The causal effect of off-campus work on time to degree

In this paper we analyze the effect of outside university work on time to first degree at German universities. The database is the 'Absolventenpanel' 2001, a panel study conducted by the 'Hochschul-Informations-System'. Aiming to estimate the causal effect correctly, we apply a matching strategy based on the approach put forward by Rosenbaum and Rubin [( 1983) "The Central Role of Propensity Score in Observational Studies for Causal Effects." Biometrika 70: 41-55]. The results of the matching approach reveal that simple prima facie results are upward biased but confirm that off-campus work has a prolonging effect on study duration.