This paper evaluates the effects of Public Sponsored Training in East Germany in the context of reiterated treatments. Selection bias based on observed characteristics is corrected for by applying kernel matching based on the propensity score. We control for further selection and the presence of...