Female, Feminine, or Feminist? Gender and Presidential Voting in Latin America

• In Event: Gender and the Executive Worldwide: Voters, Candidates, and National Leaders

Abstract

Do women and men vote differently in presidential elections in Latin America? We argue that they do, but understanding gender gaps requires exploring multiple dimensions of gender: binary sex identification, gendered personalities, and feminism. Sex gaps are commonly studied in Latin America (Lewis 2004; Morgan 2015; Layton et al. 2021; Reyes-Housholder 2019). We move this literature forward in theorizing how gendered personalities and feminism may matter.
We first argue that voters may be more drawn to candidates who demonstrate gendered leadership styles congruent with voters’ own gendered personalities (McDermott 2016). Because presidential politics is a preeminently masculine arena of power (Conroy 2016; Duerst-Lahti 2014), we argue that masculinity will matter more than femininity in determining vote choice. We also theorize the role of feminism in driving gender gaps in presidential voting. Latin America has recently experienced a resurgence of feminist protests and demands (Zaremberg and Almeida 2022; Reyes-Housholder and Roque 2019). We conceptualize feminism in terms of feminist self-identification and support for feminist policies. We argue that both measures of feminism may be correlated with voters’ sex, with women identifying more as feminists and more supportive of feminist policies. However, we hypothesize that our measures of feminism—analyzed separately—can predict vote choice better than voters’ sex.
We test our theory of the role of gendered personalities and feminism on two recent presidential elections in Latin America: the second rounds of the 2021 Chilean and the 2022 Brazilian races. In Chile, leftist Gabriel Boric competed against far-right leader José Antonio Kast; and in Brazil, center-left former president Luiz Inácio Lula da Silva vied against the far-right incumbent Jair Bolsonaro. This pair of elections is ideal for testing our theory for three reasons. First, although women had previously won presidential elections in both countries, the second rounds in 2021 and 2022 featured two men competing against other men. Much scholarship has shown the myriad ways in which gender matters when women are running for president in Latin America (Reyes-Housholder 2018; dos Santos and Jalalzai 2021; Reyes-Housholder 2019). We explore how gender matters in presidential elections when women are not competing.
Second, these match-ups feature candidates who promoted policies on women's rights and gender ideologies that were near polar opposites. Boric mainstreamed feminism throughout his platform, and upon his election, came to be known as the first feminist president in Latin America (Montes 2022). Although Lula’s 2022 campaign did not explicitly promote feminism, Brazilian feminists often considered him a potential ally, given the role of feminist thought historically within the PT (Macaulay 2006). Third, while the left-leaning candidates’ leadership styles were in some ways gendered feminine, the far-right contenders’ leadership styles were more traditionally masculine. Such contrasts are most evident in Chile where Boric portrayed himself as a leader who dialogs and prioritizes consensus and pledged to de-concentrate presidential power. Kast emphasized recovering the law and order characteristic of Chile prior to the social uprising (Montes 2021; Reyes-Housholder, Suárez-Cao, and Le Foulon Forthcoming) and ‬risk-taking.
We fielded two original surveys via Netquest. We ran the Chilean survey five months after the December 2021 election, and the Brazilian survey during the two weeks prior to the second round on October 30, 2022. This strategy enables us to study both retrospective (Chile) and prospective (Brazil) vote choice, enhancing the generalizability of our findings. Each of our samples feature about 1,500 respondents who said they had voted or had planned to vote in the respective presidential election.
We find that in Chile, a gender gap did exist with about 8% more women voting for Boric than Kast. The feminism gap was larger and wiped out the traditional gender gap when both are included in the same vote choice model. Gendered personalities mattered in that more masculine voters were less likely to vote for Boric. However, this result disappears once voter ideology enters our models. We find similar results in Brazil. About 8% more women intended to vote for Lula than Bolsonaro, but this gender gap loses significance when controlling for economic evaluations. Feminism is a much larger and more consistent predictor of voting for Lula and remains strong and positive after economic evaluations are controlled. Gendered personalities are not associated with vote intentions in Brazil.
This paper shines light on the multiple ways in which gender pervades presidential elections in the region. It has novel implications for research on gender, feminism, masculinity, and presidential vote choice in Latin America.

Authors

• Catherine Reyes-Housholder, Pontificia Universidad Católica de Chile

• Leslie A. Schwindt-Bayer, Rice University