Mexican migration to the United States: Theory and evidence

Studies on Mexican migration to the U.S. differ as to whether immigrants are drawn from the bottom or the top of Mexico's education distribution. The difficulty is explained by the use of non-representative data of the Mexican population and by the use of inappropriate econometric techniques. The first chapter of the dissertation obtains a representative sample of Mexican males age 25 to 58 in the United States and Mexico. It uses consistent and efficient methods to address the choice-based sampling problem inherent in obtaining the sample. The results show Mexican immigrants to the U.S. are drawn disproportionately from the middle-to-upper end of the Mexican education distribution.; This finding contradicts the standard migration model which predicts that, because Mexico has more wage inequality and higher returns to education, low-skill and low-education workers are more likely to migrate. The second chapter develops and estimates a model with three potential explanations: decreasing migration costs in education, different relations between education and employment in the two countries, and differential availability of informal social insurance. The results reveal that increases in the employment probability of the U.S. and decreases in Mexico raise the migration probability, especially for the more educated. However, the greater migration of high education workers is driven primarily by the fact that low-education workers suffer a much greater loss of informal social insurance when they migrate to the U.S.; Two apparently contradictory theories explain return migration. Under disappointment theory, those who fail to meet their expectations return. Under target saving theory, those who save enough fiends return, which suggests people who fail are the ones who stay. The third chapter presents a model with savings, and strong preferences for home consumption that encompasses both mechanisms. Calibration results suggest that return migration can shape entirely the nature of selection of individuals, depending on whether we look at U.S.-averse individuals or U.S.-lovers. The model is use to examine increases in migration costs, which will reduce migration of the young but may increase the number of permanent settlers.