Ethnic Segregation in Friendship Networks : Studies of its Determinants in English, German, Dutch, and Swedish School Classes

Adolescent interethnic friendship is an important indicator of social cohesion in multi-ethnic societies. Therefore, this dissertation examines individual, network, and contextual explanations for ethnic segregation in adolescent friendship networks in school classes. More specifically, the dissertation includes studies on (1) cultural and socioeconomic boundaries, (2) parents, (3) neighborhood contexts, and (4) organizational characteristics of schools in relation to the degree of ethnic segregation in friendship networks, while taking the opportunities for interethnic friendship into account. In addition, (5) the factors associated with successful interethnic friendship maintenance over time are addressed. Hypotheses are tested on the ‘Children of Immigrants Longitudinal Survey in Four European Countries’ (CILS4EU) dataset. This is a large and cross-national sample of ~18,000 students from England, Germany, the Netherlands, and Sweden in over 900 classes. The common belief is that ethnic homophily (i.e., a preference for same-ethnic friends) is the result of differences in cultural attitudes and socioeconomic inequality. However, we find no empirical grounds for this belief, as ethnic groups within school classes often do not differ significantly in the opinions and socioeconomic status indicators under study. In addition, we cannot confirm that residential segregation explains the incidence of ethnic segregation in school settings, most likely because the level of ethnic partitioning across neighborhoods is relatively low. Should residential segregation grow more common in the countries under study, then negative effects on interethnic friendship can be expected because adolescents tend to befriend peers that live nearby (who would be same-ethnic) instead of peers that live further away (who would be interethnic). Instead, ethnic homophily develops by other means. First, we find that ethnic homophily develops intergenerationally through parents. The incidence of interethnic friendships depends on how strongly their parents are committed to preserve an ethnic ingroup identity and how many interethnic friends they have themselves. Second, we demonstrate how the ethnic composition of schools affects ethnic homophily. In-depth analyses of the school’s ethnic composition suggest that immigrants cluster as soon as they can find suitable friends among their same-ethnic peers, whereas native homophily is relatively low and only increases significantly when the outgroup is unified (i.e., when immigrants befriend only amongst each other or originate from one single country instead of various countries). As such, we reveal that natives and immigrants respond differently to diversification efforts, and provide schools with a clearer understanding of why one policy can have different responses. Lastly, the results suggest that interethnic friendships are more likely to break apart in comparison to same-ethnic friendships because of lower friendship quality. Interethnic friendships that have more social support in the form of shared friends and tolerant popular peers in class have a lower dissolution risk than interethnic friendships that are isolated or experience negative interethnic attitudes from popular peers