Performance and decisions: immigrant–native gaps in educational transitions in Italy

Abstract

Following the seminal work of Boudon [5], sociological research has conceptualized immigrant–native gaps in educational transitions as deriving from children of immigrants' poorer academic performance (primary effects) and from different decision models existing between native and immigrant families (secondary effects). The limited evidence on immigrant–native gaps in Europe indicates that secondary effects are generally positive: children of immigrants tend to make more ambitious educational choices than natives with the same prior performance. In this paper we review the different decomposition methods employed so far in the literature to tackle similar research questions, and extend the existing methodology to allow including interaction effects and taking explanatory variables under control. We apply this method to data coming from a unique Italian administrative data set. We find that children of immigrants exhibit higher likelihood to opt for vocational training over more generalist and academic programs, even when controlling for socio-economic background. A large share of the immigrant–native differentials in the probability to attend the different school programs is explained by the different prior performance distribution. However, decision models differ between groups, and, contrary to the evidence on other countries, these differences contribute to widening the existing gaps. If children of immigrants had the same social background and prior performance of their native peers, they still would be more likely to enroll in shorter and less-demanding school programs. Interestingly, these results hold true only for boys, while we find no evidence of decision effects for girls.

Keywords:

• educational choices
• children of immigrants
• multinomial logit models
• decomposition methods
• socio-economic background

Acknowledgements

Data from the Anagrafe Unica degli Studenti della Provincia di Trento (AUS-PAT) are available thanks to the support of Fondazione Caritro. The authors are especially grateful to Anna Ress for making the data available to this study.

Disclosure statement

No potential conflict of interest was reported by the authors.