Under the premise of rationality, politicians behave to maximize re-election probability. Favorable macroeconomic contexts, alignment with central governance, and balanced public finance will be rewarded leading to re-election. Logit estimations applied to Portuguese municipalities in the period 2002-2017 fail to empirically support these theoretical effects, providing no incentive-controlled policy actions. Local voters do not punish mayors for the adverse economic performance, staying loyal to ideological voting geographically and over time. Only turnout punishes incumbents over the entire period. The introduction of the law of limitation of terms did not change the incentives towards wise governance; therefore, lack of electoral punishments to undesirable policy actions withstands the potential misconduct of incumbents. Existing evidence points to the need of reforms in what concerns electoral participation as when we compel voters to express their democratic rights, they become more critical to undesirable achievements.
TopIntroduction
Elections are the institutional mechanism to legitimate power and the democratic selection of candidates; in this moment the power of choice is twofold as in one hand it allows the identification of the most consensual candidate and the evaluation of incumbents. As democratic elections assume the existence of multiple candidates, and therefore the existence of opposing parties, it promotes alternance and the replacement of the incumbents. Still, political changeovers and the replacement of governants or of the incumbent party do not always happen, as electors tend to continuously vote for the same political party, regardless the performance in former terms. Under the premise of rationality of voting, this action is non-sense and deviates public management from efficiency trends, perpetuating misconducts and misuse of public resources. Consequently, it is of major relevance, in one hand, to appraise the re-election determinants in municipalities with specific detail on the public management and the economic performance to evaluate voter rationality and incumbent incentives. On the other hand, it is of worth addressing if the Law of Limitation of Terms did bring any change.
During the last 43 years, local governance has been playing a major role in the operation of the political system, being the front office of the political power towards voters, moreover, it serves people providing public services and enrolling citizens in proximity decision-making. So, analysing the determinants of re-election at the municipality level is key to address the multilevel framework of the Portuguese governance.
A milestone of the present research is the introduction, in 2005, of the law of the limit of terms in municipalities (Law 46/2005, from 29th of August). The diploma establishes the limits of successive years in governance in municipalities, setting an upper bound of three consecutive terms. Additionally, it was highlighted the incumbent re-election determinants, relying on three vectors of variables: social, economic and political.
Given that the objective was twofold: address the determinants of re-election, with emphasis on public finance, and assess if the LLT has brought any difference, it was collected information for four electoral moments, two before and another two after considering 2002-2005, 2006-2009, 2010-2013 and 2014-2017. To approach the question two sub-samples were set, expecting to find some sort of a structural break, and the determinants were kept to understand the eventual changes in terms of significance and/or magnitude of the effect. This question deserves major attention of the institutions as more than a half of the incumbents were not allowed to run for mayor in each of the electoral moments after the implementation of the LLT.
The present chapter is organised as follows: section 1 presents the literature review on determinants of re-election, section 2 describes the descriptive analysis of the database and the empirical methodology. Section 3 describes the hypotheses in test and the econometric estimation and section 4 concludes.