Electoral systems vary in terms of the choice and influence they offer voters. Beyond selecting between parties, preferential systems allow for choices within parties. More proportional systems make it likely that influence over who determines the assembly’s majority will be distributed across relatively more voters. In response to systems that limit choice and influence, we hypothesize that voters will cast more blank, null, or spoiled ballots on purpose. We use a regression discontinuity opportunity in French municipal elections to test this hypothesis. An exogenously chosen and arbitrary cutpoint is used to determine the electoral rules municipalities use to select their assemblies. We find support for our reasoning—systems that do not allow intraparty preference votes and that lead to disproportional outcomes provoke vote spoilage. Rates of vote spoilage are frequently sufficient to change control over the assembly if those votes had instead been cast validly for the second-place party.

Why do parties form pre-electoral coalitions (PECs) in executive elections? The party of the executive candidate may welcome coalition partners because of the votes that they can deliver. However, PECs often include parties that bring few votes to the coalition. I theorize that parties with a candidate in the executive election welcome small parties because of the campaign resources that they bring to the PEC. I test this argument using data from mayoral elections in Brazil, where parties do not control access to advertisement on radio/TV. The results indicate that campaign resources increase the likelihood of being admitted to a PEC when the party without an executive candidate has few votes to offer.

Electoral systems impose incentives for relationships between parties and relationships within parties. In interparty terms, weak systems encourage many parties to enter and voters to vote sincerely for their most preferred options. Strong systems discourage many parties from entering and encourage voters to think strategically about viability (the likelihood a preferred option will win seats). In intraparty terms, centralized systems empower party leaders and put an emphasis on the party’s shared reputation. Individualistic systems empower individual candidates and members of congress and put an emphasis on their personal reputations. The individual rules examined when defining system incentives include ballot type (can voters choose among copartisans), the level to which votes are pooled before seats are awarded, the number and level at which votes are cast, district magnitude (the number of seats to be decided in a district in a given election), and legal thresholds (predefined vote total barriers to being awarded seats). The electoral systems used to elect lower houses, upper houses (where they exist) and presidents in Latin America are located in a two-dimensional space based on these incentives. In interparty terms, weak systems outnumber strong ones in the region. In intraparty terms, there is a great deal of diversity with centralized systems slightly outnumbering individualistic ones. Instances of electoral reform are captured as changes in incentives or movements in this space. Reforms are frequent but no clear pattern emerges in terms of countries across the region converging toward imposing similar electoral incentives.

This dissertation has as its motivation the question: How is the Municipal Legislative Branch in Brazil? It is commonly assumed that Municipal Chambers are less important than Executive Branch to elaborate and produce policies in Brazilian municipalities. I argue this is an equivocal understanding based on studies conducted with few information and that there are not sufficient empirical evidences to sustain it. In order to present a new and more complete knowledge about Municipal Legislative Chambers, I analyze three issues. First, if the Chambers are professionalized and well structured. Second, the socio-educational and socio-occupational profiles of candidates and council members compared with the municipal population. Third, the content of the bills presented by council members. Using data from Legislative Census, Superior Electoral Court and an original dataset, I conduct a series of tests including descriptive statistics and regression models. The analysis shows that (1) Legislative Chambers are little professionalized and poorly structured; (2) candidates and council members are better educated and from more prestigious occupations than the population mean. However, they are the most accessible representative institution in Brazil for individuals from lower layers of the population; (3) council members can propose bills in important areasand if some are very regulated (such as health and education), others are not (for instance, housing and soil regulation). Therefore, it is not correct to say that there are not key policies in Municipal Legislative Branch.

The international community invests heavily in democracy promotion, but these efforts sometimes embolden leaders not interested in true democratic reform. We develop and test a formal model explaining why this occurs in the context of electoral system reform—one of the most important signals of democratic quality. Our formal model characterizes leaders as either truly reform minded or pseudo-reformers, those who increase electoral system proportionality in order to receive international community benefits while engaging in electoral fraud. We hypothesize that the international community will be more (less) likely to detect fraud when leaders decrease (increase) proportionality, regardless of whether there is evidence of numerical fraud. Using a mixed-methods approach with cross-national and case study data from post-Communist states, we find that the international community is generally less likely to detect fraud following an increase in proportionality and vice versa. We suggest that democracy promoters over-reward perceived democratic progress such that pseudo-reformers often benefit.