Women consistently report lower levels of political interest and political ambition than men. Studies have examined how women politicians as role models can make women more politically engaged, but studies have not typically examined additional shared identities beyond gender. I propose that women politicians with whom women share additional identities will be perceived as more relatable and inspiring and, consequently, have stronger role model effects on women’s political engagement. Through a conjoint experiment (Study 1), I find that having additional shared identities with a woman politician increases perceptions of her as relatable and inspiring. However, a survey experiment with novel video treatments depicting a woman politician’s story (Study 2) did not produce positive role model effects on women’s political interest, internal political efficacy, or nascent political ambition – regardless of additional shared racial and partisan identity with the woman politician. Moreover, the video treatment had remarkably different effects across White and Black women, regardless of additional shared racial and partisan identity. An observational analysis (Study 3) replicates the finding that additional shared racial and partisan identity with women political representatives is not systematically correlated with higher levels of political interest. These findings carry important implications for understanding women politicians’ role model effects and the potential of role models to increase women’s political engagement.
Despite the increasing popularity of podcasts as a source of news for many Americans, little is known about their precise impact. We conduct a mobile field experiment that examines how consumption of trustworthy news via daily podcasts affects attitudes about politics and the media over a one-week period. Our treatment, deployed at the end of the first of two survey waves, asks subjects to subscribe to a specific high-quality daily news podcast, NPR’s Up First. In addition, treated subjects are asked to turn on mobile push notifications for the podcast. The intervention successfully increased participants’ consumption of news from the podcast, with minimal substitution from other programs. We find that podcast consumption significantly increased event knowledge ( d = 0.29 ), but only for topics covered in that week’s episodes. Trust specifically in news from podcasts also increased ( d = 0.09 ), but not in mainstream news more generally. Our results shed light on outcomes linked to the unique possibilities of this medium and, more broadly, on the malleability of attitudes toward the media.
A much-cited finding in recent political science is that average citizens in established democracies appear to have little influence on public policy, while the affluent have strong influence. Yet, as critics have pointed out, average and affluent often want the same from government, meaning that even without influence, average citizens often get the policies they want ‘by coincidence’. This raises an important question: Do citizens care about ‘who governs’ if government regularly adopts their preferred policies anyway? Using a preregistered survey experiment, I show that political inequality causes just as much popular dissatisfaction with government when policy outcomes are highly consistent with a respondent’s policy preferences, as it does when policy outcomes are inconsistent with those preferences. While opinion-policy congruence remains an important predictor of satisfaction, citizens also appear to value political equality independently of policy outcomes. These findings suggest that coincidental representation may not fully compensate for perceived inequalities in political influence.
How do political elites shape public perceptions of what kinds of relationships are possible across partisan lines? While existing research has emphasized identity and contact as drivers of cross-partisan social ties, less is known about how elites influence beliefs about the feasibility of such relationships. Building on theories of social norms, I argue that elite cooperation serves as a visible signal that cross-party friendships are socially viable. By engaging in cooperation, elites model behaviors that redefine partisan boundaries and shape citizens’ second-order beliefs about inter-partisan relationships. To test this argument, I combine original 2019 surveys from Canada, Germany, and the United Kingdom with the Quantified Political Relationships (QPR) dataset, which tracks real-world episodes of elite cooperation. The results show that while elite cooperation has a modest average effect, its influence is significantly stronger among partisans of cooperating parties and among politically knowledgeable non-partisans. These findings demonstrate that elite cues shape not only policy attitudes or affection toward out-partisans but also perceptions of the viability of cross-partisan friendship. Overall, the paper highlights the reach and the limits of elite influence in sustaining the social fabric of democracy.
