Search

Polls in the first round of Brazil's 2022 presidential election diverged from the official results, leading to criticisms that they made mistakes. This paper examines an alternative explanation: that actual changes in vote intentions occurred between the polls and election day. Although unpopular, this explanation finds theoretical support in the scholarship based on two main processes: strategic voting and delayed decision-making. Using an experiment conducted a week before the election, we show how undecided voters and voters for less competitive candidacies displayed high propensity to change their minds after watching campaign videos. We also use data from one of the last polls conducted before the election to develop models that identify voters most prone to late swings in vote choice and reallocate such votes to adjust vote estimates. The results suggest that late swings in vote intentions can be a more than plausible phenomenon in recent Brazilian elections. [ABSTRACT FROM AUTHOR]

The spatial theory of voting assumes that voters will cast their votes for the candidate whose policy position is closest to their own views. The goal of this paper is twofold: the first one is to test this assumption for the Brazilian presidential elections from 2002 to 2014. In order to do that we use data from the Brazilian Electoral Study along with scaling methods. The results suggest that the probability of a voter casting her vote for the candidate whose policy position is closest to her own is pretty high. The second goal of this paper is to assess if this result holds despite voters' political sophistication. In other words, we test if ill-informed voters make their electoral decisions based on non-spatial candidate characteristics. The results contradict this idea. Politically ill-informed voters also choose the candidates whose views are closer to their own. [ABSTRACT FROM AUTHOR]

Most studies on voting behavior in Brazil agree that the left-right scale is not associated with vote choice. However, part of the scholarship often ignores or challenges such consensus. To solve this matter, this paper replicates and re-analyzes public opinion data used by studies that propose that voters choose candidates based on left-right positions. The results indicate that those studies present several methodological problems that invalidate their claims. The statistical relationship between ideology and voting is generally weak and only strong among a small group of sophisticated voters. Finally, panel data analysis does not find evidence that the statistical association reflects a causal relationship. All in all, it is time to search for new factors in the explanation of voting behavior in Brazil. [ABSTRACT FROM AUTHOR]

Interactive experiences online have been stimulating the development of public debate. Facebook, for example, has attracted citizens and institutions, as well as communication enterprises that see in this tool an opportunity for greater interaction with an audience. This paper identifies the characteristics of the debate among audiences in the official Facebook pages of the three main Brazilian newspapers (Folha de S.Paulo, O Estado de S.Paulo e O Globo) on the 2014 presidential election. Therefore, we have observed in 610.660 comments: 1) the debate's intensity over time; 2) the commentator's position (reciprocity); 3) justification; and 4) reflexivity. Among our main findings, one might highlight the predominance of monologues and the low level of justification. However, Facebook can be considered a space for persuasion, which prevailed in the comments. [ABSTRACT FROM AUTHOR]

The game between government and opposition is something that has always been widely discussed in political science. Recent studies indicate that party systems may have a strong pro-government bias, weakening opposition parties. How to limit the power of executive leaders is a key question in Brazil, as suggested by the recent ban on immediate re-election enacted by the Chamber of Deputies. In this paper, we propose a debate about the relative power of incumbent governors and challengers in state government elections. All elections since 1994 were analyzed using TSE data from all Brazilian states. The results indicate that re-election does not impose an insurmountable barrier to those who challenge the governor. At the same time, the data suggest a relative continuity of political parties at the state level. Moreover, the idea that the Executive branch has a great ability to appeal to political parties cannot be disregarded. [ABSTRACT FROM AUTHOR]

Much has been said about the role of "Bolsa Família" in the Presidential election of Lula in 2006 and the change of his voter profile. However, little has been said about the effects of Lula's support and "Bolsa Família" on Dilma's electoral performance in the 2010 Presidential elections. Using a spatial econometrics's approach in this paper we evaluate the importance of these (and other) factors on Dilma and her opponents' electoral performance in the 1st round of 2010 elections. It is interesting to understand what factors had more weight on her previous election and how this can be seen in the light of the current scenario. Aggregated data at the municipality level are used in the analysis, considering variables such as candidates' share of total votes and socioeconomics indicators (per capita income, proportion of beneficiaries of "Bolsa Família", percentage of poor, rate of illiteracy, among others). The results indicate a decisive participation of former President Lula and the Bolsa Família Program. Even controlling for Lula's share of votes in the previous election, the "Bolsa Família" Program remains significant for Dilma's election. The Lula's effects, however, was larger than the one found for the Program. It is worth noting that this results in line with the one indicated by Zucco (2013), in the sense that the "Bolsa Família" shows positive short run electoral effects, helping the incumbent party, but shows no long run effects, with the loyalties to parties and/or persons. [ABSTRACT FROM AUTHOR]

Social capital theory emphasizes interpersonal trust as a facilitator of cooperation among individuals, crucial for the formation of autonomous associations within civil society and for the engagement of citizens in public affairs. Despite the importance of trust for the concept of social capital, recent empirical research has been demonstrating that its effect on civic participation is weak, at best. This paper suggests an interactive model, where the effect of trust is conditioned by motivational aspects of participation (expectations of benefits and perceptions of costs). Using data from the 2005 British Election Study (BES), a survey conducted in the United Kingdom, I test the conditional effect of interpersonal trust on individuals' willingness to engage in different types of collective action. The findings point to a conditional effect of trust that varies not only according to motivation variables, but also depending on the type of engagement. [ABSTRACT FROM AUTHOR]