"The German dual apprenticeship system has come under pressure in recent years because enterprises have not been willing to provide a sufficient number of apprenticeship positions. An argument that is frequently put forward is that the gap could be closed if more firms were willing to incur net costs during the training period. This paper investigates on the basis of representative data whether German enterprises do indeed incur net costs on average during the apprenticeship period, i.e. whether the impact of an increase in the share of apprentices on contemporary profits is negative. The paper uses the representative linked employer-employee panel data of the IAB (LIAB) and takes into account possible endogeneity of training intensity and unobserved heterogeneity in the profit estimation by employing panel system GMM methods. An increase in the share of apprentices has no effect on profits. This can be interpreted as a first indication that most establishments in Germany do not invest more in apprentices than their productivity effects during the apprenticeship period."

Support: Digital