Mauro Borges Institute - IMB
E-mail: savio.oliveira@goias.gov.br
Federal University of Paraíba
E-mail: savio.oliveira@academico.ufpb.br
Education
M.Sc. in Economics, Federal University of Paraíba.
B.Sc. in Economics, State University of Rio Grande do Norte.
CV Lattes and Research Gate
Field
I have interests in Political Economy and Applied Micro.
Research
Note on the Bandwagon Effect in Brazilian Municipal Elections (with Wallace Farias). First draft
This study documents that the voter's decision may be influenced by his preference to be on the winning side. Using RDD in Brazilian municipal elections to compare candidates who barely won with those who barely lost the first round, we find that winning the first round increases the vote share (3.6 pp) in the second round and increases the probability of winning the election (38.7 pp). These results are robust to different econometric specifications, and competing interpretations, such as differences in campaign spending, incumbency status or being from the same parties as the state governor or president.
Political Dynasties and Corruption: Evidence from Brazil (with Wallace Farias). First draft
In this study, we investigate whether dynastic mayors are more or less likely to engage in corruption when compared to non-dynastic mayors. Using data from random government audits conducted in the Brazilian municipalities, we use RDD on close elections to compare municipalities that barely had a dynastic candidate elected as mayor to those that did not. The identification of relatives in politics is based on the matching of their last names. We find that dynastic mayors are more likely to engage in practices of over-invoicing compared to non-dynastic mayors.
Political Competition and Candidate Selection in Brazilian Municipalities (With Wallace Farias) - Economics Bulletin. Link to paper
We investigate the relationship between political competition and the quality of the candidates, as well as the elected politicians. We use the effective number of candidates for measuring electoral competition, and the level of education for measuring the quality of politicians. By employing an instrumental variable approach, we find that electoral competition improves the quality of the candidates as well as the elected politicians. At the same time, it decreases the participation and the performance of women in politics.
Do Business Candidates Make Better Politicians? New Evidence for brazil. (with Jevuks Araújo and Wallace Farias) - Revista de Economia. Link to paper
Does the profile of those who are elected matter for the type of policy implemented? This study empirically investigates whether mayors in Brazil with prior experience as entrepreneurs achieve different governance outcomes. Using a Regression Discontinuity Design (RDD) applied to closely contested municipal elections, the research compares municipalities where a businessperson candidate narrowly won with those where other candidates narrowly prevailed under similar electoral conditions. The results consistently indicate that mayors with a business background do not differ from others across a range of analyzed variables, encompassing economic development indicators (such as GDP growth), social development metrics (including spending on education and social assistance, and IDEB scores), and municipal health quality (vaccination coverage rates). It is concluded that, for the set of policies and outcomes examined within the Brazilian context, prior business experience did not prove to be a determining factor for distinct governmental performance.
Labor Market Conditions and Incumbent Performance in Brazilian Municipalities (with Wallace Farias)
Legislative Representation, Local Ties and Intergovernmental Transfers: Evidence from Brazil (with Wallace Farias)
Independence of the Brazilian Central Bank and Volatility of Iflation and Unemployment (with Wallace Farias and Jevuks Araujo)
Estatuto do Desarmamento e Suicídios no Brasil: Uma Análise de Intervenção com Séries Temporais (com Cássio Besarria) - aceito em A Economia em Revista.
Uma Análise do Efeito da Mortalidade no Trânsito sobre a Criminalidade Brasileira (com Francisco Soares, Raphael Lucena e Rodolfo Costa) - Revista Econômica. Link para o artigo
Este estudo tem como objetivo verificar a influência da mortalidade no trânsito sobre o fenômeno da criminalidade dos estados brasileiros. A hipótese é de que a violência no trânsito é uma proxy para o comportamento transgressor dos indivíduos. Foi utilizado o estimador system GMM para dados em painel, sendo essa escolha justificada dada a natureza da relação contínua no tempo que o processo da criminalidade possui com seu próprio valor passado. A amostra contempla os 27 estados brasileiros, no intervalo temporal de 1996 a 2017. As elasticidades estimadas apontam que uma elevação na taxa de mortes no trânsito em 15% resulta em aproximadamente 1,2% de aumento na taxa de homicídios. Além disso, os resultados também apontam que a criminalidade diminui com aumentos na renda da população, e se perpetua no tempo através do efeito inércia.