Why is the Roy-Borjas model unable to predict international migrant selection on education? Evidence from urban and rural Mexico

The Roy-Borjas model predicts that international migrants are less educated than nonmigrants because the returns to education are generally higher in developing (migrant-sending) countries than in developed (migrant-receiving) countries. However, empirical evidence often contradicts this prediction, even when comparing migrants from different regions of the same country. Specifically, whilst Mexicans migrating to the United States from urban areas are negatively selected based on education, those from rural areas are positively selected - despite similar returns in both regions. Leveraging this setting, we demonstrate that the inconsistency between theory and evidence can be resolved by assessing migrants' human capital using a two-dimensional measure of occupational skills rather than educational attainment. Our findings caution against using positive migrant selection on education as evidence to reject the Roy-Borjas framework for explaining international migration.