When deciding whether to support social policies, Americans often rely on a simple heuristic: whether those in need are deserving of assistance. While prior work has identified widely shared criteria for deservingness judgments, we argue that in an age of affective, identity-based polarization, these criteria are applied through a partisan lens. Though public policies are not geared specifically to benefit “Democrats” or “Republicans,” we suggest that Americans use group-based cues to identify the likely partisan composition of those who benefit from a particular policy and, in turn, evaluate beneficiaries in ways that reflect inparty bias. Using a series of survey experiments, we provide evidence for three claims. First, we show that Americans hold clear stereotypes about which social groups are associated with the Democratic and Republican parties. Second, wsse demonstrate that partisans generally view “inparty” social groups more favorably than “outparty” groups with respect to core deservingness criteria. Third, we show that these group-based perceptions shape evaluations of social policy beneficiaries, as partisans are more likely to support otherwise identical policies if told that the benefits will go to inparty groups. These results highlight the role of group identity in structuring public opinion about redistribution. They also suggest that policymakers can influence support for social programs by framing them in group-based terms, allowing partisans to infer whether inparty or outparty groups are likely to benefit.
In a recent paper, Perez Brower (2024) argues that identity framing of policy agendas shapes women’s willingness to become more politically involved. She conducts two survey experiments to examine whether intersectional identity framing—messaging by advocacy organizations that links policy issues to individuals’ multiple identity characteristics—alters self-reported political engagement and disengagement. Our re-evaluation of Perez Brower’s (2024) data and analysis reveals an omitted control group, miscoding of one dependent variable, and the selective reporting of outcomes from analyses, including omissions of results which the author preregistered. Correcting these errors undermines, either partially or completely, all of the findings from Perez Brower (2024). Our investigation underscores the importance of preregistration, data replication, and public availability of replication materials and experimental documentation for reliable social science research.
Women are underrepresented across all levels of government, yielding profound public policy consequences. Yet, policy outcomes may also shape who runs for office. To align research on women candidates and leaders, we consider both parts of the broader political ecosystem. Research shows women hesitate to run because politics do not fulfill their communal goals. But governments’ communal policy priorities vary, producing substantial differences in women’s entrance into politics. Analyzing over 70,000 Brazilian candidates and municipal spending patterns, we show that communal spending patterns at the local level and policy outputs predict women’s emergence. We provide consistent evidence that women’s interest in pursuing political office varies across place and time.
This paper provides a response to a critique of my article, “Reframing Gendered Issues: Intersectional Identity Frames and Policy Agendas,” originally published by Political Behavior in 2022 (initially online, and later in print in 2024). The critique raises concerns about choices in the use of reference groups and dependent variables that deviate from pre-registration, and a coding error in one of the dependent variables. In this response, I address these concerns by explaining my choices of reference groups and dependent variables and discussing results for analyses with the alternate choices that were pre-registered. I acknowledge the coding error and provide a discussion of the corrected findings. Throughout, I stress that changes from pre-registration were made in response to substantive feedback received during the peer review process. I close with a brief reflection on lessons learned and how to move forward.
How do local attitudes and policies influence where refugees choose to live within host countries? This study investigates the internal or secondary migration preferences of Syrian refugees in Turkey and Germany. Adopting a multi-method approach, we conducted a choice-based conjoint experiment with over 5,000 participants in both countries, complemented by a content analysis of qualitative survey data collected in Turkey. The results of the experimental study reveal that, in both contexts, local attitudes and policies shape Syrian refugees’ destination preferences more significantly than economic and welfare-state factors or social ties. The Turkish qualitative data further demonstrate that Syrian refugees associate many local policies and attitudes primarily with exclusion, social isolation, discrimination, and heightened anxiety. They fear that such policies and attitudes may prevent them from developing a strong sense of belonging, leading them to avoid hostile localities. The convergence of experimental results across both countries suggests that the deterrent effect of hostile policies and discourses is robust. From both normative and policy perspectives, we conclude that local openness is essential for promoting refugees’ integration and psychosocial well-being.
