Article Summary
Introduction
As disdain for members of the opposite party among Americans increases, researchers are increasingly interested in interventions that can effectively reduce differences in affect. One potential intervention is the priming and increasing of dispositional empathy – the sharing of others’ perspectives or emotions. While effective in other domains (such as support for stigmatized groups), can empathy reduce partisan animosity?
In their paper, Simas, Clifford, and Kirkland argue dispositional empathy may actually have the opposite effect on affective polarization, exacerbating existing partisan divides. Empathy does not function uniformly; empathic individuals may be more willing to endure contact with out-group members, but are less empathetic toward such people (and more empathetic toward in-group members). Put differently, existing research suggests empathetic individuals are more in-tune with struggles of in-group members, making them especially sensitive to perceived wrong-doing of out-group members. This leads the authors to expect empathic individuals to hold lower evaluations of out-party members and be more supportive of out-party harm. They test their hypotheses through a nationally-representative survey and an experiment, finding results supporting their theoretical expectations.
This study is one of a growing number of political science studies linking more generalized measures from the psychological literature to political attitudes and behavior. Importantly, it does not attempt to induce empathy, but to measure baseline levels of empathy and report results conditional on such levels. In doing so, the authors provide deeper insight into the influence of trait-like dispositions on political behavior, and potentially into the sorts of interventions that may be most fruitful in reducing inter-party animosity.
Analytical Approach
Simas, Clifford, and Kirkland conduct two studies to test their expectations of empathy’s role in partisan antipathy.
First, the authors field a nationally representative survey (n = 1,181) in May 2016 through YouGov. They operationalize empathy through the Interpersonal Reactivity Index (IRI), which asks respondents to rate on a five-point-scale how a series of 28 statements describes them. For their main dependent variables, the authors ask respondents to rate the Democratic and Republican parties on seven-point favorability scales, taking the difference between them as the gap in affect. Additionally, they gather measures of social distance by asking respondents how upset they would be if a family member married an out-partisan or if a neighbor posted a yard sign for an opposite-party candidate. They then regress their main dependent variables on the IRI score using an ordinal logistic regression (with several controls) to determine empathy’s relationship with partisan animosity.
Second, the authors conduct a survey experiment on an undergraduate sample (n = 1,232), again using IRI as the main independent variable. Subjects were asked to read one of two randomly assigned news articles about a protest on campus against an invited speaker who criticized either Democrats or Republicans (depending on treatment condition). The protest was held by the corresponding out-party group. A bystander is struck by the protestors and police cancel the event in both conditions. Subjects were asked about their feelings on the speaker, protestors, and bystander. The authors then regress the various resulting attitudes on censorship, punishment, sympathy, and schadenfreude on treatment condition and empathic concern using OLS and ordinal logit to determine the conditional partisan effect of empathy.
Main Findings
In their survey, the authors find dispositional empathy is strongly associated with relative in-party favoritism and negatively related to out-party favorability. That is, the increase in the difference in favorability scores for parties is not just due to an increased favorability of the in-party, but a decrease in favorability toward the out-party. Additionally, the authors find support for their hypothesis of social distancing; while empaths favor out-partisan less, they are less likely to engage in socially avoidant behaviors. Results are strongest among stronger partisans and robust to the inclusion of alternative personality trait measures (Big Five) in the model.
Turning to their experimental results, the authors find similar patterns among empaths. High-empathy subjects were more likely to support censoring an out-party speaker and more likely to feel schadenfreude for the injured bystander attempting to listen to the out-party speaker. While empaths are also more likely to support punishing protestors and more likely to feel sympathy for the bystander, these feelings were not moderated by the partisanship of those students. In summary, empathy does little to ameliorate partisan animosity; in some cases, it increases it.
The authors also consider whether an alternative to empathy as a means to decrease partisan animosity: perspective taking. Measuring one’s proclivity to see situations from another’s perspective, the authors find such a trait does not reduce affective polarization.
Implications
Simas, Clifford, and Kirkland’s results have several negative implications for interpersonal behavior between partisans in America. Namely, those who we may most expect to mend the divide between partisans – those who feel empathy for others – are actually among the most divided. Interventions designed to increase empathy, then, may backfire on practitioners, only exacerbating the affective gap between partisans.
One hopeful implication, however, is that empaths are more likely to engage in intergroup contact, which has been shown to lower negative outgroup feelings. While empathy itself may not be enough to reduce feelings of animosity, priming empathy may at least give interventions the opportunity to ameliorate such feelings through other means.
One element of empathy that should be studied further is the difference between dispositional versus primed or manipulated empathy. The authors of this study limited their focus on dispositional empathy, but many interventions attempt to prime it in a more artificials way. It is possible more intentional interventions could avoid the more negative aspects of empathy if done in a tailored way.
Questions left unanswered
One component of empathy raised by the authors is the ability of empaths to find common ground with others. While the results suggest this is limited to in-party members only, it would be interesting to know if empaths do see common ground with opponents but feel more negatively toward them regardless. This would give insight into the policy versus social aspects of empathy vis a vis affective polarization.
Additionally, the authors focus their outcome measures mostly on attitudes, but mention potential behavioral implications for empathy. Another potential analysis could determine whether empaths are merely more attitudinally polarized than non-empaths, but are behaviorally similar. While negative attitudes toward out-party members are certainly troubling, more troubling would be empaths acting upon such attitudes in more extreme ways.
Methods and Analysis
Was the study and its analyses pre-registered?: No
Did the study rely on proxy variables to measure polarization?: Yes
Instead of the typical 0-100 feeling thermometer, authors use 7-point scale (very favorable to very unfavorable)
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
The authors acknowledge their experiment was fielded on a convenience sample of undergraduates from the University of Houston. While they mention this is a diverse population with regard to both demographic and attitudinal attributes, one might expect college students to use empathy in ways distinct from adults. A large psychological literature documents changes in emotional maturity, especially during and after college, with particular work focusing on changes in empathy (Schieman and Van Gundy 2000, for instance). Beyond differential levels of empathy, the psychosocial functioning of empathy may also change with age. We may expect empathy’s effect on polarization to be greater in an experiment about a college protest, as college-aged respondents are more likely to have social connections to people similar to the ones described in the treatments.
Open Data & Analyses
Does the article make the replication data publicly available?: Yes
Does the article make the replication analysis scripts publicly available?: Yes
Article Citation
Simas, E., Clifford, S., & Kirkland, J. (2020). How Empathic Concern Fuels Political Polarization. American Political Science Review, 114(1), 258-269. doi:10.1017/S0003055419000534
Bibtex
1
2
3
4
5
6
7
8
9
10
11
12
@article{simas_clifford_kirkland_2020,
title={How Empathic Concern Fuels Political Polarization},
volume={114},
DOI={10.1017/S0003055419000534},
number={1},
journal={American Political Science Review},
publisher={Cambridge University Press},
author={Simas, Elizabeth N. and Clifford, Scott and Kirkland, Justin H.},
year={2020},
pages={258–269}
}