Returns to Company Training - Evidence from a New Approach Using Quasi Experimental Data

René Fahr 1, Sabine Simons3
1University of Paderborn, Germany, 2IZA, Germany, 3Sparkasse KölnBonn, Germany

In the present paper we estimate wage returns to company training using representative individual-level cross-sectional data for the years 2000 and 2003 for Germany. We employ the approach proposed by Leuven and Oosterbeek (Journal of Applied Econometrics 2008) to identify average treatment effects on the treated. While OLS estimates identify a 7.5% wage premium for participants in company training the positive returns vanish when restricting the comparison group to eligible employees who declined participation in training due to some random event. The fact that most of the training serves the purpose of adjustment to new job requirements might explain our findings.

View full paper