Estimating the Effect of Vocational Training on Unemployment Duration in West Germany: A Discrete Hazard Rate Model with Instrumental Variables.

This paper is concerned with the effect of vocational training on individual unemployment duration in West Germany. The database used in this study is the German Socio-Economic Panel (GSOEP) for West Germany for the period from 1984 to 1994. The econometric model is a semi-parametric single risk discrete time duration model. To take into account a possible selection bias, an instrumental variable for actual participation in vocational training is used. This instrumental variable is the predicted participation propensity obtained by estimating a probit model for the participation in vocational training measures. The main results show that prior participation does have a significant effect on the reduction of unemployment duration in the short run. However, this effect does not persist in the long run.