Submitted by Caroline Xavier (caroline.xavier@pucrs.br) on 2017-08-25T13:56:44Z No. of bitstreams: 1 TES_ANA_CAROLINA_BORGES_MARQUES_RIBEIRO_COMPLETO.pdf: 888855 bytes, checksum: 33188bddc98bb05c6cb04410adde87e1 (MD5) / Made available in DSpace on 2017-08-25T13:56:44Z (GMT). No. of bitstreams: 1 TES_ANA_CAROLINA_BORGES_MARQUES_RIBEIRO_COMPLETO.pdf: 888855 bytes, checksum: 33188bddc98bb05c6cb04410adde87e1 (MD5) Previous issue date: 2017-07-11 / Coordena??o de Aperfei?oamento de Pessoal de N?vel Superior - CAPES / This thesis presents three independent essays in Economics of Migration. At first, the aim is to measure the impact of an international migration experience on wages in the Brazilian labor market in 2010. The estimation becomes more complex due to the various potential selection biases involved. Compare returned migrants to nonmigrant individuals without regard the selection processes involved can generate biased estimates of the impact of the migration experience on wages. The main contribution of this study is to estimate the wage premium for returnees to Brazil considering a triple selection process. I estimated a model of simultaneous equations considering the occupation decision, the decision to participate in the labor market, and the decision to return to the country since the individual had previously emigrated. Even after controlling for potential biases, the results indicate that there is a positive and statistically significant wage premium for the migration experience on wages, indicating that return migration generates benefits to the country by bringing individuals with characteristics valued in the labor market of the country, and, therefore, this population can be a channel of increase of human capital and productivity in Brazil. The second essay analyzes how the human capital and the income of the spouses can be associated to the family migration in Brazil. We used microdata from the 2010 Census to estimate a logit model, where three groups of families are investigated: non-migrants, in-state migrants, and interstate migrants. The causal effect of the variables was not analyzed, but only its association with the probability of family migration. The contribution of the study is empirical; the results show that the fact that the spouses have higher education favors the propensity to migrate, but the employment relationship of the tied spouse (the one whose wage variation does not determine the migration), usually the woman, has a negative association with the probability of migration. When analyzing the income variables, the evidence shows that both a greater income dispersion between the spouses and a higher sum of the spouses' incomes have a positive association with the probability of migration. When comparing only migrant families, the results indicate that the probability of interstate migration is not associated with a higher educational level of the spouses. Finally, the third essay analyzes the role of migration networks in migratory flows in Brazil for different levels of schooling. A network of migrants can be seen as a social network that attracts new migrants because of the network's informational and financial support that can reduce migration costs and facilitate new migration flows. As migration costs are relatively higher for low-skilled individuals, one hypothesis is that migration networks reduce the qualification levels of new migration flows, making negative self-selection more likely. The objective of this study is to analyze the role of migration networks in migratory flows within Brazil. The results show a positive and significant impact of migrant networks on migratory flows of different levels of schooling. In addition, the selection of migrants is also influenced by migrant networks, which particularly attract new low-skilled migrants. The results are robust to various econometric specifications, including the treatment of selection bias and endogeneity. / Esta tese apresenta tr?s ensaios independentes em Economia da Migra??o. O objetivo do primeiro ensaio ? mensurar o impacto da experi?ncia de migra??o internacional sobre sal?rios no mercado de trabalho brasileiro em 2010. A estima??o se torna mais complexa devido aos v?rios potenciais vieses de sele??o envolvidos. Comparar emigrantes retornados a indiv?duos n?o migrantes sem levar em conta os processos de sele??o envolvidos pode gerar estimativas distorcidas do impacto da experi?ncia de migra??o sobre sal?rios. A principal contribui??o deste estudo ? estimar o pr?mio salarial para retornados ao Brasil considerando um triplo processo de sele??o. Foi estimado um modelo de equa??es simult?neas considerando a decis?o de ocupa??o, a decis?o de participa??o no mercado de trabalho, e a decis?o de retorno ao pa?s dado que, o indiv?duo emigrou anteriormente. Mesmo ap?s controlar os potenciais vieses, os resultados mostram que h? um pr?mio salarial positivo e estatisticamente significante da experi?ncia de migra??o sobre sal?rios, indicando que a migra??o de retorno gera benef?cios ao pa?s ao trazer indiv?duos com caracter?sticas valorizadas no mercado de trabalho local, e, portanto, esta popula??o pode ser um canal de aumento de capital humano e produtividade no Brasil. O segundo ensaio analisa como a renda e o capital humano dos c?njuges podem estar associados ? migra??o familiar no Brasil. Foram utilizados microdados do Censo de 2010 para estimar um modelo logit, onde se investiga a migra??o de tr?s grupos de fam?lias: as n?o migrantes, as migrantes intra-estaduais e as migrantes interestaduais. N?o foi analisado o efeito causal das vari?veis, mas apenas sua associa??o com a probabilidade de migra??o familiar. A contribui??o do estudo ? emp?rica; os resultados mostram que o fato dos c?njuges possu?rem ensino superior favorece a propens?o a migrar, mas o v?nculo laboral do c?njuge vinculado (aquele cuja varia??o salarial n?o determina a migra??o), em geral a mulher, tem uma associa??o negativa com a probabilidade de migra??o. Quando se analisa as vari?veis de renda, as evid?ncias mostram que tanto uma maior dispers?o de renda entre os c?njuges, quanto uma maior soma das rendas dos c?njuges, tem uma associa??o positiva com a probabilidade de migra??o. Quando s?o comparadas apenas as fam?lias migrantes, os resultados indicam que a probabilidade de migra??o interestadual n?o est? associada a um maior n?vel educacional dos c?njuges. Por fim, o terceiro ensaio apresenta uma an?lise do papel das redes de migra??o nos fluxos migrat?rios no Brasil para diferentes n?veis de escolaridade. Uma rede de migrantes pode ser vista como uma rede social que atrai novos migrantes. O suporte informacional e financeiro desempenhado pela rede pode diminuir os custos de migra??o e facilitar novos fluxos migrat?rios. Como os custos de migra??o s?o relativamente maiores para os indiv?duos de baixa qualifica??o, uma hip?tese ? que as redes de migra??o reduzem os n?veis de qualifica??o dos novos fluxos migrat?rios, tornando uma autossele??o negativa mais prov?vel. O objetivo do estudo ? analisar o papel das redes de migra??o nos fluxos e na sele??o de migrantes dentro do Brasil. Os resultados mostram um impacto positivo e significante das redes de migrantes nos fluxos migrat?rios de diferentes n?veis de escolaridade. Al?m disso, a sele??o de migrantes tamb?m ? influenciada pelas redes de migrantes, que particularmente atraem novos migrantes de pouca qualifica??o. Os resultados s?o robustos a diversas especifica??es econom?tricas, incluindo o tratamento de vi?s de sele??o e endogeneidade.