The Effects of Political Attitudes on Affective Polarization: Survey Evidence from 165 Elections
| Author | João V Guedes-Neto |
| DOI | http://doi.org/10.1177/14789299211067376 |
| Published date | 01 May 2023 |
| Date | 01 May 2023 |
| Subject Matter | Articles |
https://doi.org/10.1177/14789299211067376
Political Studies Review
2023, Vol. 21(2) 238 –259
© The Author(s) 2022
Article reuse guidelines:
sagepub.com/journals-permissions
DOI: 10.1177/14789299211067376
journals.sagepub.com/home/psrev
The Effects of Political
Attitudes on Affective
Polarization: Survey Evidence
from 165 Elections
João V Guedes-Neto
Abstract
How do individual-level political attitudes influence affective polarization on a global scale?
This article contributes to the debate on the social distance of party affect by testing a set of
hypotheses in 165 elections across the world. With a sample of over 170,000 voters, the results of
multilevel mixed-effects regressions demonstrate that ideological radicalism, political knowledge,
and external efficacy substantively affect how voters see the main political parties in electoral
disputes taking place in 52 countries from 1996 to 2019. Satisfaction with democracy, however, is
context-dependent; it positively influences affective polarization only when generalized democratic
satisfaction is low. Furthermore, I show that these correlations remain stable regardless of the
operationalization of affective polarization—that is, based on two dominant parties and weighted
for multiparty competition. These findings provide robust inputs to the study of party preferences
and social distance in a cross-national longitudinal perspective.
Keywords
affective polarization, comparative political behavior, political attitudes, survey analysis, political
polarization
Accepted: 26 November 2021
Introduction
In Brazil, presidential elections take place every 4 years in October. When no candidate
can reach a simple majority, the people must choose one out of the two front-runners in a
second round that often takes place in November. Since most of the population is
Christian, the following month is marked by some of the main family gatherings in the
year—Christmas and New Year’s Eve. Every 4 years, the recency of the election makes
politics one of the hot topics of these celebrations. In 2018, as the country faced one of
Department of Political Science, University of Pittsburgh, Pittsburgh, PA, USA
Corresponding author:
João V Guedes-Neto, Department of Political Science, University of Pittsburgh, Pittsburgh, PA 15261, USA.
Email: jog118@pitt.edu
1067376PSW0010.1177/14789299211067376Political Studies ReviewGuedes-Neto
research-article2022
Article
Guedes-Neto 239
the most polarized campaigns of the century, most newspapers offered tips on how to
navigate the “minefield” of discussing politics while keeping family bonds unshattered in
the year to come. To avoid labeling people (Folha de São Paulo, 2019) and to talk about
unifying topics (Maakaroun, 2018) were among the most frequent recommendations.
This anecdotal illustration proposes that, whereas most scholarship has been dedicated
to studying affective polarization in the American context (Iyengar et al., 2018), it is not
unlike to identify this phenomenon in comparative settings (for other examples, see
Hernández et al., 2020; Reiljan, 2020; Wagner, 2020; Ward and Tavits, 2019; Westwood
et al., 2018). In fact, Gidron et al. (2020) show that several countries have higher levels
of affective polarization than the United States, for example, Portugal, Greece, and Spain.
In this article, I aggregate knowledge to this debate by asking, from a cross-national time-
series perspective, how individual-level political attitudes influence affective polarization
in the world.
This means that rather than investigating how attitudinal variables correlate to inter-
party social distance exclusively in the United States, I demonstrate that there is a certain
type of voter who is more affectively polarized regardless of context. This voter holds
more extreme ideological views, a higher degree of external efficacy, and less political
knowledge. Yet, my data analysis also shows that the effect of satisfaction with democ-
racy is context-dependent, as it is connected to higher levels of inter-party social distance
only in countries with generalized low levels of democratic satisfaction.
I benefit from the large database of the Comparative Study of Electoral Systems
(CSES), which surveyed over 170,000 voters in 165 elections from 52 countries ranging
from 1996 to 2019. Relying on these data, I use multilevel hybrid mixed-effects regres-
sions to test the effects of individual-level political attitudes on voters’ affective polariza-
tion when considering, as originally done by Iyengar et al. (2012), each country’s two
main political parties.
This is not the first study of affective polarization in a cross-national setting. Westwood
et al. (2018) use a set of games with participants from Great Britain, the United States,
Belgium, and Spain to find that voters assign more resources to co-partisans and those of
ally parties than to opponent partisan groups. Ward and Tavits (2019: 6) consider affective
polarization as an independent variable and use a subsample of CSES to calculate it based
on “the standard deviation of a respondent’s affect towards parties.” The rationale is that
while it is reasonable to calculate affective polarization in two-party systems by compar-
ing attitudes to the two main political parties, this method may not be adequate in multi-
party countries.
Reiljan (2020: 380) adds complexity to this formulation considering “the average
divergence of partisan affective evaluations between in-party and out-parties, weighted
by the electoral size (vote share) of the parties.” His study is based on the subsample of
European countries surveyed by CSES, identifying that voters in Central Eastern and
Southern Europe have higher levels of polarization than in the United States. A similar
method is adopted by Wagner (2020) to study a larger subsample of the same data—an
approach that was used by Hernández et al. (2020) to find a positive relationship between
affective polarization and election salience.
The argument that the differences between two-party and multiparty systems should
be considered in the study of affective polarization is valid. Yet, in many multiparty
democracies, two parties end up controlling most electoral disputes. Consider, for
instance, the German case, where the center-right Christlich Demokratische Union (CDU)
and the center-left Sozialdemokratische Partei Deutschlands (SPD) dominated most
To continue reading
Request your trial