Os efeitos dos ciclos político-orçamentários na decisão de gastos em municípios brasileiros

Abstract

This thesis aimed to evaluate whether municipal managers increase investment spending in an election year and the relationship with the size of the municipality, whether this increase is related to the geographic location of each municipality (State to which it belongs), and the Current Operating Result - ROC. It was also verified whether there was a significant increase in the current expenditure account in a post-election year. The population analyzed was all Brazilian municipalities from 2013 to 2021, covering two complete electoral cycles. The model used was multiple linear regression with panel fixed effects, using the variation within the same municipality over time and the OLS - Ordinary Least Squares approach to estimate the coefficients. For municipalities with up to 10,188 thousand inhabitants, the election year has a positive and significant coefficient about the increase in investment spending. The state to which the municipality belongs did not present a significant coefficient in isolation, which demonstrates that the effect of the election year on investment spending can vary from one state to another. Municipalities with an ROC greater than 1, i.e., in surplus, tend to increase investments in election years, but the same occurs in municipalities with an ROC between 0 and 1. Current expenses in post-election years increase in practically all municipalities in the country. For municipalities with 10,189 to 50 thousand inhabitants, the increase in investment spending in election years is confirmed and the relationship with the population is inverse, i.e., the smaller the municipality, the greater the number of cases. The state remains without influence and the ROC of the same form, although surplus municipalities increase investment spending, the same occurs with deficit municipalities. Municipalities with 50,001 to 100 thousand inhabitants and those with 100,001 to 142,633 inhabitants also increase investment spending in election years, but the population variable does not present a significant p-value for these municipalities, indicating that the population does not influence the increase in spending. The state to which they belong also does not have a significant influence, the same occurs about the fiscal situation. For the municipalities considered to be part of the reserve, with more than 142,634 inhabitants and the capitals, the increase in investment spending in an election year was confirmed and the relationship with the population is inverse, that is, the smaller the municipality, the more cases of increased investment spending in an election year. The state and ROC continue to have no relation to this increase in investment spending.