Date created: | Last Updated:

Category: Project

Description: This study investigates the electoral effects of polarization in Latin America, distinguishing by two key types of polarization: that in ideology and that in affects. Traditionally, polarization was primarily understood in terms of ideology, and the study of mass polarization was mainly based on citizens’ positions on public policy issues (e.g., Fiorina and Abrams 2009). While ideology remains a key dimension of polarization, there is increasing interest in affective polarization, understood as the animosity between parties, or the tendency to dislike and distrust those from the out-group (Iyengar et al. 2012). Although various studies find that affective polarization may be driven by ideological differences (Rogowsky and Sutherland 2015; Webster and Abramowitz 2017; Lelkes 2019), others find that these two types of polarization do not always go hand in hand (Levendusky and Malhotra 2016), and we know less about the effects of affective polarization on ideological polarization. Thus, this study aims to better understand the relationship between polarization in ideology and affects, and how each of them maps into electoral behavior. To this end, we conducted an online survey experiment fielded in a highly polarized context in a developing country: the presidential runoff in Chile (December 2021). We first ask respondents to declare their preferred candidate and their position toward an ideological policy issue: taxes. Next, we randomize a 2x2 treatment. On one arm, the “affective depolarization treatment”, we randomize whether respondents receive a video showing positive non-ideological biographical information from their opposing candidate, intended to reduce their affective polarization (Rogowski and Sutherland 2015), while not affecting their ideological polarization. The second arm, the “ideological depolarization treatment,” randomizes whether they receive a video providing information supporting the opposite position they have toward taxes, designed to affect their ideology, but not their affective polarization. If randomized to treatment, respondents without a preferred candidate or position toward taxes receive a video of a randomized candidate/position. After the treatment, we measure ideology, affective polarization, and intended voting behavior. With this research design, we will analyze the roles of ideology and affect in polarization, and their relationship with electoral outcomes.