Discovering migrant types through cluster analysis: changes in the Mexico-U.S. streams from 1970 to 2000

Filiz Garip, Harvard University

This paper explores types of migrants from Mexico to the United States in the period 1970-2000. Prior research treats migrants as a homogenous group, assuming they respond to changes in the migration context in the same way. This paper argues that variations in the social, economic and political context of sending and receiving regions create different conditions for migrating. These conditions are heightened or lessened by migrants' demographic characteristics and family networks. These elements, altogether, help identify different types of migrants. A cluster analysis, informed by theories of migration, finds five distinct types of migrants from Mexico to the U.S: network migrants, income-maximizing migrants, risk-diversifying migrants, push migrants, and pull migrants. The relative presence of each migrant type follows a clear time pattern. Migrant types also seem to influence several outcomes (legal or illegal entry, subsequent trips, length of stay). These results suggest that different theories about why individuals migrate may each be correct in different contexts.