Egalitarian Equivalent Treatment Effects: The Econometrics of Inequality-Sensitive Treatment Effects Estimation

Decision makers often have a strong interest in understanding the distributional effects of different practices, but existing analyses of these experiments often either fail to adequately consider these issues or lack a clear economic framework for evaluating them. The proposed research seeks to bridge this gap by developing a statistical framework for analyzing distributional impacts in interventions, building on modern welfare economic theory. This will provide a more comprehensive and rigorous approach to evaluation that brings both efficiency and distributional concerns to the analysis. The proposed work will consist of three components. The first component, written in collaboration with Marc Fleurbaey, involves developing a methodology for determining an evaluator’s social preferences based on the tradeoffs between the well-being of different individuals that the evaluator considers acceptable. The second component involves incorporating this social welfare information into a statistical estimation framework, using an egalitarian equivalent representation of the evaluator’s preferences, a concept akin to the certainty equivalent representation in expected utility theory. The third component involves applying this framework to the analysis of optimal treatment rules in randomized controlled trials, using Bayesian, maximin, and minimax regret criteria, and testing the developed methodologies on several well-known trials that exhibit considerable treatment effect heterogeneity. Interventions that exhibit treatment effect heterogeneity can be difficult to evaluate and summarize in terms of welfare. The methods developed in this project provide specific, quantitative guidance for how to do this evaluation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.