The impact of earnings gaps and networks on migration decisions: an empirical study of undocumented Mexican migrants

Abstract

This study shows that the effects of earnings differentials, ethnic networks, and herd behavior on migration decisions are more nuanced than the existing literature suggests. Our conclusion stems from the analysis of the geographic and occupational choices of undocumented Mexican migrants to the USA. We apply mixed logit and latent class methods to data from a survey based on interviews with Mexican nationals at border towns inside Mexico. While better-educated migrants favor occupations and places with higher earnings gaps, migrants with little schooling behave the opposite way. We also find that educational attainment is an important driver of network effects. The preferences of less educated migrants with regard to network size show the traditional inverted U-shape pattern, but the behavior of individuals with more education is consistent with a U-shape pattern instead. Another conclusion that stems from our empirical results is that the herd effect can be positive or negative, depending on migrants’ occupational choices.