演講者簡介 : Professor Tong Li received his Ph.D. in Economics from the University of Southern California in 1997. He is currently the Gertrude Conaway Vanderbilt Chair in Social and Natural Sciences and Professor of Economics at Vanderbilt University. His research interests are microeconometrics and empirical microeconomics.
演講摘要 : This paper studies the identification problem for Bayesian games within the private type paradigm when the researcher cannot perfectly know players’ payoff structures. Under the exclusion restriction in the form of an exogenous players’ participation, we show that point identification is not feasible in general but partial identification with sharp nonparametric bounds can be established under weak assumptions. If and only if the distributions of actions intersect with each other as the number of players varies, our partial identification reduces to point identification. Our results can be extended to allow for corner solutions, asymmetric players, unobserved heterogeneity, and endogenous participation. As such, we have presented positive identification results and a general econometric framework for the structural analysis of general Bayesian games.