We build on earlier work by Baert and Cockx (2013) that explored second generation-native gaps in educational attainment and school-to-work transitions in Flanders, Belgium. They  innovated by taking into account delays in educational attainment, by identifying the moments at which the residual immigrant-native gaps arise, by disentangling the decision to continue schooling at the end of a school year from the achievement within a particular grade, and by integrating the language spoken at home among observed family endowments. They found gaps in educational transitions are small if delays are neglected, but large if not and for school-to-work transitions. 

New analyses focus on how native-immigrant differences in educational trajectories and school-to-work transitions vary by gender, specifically for Moroccan and Turkish second-generation youths who are substantially more disadvantaged than natives and other immigrants.

Adjusting for family background and educational sorting, unlike their male co-ethnics, Turkish and Moroccan second-generation women lag native women in educational attainment, ignoring delay. However, both male and female second-generation Turks and Moroccans are substantially less likely to finish secondary education without delay, or begin tertiary education without delay, than their native counterparts, though the female gap is much larger. Residual gaps in the transition from school to work are also large for both male and female Turkish and Moroccan second-generation youths, though again the female gap is larger.  

The literature has interpreted larger gaps among females than males as evidence of the importance of culture (e.g., traditional gender-role norms), rather than discrimination, since anti-immigrant discrimination is thought to be stronger against males than females. To explore cultural hypotheses more directly, we study whether gaps in demographic behaviors such as earlier marriage among Moroccan and Turkish immigrants are consistent with gender differences in native-immigrant gaps in schooling and the labor market.  

While unadjusted ethnic gaps in demographic behaivors at age 23 are large, after controlling for socioeconomic background, ethnic gaps among females are small and not statistically significant, except for much higher rates of marriage among second-generation Turkish and Moroccans. For men, the pattern of gaps is similar, except that, after SES controls, Turkish and Moroccan men are less likely than natives to have left the parental home and more likely to have married. Proportional hazard models show that, after SES controls, the first-marriage hazard for female Moroccan or Turkish immigrants greatly exceeds that of natives. But there is no evidence of significant or meaningful adjusted differences in first-births or in cohabitation (marital or non-marital).  Thus, the demographic evidence for cultural explanations of ethnic gaps in schooling and the labor market is mixed.