Elections are not only about how many votes candidates win, but also about where those votes come from. Yet we know less about how the geography of electoral support changes when candidates coordinate across offices through electoral alliances. I study Brazil’s 2020 municipal elections, where mayoral and city-council races are held on the same day under distinct electoral rules, and argue that these alliances can reorganize executive support through a coalition-coordination mechanism: allied council candidates with geographically concentrated personal-vote bases increase mayoral vote concentration within their local strongholds, making legislative dominance areas partially overlap with the supported mayoral candidate’s dominance areas. Using polling place election returns, I construct fine-grained measures of within-municipality vote concentration and estimate polling place fixed-effects models linking mayoral concentration to the concentration of allied council candidates. To address endogeneity, I instrument allied concentration with a Friends-and-Neighbors measure based on where allied council candidates are registered to vote. The results show that increases in allied council candidates’ local overrepresentation translate into greater local overrepresentation for the supported mayoral candidate. The findings are robust to alternative spatial measures, controls for localized mayoral campaign spending, and placebo coalition assignments. Overall, the analysis shows that coalitions can shape not only aggregate electoral performance but also the territorial structure of representation by aligning executive and legislative strongholds.
Although extensive literature examines the relationship between descriptive and substantive representation in the legislatures of liberal democracies, it remains under-researched whether this relationship depends on multiparty competitive elections or can also be observed in single-party legislatures. Drawing on original datasets containing 75,797 proposals submitted to China’s National People’s Congress between 2008 and 2017, this article reveals that descriptive representation in single-party legislatures partly facilitates substantive representation. Specifically, deputies who are women, members of ethnic minorities, workers, farmers, or professionals tend to submit significantly more proposals related to their respective group-based issues. For certain nongroup-based issues, differences in the numbers of relevant proposals between specific pairs of deputy groups are largely consistent with the corresponding differences in citizen preferences. Additionally, shared experience, rather than top‑down directives from the Communist Party, social pressure, resource constraints facing part‑time deputies, leadership positions in women’s affairs, or mobilization in ethnic autonomous regions, is the mechanism most likely to explain the link between descriptive and substantive representation. These findings suggest that single‑party regimes can draw on this link to make legislatures more representative, thereby helping to gather public opinion information without introducing multiparty elections.
Multiracial populations increased faster than any single race in the most recent U.S. census. However, we know little about how this demographic shift might impact the political and racial attitudes of monoracial Americans. This paper addresses this question by examining how learning about the proportion of African Americans with Irish ancestry impacts the racial attitudes of white Irish Americans. Experimental data from the largest online survey of Irish Americans reveal that learning about African Americans with Irish ancestry reduces racial prejudice among white Irish Americans who identify with their Irish American identity, and that these effects are mostly driven by Republicans—i.e., those Irish Americans with higher baseline levels of racial prejudice. Open-ended responses highlight the potential for multiracialism to reduce the perceived distance between racial groups and undermine racially essentialist beliefs.
White evangelical Protestants support Republican Party presidential candidates at rates much higher than White Catholics and White mainline Protestants. I argue that local evangelical Protestant sermon content plays an important role in reinforcing this coalition. To explore this, I develop and analyze a novel random sample of 450 U.S. sermons from throughout 2022. Relative to Catholic and mainline Protestant sermons, evangelical Protestant sermons consist of cultural and elite threats that more closely map to Republican rhetoric. Using a series of survey experiments, I find that exposure to a sermon with a cultural and elite threat targeting an out-group decreases support for policies that would protect the targeted group’s rights. Moreover, this rhetoric is highly mobilizing for White evangelical churchgoers who identify as Independents or Republicans and who regularly attend Bible studies. These findings suggest that local evangelical sermon rhetoric plays an important role in reinforcing one of the most influential voting blocs in American politics.
