTs have responded to losing the first round decisively by changing significantly both their claims and their practice.
Lant Pritchett is a Professor of the Practice of International Development at the Harvard Kennedy School. He will be moving in 2018 to Oxford's Blavatnik School of Government where he is working on a large research project on how to improve systems of basic education in developing countries. Read more here