Friends at the Party
Durably reducing partisan animosity through multiple scalable treatments
Matthew Hall et al.
Proceedings of the National Academy of Sciences, 13 May 2025
Abstract:
Recent research has identified several effective strategies for reducing Americans’ animosity toward supporters of opposing political parties. However, whether these strategies can durably reduce partisan animosity in a scalable manner and in everyday life remains unclear. We bridge the gap between prior research and useful application by assessing whether exposure to multiple, scalable treatments that portray inparty and outparty members interacting positively, receiving accurate information about one another’s views, and learning about cross-party similarities can a) durably shift partisans’ sentiments and b) influence partisans’ sentiments toward specific, personally known others encountered in everyday life -- not only general, hypothetical, or one-off rival partisans. In a longitudinal survey experiment, we find that exposure to three brief, scalable treatments over a week reduces partisan animosity, with effects persisting for at least a month. Moreover, the treatments durably ameliorate animosity toward both general outparty members and a personally known outparty member, specified prior to the treatments. These findings suggest promising avenues for redressing social divisions in real-world contexts.
Asymmetric polarization: The perception that Republicans pose harm to disadvantaged groups drives Democrats’ greater dislike of Republicans in social contexts
Krishnan Nair et al.
Journal of Personality and Social Psychology, forthcoming
Abstract:
Given growing political polarization in recent years, partisan dislike -- defined as the negativity that individuals display at the prospect of having close social relations with supporters of the other party -- has received increasing attention. While traditional work in social and political psychology has held that conservatives display greater outgroup hostility than liberals, the worldview conflict perspective suggests that both groups similarly express hostility toward value incongruent outgroups. Contradicting both established perspectives, we present evidence across five preregistered studies (and two additional studies reported in the Supplemental Materials) conducted between 2022 and 2023 -- two social media field experiments (N = 10,000) examining actual behavior and five survey-based studies (N = 2,443) operationalizing partisan dislike in various ways (e.g., blocking on social media, rating the likability of various targets, and evaluating hiring suitability) -- that Democrats (i.e., liberals) dislike Republicans (i.e., conservatives) more than vice versa. We provide a potential explanation for this phenomenon by extending the worldview conflict perspective to account for asymmetries in how moralized specific values are among two conflicting groups at a given point in time. Specifically, we theorize that in light of recent social trends in the modern-day United States, the moralized belief that counter-partisans pose harm to disadvantaged groups, particularly racial/ethnic minorities, has become an asymmetric contributor to partisan dislike among Democrats. We found support for our theory across both measurement-of-mediation and experimental-mediation approaches and in both field experimental and survey data. Overall, this work advances research on ideology and outgroup hostility and extends the worldview conflict perspective to better explain partisan dislike.
Double standards in judging collective action
Nils Reimer et al.
Journal of Experimental Psychology: General, forthcoming
Abstract:
Collective action is a powerful force driving social change but often sparks contention about what actions are acceptable means to effect social change. We investigated double standards in judging collective action -- that is, whether observers judge the same protest actions to be more acceptable depending on who the protesters are and what they are protesting. In two studies, we used item response theory to develop an instrument of 25 controversial protest actions to measure where people draw the line between acceptable and unacceptable forms of collective action. In three preregistered experiments (N = 2,776), we found no consistent evidence for ingroup bias in terms of social class when judging protests for workers’ rights (Experiment 1), in terms of race when judging protests for and against defunding the police (Experiment 2), and in terms of gender when judging protests for and against restricting abortion (Experiment 3). Instead, we found that progressive participants (Experiments 1–3) who rejected system-justifying beliefs (Experiments 1 and 2) considered the same protest actions more acceptable when a cause aligned with their ideological orientation (for workers’ rights, for defunding the police, against restricting abortion) than when it did not (against defunding the police, for restricting abortion). Conservative participants considered the same actions somewhat more acceptable when protesters supported, rather than opposed, restricting abortion (Experiment 3) but considered all protest actions, for and against defunding the police, equally unacceptable (Experiment 2). Our findings have theoretical and practical implications for understanding the often-divided response to social movements.
