Article Summary
Introduction
Research into affective polarization employs a range of measures to assess how voters feel about those who support other parties. To what extent are these measures interrelated? And do respondents think of voters, politicians, or the party in general when they are asked to evaluate “the Democratic Party” or “the Republican Party”? The authors address these two questions by implementing a survey that incorporates various measures of affective polarization and randomly varies whether the questions ask about evaluations of supporters, elites, or the party. In presenting evidence from their survey, the authors advance research into affective polarization by evaluating how different survey measures are interrelated with each other and by examining whom survey respondents exactly think of when asked about other parties. More generally, the study offers insights into whether people feel more or less negative about out-party supporters than out-party elites.
Analytical Approach
The authors identify four types of measures that researchers have used to assess affective polarization in their surveys:
- The general feeling thermometer
- Questions about party traits, such as patriotism, intelligence, and honesty (positive traits), and hypocrisy, selfishness, and meanness (negative traits)
- Trust in other parties
- Social distance from party supporters, including whether respondents would be comfortable with their child marrying someone from the other party
The implemented survey incorporates these various measures, allowing for an evaluation of the associations between them. Furthermore, the authors randomly vary whether respondents encounter survey items that refer to voters, party elites, or the party itself. However, they choose not to vary the groups for the social distance measure, as this could lead to unrealistic scenarios (e.g., a child marrying an out-party politician).
| Condition | Group mentioned in survey questions |
|---|---|
| Voters | Democratic (Republican) Party voters |
| Elites | Democratic (Republican) Party candidates and elected officials |
| Party | Democratic (Republican) Party |
Main Findings
The study reveals that all measures of affective polarization are highly correlated, except for the social distance measure. Furthermore, respondents expressed significantly more negative sentiments about elites from the other party than about their supporters. There is no evidence to suggest that respondents evaluate out-party elites more or less negatively than the party as a whole.
Implications
While many measures of affective polarization show strong correlations, researchers should select their measurements based on the specific research questions they aim to address. For instance, when investigating whether citizens are willing to engage with supporters of a particular party, social distance measures and other behavioral indicators are preferable. If the objective is to explore how affective polarization influences political evaluations, items related to partisan animosity involving elites might be a more suitable choice. Therefore, even though these measures exhibit associations with each other, the selection of measurements should be tailored to the specific research question. In addition to these methodological implications, the authors emphasize that people tend to harbor more negative feelings towards out-party elites than towards their supporters. This finding holds substantial significance when assessing the prevalence of affective polarization in the US: partisan hostilities are predominantly directed at the leaders of political parties rather than at ordinary supporters. Most citizens remain open to interacting with supporters of the opposing party, even if they harbor negative sentiments towards their party’s leaders and the out-party in general.
Questions left unanswered
Do highly polarized citizens also hold more negative views of other out-groups, or are their animosities towards different out-groups unrelated to one another? The authors emphasize the significance of other identities in people’s social and political interactions, such as gender or race. While the authors establish that partisan affiliation is not the primary identity for citizens, the question remains: how is affective polarization towards other out-groups, such as different races, urban/rural residents, and various other groups, interconnected with partisan hostilities? Methods & Analysis
Methods and Analysis
Was the study and its analyses pre-registered?: Yes
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
The authors do not discuss potential differences in how affective polarization measures are conditional on partisan affiliation. It might be the case that these measures are more or less associated with one another in one partisan group than in another. Furthermore, the less negative evaluation of out-party supporters than their elites might be mainly driven by one partisan group and might not be present in the other.
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
Druckman, J. N., & Levendusky, M. S. (2019). What Do We Measure When We Measure Affective Polarization? Public Opinion Quarterly, 83(1), 114–122.
Bibtex
1
2
3
4
5
6
7
8
9
10
11
12
Bibtex entry:
@article{Druckman.2019,
author = {Druckman, James N. and Levendusky, Matthew S.},
year = {2019},
title = {What Do We Measure When We Measure Affective Polarization?},
pages = {114--122},
volume = {83},
number = {1},
issn = {0033-362X},
journal = {Public Opinion Quarterly},
doi = {10.1093/poq/nfz003},
}