[pt] Este estudo analisa os efeitos de um programa governamental de educação técnica no Brasil. Explorando o fato de que a seleção para o programa foi aleatorizada, com estudantes em classes com excesso de inscrições selecionados através de uma loteria, nós olhamos para os efeitos de uma oferta para o programa sobre emprego formal, tipo de ocupação e apego ao mercado de trabalho. Nossos resultados não mostram efeitos significantes sobre emprego formal, salários ou apego ao mercado de trabalho após o programa. Por outro lado, o programa melhorou significativamente o match entre estudantes e empregos, com um aumento de 1 p.p. (100 porcento) na probabilidade de um indivíduo estar empregado em uma ocupação relacionada com o curso escolhido. Nós também encontramos evidência de que receber uma oferta para o programa levou a uma maior probabilidade de estar empregado em uma ocupação com alta demanda por habilidade. Quando examinamos como esses resultados variam com o gênero dos aplicantes, encontramos que, enquanto ganhos em ocupações associadas são similares para ambos os grupos, os efeitos sobre empregos de alta habilidade aparecem apenas para homens e que nenhum experimenta melhoras sobre emprego ou salário após o término dos cursos. / [en] This study analyzes the effects of an extensive public technical education program in Brazil. Exploring the fact that acceptance to the program was randomized, with students in over-subscribed classes selected through a lottery, we look at the effects of an offer for the program over formal employment, type of occupation and attachment to the labor market. Our findings show no significant effect over formal employment, earnings or the level of attachment after the program. On the other hand, the program significantly improved the match between students and jobs, with an increase of 1 p.p. (100 percent) in the probability that an individual is employed in an occupation associated with the course chosen. We also find evidence that receiving an offer for the program led to a higher probability of being employed in an occupation with a higher skill demand. When we examine how these results vary by the gender of the applicant, we find that, while gains in associated occupations are similar for both groups, the effects for high skill jobs appear only men and neither experience improvements on employment or wages after the end of the courses.

This thesis consists of four self-contained essays. Essay 1 (with Olof Åslund and Matz Dahlberg): In this essay we investigate the impact of commuter train access on individual labor market outcomes. Our study considers the exogenous introduction of a commuter train linking locations in the northern part of Uppsala County (Sweden) to the regional employment center, considerably decreasing commuting times by public transit to the center for those living close to the pre-existing railroad. Using difference-in-differences matching techniques on comprehensive individual panel data spanning over a decade, our intention-to-treat estimates show that the reform had mainly no impact on the earnings and employment development among the affected individuals. Essay 2: In this essay I look into the role of public transit for residential sorting by studying how the introduction of a commuter train linking locations in the northern part of Uppsala County (Sweden) to the regional employment center affected migration patterns in the areas served. Using a difference-in-difference(-in-difference) approach and comprehensive individual level data, I find that the commuter train had a positive effect on overall in-migration to the areas served and no effect on the average out-migration rate from these areas. With regards to sorting based on labor market status, I find no evidence of sorting based on employment status but some evidence that the train introduction increased the probability of moving out of the areas served for individuals with high labor incomes relative to the probability for individuals with lower income. Considering sorting along other lines than labor market status, the analysis suggests that people born in non-western countries came to be particularly attracted towards the areas served by the commuter train as compared to other similar areas. Essay 3: In this essay I look into the relation between housing mix and social mix in metropolitan Stockholm (Sweden) over the period 1990-2008. Using entropy measures, I find that although the distribution of tenure types over metropolitan Stockholm became somewhat more even over the studied period, people living in different tenure types still to a large extent tended to live in different parts of the city in 2008. The degree of residential segregation was much lower between different population groups. I further find that the mix of family types, and over time also of birth region groups and income groups, was rather different between different tenure types in the same municipality. The mix of different groups however tended to be similar within different tenure types in the same neighborhood. While the entropy measures provide a purely descriptive picture, the findings thus suggest that tenure type mix could be more useful for creating social mix at the municipal level than for creating social mix at the neighborhood level. Essay 4 (with Matz Dahlberg): The last decade’s immigration to western European countries has resulted in a culturally and religiously more diverse population in these countries. This diversification manifests itself in several ways, where one is through new features