Search for a command to run...
Although the school performance of children of migrants in the Netherlands continues to improve, it still lags behind compared to that of children of Dutch origin, which contributes to inequality in socio-economic outcomes. To reduce this performance gap it is of great societal importance to gain insight into the factors influencing language proficiency of children of migrants. This paper investigates to what extent the share of people from children’s own origin group in the network – network isolation – influences their Dutch language proficiency and subsequent sorting in educational secondary school tracks. To measure network isolation of children we used a whole-population network from Statistics Netherlands containing family members, household members, neighbours, classmates and colleagues. The score on the language component of the final-year primary school exam is used as an indicator of Dutch language proficiency. Results show that higher levels of network isolation, i.e., a higher share of the own origin group in the network, are associated with lower Dutch language proficiency. Furthermore, higher levels of network isolation are related to lower teacher track recommendations for secondary school, and this relationship is fully mediated by language proficiency. These findings highlight the importance of considering the broader social context —beyond language exposure in the household — when promoting social mobility of children of migrants. ● A whole-population network is used to measure network isolation – the rate of exposure to the own origin group – in children’s networks. ● Higher levels of network isolation are related to lower levels of Dutch language proficiency for children of migrants at primary school. ● In turn, higher isolation is related to lower teacher’s track recommendation for secondary school ● The total effect of network isolation on language proficiency and secondary school tracking is negative
Published in: Research in Social Stratification and Mobility
Volume 103, pp. 101138-101138