ABSTRACT
Career mobility theory suggests that given a certain occupation, schooling improves upward mobility in terms of promotion and wage growth. We are the first to test the implications of this theory for over- and undereducation by means of direct information about promotions to managerial positions. Using German administrative data entailing an employer-reported – and hence objective – measure of educational requirements, we show that overeducated workers are indeed more likely to be promoted and that this career mobility advantage is more pronounced in the early stages of their working lives. By contrast, undereducated workers are less likely to be promoted to managerial positions. Moreover, in terms of wage growth, while overeducated workers benefit more, undereducated workers benefit less from promotions than their well-matched educational peers. Altogether, these findings strongly support the career mobility theory. Furthermore, by differentiating between internal and external promotions, we provide evidence that promotions are more likely for overeducated workers within the establishment, whereas the opposite applies for undereducated workers. This finding indicates the relevance of both over- and undereducation as signals of true ability to other employers.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 Most of the U.S. evidence on the career mobility hypothesis relies on data from the Panel Study of Income Dynamics (the one exception being the Current Population Survey data used by Rubb Citation2006).
2 Other studies examined the relationship between educational mismatch and promotion opportunities (e.g. Korpi and Tåhlin Citation2009; Pecoraro Citation2014), the latter outcome being based on expectations and not on actual events.
3 Groeneveld and Hartog consider 14 job scales, ranked from unskilled (1) to academic, management (14).
4 Consequently, we do not observe those who are self-employed or civil servants. As promotions are irrelevant for the self-employed, this should not affect the validity of our results.
5 The social security reports are due at least once a year. We simply argue that it is more likely that the employer forgets to update the occupational classification when a worker stays with the current establishment than when a worker changes establishments, where a new occupational code has to be set for this worker.
6 By construction, any unsystematic overrepresentation of an outcome category is purged by the relative risk ratios we use in our estimations.
7 These four educational schooling states are no educational degree, secondary school leaving certificate, high school diploma and unknown educational degree.
8 These five vocational training states are no vocational degree, apprenticeship training certificate, Meistermeister/technician, university/college degree and doctoral degree.
9 By construction, for workers who do not change employers during our study period, the initial and destination establishments are the same, just at different points in time.
10 We draw information about unemployment benefits receipts from the benefit recipient histories of the IAB.
11 As we only observe information about the destination establishment for workers who remain employed, the specification in Panel B only controls for workplace heterogeneity of the initial establishment.