This paper measures the quality of apprenticeship training by comparing the wage difference between apprentices who change their employer or occupation directly after the training period with apprentices who stay in the training firm and their occupation. It tests empirically which individual, occupation, and employer characteristics have an impact on these differences. A little wage advantage for those who change their employer after the apprenticeship training is an indication that the German dual apprenticeship training system provides generally usable human capital. There are small average wage losses for occupation changers who do not simultaneously change their employer. In general, a change of occupation does not have a wage impact. These average effects mask large differences by occupation groups, however. For apprentices in commerce and trading occupations, changing the employer is associated with a large positive wage mark-up. Apprentices in industrial occupations suffer from wage disadvantages when they have to change their employer or occupation. These results support the findings that industrial occupations are relatively specific and establishments that offer industrial occupations invest in apprenticeship training, try to keep their apprentices after the training period and offer attractive internal labour markets. Apprentices with an upper secondary education who change their employer get a higher wage mark-up than the stayers – this indicates that better qualified apprentices are privileged because they have the interesting outside option to acquire an academic qualification. The distinction between occupation groups and apprentices with different schooling back-grounds is new. The previous literature mainly concentrates on differences in economic sectors and firm size. We do not find large wage differences for occupation and/or employer changers in East-Germany, between the manufacturing and services sectors or from an apprenticeship training establishment which is smaller than the first skilled employer, however. A series of robustness checks demonstrates that occupation selection, the decision to change the employer and/or occupation, unobserved heterogeneity between apprentices, and the homogeneity of the sample regarded all have a strong impact on the estimation results. This indicates that previous results might be biased.
Göggel, Kathrin and Thomas Zwick (2009), Good Occupation - Bad Occupation? The Quality of Apprenticeship Training, ZEW Discussion Paper No. 09-024, Mannheim, published in: The Scandinavian Journal of Economics.