Case Study 1: Affective Polarization in the United States — Causes, Consequences, and Measurement

Overview

Affective polarization — the growing dislike, distrust, and animosity that Americans feel toward members of the opposing political party — has become one of the most extensively studied phenomena in contemporary American political science. Unlike ideological polarization (disagreement about policy), affective polarization operates at the level of identity, emotion, and social perception. This case study examines the research on affective polarization in depth, with particular attention to its measurement, causes, consequences for democratic epistemics, and implications for misinformation susceptibility.

Research Background: The Iyengar Research Program

The systematic study of affective polarization in the United States has been most associated with Shanto Iyengar of Stanford University and his collaborators, particularly Sean Westwood (Dartmouth), Neil Malhotra (Stanford), Yanna Krupnikov (Stony Brook), and Eli Finkel (Northwestern). Their research program, spanning from approximately 2010 to the present, has established the empirical reality of dramatic affective polarization and has begun to unpack its causes and consequences.

Iyengar and Westwood's landmark 2015 paper "Fear and Loathing Across Party Lines: New Evidence on Group Polarization" in the American Journal of Political Science established several key findings:

1. Partisan animus now rivals, and on some measures exceeds, racial prejudice as a source of discriminatory behavior in the United States.
2. Affective partisan polarization is driven by negative affect toward the out-party as much as or more than positive affect toward one's own party.
3. Partisan identification has become a strong predictor of discriminatory decision-making in everyday settings.

Measurement Methods

The American National Election Studies (ANES) Feeling Thermometers

The primary longitudinal data source for tracking affective polarization is the American National Election Studies, conducted before and after each presidential and midterm election since 1948. The ANES includes "feeling thermometer" items asking respondents to rate their feelings toward various groups, candidates, and parties on a 0-to-100 scale, where 0 represents "very cold or unfavorable" and 100 represents "very warm or favorable."

ANES data from 1978 to 2020 shows a striking monotonic decline in partisan warmth toward the opposing party:

These numbers represent a dramatic collapse in cross-partisan warmth: from slightly below neutral in the 1970s and 1980s to well into cold territory by the 2010s and 2020s.

The Party Dislike Question

In addition to feeling thermometers, the ANES has periodically asked about specific attributes respondents associate with party members. The data show that by 2016, large majorities of partisans rated the opposing party's members as:

• Unintelligent: 62% of strong Democrats and 64% of strong Republicans
• Dishonest: 71% of strong Democrats and 72% of strong Republicans
• Closed-minded: 72% of strong Democrats and 76% of strong Republicans

These attribute-based assessments are particularly significant for understanding misinformation dynamics: if you believe the opposing party is systematically dishonest and closed-minded, you are pre-disposed to discount information from that party or from media associated with it.

Experimental Methods: Cross-Partisan Discrimination

Iyengar, Westwood, and colleagues have moved beyond survey self-reports to use behavioral experiments to measure discrimination based on partisan identity. In one widely cited paradigm, participants are asked to allocate scholarship funds between two equally qualified student applicants, where one is identified as a Democrat and one as a Republican (or where GPA and other credentials are varied alongside partisan identity).

Key findings from these experiments:

• Partisan identity is a stronger predictor of scholarship allocation than race, supporting the claim that partisan discrimination now rivals racial discrimination in magnitude.
• Both Democrats and Republicans discriminate in favor of co-partisans, but the discrimination increases with the strength of partisan identification.
• The discrimination operates through motivated reasoning: partisans find in-party applicants more "qualified" despite identical objective credentials.

Causes of Affective Polarization

Researchers have proposed multiple, complementary explanations for the rise of affective polarization. The causal story is complex because these explanations are not mutually exclusive; most likely contribute.

Partisan Sorting

One major driver is "partisan sorting" — the process by which Americans have increasingly sorted themselves into the "correct" party based on their racial, religious, and regional identities. As recently as the 1970s, the Democratic Party contained large numbers of conservative white Southerners and the Republican Party contained significant numbers of liberal Northeasterners. As these anomalous combinations dissolved, the parties became more internally homogeneous and externally distinctive.

