Three Essays on Matching Estimators

This dissertation is a collection of three stand-alone research papers. Thereby, the class of local polynomial matching estimators is the central object of investigation. The first essay concentrates on applying local polynomial matching methods in order to account for missing data when estimating parametric conditional mean functions. The second and the third essay provide a closer link at the nonparametric regression literature by developing data-driven smoothing parameter selection algorithms for the matching estimator.