Purpose: The paper seeks to analyse the self-selection of international migrants on observable skills. Design/methodology/approach: Based on an extended version of the Roy model, which considers random migration costs, the authors analyse the self-selection of migrants on observable skills empirically. For this purpose, the authors employ a new panel data set on the educational attainment of migrants, which covers migration from 143 sending countries into the six main receiving countries in the OECD from 1975 to 2000. Findings: Migrants tend to be positively self-selected on observable skills, although the inequality in earnings is larger in the sending country relative to the destination countries. The estimation results indicate that a higher inequality in the distribution of earnings in both the receiving and the sending country affects the skill bias of the migrant population favourably. Moreover, higher migration costs and selective immigration policies increase the skill level of migrants relative to those of stayers in the sending countries. Research limitations/implications: The results may be affected by measurement error, since it was necessary to approximate the returns to education by measures for the inequality of earnings. Practical implications: The paper provides, inter alia, insights as to how immigration and other policies affect the self-selection of migrants on observable skills, which may be relevant for policy makers. Originality/value: To the authors' knowledge, this is the first paper to analyse the self-selection of migrants on the basis of a panel data set. © Emerald Group Publishing Limited.

Europe skills Human capital Migrant workers Income

[No Keywords available]