Who should be Treated? Empirical Welfare Maximization Methods for Treatment Choice
2015/01/06
研討會日期 : 2015-01-06
時間 : 14:30
主講人 : Professor Toru Kitagawa
主持人 : Professor Le-Yu Chen
地點 : Conference Room C103
演講者簡介 : Professor Kitagawa received his Ph.D. in Economics from Brown University in 2009. He is currently an Assistant Professor at Department of Economics, University College London. His research fields are Econometric Theory, Causal Inference, Bayesian Econometrics, and Applied Econometrics.
演講摘要 : One of the main objectives of empirical analysis of experiments and quasi-experiments is to inform policy decisions that determine how treatments are allocated to individuals with different observable covariates. We propose the Empirical Welfare Maximization (EWM) method, which estimates a treatment assignment policy by maximizing the sample analog of average social welfare over a class of candidate treatment policies. We show that, when propensity score is known, the average social welfare attained by EWM rules converges at least at n^{-1/2} rate to the maximum obtainable welfare. This holds uniformly over a minimally constrained class of data distributions, and this uniform convergence rate is minimax optimal. In comparison with this benchmark rate, we examine how the uniform convergence rate of the average welfare improves or deteriorates depending on the richness of the class of candidate decision rules, on the distribution of conditional treatment effects, on the lack of knowledge for the propensity score, and on additional smoothness assumptions for the regression functions or propensity scores. We also discuss practically implementable computation for the EWM rule. As an empirical application, we derive an EWM rule for the a training program using the experiment data analyzed in La Londe (1986).
