Title:Treatment Effect Models with Strategic Interaction in Treatment Decisions

Abstract:This study considers treatment effect models in which others' treatment decisions can affect both one's own treatment and outcome. Focusing on the case of two-player interactions, we formulate treatment decision behavior as a complete information game with multiple equilibria. Using a latent index framework and assuming a stochastic equilibrium selection, we prove that the marginal treatment effect from one's own treatment and that from the partner are identifiable on the conditional supports of certain threshold variables determined through the game model. Based on our constructive identification results, we propose a two-step semiparametric procedure for estimating the marginal treatment effects using series approximation. We show that the proposed estimator is uniformly consistent and asymptotically normally distributed. As an empirical illustration, we investigate the impacts of risky behaviors on adolescents' academic performance.