Self-selection and migrants\u27 destination choice: A study of Puerto Ricans in the United States and Puerto Rico

The long-standing claim that migrating individuals are uniformly positively selected (i.e. possessing labor market skills of higher economic value than those who do not migrate) has been challenged in recent years. Research suggesting that some immigrants to the United States possess below average skills has generated debate over both the nature of migrant self-selection as well as the impact of immigration policy on the stock of immigrant flows. This debate has raised important theoretical questions about the relationship between the attributes of sending and receiving areas and the characteristics of migrants. This dissertation addresses the selectivity debate using a conditional logistic regression model of migrants\u27 location choice. The model identifies the individual characteristics and location attributes that determine location choice and provides a unique approach to the issue of selectivity. The analysis tests two theoretical perspectives that have figured prominently in the economic and sociological literature on migration. The data are taken from the 1980 and 1990 U.S. Census Public Use Microdata Samples on Puerto Ricans living in the United States and Puerto Rico. A review of the growth and geographic dispersal of migration originating from Puerto Rico suggest a notable amount of human capital selectivity across migrant destinations. Recent data on metropolitan populations of Puerto Ricans in the United States and Puerto Rico reveal that the migratory process has lead to wide variation in the characteristics of individuals at different locations. The location choice analysis finds support for both theories of migrant self-selection. The results indicate that more competitive labor markets encourage positive selection and that larger migratory social networks encourage negative selection. This research makes important methodological and theoretical contributions to the literature. The analytical strategy taken in this study represents an original approach to the question of migrant self-selection. The location choice model provides the first simultaneous assessment of multiple causes of selectivity. The conceptualization of self-selection as the interaction between individual characteristics and location attributes is also innovative. The policy implications of this research are far-reaching. The results suggest that immigration criteria based on family reunification have a negative impact on immigrant selectivity by reinforcing the operation of migratory social networks. Further, the migration process examined in this analysis suggests that selectivity contributes to socio-economic segregation across geographic space