Article Summary
Introduction
Democratic norms are values and practices central to well-functioning democracies, but recent research suggests that public support for these norms is declining. These authors ask whether (and how) affective polarization plays a role in this decline, encouraging people to undermine democratic norms. The authors focus on two broadly important norms: constitutional protections (support for checks and balances, separation of powers, etc.) and political tolerance (the belief that all citizens should have equal rights to voting, free speech, etc.). They argue that affective polarization may shape support for these norms in two ways. First, people who are affectively polarized may be politically intolerant because they want to limit the other party’s power by depriving them of the chance to speak or vote. Second, political elites in power (congresspeople, the President, etc.) tend to dislike democratic norms like constitutional protections because those norms limit their political power; elites not in power tend to like those norms for the same reason—they don’t want their opponents in power to use that power. Party voters, especially those who are affectively polarized, tend to follow cues from political elites. As such, people who are affectively polarized will oppose democratic norms when their party is in power—but they will support those same norms when their party is not in power. The authors argue that the first process (partisan bias) primarily influences support for political tolerance, and the second process (taking cues from elites) primarily influences support for constitutional protections.
Analytical Approach
The authors test their expectations with a nationally representative survey of 2,815 Americans in 2019, a year in which Republicans controlled the presidency and were thus the party in power. The survey measured affective polarization with the standard set of measures (feeling thermometers, stereotype ratings of the parties, partisan trust, and social distance measures) combined into a single indicator of affective polarization. Then, respondents were asked sets of items capturing support for constitutional protections and political tolerance. These items did not mention parties or partisanship, so responses represent general support for the norm. The authors then conducted ordinary least squares regressions to examine whether affective polarization was associated with responses to each of the norm questions. These regressions controlled for partisanship, political knowledge, policy conservatism, education, race, gender, and religion. Most importantly, the authors interacted affective polarization with partisanship, allowing the relationship between affective polarization and norms to vary for members of different parties. This last step was crucial to assess the expected politicization dynamic (wherein only members of the party in power would support norm violations).
The authors then extend this analysis in three ways. First, they explore if the relationship between affective polarization and support for norms depends on respondents’ knowledge about politics. Conceptually, knowledgeable people are more likely to know the parties’ positions, strengthening the relationship between affective polarization and support for constitutional protection norms.
Second, they consider whether the same dynamics occur during a Democratic presidential administration. Using data from a large survey conducted in 2012, they examine a norm question similar to two of the items they used in their 2019 survey. The survey does not have a direct measure of affective polarization, so they rely on the strength of respondents’ party affiliation as an indicator of it instead.
Third, reanalyzing data from another study conducted in 2013 (i.e., also under a Democratic presidency), they examine the relationship between affective polarization and people’s support for tear gassing peaceful protestors from the opposing party.
Main Findings
The authors find that affective polarization is associated with support for democratic norms, and the relationship seems consistent with the elite-cue-taking account. When a Republican was president in 2019, Republicans who were affectively polarized supported democratic norms less, and Democrats who were affectively polarized supported democratic norms more. In both cases, this was particularly true for norms relating to constitutional protections. The partisan dynamic reversed in the 2012 data, when a Democrat was president. Then, affective polarization was positively associated with democratic norm support for Republicans (although it was not significantly associated with democratic norm support for Democrats). The norm question from 2012 primarily involved constitutional protections, so this provides further evidence for the cue-taking account.
There was no support for the partisan bias account. Affective polarization was largely unrelated to support for norms of political tolerance—and when it was, the pattern of results was apparently driven by the party in power (and thus more consistent with cue-taking than partisan bias). This was true in both the 2019 data and the 2012 data. However, the 2013 data complicate this conclusion slightly. There, the authors analyze responses to a question about political tolerance, finding that the party in power is not always the one less supportive of democratic norms. In 2013, affective polarization decreases Republicans’ support for a norm of allowing peaceful protests (i.e., a dynamic opposite to the cue-taking account, with the party not in power showing less support for a norm). This was basically similar for tolerance for protest in 2019, where more affectively polarized Republicans were less supportive of protest. Thus, the authors suggest that there may be a ‘partisan slant’ to some norms, where one party tends to support or oppose some norms more than the other party regardless of the party in power.