Matthew Levendusky's "The Partisan Sort" (2009) documents this process extensively. Sorting matters for affective polarization because it makes partisan identity more cognitively accessible and socially meaningful — when your partisan identity aligns with your racial, religious, and geographic identity, partisan conflict feels like social identity conflict, which generates stronger emotional responses.

Geographic Sorting

Americans have also sorted geographically, with Democrats increasingly concentrated in urban areas and Republicans in rural and exurban areas. Bill Bishop's "The Big Sort" (2008) documented this geographic sorting, which has reduced the frequency of cross-partisan social contact. As fewer Americans have meaningful friendships, work relationships, or family ties with members of the opposing party, partisan opponents become increasingly abstract "the Other" rather than known individuals — making negative stereotyping more persistent.

Negative Partisanship

Alan Abramowitz and Steven Webster's research on "negative partisanship" documents that contemporary American partisan identification is increasingly driven by dislike of the opposing party rather than positive attachment to one's own party. Americans vote for their party primarily to prevent the other party from winning, not out of enthusiasm for their own. This "negative" motivational structure intensifies affective polarization because it makes inter-partisan conflict feel existential: it's not just that my party should win but that their party must be stopped.

Media Environment Changes

The transformation of the media environment from a centralized broadcast system (three national TV networks, local newspapers with broad audiences) to a fragmented, partisan, digital ecosystem has created information environments in which partisans can largely avoid exposure to information that challenges their worldviews or humanizes opposing partisans.

Matthew Levendusky and Neil Malhotra's research shows that partisan media exposure increases affective polarization. Eli Finkel et al.'s 2020 meta-analysis in the Proceedings of the National Academy of Sciences found that social media use is associated with increased affective polarization, though the magnitude of this effect is debated.

Political Elite Behavior

Finally, the behavior of political elites — politicians, party activists, media figures — contributes to affective polarization by modeling and incentivizing partisan hostility. When prominent politicians demonize the opposing party, their supporters observe and internalize this behavior. Research on contagion of extremism suggests that elite hostility "permissions" and amplifies mass-level animosity.

Consequences for Democratic Epistemics and Misinformation

The Credibility Effect

The most direct consequence of affective polarization for misinformation is its effect on source credibility judgments. When people dislike the opposing party intensely and believe its members are dishonest and unintelligent, they become systematically less likely to credit information that comes from sources associated with that party — including news organizations, fact-checkers, and experts who are perceived as politically aligned with the opposition.

Nyhan and Reifler's research on "backfire effects" (though the robustness of this specific effect has been questioned by subsequent research) documented that corrections to political misinformation could sometimes increase rather than decrease belief in the false claim, particularly for strongly identified partisans. Even where backfire effects are not systematically observed, corrections from politically misaligned sources are systematically less effective than corrections from same-party sources.

Partisan Fact-Checking

Affective polarization has created a "partisan fact-checking" dynamic in which audiences evaluate fact-checks not by the quality of the evidence but by the perceived political alignment of the fact-checker. PolitiFact, FactCheck.org, and mainstream newspaper fact-checking operations are widely perceived by Republican partisans as liberal-aligned, substantially reducing their effectiveness with this audience.

This dynamic creates a structural problem for fact-checking as a misinformation intervention: in a highly affectively polarized environment, fact-checkers can only be effective within the partisan communities that trust them, while being dismissed by the communities whose misinformation beliefs they most need to reach.

The "Liar's Dividend"

Kathleen Hall Jamieson and others have noted that affective polarization enables what might be called a "liar's dividend": when audiences are pre-disposed to believe that the opposing party is dishonest, true unflattering claims about opposing politicians are treated as credible while false unflattering claims are believed for the same reason. Conversely, true flattering claims about one's own party are accepted while false ones are also accepted. The distinguishing mechanism — the evidence — is overridden by partisan prior expectations.

