Participation in social programs is often misreported in survey data, complicating the estimation of the effects of those programs. In this paper, we propose a model to estimate treatment effects under endogenous participation and endogenous misreporting. We show that failure to account for...