Sampling international migrants with origin-based snowballing method. New evidence on biases and limitations

Abstract

This paper provides a methodological assessment of the advantages and drawbacks of the origin-based snowballing technique as a reliable method to construct representative samples of international migrants in destination areas. Using data from the MAFE-Senegal Project, our results indicate that this is a very risky method in terms of quantitative success. Besides, it implies some clear selection biases: it overrepresents migrants more strongly connected to their home country, and it tends to overestimate both poverty in households at origin and the influence of previous migration experiences of social networks on individuals’ out-migration.