Explaining differences in gender role attitudes among migrant and native adolescents in Germany: intergenerational transmission, religiosity, and integration

ABSTRACT

This study examines gender role attitudes of native and migrant adolescents in Germany and attempts to explain why adolescents of Turkish, former Yugoslavian, and Eastern European origin tend to have more traditional attitudes than their native peers. In order to do so, it combines a migrant–native comparative approach that highlights the impact of religiosity and host society integration with an intergenerational transmission perspective that emphasises the continuity of gender role attitudes across generations. The empirical analysis relies on dyadic parent–adolescent data (N = 2744) from the first wave of the Children of Immigrants Longitudinal Survey in Four European Countries. It demonstrates the importance of incorporating intergenerational transmission processes to fully understand attitude differences between natives and migrants: a substantial part of native–migrant gaps in gender role attitudes can be attributed to migrant parents’ more traditional attitudes and a strong transmission of attitudes across generations. Once intergenerational transmission and the influence of religiosity and integration have been accounted for, the remaining differences between gender role attitudes of native and migrant adolescents are small.

KEYWORDS:

• Gender role attitudes
• transmission
• integration
• native–migrant differences
• Germany

Acknowledgements

For helpful comments, I am indebted to Sarah Carol, Julia Ingenfeld, Regina Keller, Zehra Mermer, Paul Schäfer, Lea Waniek, and the participants of the Frühjahrstagung Migration und Familie, Universität Duisburg-Essen, 18–19 February 2016. I further thank the anonymous reviewers for their constructive feedback, which substantially improved the paper.

Disclosure statement

No potential conflict of interest was reported by the author.

ORCID

David Kretschmer http://orcid.org/0000-0002-8702-3007

Notes

1 For example, if transmission effects were only at work in native but not in migrant families, estimating a common transmission effect would overestimate transmission in migrant families. If, additionally, migrant parents were more traditional, this situation would lead to the erroneous conclusion that transmission partially explains native–migrant attitude differences.

2 I focus on the German CILS4EU data and not on the three remaining countries – England, the Netherlands, and Sweden – mainly because of data limitations. In the English and Swedish samples, information on parental characteristics is missing for a large proportion of the sample. In the Dutch sample, on the other hand, there are too few observations per migrant group. Small migrant sample sizes are particularly problematic for an empirical approach that attempts to explain native–migrant gaps in attitudes: for small sample sizes, not detecting significant gaps between groups may be a consequence of low statistical power rather than a result of differences being actually absent.

3 The largest source of attrition is parental dissent to participate in the study (20% of the gross sample). Non-participation is above-average among parents of Turkish (29%) and former Yugoslavian (28%) origin. In general, the children of participating and non-participating parents are similar in their characteristics, though the latter are slightly more traditional and slightly less integrated than the former. However, differences between origin groups (including Germans) among adolescents whose parents do not participate are minor. To assess robustness of the empirical findings to listwise deletion of observations with missing information, I re-estimated the models using multiple imputations with chained equations. Results from these models are substantively identical to those based on listwise deletion, and I present the latter in the main text for consistency with the descriptive analysis.

4 Strictly speaking, an allocation that assigns most of the responsibility for these tasks to the male partner is not modern (in the sense of egalitarian attitudes). However, there are very few respondents in this category and all results are robust both to excluding these respondents from the analysis and to using alternative classifications of response categories for the affected respondents.

5 Principal component analysis also suggests a single-component solution. Furthermore, all substantive results are replicated no matter whether analyses are conducted with the summary score introduced here or with the individual gender-role items as dependent variables. Furthermore, the results are robust to using alternative plausible classifications of response categories.

6 However, analyses that differentiate the sample according to migrant status and religious affiliation rather than country of origin produce comparable results for the impact of intergenerational transmission and the remaining background characteristics.

7 I also investigated the role of family structure in supplementary analyses. Family structure was not associated with adolescents’ attitudes and did not moderate the impact of transmission, either.

8 One reason for this counterintuitive finding may be the way gender-role attitudes are assessed in CILS4EU: As the gender-role items refer to the preferred distribution of tasks within families, parental assessment may be shaped by the experience that many tasks have to be shared between men and women in order to sustain a healthy family life. Adolescents have not yet been involved in comparable negotiation processes such that their attitudes are affected less by differences between the demands of relationship management and ideal conceptions. An alternative explanation may be stronger social desirability bias at the parent level. Furthermore, there may be gender-based selection effects in that those fathers participating in the survey may be particularly egalitarian. Such selection effects may explain the absence of gender differences in parental attitudes, and may lead to an overestimation of modern gender-role attitudes in the parent generation.

9 Additional analyses suggest that parental characteristics also exert indirect effects on adolescents’ attitudes by shaping adolescents’ religiosity and social/structural integration into the host society.

Additional information

Funding

Data collection for CILS4EU was funded by the NORFACE research programme on Migration in Europe - Social, Economic, Cultural and Policy Dynamics.