Abstract
This thesis assesses the role immigrant school segregation plays for inequalities in education and provides insights into mechanisms that may aid our understanding of social stratification, using primarily Norwegian register data. The main body of the thesis consists of five standalone papers.
The first study investigates whether the share of immigrants at school affects the likelihood of completing the academic track in upper secondary school. The paper’s findings suggest that a higher proportion of immigrant students reduces the likelihood of completion. However, this negative effect seems not to stem from exposure to immigrant peers but rather from other traits
typical for schools with many immigrant students. The main contribution of this study is the conceptual and empirical distinction between immigrant peer effects and the effect of attending schools with a high share of immigrant peers. The results warrant a question of whether the education offered to students in immigrant-dense upper secondary schools is of lower quality than the education provided at other schools.
The second study investigates whether the share of immigrants at lower secondary schools affects student outcomes. Similar to the first study, it distinguishes between immigrant peer effects and the effects of attending schools with immigrant peers. The paper’s main contribution, however, is investigating whether these effects vary across the outcome distributions. A conceptual takeaway from this paper is that conventional mean estimates may mask different effects across the outcome distribution. The findings suggest that low achievers improve their objectively rated national test scores from having immigrant peers. In contrast, high achievers show no clear sign of improvement on objectively rated national tests, yet teachers award them better grades if they attend schools that host many immigrant peers.
The third study investigates whether estimates of peer effects may be composites of several and potentially contradicting peer influences. The results show that children get lower grades if they have high-achieving peers. This peer effect conceals that the presence of high-achieving peers improves the learning environment but simultaneously harms students’ academic selfconfidence and work effort. Our findings demonstrate that the total effect of peers on student outcomes consists of several partly contradicting influences and illustrates that only focusing on the total effects of peers may hinder insights into how peers influence one another.
While papers 1 to 3 concern the mechanisms behind peer and school effects and, at large, the consequences of immigrant school segregation for children’s life chances, paper 4 examines the causes of segregation. Specifically, it assesses whether there is a so-called ‘native flight’ from neighborhoods motivated by parents’ school preferences. The paper’s results indicate that native-origin families systematically move away from schools with high shares of students with non-Western immigrant backgrounds. This ‘native flight’ process likely contributes to increased neighborhood segregation and school segregation.
The fifth and final paper relates to social stratification between children of immigrants and majority children. The paper investigates the surprisingly high educational ambitions among children of immigrants considering their relatively low academic achievements. It documents such ‘immigrant optimism’ in post-secondary education as well, where immigrant descendants choose more prestigious and better-paying fields of study than their majority counterparts. The paper also assesses whether this ‘immigrant optimism’ may be explained by immigrant parents being positively selected on education from their origin-country. Parents’ pre-migration educational status plays a role in immigrant descendants’ academic achievements and upper secondary completion but contributes less to immigrant descendants’ tendency of selecting more prestigious and better-paying fields of study in post-secondary education relative to majority children.