Emigrant or sojourner? The determinants of Mexican labor migration strategies to the United States

This dissertation examines migration behavior with a focus on male labor migrants from Mexico to the United States. I develop the concept of migration intensity, defined as the degree to which a migrant shifts his attachment, association and engagement from the place of origin to the migration destination. Using data for male Mexican migrants in the years 1950 to 2005, I find strong complementarities among remittances, migration patterns, and investment decisions, allowing me to derive an Index of Migration Intensity (IMI). The IMI shows that male Mexican migrants use a continuum of labor migration strategies.;Augmenting a conventional Harris-Todaro model, I develop a simultaneous model for the initial migration, return, repeat migration, and remittance decisions of migrant workers. This model can incorporate various migration strategies, including "circular migration," "target earning," and "emigration." Modeling the effects of immigration policies, I find that stricter border enforcement has two consequences: an intended deterrence effect, and an unintended intensification effect whereby stricter border controls lead migrants to make fewer return trips, prolong total U.S. time, and reduce remittances. The impact of the latter on origin-country incomes may induce others to migrate as well.;I then examine the determinants of Mexicans' propensity to migrate illegally (extensive migration behavior) and their degree of socio-economic detachment from home (intensive migration behavior), using instrumental variables estimation with individual fixed effects. My findings support the hypotheses that stricter U.S. border enforcement leads to higher migration intensity, which in turn leads to a net increase in the volume of illegal Mexican migration. My results also indicate that reducing the U.S.---Mexican wage gap would curtail both the extent of illegal migration and migration intensity.;The dissertation also investigates the significance of social networks in facilitating undocumented Mexican migration to the U.S. I argue that the importance of social network assistance arises from problems of asymmetric information. Drawing on secondary data sets as well as field research, I quantify the extent of social network assistance, disaggregated by type of assistance and helper.