Ensaios sobre a descentralização fiscal e a dinâmica eleitoral nos municípios brasileiros

Abstract

This Doctoral thesis aimed at analyzing the effects of interaction between fiscal decentralization and electoral dynamics in Brazilian local governments from 2005 to 2017. It was assumed the thesis that fiscal decentralization is influenced by electoral dynamics to achieve fiscal balance, technical efficiency and public service provision for Brazilian municipalities. This investigation is innovative in its analyses of the set of elements from electoral dynamics concerning all three aspects of municipal management: fiscal balance, efficiency and public services. In addition, it proposes a new method of outlier treatment to measure efficiency through Data Envelopment Analysis (DEA). The thesis is structured in five chapters. In Chapter 2, the analysis of effects from the interaction between fiscal decentralization and electoral dynamics variables on fiscal balance showed mainly that: (i) reelections have negative effect; (ii) the partisan alternation has positive effect; (iii) greater electoral competition impairs the effect of fiscal decentralization when it occurs by unconditional transfers and expenditure per capita, contrary to tax collection which has positive effect; (iv) partisan alignment between the mayor and the governor produces negative effects when evaluated by expenditure per capita, whereas alignment with the president has negative effect; (v) municipal election generates negative effects, while state elections might affect either negatively or positively, depending on populational size. Chapter 3 allowed the identification that the effect of fiscal decentralization measured by the participation of tax collection in the total revenue for the technical efficiency of health services for municipalities in the state of Minas Gerais was negative, regardless of the treatment method for outliers to measure efficiency. In Chapter 4, the study of the interaction between fiscal decentralization and electoral dynamic variables allowed to observe that: i) tax collection is more likely to suffer the influence of political factors in order to achieve efficiency; ii) budget political cycle displayed different effects. In the state elections, municipalities displayed negative effects for the interaction between fiscal decentralization and state elections which was dependent on populational size and efficiency level; however, municipal elections displayed the opposite effects; iii) partisan alternation had negative effects on the interaction between tax collection and efficiency; iv) electoral competition and reelection had different effects on fiscal decentralization to achieve efficiency; and; v) partisan alignment between the mayor either with the president or the governor had different effects on fiscal decentralization in order to reach higher efficiency scores which depended on the municipality size. Chapter 5 revealed that the effect of the interaction between fiscal decentralization and electoral dynamics was negative statistically significant for the majority of the estimated coefficients, especially for electoral cycles, and that the effect depended on populational size and educational efficiency levels. Chapter 5 confirmed the hypothesis that education and healthcare public service provision are influenced by the interaction between fiscal decentralization and the electoral dynamics of Brazilian municipalities. The best way for fiscal decentralization to improve public services includes: i) primary education and vaccine coverage by unconditional transfers; ii) infant mortality by conditional transfers, and iii) the potential years of life lost (PYLL) by taxation collection. Results corroborate the thesis, although the interaction effect depends on populational size and concerning efficiency, it also depends on its own level.