in the cityscape. Using a quasi-experimental approach, essay 4 examines how one such new feature, public calls to prayer, affects neighborhood dynamics (house prices and migration). The quasi-experiment is based on an unexpected political process that lead way to the first public call to prayer from a mosque in Sweden combined with rich (daily) information on housing sales. While our results indicate that the public calls to prayer increased house prices closer to the mosque, we find no evidence that the public calls to prayer served as a driver of residential segregation between natives and people born abroad around the mosque in question (no significant effects on migration behavior). Our findings are consistent with a story where some people have a willingness to pay for the possibility to more fully exert their religion which puts an upward pressure on housing in the vicinity of a mosque with public calls to prayer.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-02-12T14:16:05Z No. of bitstreams: 1 antoniojosenegreirosaraujo.pdf: 1895732 bytes, checksum: b14b0a28de5965d826a7eb8b2f58867c (MD5) / Rejected by Adriana Oliveira (adriana.oliveira@ufjf.edu.br), reason: Renata, só confirme, por favor, se Enem se escreve assim mesmo porque eu sempre pensei que fosse sigla ENEM. Obrigada. on 2016-02-26T12:16:23Z (GMT) / Submitted by Renata Lopes (renatasil82@gmail.com) on 2016-02-26T12:35:05Z No. of bitstreams: 1 antoniojosenegreirosaraujo.pdf: 1895732 bytes, checksum: b14b0a28de5965d826a7eb8b2f58867c (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2016-03-03T14:01:06Z (GMT) No. of bitstreams: 1 antoniojosenegreirosaraujo.pdf: 1895732 bytes, checksum: b14b0a28de5965d826a7eb8b2f58867c (MD5) / Made available in DSpace on 2016-03-03T14:01:06Z (GMT). No. of bitstreams: 1 antoniojosenegreirosaraujo.pdf: 1895732 bytes, checksum: b14b0a28de5965d826a7eb8b2f58867c (MD5) Previous issue date: 2014-03-26 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Uma das metas do milênio para o Brasil é a melhora dos indicadores educacionais. Embora nas últimas duas décadas o acesso à educação em todos os níveis tenha aumentado, a qualidade desse ensino ainda é considerada baixa segundo as avaliações de desempenho internas e externas. Nesse contexto, insere-se a educação profissional e tecnológica (EPT) de nível médio. Este tipo de formação pode ser realizada integrada ou subsequente ao ensino regular, assumindo, na maioria das vezes, um caráter de transição para as universidades, no primeiro caso, e, para o mercado de trabalho, no segundo. Nessa dissertação, a EPT será caracterizada quanto à composição de gênero, cor, modalidades de ensino, dependência administrativa, unidades da federação, oferta, demanda e número de residentes e não residentes, por meio da base de dados do Censo Escolar de 2007 a 2012. Estima-se, com base no método de pareamento com escore de propensão, o efeito médio da EPT sobre o desempenho escolar e a inserção produtiva para os tratados, ou melhor, para aqueles que optaram pela EPT. O desempenho escolar é medido pela proficiência nas provas de Ciências da Natureza, Ciências Humanas, Linguagem e Códigos, Matemática e Redação presentes na avaliação do ENEM em 2009 e 2010. A inserção produtiva é medida pelas probabilidades de estar trabalhando e de trabalhar na área para a qual se preparou. Essas variáveis foram criadas a partir do questionário socioeconômico do ENEM de 2009 e 2010. Pelo método de pareamento do “vizinho mais próximo”, o efeito médio de realizar EPT para os alunos que realizaram, em relação aos alunos da escola regular, é positivo e significativo nas notas de Linguagens e Códigos e Redação, respectivamente, de 4,48 e 13,44 pontos no escore padronizados. O método de estratificação confirma o efeito médio positivo, respectivamente, de 4,28 e 13,08 . O efeito médio de realizar EPT para aqueles que realizaram sobre a inserção produtiva está entre 1,2 e 1,3 pontos percentuais para a probabilidade de trabalhar e de 17 pontos percentuais para a probabilidade de trabalhar em área para o qual se preparou. Esses resultados referem-se ao modelo completo para 2009. / One of the millennium goals for Brazil is improving educational outcomes. Although the access to education at all levels has increased in the last two decades, the quality of Brazilian education is still considered low by internal and external evaluations. In this context, it is important to discuss the role of the vocational and technical education and training (VTET) of middle level. This type of training can be integrated or made subsequent to regular education. It the first case, it is characterized as a transition to the university, while in the second case, it improves the access to the labor market. In this dissertation, we explore the Censo Escolar database from years 2007 to 2012, and ENEM database, from years 2009-2010, both collected by Inep. This study evaluates the effects of technical and vocational education and training (TVET) on the mid-level proficiency and productive insertion of the student. The results of the estimates with propensity score matching show an average treatment effect between 4 and 13.5 points in the standardized score of Language and Writing, respectively. There are also significant effects on the productive insertion, on average, there is an increase between 1.2 and 1.3 percentage points in the probability of being employed, and 17 percentage points in the probability of being employed in their area of technical education and training, for TVET students comparing to other students.

