Being Partisan
Reluctant Partisans, Not Undercover Partisans: Why Americans Increasingly Identify as Independent
Alex Tolkin
Public Opinion Quarterly, forthcoming
Abstract:
In the United States, partisanship has a greater impact on political behavior than any other social identity. However, the proportion of Americans who actively identify with a party, as opposed to Independents who “lean” toward one, is at its lowest point in decades. Existing research offers two explanations for why: First, being seen as partisan is not socially desirable, so people identify as Independent despite covertly preferring one party. Second, nonidentification may be a result of viewing both parties negatively. Using eight waves of panel data from 2016 to 2022, I compare the strength of these competing explanations. I only find support for the explanation-based evaluations of the parties. People switch from partisans to Independents when they view their former party more negatively, not when they seek to make a good impression.
Just a Little Melancholic, Maybe a Little Blue: Mental Health as an Emerging Political Identity
Lauren Van De Hey
Political Behavior, forthcoming
Abstract:
Is mental health an emerging political identity? In the first study that investigates experiencing mental illness as a political identity, I find that it is. Using a nationally representative survey of Americans fielded in the 2022 CES (N = 1,000), I answer the question: “For whom is mental illness a political identity?” I adapt Jardina’s work (White Identity Politics. 1st ed. Cambridge University Press, 2019) to create mental health identity and mental health alienation batteries that examine closeness with the ingroup, importance of identification to self, strength of identification within the ingroup, and alienation. I find that people who have experienced mental illness feel close to others who have experienced mental illness. They are also likely to self-categorize as having or having had a mental illness, share a sense of group consciousness with others who have or had mental illness, and recognize the need to work together to change laws that are unfair to people with mental illness. I further find that there is an emerging mental health political identity that is most pronounced among younger (Gen Z) and more liberal Americans. I also find that the emerging mental health identity has political predictors and political consequences. Those who self-categorize and have high scores on the mental health identity and/or alienation scales are just as likely to participate politically and use (social) media, on average, as those who do not self-categorize and have low scores on the mental health identity and/or alienation scales. In addition, there is a strong association between mental health categorization, identification, and alienation and the expressed desire for increased healthcare, education, and welfare spending. Finally, I find that the political predictors and political consequences for the emerging mental health identity differ from those for physical disability and serious physical illness categorization and identification. These findings have far-reaching consequences for mental health advocacy and the role mental health identity will play in the political sphere -- especially as Gen Z matures as a cohort.
DC On My Mind: Voter Evaluations of Nationalized Policy Positions in State and Local Elections
Derek Edward Holliday & Aaron Rudkin
Journal of Politics, forthcoming
Abstract:
Recent research demonstrates that national, state, and local election outcomes are becoming closely correlated, raising concerns that political accountability may be undermined by voters making electoral decisions using information irrelevant to subnational governance. The observational and aggregate data used to make these conclusions, however, are consistent with an alternative: the dimensions of national and subnational politics are correlated, and voters make informed decisions with imperfect but useful information. We distinguish between these possibilities using conjoint experiments varying the contested office level and candidate policy positions across national, state, and local politics. We find national issues more strongly affect candidate selection, but do not eliminate the effects of state and local issues. Additionally, national, state, and local issues affect candidate selection for all levels of office. Our results suggest U.S. politics is nationalized, but not fundamentally broken, with meaningful opportunities for the enforcement of political accountability in low-information settings.
Direct democracy and political extremism
Nicolas Schreiner & Alois Stutzer
Economica, forthcoming
Abstract:
We study how citizens' right to directly decide on policies through popular initiatives affects the attractiveness of extreme candidates in representative elections. In our theoretical framework, single prominent policy issues on which individual voters hold extreme views get a large weight in their assessment of candidates, thereby favouring ideologically extreme ones. If citizens can decide the controversial policy issues separately on the ballot, then this decouples the issues from legislative politics, and moderate candidates become relatively more attractive to voters. We apply our theory to US state legislative elections, and find that ideologically extreme candidates receive significantly lower voter support in initiative than in non-initiative states. This holds in particular for states with low qualification requirements for initiatives. In concurrent elections for the US House of Representatives, we do not observe this difference in the electoral success of extreme candidates between initiative and non-initiative states. The effect seems to be partly mediated by lower campaign donations to extreme candidates.
Will the Truth Free Us from Misinformation?
David Godes
Management Science, forthcoming
Abstract:
We study the impact of truthful revelation mechanisms, like fact checking, on the incentives for media outlets to produce misinformation. The existing literature suggests that such interventions will always reduce this incentive. In this paper, we consider a context in which consumers benefit both, and to varying degrees, from the truth contained in news reports and from sharing, discussing, and reacting to news stories that, independent of their truthful content, reinforce and enhance their social identity. In such a setting, we find that truthful revelation will not always decrease, and may in fact increase, the incentives for biased reporting. We identify two distinct mechanisms behind this effect. First, we show that truthful revelation may improve the ability of a biased media outlet to signal its type to consumers. Second, we find that heterogeneous preferences for misinformation endow the outlet with a demand function that may imply an increase or decrease in the incentives to produce biased news due to the stochasticity implied by the fact checking process. Finally, we provide two empirical examples that offer results that are consistent with the idea that media outlets are unlikely to reduce, and may increase, bias when they expect the truth to be revealed imminently.
