Abstract
About 10% of 1081 adolescent ethnic German immigrants aged between 10 and 20 from Russia and Kazakhstan experienced severe difficulties during acculturation. They felt discriminated against in various domains of their lives and failed at school. Not surprisingly, this group at risk also reported higher rates of delinquent activities than the others. Yet, there were a few young people who seemed to function well even under such bad conditions. The present paper aims at shedding light on the processes behind such positive developmental outcomes in spite of discrimination and school failure. We computed logistic regressions (1 = delinquent, 0 = not delinquent) separately for the risk group and the normative group, with protective factors (personal and family assets) as predictors. Financial assets, language competence, or shorter residence predicted membership in the non-delinquent group for those at risk and not at risk alike. Educational attainment of the father, however, worked differently for the two groups, but opposite to what was expected: Among those at risk, a higher education of the father posed an additional problem, whereas those whose fathers had a lower education often showed less problematic outcomes. Given that there is ample evidence that high parental educational levels relate to positive adolescent outcomes in non-immigrant groups, it seems that intergenerational transmission does not work under conditions of a “glass ceiling”, which may be particularly problematic for immigrant families from higher educational backgrounds.
Notes
1“The impact of social and cultural adaptation of juvenile immigrants from the former Soviet Union in Israel and Germany on delinquency and deviant behaviour”; Principal Investigators in Germany: Rainer K. Silbereisen & Eva Schmitt-Rodermund; funded by the German – Israeli Project Coordination, Federal Ministry of Education and Research, Germany.
2Exp B is the odds ratio corresponding to one unit change of the independent variable. An Exp B value below 1 reveals a relationship between higher levels of the independent variable and a decreased likelihood of delinquency. Exp B above 1 stands for a relationship of higher levels of the independent variable with a higher likelihood of delinquency. Thus, bars pointing to the left reveal protective effects of higher levels of the independent variables (i.e., higher age relates to lower likelihood of delinquency) and bars pointing to the right reveal vulnerability factors (i.e., longer residence is accompanied by a higher likelihood of delinquency).
3Four different structural equation models were computed using Amos: (1) default model with none of the paths fixed; (2) model with fixed paths, all pairs required to be equal; (3) model with the paths of family cohesion required to be equal; and (4) a model with father education required to be equal. Whereas the first two models fitted the data, the other two did not (family cohesion: χ2 = 3.821, df = 1, p = .051; father's education: χ2 = 4.511, df = 1, p = .034).
4Indeed, in none of the first-generation immigrant groups of the present sample nor of the additional groups of immigrants that were studied as controls was there a relationship between fathers' education and fathers' occupational attainment in Germany (ethnic Germans from Russia r = .00; ethnic Germans from Kazakhstan r = .03; ethnic Germans from Poland r = .10; first-generation immigrants from Turkey, former Yugoslavia and other countries r = −.14, all by far non-significant; but second-generation immigrants from Turkey and other countries r = .18, p < .05, and local Germans r = .21, p < .0001). Also, for the most part, fathers' education and college-bound adolescent school track are unrelated in the immigrant groups (ethnic Germans from Russia r = .19, p = .0001; ethnic Germans from Kazakhstan r = .06; ethnic Germans from Poland r = .057; first-generation immigrants from Turkey, former Yugoslavia and other countries r = .16, second-generation immigrants from Turkey and other countries r = .14, with the exception of Russia all non-significant; but local Germans r = .29, p < .0001). The relationships between first-generation fathers' education and the jobs they held in their countries of origin, however, are significant and positive, just as among the fathers in the local German sample.