Implications
Overall, the paper finds suggestive evidence that affective polarization may be related to declining support for democratic norms. Even though political tolerance and constitutional protections are principles fundamental to U.S. democracy, many people seem to support them only when it benefits their own party. However, people are not uniformly biased—rather, they seem to follow the cues of political elites from their party. If elites stand to benefit from norms, as they do when they are not the party in power and those norms constrain the opposing party, the public tends to support norms more. But for elites in power, norms are constraining, and the public tends to oppose those norms more.
Questions left unanswered
The authors identify what they describe as a “partisan slant” in norms. In their examination of knowledge and of the 2013 data, they find that dynamics related to political tolerance seem to differ between parties and to follow the cue-taking account is some cases (rather than the expected bias account). The authors discuss this somewhat more in the supplementary materials, suggesting that Republican elites may be more likely to send cues about protests given that protests are a historically Democratic tactic. However, more work is necessary to examine the differences between types of norms.
The authors primarily discuss the relationship between affective polarization and support for norms as a correlational one. Some of the evidence the authors provide is more consistent with affective polarization causing support for norms: the cue-taking argument focuses on responsiveness to political elites, and they find that results tend to be stronger for politically knowledgeable respondents (who are presumably more exposed to elite cues). Similarly, they find that the pattern of results flips (responding to the party in/out of power) with data from different presidential administrations, which supports their argument that changes in the elite political environment drive public opinion. However, all of this evidence is indirect and whether affective polarization causes support for norms or the other way around—remains unclear, and future research must be done to explore it further.
Methods and Analysis
Was the study and its analyses pre-registered?: No
Did the study rely on proxy variables to measure polarization?: Yes
In the authors’ robustness test with the 2012 survey data, they use partisan strength as an indicator 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?: Yes
The title of the article suggests a causal direction from affective polarization to support for democratic norms, but the data is ultimately correlational and cannot provide much evidence about the direction of causality. The authors are otherwise appropriate in their use of correlational/causal language.
Limitations / Weaknesses
The title of this paper, “How Affective Polarization Undermines Support for Democratic Norms,” suggests a causal dynamic that the evidence in this paper is ultimately unable to support. As such, readers should be cautious in interpreting the findings.
The authors do not clearly report statistical significance for the interactions between affective polarization, partisanship, and knowledge in their first robustness test. In the supplementary materials, the interaction between affective polarization and political knowledge is significant for all four constitutional protection items. For Democrats, political knowledge strengthens the relationship between affective polarization and support for democratic norms. The evidence appears less clear for Republicans. The three-way interaction (between political knowledge, affective polarization, and a binary indicator for Republicans) appears to be significant for only one of the norm items. Political knowledge only strengthens the relationship between affective polarization and norms for Republicans very inconsistently. Moreover, this kind of three-way interaction is highly prone to statistical error, and the authors should discuss their findings explicitly in the context of this potential for error.
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., Druckman, J. N., Klar, S., Krupnikov, Y., Levendusky, M., & Ryan, J. B. (2021). How affective polarization undermines support for democratic norms. Public Opinion Quarterly, 85(2), 663–677. https://doi.org/10.1093/poq/nfab029
Bibtex
1
2
3
4
5
6
7
8
9
10
11
12
13
@article{10.1093/poq/nfab029,
author = {Kingzette, Jon and Druckman, James N and Klar, Samara and Krupnikov, Yanna and Levendusky, Matthew and Ryan, John Barry},
title = "{How Affective Polarization Undermines Support for Democratic Norms}",
journal = {Public Opinion Quarterly},
volume = {85},
number = {2},
pages = {663-677},
year = {2021},
doi = {10.1093/poq/nfab029},
url = {https://doi.org/10.1093/poq/nfab029},
eprint = {https://academic.oup.com/poq/article-pdf/85/2/663/40810644/nfab029.pdf},
}