Educational Polarization in American Politics: More than Just a Diploma Divide
Joshua Zingher
American Politics Research, forthcoming
Abstract:
The American electorate is divided along educational lines. The 2024 presidential election was marked by the largest-ever educational divide in vote choice. Typically, we frame this divide as a split between people with college degrees and those without. In this paper, I argue that there is significant polarization along the entire educational attainment spectrum. I show that the divide between BA and graduate degree holders is just as pronounced on many dimensions as the divide between BA holders and high school graduates. This point is especially critical because graduate degree holders are the fastest-growing educational category, moving from 9% of the population in 2000 to 14.5% today. I use ANES and US Census data to show that educational polarization encompasses political attitudes, engagement, and vote choice. As people become more educated, they become more liberal on social, cultural, and economic issues, become more likely to participate in politics, and support the Democrats. I also show that educational polarization combined with demographic changes has dramatically reshaped the party coalitions, with the Democratic Party’s base becoming more educated while the Republican coalition is little changed. I also show that the proportion of the population with a graduate degree was one of the strongest demographic predictors of county-level vote share in the 2024 election. All told, the American electorate is polarized along the entire educational attainment spectrum.
The misery of misbelief: People are more disturbed by others' false beliefs than by differences in beliefs
Andras Molnar & George Loewenstein
Political Psychology, June 2026
Abstract:
Belief homophily -- the tendency to associate with others who hold similar beliefs and the distaste for different beliefs -- is often seen as a major cause for belief-based social segregation and polarization. We question, however, whether social scientists have been correct in identifying belief-homophily as the primary force driving these pernicious social effects. We argue that when people face others who hold beliefs different from their own, they find these encounters disturbing, primarily when they are convinced that others' beliefs are false. In four pre-registered online studies (N = 2027 U.S. adults) featuring self-recalled experiences and vignette scenarios, we find that participants express stronger negative feelings when others hold false beliefs, compared to when others' beliefs are merely different from their own. We also document that higher confidence that others hold false beliefs evokes more negative emotions, triggers stronger avoidance behaviors, and reduces people's desire to form any kind of relationship with others. These findings highlight the possibility that many of the effects that have been previously attributed to belief homophily may be better explained by the desire to avoid others holding false beliefs.
Support for Political Violence among the American Public: A Split-ballot Experiment
Christopher Bader, Joseph Baker & Peter Simi
Terrorism and Political Violence, forthcoming
Abstract:
Assessing support for political violence in the United States remains a major point of focus. However, generating reliable population-level estimates of support for political violence has proven difficult. In this study, we fielded a split-ballot survey experiment comparing two measures of support for political violence. Specifically, we examine support for abstract and unspecified forms of political violence, as well as the willingness to engage in concrete actions, such as damaging property, to support political goals. As expected, we find greater support for abstract political violence than for a personal willingness to engage in political violence. In our multivariable analyses, we found that younger people were more likely to endorse both expressions of political violence. Regarding political ideology and party identification, we found a complex set of results. While the results related to ideology suggest people who identify as “very liberal” were more likely to endorse both expressions of political violence, party identification showed that independents, not Democrats or Republicans, were more likely to endorse both expressions of political violence. As such, we argue that alienation from conventional political parties is an important feature that deserves more attention.
Has Political Violence Affected Americans’ Commitment to Democratic Participation? The Mitigating Role of Neighborhood Social Cohesion
James Piazza
Social Science Quarterly, May 2026
Methods: The study employs an original online survey experiment of US residents that exposes treated subjects to information showing that political violence is increasing in the United States. It then measures subjects’ commitment to core democratic participatory behaviors. Treatment exposure is also interacted with an index measuring neighborhood social cohesion to test for moderation effects using ordinary least squares estimations. Demographic and attitudinal confounders are also considered in the analysis.
Results: The findings reveal that subjects treated with information showing that political violence is increasing in the United States are around 5.1% less likely to commit to democratic participation, compared with control group subjects. However, treated subjects from socially cohesive neighborhoods are around 162% more likely to respond to increased political violence by continuing to commit to democratic political participation.
The structural origins of the conservative online media niche, US Twitter 2022
Martin Arvidsson, Pablo Bello & Marc Keuschnigg
Network Science, April 2026
Abstract:
We investigate why conservative online news media are often seen as niche, whereas liberal outlets have ideologically broader audiences. We examine two explanatory mechanisms for this asymmetry. The behavioral explanation focuses on differences in homophily, where one ideological camp would be exposed to more cross-cutting content due to more diverse networking preferences. The structural explanation highlights how a platform’s user base places some in the minority, naturally exposing them to more cross-cutting content. We analyze network exposure and sharing of news media content among 420,000 US Twitter users in 2022, prior to Musk’s acquisition of the platform. We find that conservative users, as the minority, were overexposed to cross-cutting media content through their network contacts, while liberal users, as the majority, were underexposed. Consequently, liberal media were shared across party lines, while conservative media were overlooked by liberals and circulated mostly within a tight network of conservative accounts. This apparent paradox suggests that although conservatives primarily engage with their own media, liberal outlets attract a broader audience, including many conservatives. By combining observational data with simulated benchmarks, we find that the structural mechanism plays a primary role in the observed asymmetry, as exposure to liberal content extends farther into conservative online communities.