Paper: Power and Pitfalls of Experiments in Development Economics

Development economists increasingly employ experimental methods in seeking to answer key questions that inform development policies and programs. Impact evaluation based on randomized controlled trials (RCTs) has in particular fundamentally altered the discourse of development economics over the past decade.

While these methods have power, their use in development economics facing a number of intrinsic and extrinsic problems, including ethical dilemmas, faux exogeneity and what we term the ingénue effect. This paper illustrate these points with concrete examples concerning capital access and productivity growth, and argues for greater use of behavioral experiments that can help resolve the identification problems that motivate RCTs by eliciting otherwise unobservable, but fundamental economic characteristics such as rates of time preference and risk aversion.

In the end, this paper argues, economic development and development economics will be better served if we can arrive at a balanced appreciation of the essential but limited role of RCTs and experiments.