Human Capital or Discrimination? Labor Market Entry Disadvantages of Second-Generation Turkish Migrants in Germany
Faculty of Social Sciences Mannheim University Mannheim, Germany
Earlier studies disagree over whether differences in the human capital configuration or employer discrimination explain second-generation migrants? disadvantages when entering German labor markets. While the human capital explanation has been tested extensively, less convincing research explores employer discrimination. Furthermore, past research understood the successful completion of a vocational education as part of the human capital configuration and identified it as the major predictor of a successful transition into the labor market. This disregards, however, that for the most part companies are the providers of access to vocational education in Germany, and hence discrimination may occur when companies make their enrollment decisions for these programs. Importantly, this suggests investigating an earlier time point in the process when discrimination may occur than previous studies have considered. Therefore, using data from the German Socioeconomic Panel Study, I analyze the transition from secondary school into the labor market in two steps: first, the transition into vocational education, and second into employment. The GSOEP allows a comprehensive specification of human capital and testing of corollary hypotheses derived from statistical discrimination and taste discrimination.
Using discrete event history models for access to and completion of vocational education, I find significant and substantial ethnic residuals especially for young Turkish men, even when controlling for receiving country specific capitals. This raises serious doubts in specifying vocational education as part of human capital. For the second part of the process, the actual transition into regular employment, I use hybrid estimation models. Human capital, including receiving country specific resources, such as German language fluency, does not fully explain the ethnic penalties young males with Turkish migration background experience. Finally, interaction models show that a completed vocational education pays off less for Turkish as compared to Germans, again the effect is pronounced for Turkish men. The latter finding is direct evidence for statistical discrimination. Regarding taste discrimination I find no evidence, although, this is conducted through an indirect test. In conclusion, the inclusive human capital measures available in the GSOEP do not fully explain Turkish second-generation migrants? disadvantages; partly it can be attributed to statistical discrimination.