Disagreement Spillovers
Giampaolo Bonomi
University of California Working Paper, March 2025
Abstract:
As US political party leaders increasingly take stances both on economic and cultural (i.e., social policy) issues, the economic views of opposite cultural groups are growing apart. This paper explores a novel explanation for this phenomenon. I provide experimental evidence that adding social policy content to a policy message pushes those disagreeing with the social policy to disagree also with the economic content of the message. In contrast, I find that agreement with the social policy does not trigger agreement on economics, suggesting systematic deviations from Bayesian explanations. I propose a model of identity-based belief updating that predicts the main regularities found in the experiment. Finally, I shed light on opinion leaders’ incentives to strengthen the association between social policy and economic policy views.
Measuring Perceived Slant in Large Language Models Through User Evaluations
Sean Westwood, Justin Grimmer & Andrew Hall
Dartmouth College Working Paper, May 2025
Abstract:
As LLMs become the default interface for search, news, and everyday problem-solving, they may filter and frame political information before citizens ever confront it. Identifying and mitigating partisan “bias” -- output with a systematic slant toward a political party, group, or ideology -- is therefore a growing concern for researchers, policymakers, and tech companies. Existing methods often treat political slant as an objective property of models, but it may vary depending on the prompt, reader, timing, and context. We develop a new approach that puts users in the role of evaluator, using ecologically valid prompts on 30 political topics and paired comparisons of outputs from 24 LLMs. With 180,126 assessments from 10,007 U.S. respondents, we find that nearly all models are perceived as significantly left-leaning -- even by many Democrats -- and that one widely used model leans left on 24 of 30 topics. Moreover, we show that when models are prompted to take a neutral stance, they offer more ambivalence, and users perceive the output as more neutral. In turn, Republican users report modestly increased interest in using the models in the future. Because the topics we study tend to focus on value-laden tradeoffs that cannot be resolved with facts, and because we find that members of both parties and independents see evidence of slant across many topics, we do not believe our results reflect a dynamic in which users perceive objective, factual information as having a political slant; nonetheless, we caution that measuring perceptions of political slant is only one among a variety of criteria policymakers and companies may wish to use to evaluate the political content of LLMs. To this end, our framework generalizes across users, topics, and model types, allowing future research to examine many other politically relevant outcomes.
How Cross-Cutting Ties Reduce Affective Polarization: Evidence from Latino Americans
Rongbo Jin et al.
Political Behavior, forthcoming
Abstract:
Affective polarization -- that is, personal dislike and distrust between Democrats and Republicans -- is argued to arise, at least in part, from fewer cross-cutting ties that bridge Democrats and Republicans. We argue that this phenomenon might be specifically relevant to non-Latino white Americans, but less so to Latinos who form a politically diverse group with strong social ties that unite them. Across six years of American National Election Studies data and original survey data from twelve different states across the country, we first show that cross-cutting group memberships predict warmer out-party affect. We then show, across our multiple datasets, that Latinos hold more cross-cutting ties than do non-Latino whites. Further, our data reveal that Latinos consistently hold warmer views of the out-party. Finally, we show with novel survey data that Latinos are especially warm toward out-partisans from within the Latino community. We conclude that affective polarization is less prevalent among Latino Americans who make up the fastest-growing proportion of the American electorate. More broadly, we argue that political behavior among white Americans should not be generalized across ethnic groups without considering unique characteristics of each group.
Why can't we be friends? Untangling conjoined polarization in America
Julie Norman & Beniamino Green
Political Psychology, forthcoming
Abstract:
Party affiliation in the United States is increasingly aligned with multiple social identities. In this era of social sorting, to what extent is polarization motivated by partisan identities compared to other overlapping identities in non-political settings? Using a conjoint survey experiment to examine the multivariate nature of affective polarization, we find that political identity outweighs all other social identities in informing citizens' attitudes and projected behaviors towards others. In addition, we find that partisanship usually outweighs ideology, but ideology matters in driving polarization; Democrats dislike Republicans more than Republicans dislike Democrats; and partisan identity has a particularly strong effect on racial and religious biases and preferences. Contrary to assumptions, cross-cutting identities do not appear to dampen social polarization. We also find that, while out-group animosity is stronger than in-group sentiment in abstract attitudinal measurements, the results are mixed in interpersonal behavioral measurements, suggesting that partisan animus between citizens at the community level may be more nuanced than often suggested.
