Public Policy 821 is intended as a complement (and not a substitute) for Applied Econometrics 675 and similar courses in other schools and departments. Students should have basic proficiency in STATA and be able to learn some commands on their own. (An example is the "simulate" command). Topics may include "The Non-Central Role of the Propensity Score in Observational Studies", "average treatment effect" estimators(ATE), the metaphysics of Bayesian versus non-Bayesian views of the DGP, Monte Carlo simulation, Varieties of "treatment effect heterogeneity", "decomposition methods", non-parametric and semi-parametric estimation, density estimation, going from local ATE (LATE) to ATE and the Regression Discontinuity Design. Other topics will be chosen by the students who will be expected to collaborate with me and fellow students on presenting the material. (Potential examples include, Heckman-style sample selection, multinomial logits, etc.) Prerequisite: Economics 671 and/or Economics 675 and/or Political Science 699 and/or Political Science 787 or the equivalent. Possibly Econ 672 or equivalent with instructor permission.