Social Isolation and Information Ecosystem Effects

The social distance created by affective polarization — the declining inter-partisan friendships, family ties, and social contacts — reduces the informal social correction mechanisms that historically played an important role in countering misinformation. In communities with cross-partisan social networks, false claims spread by one partisan community are more likely to encounter correction from cross-partisan social contacts. As these networks thin, misinformation spread within partisan communities faces less resistance.

What Can Reduce Affective Polarization?

Correcting Perceptual Misperceptions

A significant component of affective polarization is based on inaccurate perceptions of the opposing party — specifically, the perception that the opposing party is more extreme than it actually is. Research by Douglas Ahler and Gaurav Sood, and separately by More in Common's "Hidden Tribes" project, has documented that both Democrats and Republicans dramatically overestimate the proportion of the opposing party that holds extreme positions.

Several studies have found that correcting these misperceptions — showing partisans the actual distribution of the opposing party's views — can reduce affective polarization, at least temporarily. This finding suggests that some portion of affective polarization is knowledge-based and potentially addressable through accurate information.

Cross-Partisan Contact

Contact theory (Allport, 1954), applied to partisan polarization, suggests that meaningful contact with members of the opposing party can reduce partisan animosity, particularly when that contact involves equal status, shared goals, institutional support, and acquaintance potential. Organizations like Braver Angels and Living Room Conversations have developed structured cross-partisan dialogue programs based on these principles.

Initial evaluations of these programs are cautiously positive: participants show reduced partisan animosity immediately after structured cross-partisan encounters. Whether these effects persist and generalize is less clear, and the populations reached by voluntary cross-partisan dialogue programs may be systematically different from those with the highest levels of polarization.

Elite Modeling

If elite behavior contributes to mass affective polarization, then changes in elite behavior — politicians who model cross-partisan respect, media figures who portray out-partisans as human, etc. — might reduce it. This is a hopeful hypothesis with limited empirical testing, but the logic follows from the social learning literature: people update their social norms based on what they observe elites doing.

Implications for Media Literacy Education

The affective polarization research has several important implications for media literacy education:

1. Teaching partisan identity awareness: Media literacy education that helps students recognize how their own partisan identity affects their credibility judgments and information processing is more important than ever.

2. The limits of fact-checking: Educators should help students understand why fact-checking is less effective in polarized environments, without drawing the conclusion that truth is entirely relative.

3. The misperception curriculum: Teaching students about systematic misperceptions of out-party positions — and providing them tools to check actual positions against perceived ones — addresses a specific, correctable component of affective polarization.

4. Humanizing the other: Media literacy education that includes authentic exposure to well-reasoned arguments from the opposing political tradition, presented charitably, can partially counteract the dehumanizing effects of partisan media exposure.

Discussion Questions

1. The data shows that partisan animosity now rivals racial prejudice as a source of discriminatory behavior. What are the implications of this for our understanding of American democracy and civil society?

2. Geographic sorting and media environment changes are both associated with increased affective polarization. Which do you think is more amenable to policy intervention, and why?

3. If correcting misperceptions about the opposing party can reduce affective polarization, why hasn't this approach been more widely implemented? What are its limits?

4. Negative partisanship — voting against rather than for — is associated with increased partisan hostility. What would it take to shift partisan motivation from negative to positive?

Sources and Further Reading

• 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.
• Finkel, E. J., Bail, C. A., Cikara, M., Ditto, P. H., Iyengar, S., Klar, S., ... & Druckman, J. N. (2020). Political sectarianism in America. Science, 370(6516), 533–536.
• Levendusky, M. S., & Malhotra, N. (2016). Does media coverage of partisan polarization affect political attitudes? Political Communication, 33(2), 283–301.
• Ahler, D. J., & Sood, G. (2018). The parties in our heads: Misperceptions about party composition and their consequences. Journal of Politics, 80(3), 964–981.
• Abramowitz, A. I., & Webster, S. (2016). The rise of negative partisanship and the nationalization of US elections in the 21st century. Electoral Studies, 41, 12–22.
• More in Common. (2018). Hidden Tribes: A Study of America's Polarized Landscape. More in Common.