Article Summary
Introduction
When asked to report their feelings on the Democratic and Republican parties, Americans increasingly show large differences in warmth between the two. But who are respondents thinking of when they are asked to evaluate their feelings toward parties: politicians or ordinary people? Such a difference is important, as it distinguishes a more deep-rooted, ubiquitous disdain for another group of people from a more constrained disdain for those with political power. If respondents conflate party with politicians in their responses, affective polarization in everyday life may be overstated for most Americans. In a conceptual replication of Druckman and Levendusky (POQ 2019), Kingzette attempts to differentiate between levels of affective polarization displayed for politicians versus ordinary people.
Analytical Approach
Kingzette fields a nationally representative survey of 1,000 respondents (460 of which are Democrats and 399 Republicans). Each respondent is first asked to rate both the Democratic and Republican parties on a feeling thermometer from 0-100 to set a baseline for later responses. Then, in a randomized order, respondents are asked to rate Democratic/Republican politicians and ordinary members of the Democratic/Republican Party on the same scale. Randomization is meant to prevent differences in evaluations to occur as a result of anchoring responses to previous answers. Kingzette can then compare average evaluations of parties, politicians, and ordinary people to determine if any differences exist.
Kingzette uses a within-subjects design, where each subject receives all possible treatments. This is in contrast to Druckman and Levendusky’s (2019) between-subjects design, where respondents only receive one treatment. The latter relies on the expectation of treatment groups to be indistinguishable with regard to confounding variables, but Kingzette’s approach is able to make comparisons within each respondent, making imbalance concerns moot and increasing statistical power with the ability to leverage greater sample size.
Main Findings
Kingzette finds respondents hold consistently higher evaluations of ordinary people from the opposite party than of politicians from the opposite party and of the opposite party itself, with differences between 6.5% and 10.6% for Democrats and Republicans, respectively. When asked to evaluate parties, respondents seem to have politicians in mind, as evaluations of parties and politicians are statistically indistinguishable. These results are robust to analyzing respondents with different question ordering separately; evaluations of parties, politicians, and ordinary people stay the same regardless of whether respondents were asked to evaluate politicians or ordinary people first.
In addition to the main analysis of the paper, Kingzette also presents findings and how evaluations of parties, politicians, and ordinary people vary within one’s own party as well. Democrats rate ordinary Democrats and Democratic politicians about 5 to 6 points lower than the Democratic party. Republicans have similar results, although they rate ordinary Republicans about 3 points more favorably than Republican politicians.
Implications
These results help clarify who Americans have in mind as they demonstrate higher levels of disdain for the opposite party; not ordinary people, but politicians. This suggests the scope of affective polarization in everyday social interactions may be overstated. These differences are nuanced, but have important implications. Dislike for opposing politicians is a somewhat expected if not a normal part of democratic politics, but dislike for opposing partisans could have deleterious effects on civil society. Of course, Kingzette still finds some dislike for opposing partisans, but such dislike is not as intense as dislike for politicians.
Questions left unanswered
In discussing the different levels of in-party affect among Democrats and Republicans, Kingzette briefly raises the notion that such partisans may have different conceptions of who an ordinary member of the party is. Why this might be the cause, however, is left unanswered, but could provide additional context to the parties voters have in their minds (see Ahler and Sood 2018). Furthermore, Kingzette leaves open to discussion the level at which dislike for opposing partisans becomes concerning for civil society. As was mentioned previously, while dislike for opposing ordinary partisans is not as intense as dislike for opposing politicians, such dislike is not zero. As affective polarization is often measured with regard to its relation to more concrete social ills (democratic backsliding, social isolation, etc), it would be interesting to see if dislike of rival ordinary partisans has a different relationship to such negative consequences.
Methods and Analysis
Was the study and its analyses pre-registered?: No
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
This paper is a brief and very focused study of a particular measure of affective polarization, and that brevity lends itself to very clear and intentioned analyses. One concern the author raises, however, is related to survey learning being the root cause of any differences between conditions; “learning could entail respondents anchoring their responses to thermometer items based on their responses to previous thermometer items.” While Kingzette addresses this issue by randomizing the thermometer prompts for politicians and ordinary people, he does not do so for the question about parties – that question always appears first. If the “party” respondents have “in their head” is about politicians, as the author gives evidence of, this would fall victim to the same anchoring that Kingzette tries to avoid. If respondents were first asked to evaluate ordinary party members, the results of subsequent questions may be slightly different.
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
Kingzette, J. (2021). Who Do You Loathe? Feelings toward Politicians vs. Ordinary People in the Opposing Party. Journal of Experimental Political Science, 8(1), 75-84. doi:10.1017/XPS.2020.9
Bibtex
1
2
3
4
5
6
7
8
9
10
11
@article{kingzette_2021,
title={Who Do You Loathe? Feelings toward Politicians vs. Ordinary People in the Opposing Party},
volume={8},
DOI={10.1017/XPS.2020.9},
number={1},
journal={Journal of Experimental Political Science},
publisher={Cambridge University Press},
author={Kingzette, Jon},
year={2021},
pages={75–84}
}