Refugees’ and irregular migrants’ self-selection into Europe

Cevat Giray Aksoy and Panu Poutvaara

Journal of Development Economics, Volume 152


This paper assesses the self-selection—in terms of education and predicted income—of refugees and irregular migrants from African and Asian countries who arrived in Europe in 2015 or 2016. The authors theorize that the decision to migrate is a function of an individual’s: educational attainment; gender; wage in their country of origin and predicted earnings in the destination country; loss of income and risks associated with conflict or persecution in their country of origin; and costs and risks associated with the journey to the destination country. The analysis is disaggregated by the level of conflict intensity (battle deaths) in origin countries over the sample period, as follows: Afghanistan, Iraq, Nigeria, Pakistan, Sudan, and Syria are categorized as major conflict countries; Algeria and Iran are classified as minor conflict countries; and Bangladesh, Cameroon, Cote d’Ivoire, Morocco, and Senegal are categorized as ‘no conflict’ countries.

The authors predict that:

  • If returns to skills are higher in the country of origin and the country of origin is relatively safe, then migrants tend to be negatively self-selected, i.e. they tend to have lower human capital.
  • If, however, the country of origin suffers from severe conflict, then migrants tend to be positively self-selected, i.e. they tend to have higher human capital. This is because a sufficiently high risk of staying in an unsafe country lowers expected returns to human capital.
  • Given that there is more gender discrimination in African and Asian countries than in European destination countries, it is likely that even though average returns to human capital among men are higher in origin countries than in Europe, the reverse is the case for women, i.e. while the self-selection of men is expected to be negative from countries with low risk of conflict or persecution and positive from countries with a high risk of conflict or persecution, the self-selection of women is expected to be positive from both low-risk and high-risk countries.
  • Negative self-selection of male irregular migrants is stronger from more inegalitarian countries of origin.


The authors estimate a series of multivariate regression models to test these predictions. The analysis is based on data from: Flow Monitoring Surveys, 2015 and 2016 (waves 1–3), which include rich data on the profile, motivations, experiences, and intentions of migrants arriving in Europe in 2015 and 2016; the IAB-BAMF-SOEP Survey of Refugees in Germany, which includes data on refugees’ country of origin, demographic characteristics, and employment histories; 2009-2014 Gallup World Polls (GWP), which cover demographic and socioeconomic data on populations in origin countries; the Uppsala Conflict Data Program (UCDP), which provides information on conflict intensity in origin countries; and the World Bank’s World Development Indicators (WDI). The authors restrict their analysis to individuals aged between 25 and 64.

Main findings:

  • Refugees, both male and female, who arrived in Europe are positively self-selected in terms of education and predicted income. This suggests that if the risk of being a victim of conflict or persecution increases, the probability of emigration becomes eventually increasing in human capital even if returns to human capital would be higher in the country of origin in the absence of conflict or persecution.
  • Male irregular migrants who arrives in Europe are negatively self-selected in terms of education, and do not differ much from non-migrants in origin countries in terms of their income distribution. The negative self-selection of male irregular migrants compared with non-migrants in origin countries is driven by fewer migrants with secondary education. This result reflects the much wider income differences in African and Asian countries compared to European countries.
  • Female irregular migrants who arrived in Europe are positively self-selected in terms of their education and predicted income. This result may reflect: (1) more pervasive gender discrimination in Africa and Asia than in Europe, which combined with low female labor force participation rates in most origin countries, can lower expected returns to education for women below expected returns in Europe; or (2) a subset of women could be escaping gender-based violence and repression even from “safe” countries—producing a self-selection pattern that is similar to refugees who escape conflict and persecution.
  • These patterns hold whether analyzing individual responses to the “main reason to emigrate” in FMS, when comparing migrants from major conflict countries (most likely refugees) with migrants from minor or no conflict countries, and when conducting the analysis at the sub-regional level.
  • Even male irregular migrants from the poorest major conflict countries are negatively self-selected. This result suggests that liquidity constraints are not driving differences in self-selection.

The authors conclude that the positive self-selection of refugees benefits host countries, especially if they support the integration of refugees into the labor market (e.g. through language training and job assistance programs), but represents a human capital drain for origin countries that undermines their post-conflict reconstruction and economic growth.