The effects of family networks on Mexico-United States migration

My dissertation studies the influence of migration networks and other factors on the migration decisions of Mexican households. Migration networks may be broadly defined as any socioeconomic link between the origin and destination of the prospective migrant, but my dissertation primarily focuses on previous U.S. migration experience among the head of household's relatives and the presence of family and friends in the United States.;The dissertation contains three sets of econometric models. The first is a group of logit models that look at the head of household's decision regarding whether to participate in Mexico-U.S. migration. I find that migration networks increase the probability that the head of household migrates to the United States, although other factors such as age and gender exercise a more profound influence.;The second is a set of two-state duration models. I use these models to study the factors affecting the length of migration spells for Mexican heads of households both in the United States and Mexico. My results indicate that migration networks shorten the length of migration spells on both sides of the border and that the head of household's father and siblings play a key role in the head of household' s decisions regarding how quickly to end a Mexican spell of migration.;The third set examines the effect of U.S. migration experience on the earnings of heads of household who return to Mexico. Regressing the log of income on a number of explanatory variables, I find that there are indeed positive returns to such experiences in the Mexican labor market, even after correcting for the endogeneity of income and the amount of migration experience.