Social identity shapes antecedents and functional outcomes of moral emotion expression
William Brady & Jay Van Bavel
Journal of Experimental Psychology: General, June 2025, Pages 1505-1522
Abstract:
There is increasing evidence that moral and emotional rhetoric spreads widely on social media and is associated with intergroup conflict, polarization, and the spread of misinformation. However, this literature is largely correlational, making it unclear why moral and emotional content drives sharing and conflict. In this research, we examine the causal impact of moral–emotional content on sharing decisions and how social identity shapes the antecedents and functional outcomes of decisions to share. Across five preregistered experiments (N = 2,498), we find robust evidence that the inclusion of moral–emotional expressions in political messages increases intentions to share the messages on social media. Moreover, individual differences in the strength of partisan identification and ideological extremity are robust predictors of sharing messages with moral–emotional expressions, even when accounting for attitude strength. However, we only found mixed evidence that brief manipulations of identity salience increased sharing. In terms of functional outcomes, when partisans choose to share messages with moral–emotional language, people perceive them as more strongly identified among their partisan ingroup but less open minded and less worthy of conversation with outgroup members. These experiments highlight the causal role of moral–emotional expression in online sharing intentions and how such expressions in online networks can serve ingroup reputation functions while hindering discourse between political groups.
Affect, Not Ideology: The Heterogeneous Effects of Partisan Cues on Policy Support
Sam Fuller, Nicolás de la Cerda & Jack Rametta
Political Behavior, forthcoming
Abstract:
How do individuals process political information? What behavioral mechanisms drive partisan bias? In this paper, we evaluate the extent to which partisan bias is driven by affect or ideology in a three-pronged approach informed by both psychological theories and recent advances in methodology. First, we use a novel survey experiment designed to disentangle the competing mechanisms of ideology and partisan affect. Second, we leverage multidimensional scaling methods for latent variable estimation for both partisan affect and ideology. Third, we employ a principled machine learning method, causal forest, to detect and estimate heterogeneous treatment effects. Contrary to previous literature, we find that affect is the sole moderator of partisan cueing processes, and only for out-party cues. These findings not only contribute to the literature on political behavior, but underscore the importance of careful measurement and robust subgroup analysis.
Polarization and state legislative elections
Cassandra Handan-Nader, Andrew Myers & Andrew Hall
American Journal of Political Science, forthcoming
Abstract:
US state legislatures are critical policymaking institutions that are increasingly polarized, yet data and measurement limitations have prevented researchers from understanding how state legislative elections contribute to this polarization. To address this gap, we construct new measures of candidate ideology based on campaign contributions and roll-call votes, and we use them to offer the first systematic study of the relationship between candidate ideology and electoral outcomes in primary and general elections in state legislatures, 2000–2022. We find that the set of people running for state legislature has polarized substantially in recent decades. More-moderate candidates enjoy a meaningful advantage in contested general elections, but that advantage has declined somewhat in recent years. At the same time, more-extreme candidates are favored in contested primary elections. These new measures and data will allow researchers to build on these basic findings to understand how elections function in lower information, lower salience environments like American state legislatures.
How the Engagement of High-Profile Partisan Officials Affects Education Politics, Public Opinion, and Polarization
David Houston & Alyssa Barone
Education Finance and Policy, forthcoming
Abstract:
What happens to public opinion when prominent partisan officials intervene in education policy debates? We analyzed the results of 18 survey experiments conducted between 2009 and 2021 with nationally representative samples of U.S. adults. In each experiment, some respondents were randomly assigned to receive the current U.S. president's position on a specific education policy before all respondents were asked to indicate their support or opposition to that policy. Our results indicated that the engagement of high-profile partisan officials typically did little to move public opinion in the direction of the cue-giver's preferred policies. Instead, the chief consequence was increased polarization among the public along partisan lines. A key exception applied to endorsements of policies that diverged from the traditional position of the cue-giver's own party, which tended to shift aggregate public opinion modestly in favor of those policies. Such cross-party cues also had nontrivial de-polarizing consequences.