Democratic elections play a fundamental role in solving agency problems between citizens and the government. The literature largely recognizes the advantages of elections as a system of political incentives, but holds little discussion about the potential costs of providing such incentives. This thesis seeks a better understanding of the issue via an empirical exercise using Brazilian data. Its objective is to investigate if the political discontinuity – a characteristic inherent to the democratic electoral system, which presupposes the periodic turnover of elected politicians – interferes with the provision of public services in Brazil. A difference-in-differences approach is used to assess how the change of mayors and parties resulting from elections affects the municipal control of dengue from 2001 to 2009. Results indicate that, as of the third year of a new mayor’s mandate, there is a positive significant impact of the change of elected politician on the number of dengue cases confirmed within a municipality. This finding corroborates the hypothesis that the rupture caused by the political discontinuity jeopardizes the provision of municipal public health services in Brazil. The timing of the effect is attributed to both dengue’s seasonality and the dynamics of virus transmission. Evidence suggests that political proximity and institutional strength operate as inhibitors of this harmful effect. Continuity throughout consecutive mandates also appears to improve the municipal state of dengue. Results regarding the influence of climate on the relationship of interest, however, remain inconclusive.