Três ensaios em finanças públicas

• Ano de defesa: 2014
• Autor(a) principal: Nojosa, Glauber Marques
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Capital público; Regressão por Splines; Sustentabilidade fiscal
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/23059

This essay is composed of three articles. The first article analyzes the presence of the flypaper effect and its variability by using data from Brazilian municipalities. I implemented threshold model proposed by Hansen (2000) in which political strength variable assumes the role of threshold to infer about the variability of the flypaper effect on public finance of Brazilian municipalities. Results indicate the presence of the flypaper effect on those municipalities. Moreover, it proves empirically the relationship between political and budget decisions, in which the flypaper effect is stronger in those municipalities with larger political allies base in the City Council. Municipalities whose mayor is politically strong and various parties divide the City Council there are clientelistic practices and intensification of the flypaper effect. Second article analyzes the sustainability of the Brazilian public debt trajectory, from 1991 to 2009, by considering a reaction function to the debt with parameters varying over time. Estimation was conducted by using a penalized spline regression adapted to tax rule developed by Bohn (1998). Main results indicate that Brazil did not follow a sustainable debt policy during the period analyzed. Third article evaluates the contribution of public capital on the product of Brazilian economy, considering the possibility of changes in the coefficients of the aggregate production function. Data for GPD and public and private capital was obtained from 1950 to 2008. Cointegration with structural breaks approach developed by Kejriwal and Perron (2006) was used. Cointegrated vectors were estimated by Dynamic Ordinary Least Squares (DOLS) method. Engle-Granger cointegration test (1987) was used to confirm long-run relationship and the Hansen test (1992) verified model stability. According to the results, aggregate production function was stable with two structural breaks in its deterministic structure, 1958 and 1966. Public capital effects on the product are inconstant, from 0.25 to 0.48, among the selected regime.