Do immigrants benefit from selection? Migrant educational selectivity and its association with social networks, skills and health

Recent scholarship suggests that immigrant selectivity – the degree to which immigrants differ from non-migrants in their sending countries – can help us understand their labour market outcomes in the receiving country. The selectivity hypothesis rests on three assumptions: first, that immigrants differ from non-migrants in their observed characteristics, such as education; second, that there is an association between such observed selection and (usually) unobserved characteristics, and third that this association drives positive relationships between observed selection and immigrant outcomes. While there is some evidence for a relationship between the degree of immigrants' selectivity and their children's outcomes, a comprehensive assessment of these assumptions for immigrants' own labour market outcomes remains lacking. We use high-quality, nationally representative data for the UK, with large numbers of immigrants from a wide range of different origins and with a rich set of measures of networks, traits and characteristics, as well as economic outcomes, not typically found in surveys of immigrants. This enables us to conduct a comprehensive assessment of the selectivity hypothesis and its assumptions. We find that immigrants to the UK are on average positively selected on educational attainment. However, counter to theoretical assumptions, educational selection has little association with labour market outcomes: it is not or negatively associated with employment; and it is only associated with pay for those with tertiary qualifications and with occupational position for women. We show that the general lack of economic benefits from selection is consistent with an absence of association between educational selectivity and (typically unobserved) mechanisms assumed to link selection and labour market outcomes: social networks, cognitive and non-cognitive skills, and mental and physical health. We contextualise our findings with heterogeneity analysis by migration regime, sending country characteristics, level of absolute education and location of credential.