"Short-term training measures are the most important intervention of German active labor market policy in terms of persons promoted. However, evidence on the impacts of programs is missing. This study analyzes the effects of these programs on the individual unemployment duration in West Germany. By applying a multivariate mixed proportional hazards model, we are able to consider information of the timing of treatment in the unemployment spell as well as observable and unobservable factors to control for selectivity. Moreover, we allow treatment effects to vary over time and take account of heterogeneity in the effects due to individual differences."

Support: Digital