Reliable Estimators for Propensity Score Matching Estimation

Research Seminars

We investigate the finite sample properties of a large number of estimators that are suitable when a selection-on-observables identification strategy is plausible, like inverse probability weighting, kernel and other variants of matching, as well as different parametric models. We also investigate the various tuning parameters that are used in connection with many of those methods. To avoid an arbitrary design dependence on the results, the Monte Carlo simulations are based on real data used for the evaluation of labour market programmes. We vary several dimensions of the design that are of practical importance, like sample size, outcome variable, and the selection process. We find that trimming individual observations with too much weight as well as the choice of tuning parameters is important for all estimators. The key conclusion from our simulations is that the choice of different classes of estimators is less important than optimally tuning estimators within a chosen class.

People

Directions

Address

maps

Click the button below to reload the content. (I agree to external content being displayed to me. Read more in our privacy policy).

,
  • Room Heinz König Hall