The Economic literature which studies migration has always been concerned about its impact on welfare. Two different lines of research in this field focus on impacts of immigration and brain drain. The two articles which comprise the thesis are related to these subjects. Although Roy s model claims that emigrants are negatively self- selected if the rate of return is higher in the origin economy, empirical works have found positively selected emigrants. The first article uses a model to argue that both investments in education and the decision to emmigrate depend on wealth if credit markets are imperfect. This argument allows us to explain the controversy between the theoretical and empirical literature as well as why the middle-class is the most mobile one in some countries. The second part of the thesis is directly related to the beneficial brain drain literature. Works in this field claim that the possibility for an educated worker of emmigrating to another country in which skilled labor is better paid raises the rate of return to education in the origin country and consequently the investments in human capital. The article uses as an experiment the creation of Palmas, a state capital in Brazil, to investigate this hypothesis. The empirical results obtained from microdata evidence a negative relation between investments in human capital and the distance to the capital-used as a proxy to emmigration costs-in the period after the creation of the capital and no relation in the period before. Those findings are interpretated as favorable to the brain effect hypothesis, once the increase in education was greater for individuals who benefited the most from the foundation of the capital.