Article Summary
Introduction
Previous research has documented the increase in partisan animosities within the American public over the last couple of decades. However, our understanding of how affective polarization compares to other social divides, and to what extent it influences attitudes and behavior in the nonpolitical realm, remains limited. Consequently, the authors pose the following questions: How does partisan affect compare to affect based on other social divides? To what extent are partisans willing to discriminate against opposing partisans in nonpolitical decisions? Addressing these questions will help assess the pervasiveness of partisan animosities in the American public.
Analytical Approach
The article presents findings from four experimental studies. In Study 1, the authors utilized a condensed version of the Implicit Association Test (IAT), a tool developed in social psychology to assess affective prejudices. Participants were instructed to complete the test rapidly, focusing on associations between partisan groups (Democrats and Republicans) and positive or negative attributes and between racial groups (African American and European American) and positive or negative attributes. Respondents engaged in both tests, and the order of the tests was randomized for each individual.
In Study 2, the authors designed choice tasks involving two high school graduates with varying characteristics. Participants were randomly assigned to either a partisan or a racial cue choice task (see table below for details). In the partisan cue condition, respondents viewed two profiles of high school graduates differing in the partisan leaning of their extracurricular activities. In the race cue condition, the profiles varied based on extracurricular involvement in two groups (Future Investment Banker Club vs. African American Student Association). Alongside these distinctions, the authors included additional extracurricular information and school performance in the graduate profiles.
| Treatment condition | Choice task |
|---|---|
| Partisan cue | Different partisan affiliations signaled through extracurricular activity (President of the Young Republicans vs. President of the Young Democrats); a number of other candidate characteristics (including high school performance) |
| Race cue | Different races signaled through extracurricular activity (President of the Future Investment Banker Club vs. President of the African American Student Association) and name; a number of other candidate characteristics (including high school performance) |
In Study 3, participants engaged in either a trust game or a dictator game. In the trust game, respondents were tasked with allocating a sum of money between themselves and a second player. The allocated amount to the other player was tripled, and the second player had the option to reciprocate by allocating money back to the respondent. In contrast, the dictator game followed a similar setup, but the money allocated to the second player was not tripled, and no money was reallocated to the respondent. The key difference between the two games is the information about the other player’s willingness to cooperate in the dictator game, unlike the trust game where such information is available. Consequently, differences in allocation behavior in the dictator game, when varying information about the other player’s group affiliation, can be attributed to discrimination. Participants were randomly presented with information about the second player’s political affiliation (Republican or Democrat) and race (White or African American) before each allocation task. Each respondent completed four rounds of the game.
In Study 4, the trust and dictator games were repeated, but additional partisanship characteristics were introduced by including Independent and no information about partisan identification as a pure control condition. Respondents completed four rounds of the game, with the type of game (trust vs. dictator game) randomized between individuals. Information about the other player’s partisanship was also randomized within individuals.
Main Findings
Based on their studies, the authors report four main findings. First, analysis of two variants of the Implicit Association Test (IAT) reveals that partisan animosities appear to be stronger than racial prejudices. Second, when examining the impact of partisan animosities on nonpolitical behavior, the high school graduate choice task indicates a significantly higher bias based on partisanship than race. Third, results from the trust and dictator games show that individuals exhibit biased behavior toward out-partisans but not toward individuals of a different race. Lastly, the inclusion of a pure control condition in the trust and dictator games suggests that negative evaluations toward the out-group have more profound behavioral consequences than in-group identification.
Implications
The findings of the study suggest that partisanship triggers even stronger outgroup and ingroup responses compared to one of the most prevalent divisions in the US—race. While partisan identification was relatively weak in comparison to other advanced democracies a couple of decades ago, it has since evolved into a primary divide among citizens, with ramifications extending into nonpolitical realms. These growing partisan divides, in turn, diminish the incentives for politicians to moderate their attacks on politicians from the opposing party, let alone cooperate with them.
Questions left unanswered
While the authors acknowledge the potential influence of elite rhetoric and the information environment on the escalating partisan divides in the US, a lingering question pertains to the extent to which the profound consequences of partisan animosities can be attributed to developments within the American public and political leadership. Although this question has been explored in some prior research, there remains a need for future studies that specifically investigate the role of elite appeals and news reporting in accounting for the intense partisan animosities identified in the studies conducted in the article.
Methods and Analysis
Was the study and its analyses pre-registered?: Study was conducted before 2015
Did the study rely on proxy variables to measure polarization?: No
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
As highlighted by the authors, unlike race, individuals need to disclose their partisan identities. Therefore, a crucial question regarding the external validity of the results is whether partisan identifications are manifest in individuals’ everyday actions and lead to behavioral consequences. While the authors acknowledge the significance of partisan identities in daily interactions, it would be valuable for future studies to assess the accuracy of individuals’ perceptions of others’ partisan affiliations.
Open Data & Analyses
Does the article make the replication data publicly available?: Yes
Does the article make the replication analysis scripts publicly available?: No
Article Citation
Iyengar, S., & Westwood, S. J. (2015). Fear and Loathing across Party Lines: New Evidence on Group Polarization. American Journal of Political Science, 59(3), 690–707.
Bibtex
1
2
3
4
5
6
7
8
9
10
11
12
13
@article{iyengar2015,
title = {Fear and Loathing across Party Lines: New Evidence on Group Polarization},
shorttitle = {Fear and Loathing across Party Lines},
author = {Iyengar, Shanto and Westwood, Sean J.},
date = {2015},
journaltitle = {American Journal of Political Science},
volume = {59},
number = {3},
pages = {690--707},
issn = {1540-5907},
doi = {10.1111/ajps.12152}
}