Does affective polarization change how citizens view and participate in democracy? We posit that the answer to this crucial question depends on how affectively unpolarized citizens view and participate in democracy. Yet, the most common operationalization of affective polarization––measuring sympathy ratings for one party relative to another––does not differentiate between people’s absolute sympathy ratings for the parties, which may mask essential heterogeneity among unpolarized citizens. This paper theorizes the methodological origin of this heterogeneity, describes it, and highlights its implications. Cluster analyses and regressions from three nationally representative surveys in the US provide three conclusions. First, substantial shares of the US population are affectively unpolarized. Second, unpolarized citizens cluster into two groups: one feeling lukewarm and one feeling cold toward both parties. Third, these unpolarized groups report diverging democratic attitudes and participation patterns. We discuss the methodological implications and the need for future research on affectively unpolarized citizens.
How do Muslim citizens respond to immigrants who hold fundamentalist interpretations of Islam? Although the social science literature has examined public perceptions of Muslim fundamentalists, it has primarily focused on Western liberal democracies. Consequently, we know relatively little about how Muslim citizens in Muslim-majority societies evaluate immigrants with extreme religious beliefs. To address this gap, this paper provides empirical evidence on how Muslim nationals assess immigrants based on variations in religion and religious intensity. We conducted a conjoint experiment in Kazakhstan, a secular state with a Muslim-majority population, in which respondents evaluated profiles of hypothetical immigrants with randomly assigned characteristics. The results show that Muslim citizens express unfavorable attitudes toward fundamentalist Muslim immigrants. Moreover, more religious respondents are even more likely to disapprove of such immigrants. These findings suggest that Muslim citizens do not simply draw social boundaries based on shared religious affiliation. Rather, they distinguish between in-group and out-group members according to competing interpretations of what constitutes “Islam.” Higher levels of religiosity therefore do not necessarily translate into greater support for fundamentalist beliefs.
Mass opposition to governments implementing economic liberalization reforms remains a key puzzle in the political economy of transitions. This article investigates this puzzle by examining the political consequences of China’s large-scale privatization of state-owned enterprises in the late 1990s, which displaced over forty million workers while enriching politically connected elites. Combining firm-level layoff data with nationally representative surveys, we find that citizens in counties experiencing greater privatization layoffs exhibit persistently lower political trust and more contentious behavior more than a decade after the peak of privatization. This trust deficit does not dissipate even among laid-off workers whose material conditions substantially improved during China’s subsequent economic boom. Further analyses show that elite capture, social contract violation, and sociotropic anxiety jointly sustained this distrust. Short-term pain from economic reform can leave a lasting stain on political trust–even when the promised growth materializes.
The 2024 election cycle offers an opportunity to examine the influence and representation of Black women in electoral politics. Black women have been the most consistent and mobilized base for the Democratic party, playing a vital role in shaping the party's electoral outcomes. However, despite partisan loyalty to the Democratic Party, Black women have a variety of public opinions and political ideologies that differ from the mainstream and prevailing ideology within the Democratic Party. This article attempts to position qualitative research, specifically focus groups, among engaged and politically active Black women to best understand the contours of Black women's political views regarding the 2024 campaign. Despite 92% of Black women supporting the Democratic ticket, Black women voters are a politically heterogeneous group. By engaging with Black women, we can center their political agency, their organizing strategies, and their support for Black women candidates. This work relies on an intersectional framework, and the race-gendering of American politics, which best captures the intricacies within Black women's politics that become unclear in racial or gendered analyses. Our discussions reveal that focus groups provide valuable insight into the complexity of political attitudes and opinions among Black women voters.
How does the threat of adverse policy change influence affected individuals? To answer this question, we explore the responses of gun owners following salient mass shootings and after the election of Democratic presidents. We gather and merge data on the timing and location of mass shootings (2006–2020) and data on the timing of gun purchase background checks from (2006–2020). We exploit the exogenous timing of elections and the resulting determination of future partisan control of the White House over three closely contested presidential elections as well as the timing of mass shootings under a Democratic versus Republican administration. Using regression discontinuity designs, we find that the election and re-election of Barack Obama led to increases in gun sales of 70–210 per 100,000, while the election of Trump was not related to similar increases. Further, we find that salient mass shootings led to substantially higher gun sales during times with Democratic presidents.