本文主要探討居住地區劣勢、社會資本與勞動市場表現的問題，主要分成三個部份：臺灣鄉鎮市區劣勢程度的長期變遷，居住地區劣勢、觸及社會資本與勞動市場表現的關係，以及居住地區劣勢、動員社會資本與勞動市場表現的關係，以便瞭解臺灣鄉鎮市區劣勢程度的變遷情形、觸及社會資本與勞動市場表現的關係是否受到居住地區劣勢程度的影響，以及動員社會資本與勞動市場表現的關係是否受到居住地區劣勢程度的影響。 在臺灣鄉鎮市區劣勢程度的長期變遷方面，本文採用1956-2010年等六年鄉鎮市區人口普查資料來建構地區劣勢指標。在地區劣勢指標項目的選擇上，本文選擇了農業從業人口比例、高中以下教育人口比例，以及人口密度等三項指標，並且透過因素分析的最大變異法構成一個名為「地區劣勢」的因子。除了1956年的農業從業人口比例之因素負荷量為最大之外，其餘年份均以高中以下教育程度人口比例的因素負荷量為最大。這樣因素萃取的結果顯示出臺灣地區劣勢的形成有很大的部份是與人口的教育程度有關。另方面，根據GIS所描繪的地圖，可以清楚的看到發展相對優勢的地區主要是集中在北部地區，並且繼續擴展中，而這些地區主要涵蓋隸屬臺北市、新北市、基隆市、桃園市及新竹市的鄉鎮市區。 在居住地區劣勢、觸及社會資本與勞動市場表現的關係上，本文主要採用2004年臺灣地區第一波社會資本調查資料來探討的問題為居住地區劣勢、個人背景與觸及社會資本（擔任現職時）和居住地區劣勢、觸及社會資本（擔任現職時）與勞動市場表現兩項問題。在居住地區劣勢、個人背景與觸及社會資本（擔任現職時）方面，本文採階層線性模型來估計之。估計結果顯示，觸及社會資本（擔任現職時）確實有地區不平等的現象，並且與居住地區劣勢有關，因為當僅放入居住地區劣勢時，居住地區劣勢對觸及社會資本（擔任現職時）有顯著的負面影響，即居住地區劣勢程度越高，觸及社會資本（擔任現職時）越不好，並且截距項的顯著性就消失了，顯示觸及社會資本的不平等是與居住地區劣勢程度有關。然而，當放入個人背景時，個人的年齡、教育程度和工作經驗對觸及社會資本（擔任現職時）有顯著的正面影響，並且居住地區劣勢的顯著性消失了。在居住地區劣勢、觸及社會資本（擔任現職時）與勞動市場表現方面，居住地區劣勢和觸及社會資本（擔任現職時）均對薪資有顯著的正面影響，然而居住地區劣勢和觸及社會資本（擔任現職時）的交互作用項對薪資沒有任何的顯著影響，即觸及社會資本（擔任現職時）對薪資的影響，不因居住地區劣勢程度不同而有所差異。 在居住地區劣勢、觸及社會資本與勞動市場表現的關係上，本文主要採用2004年臺灣地區第一波社會資本調查資料來探討之。按照居住地區劣勢程度，本文將分析樣本分成四份（第一等級、第二等級、第三等級和第四等級），並且發現到居住地區劣勢對觸及社會資本有負面影響，特別是第四等級。另外，本文發現觸及社會資本與動員介紹人、弱聯繫與介紹人資源和動員社會資本與勞動市場表現等關係，均受到居住地區劣勢程度的不同而有差異，可見林南的動員社會資本模型並非一體適用，僅有在特定的鄰里/地區脈絡下，才有成立的可能性。