Article Summary
Introduction
How can we reduce affective polarization in the United States? Much of the existing work on affective polarization argues that this hostility is rooted in partisan identity: people like their own party and dislike the other party because they feel their party affiliation is part of who they are. Thus, weakening partisan identity may help to soothe affective polarization. These authors suggest that emotions may be an important way to weaken this partisan identity and, thus, reduce affective polarization. They argue that enthusiasm toward the in-party or anxiety toward the out-party will reinforce partisan identification, promoting affective polarization. Conversely, enthusiasm toward the out-party or anxiety toward the in-party may weaken partisan identification and reduce affective polarization. The authors explore these emotional dynamics in the context of public responses to presidential candidates. Because presidential candidates are important, highly visible members of the parties, people’s emotional responses to them may shape the way those people feel about the parties more broadly.
Analytical Approach
The authors conducted two studies to test their expectations about the link between emotions and affective polarization.
In Study 1, they used data from the 2008-2009 American National Election Studies. They focus on surveys conducted every month from January 2008 to November 2008. Each month, the surveys asked people how favorable they felt toward the Democratic and Republican parties, as well as how frequently they felt pride, hope, or fear about each party’s presidential nominee. They used these questions to generate items measuring in-party favoritism (a proxy for affective polarization), enthusiasm about the in-party candidate, anxiety about the in-party candidate, enthusiasm about the out-party candidate, and anxiety about the out-party candidate. They analyzed the 11 months of survey data examining how participants’ emotions about the candidates in September predicted their in-party favoritism in November, controlling for participants’ prior in-party favoritism from February.
In Study 2, they conducted an original survey with two waves in October and November 2016, before and immediately after the election. Like in Study 1, they measured affective polarization (this time directly rather than with a proxy measure) as well as enthusiasm and anxiety about the consequences of each party’s candidate winning the election. They then examined how participants’ emotions about the candidates in October predicted their affective polarization in November, controlling for participants’ affective polarization in October.
Main Findings
In both Study 1 and Study 2, the authors find that enthusiasm about the in-party predicts greater affective polarization and that enthusiasm about the out-party predicts less affective polarization. These relationships are not causal—the authors do not claim that enthusiasm increases or decreases polarization— but the findings are suggestive that enthusiasm may be important for understanding affective polarization.
The authors find little evidence of a connection between anxiety toward either party and affective polarization, however. They suggest that this may be because anxiety toward the in-party or the out-party does not fundamentally push people to reevaluate their feelings about the parties. When people feel anxious about their own party’s candidate, they may justify their continued party affiliation by feeling that the out-party is even worse. Conversely, when people feel anxious about the other party’s candidate, nothing much changes—they already don’t like the other party!
To sustain their argument about in-party anxiety, the authors conduct some supplementary analyses focused on questions in both studies asking about participants’ intended vote choice. In both studies, they found that in-party anxiety was negatively related to participants’ willingness to vote for the out-party candidate. There was some evidence that in-party anxiety reduced willingness to vote for the in-party candidate, but this was only true in 2008, not in 2016. In other words, participants’ anxiety about their own party’s candidate seemed to primarily make them more skeptical of the other party’s candidate.
Implications
This study’s findings highlight three major implications. First, emotions may play an important role in shaping affective polarization. Second, emotions’ relevance to affective polarization depends on the source of the emotions: emotions about the in-party are quite different from emotions about the out-party. Third, responses to presidential candidates may be important sources (and a potential way to overcome) affective polarization. Putting these three implications together, the authors argue that emotional responses to presidential candidates may play important roles in shaping affective polarization and, in particular, that enthusiasm about out-party candidates may disrupt affective polarization. They recognize that the presidential nomination process favors polarizing candidates, but are nonetheless optimistic that a less divisive candidate who both parties are enthusiastic about could seriously reduce affective polarization in the US.
Questions left unanswered
Although the authors argue that the data they present are useful for showing potential causal relationships between emotions and affective polarization, the authors explicitly highlight that their panel data cannot provide direct evidence of causality. Thus, it remains a partially open question whether emotions cause affective polarization or whether the relationship is driven by some other factor (e.g., affective polarization causing emotions about candidates).
The authors also acknowledge that they did not measure anger. Anger is very politically influential in other contexts and seems conceptually linked to affective polarization and partisan hostility. The authors argue that anxiety and enthusiasm may be more directly related to partisan identity, but anger nonetheless represents an important next step for understanding emotions and affective polarization.
The authors also focused their investigation of affective polarization only on favoritism toward the in-party over the out-party. They suggest that future research in the area of emotions and affective polarization might also explore other implications of affective polarization such as explicit hostility, discrimination, and social distance.
Methods and Analysis
Was the study and its analyses pre-registered?: No
Did the study rely on proxy variables to measure polarization?: Yes
In Study 1, the authors measured affective polarization using two ANES questions asking about favorable and unfavorable thoughts about the Democratic and Republican parties. They subtracted the measure of out-group favor from the measure of in-group favor to generate a measure of in-group favoritism.
Study 2 used the traditional feeling thermometer measure of affective polarization.
Were standard p-value thresholds used (p<.05 or 95% Confidence Intervals that don’t overlap zero)?: Yes
- Largest p-value presented as significant: 0.05
Were correlational results interpreted with causal language?: No
Limitations / Weaknesses
Although the authors highlighted that their results should not be taken as direct causal evidence, it is worth re-emphasizing this limitation. Their analyses examine whether pre-election emotions predict post-election affective polarization, controlling for pre-election affective polarization. But it is entirely possible that the election increased affective polarization itself and that the relationship between emotions and affective polarization is reversed—that people who are more affectively polarized report more enthusiasm toward the in-party candidate and less toward the out-party candidate. Or that the relationship between the two variables is driven by some other process entirely. Without an experiment or a more comprehensive set of evidence, causality remains unclear.
On this point, both studies were conducted during highly polarizing presidential elections. It is not clear that these results might generalize to other, less polarized elections (e.g., midterm elections) or to non-electoral contexts in which the public certainly still has emotional responses to politicians’ actions.
Finally, the Ns for both studies were relatively small. Study 1 consisted of 439 people and Study 2 consisted of 354. Neither is so small as to be especially concerning, but the authors do not report power analyses and the likelihood of findings being driven by chance increases for small samples like this.
Open Data & Analyses
Does the article make the replication data publicly available?: No
Does the article make the replication analysis scripts publicly available?: No
Article Citation
McLaughlin, B., Holland, D., Thompson, B. A., & Koenig, A. (2020). Emotions and affective polarization: How enthusiasm and anxiety about presidential candidates affect interparty attitudes. American Politics Research, 48(2), 308–316. https://doi.org/10.1177/1532673X19891423
Bibtex
1
2
3
4
5
6
7
8
9
10
11
12
@article{doi:10.1177/1532673X19891423,
author = {Bryan McLaughlin and Derrick Holland and Bailey A. Thompson and Abby Koenig},
title ={Emotions and Affective Polarization: How Enthusiasm and Anxiety About Presidential Candidates Affect Interparty Attitudes},
journal = {American Politics Research},
volume = {48},
number = {2},
pages = {308-316},
year = {2020},
doi = {10.1177/1532673X19891423},
URL = {https://doi.org/10.1177/1532673X19891423},
eprint = { https://doi.org/10.1177/1532673X19891423}
}