Many empirical questions in economics require estimating causal parameters. Regression models provide correlations which only have a causal interpretation if the zero conditional mean assumption holds. This assumption is often violated, for example when there are omitted variables, non-random sampling, reversed causality or measurement errors in regressors. In this course we discuss methods dealing with these confounding factors. In particular, we consider limited dependent variable models, instrumental variables estimation, panel data models and sample weighting. In this course, we introduce the potential outcomes model, which is the most general model for defining treatment effects such as average treatment effect, average treatment effect on the treated, quantile treatment effects and local average treatment effects. The emphasis of the course is on identification, estimation and interpretation rather than a thorough treatment of the asymptotic properties of the estimators. During the course applications of the different methods are discussed, mainly in the fields of labor economics, health economics, and the economics of education.
Cameron, A.C. and P. Trivedi (2005). Microeconometrics: Methods and applications, Cambridge University Press.
Verbeek, M. A Guide to Modern Econometrics, Wiley and Wooldridge, J.M. Econometric Analysis of Cross Section and Panel Data.