Publications and Working Projects

Partisanship and the Pandemic: How and Why Americans Followed Party Cues on COVID-19

Mehlhaff, I. D.; Tarillo, M.C.; Vanegas, A. and Hetherington, M.J. (2024). "Partisanship and the Pandemic: How and Why Americans Followed Party Cues on COVID-19". Journal of Health Politics, Policy and Law, 49 (3): 351-374. https://doi.org/10.1215/03616878-11066336 

Is he/she/they Conservative? Political Implications of the Use of Gender-Fair Language

Under review. Available here.

Existing research shows that gender-fair pronouns can reduce discrimination instilled by language, but less is known about their political implications. I argue that the pronouns candidates use to refer to people of unknown gender function as heuristics of ideology. In contexts where gender-related issues are central to political divides, pronouns can also signal partisan identity. To test this, I use survey experiments in the United States and Mexico. Results show that gender-fair pronouns convey information about ideology in both countries. However, they are associated with a specific party only in the United States, where gender debates are closely tied to political identities. There, citizens’ willingness to support candidates using such pronouns depends on partisanship. In Mexico, where political groups do not have fundamental disagreements around gender, these effects are absent. Additional evidence shows that in the U.S. pronouns serve as partisan cues and have an impact even beyond political contexts.

Just Don't Talk About It: The Effects of Blocking Discussions About Controversial Political Topics In and Outside Classrooms

Coauthored with Bryce Hecht (UNC Chapel Hill)

Under review. Available here.
Presented at MPSA 2025. 

Deliberation is considered a cornerstone of democracy, with recent evidence showing that facilitating cross-party discussions reduces affective polarization and bolsters support for democratic norms. Yet little is known about the consequences of having an authority figure block such discussions, a practice that has become increasingly common. We examine this question in the university classroom, a setting often viewed as a marketplace of ideas but where professors face growing incentives to avoid political controversy. We argue that universities are uniquely positioned as forums for cross-partisan dialogue, in contrast to other settings such as the workplace. We field two pre-registered vignette experiments in a student sample, varying whether an authority figure blocks, fosters, or allows debates on controversial topics. Results show that students are more likely to disapprove of blocking in the classroom than in the workplace, and that blocking elicits out-partisan perceptions only among Democrats. Political interest and worldview further shape responses, with politically interested and more fluid (liberal) individuals especially critical of blocking. Finally, blocking reduces positive emotions in the university setting but lowers negative ones in the workplace. Our findings show how institutional context structures attitudes toward political discussion and underscore the university’s distinctive role in sustaining democratic discourse.

Not Your Party, But Something Else: Political Stereotyping in Low-Partisanship Contexts

Working paper (included in dissertation).
Presented at LASA 2025. 

People form negative social views of others based on perceived political orientation, expressing attitudes like dehumanization or unwillingness to engage with out-group members. One way they infer political orientation is through cues that are not inherently political but gain significance via association with a party, group, or ideology. I focus on everyday political stereotypes—such as food, hobbies, music, and sports—which have been studied in countries with strong party systems like the United States. However, their role is less clear in contexts with weaker partisanship, where other cleavages dominate. In Argentina, for example, a key divide centers on support for or rejection of a charismatic movement: Peronism. This raises important questions: Do everyday political stereotypes exist in such settings? What are their consequences for social attitudes? And how accurate are they? I address these questions using expert surveys, public opinion data, and a within-subject survey experiment that allows me to test mediation using a parallel design logic. I find that these stereotypes exist, are linked to both Peronists and anti-Peronists, and shape negative social perceptions of political out-groups by conveying political information—with perceived political orientation acting as a key mediator. However, they are often inaccurate, suggesting that hostile social views may be driven by imprecise cues.

You Sound Angry: Populist Candidates' Voice Tone as a Heuristic

Working paper (included in dissertation).
Presented at MPSA 2025. 

Candidates who use aggressive rhetoric toward Congress, the judiciary, and the media are increasingly successful worldwide. These candidates do not just use hostile words—they yell, sound angry, and convey aggression through their delivery. While the content of their speeches has been extensively studied, less is known about the effects of their tone of voice. I argue that an aggressive tone signals two key features aligned with their political profiles: strong and personalistic leadership, as well as anger toward the opposition and institutions that provide checks and balances. These characteristics are expected to increase support for such candidates among voters with populist attitudes, who should like those characteristics. Using two pre-registered survey experiments in which professional voice actors recorded audio for different treatment conditions, I find that an aggressive tone cues perceptions that candidates are more willing to veto bills, disobey the judiciary, and criticize the free press—but only when they do not state this explicitly. It also makes their leadership style appear stronger and their decision-making process seem more personalistic. However, aggressive voice tone has a negative impact on voters' support or perceptions of their governing competence. Voters are not divided by abstract populist attitudes, but by strong support for the current president. The effect is null for those that do support him the most, suggesting that support for populist candidates might be motivated by political leanings. In sum, aggressive tone serves as a heuristic for candidates' political positions and characteristics but mostly harm their electoral prospects, especially among independents.

Political Food: Using Politics as a Marketing Tool

Working paper (included in dissertation). 

Affective polarization extends beyond politics and influences various aspects of everyday life, including economic behavior. While existing research primarily examines how individuals’ political orientation can influence others’ decisions about them even when they are not actively seeking such an outcome, this study explores the opposite phenomenon: what happens when a business deliberately signals a political stance to generate a favorable economic response? To investigate this, I conducted a pre-registered vignette survey experiment on Peronist-themed restaurants in Argentina, a country deeply polarized between Peronists and anti-Peronists. My findings show that Peronist theming deters both anti-Peronist and neutral citizens from patronizing these restaurants and negatively affects their perceptions of them. This holds true not only for aspects that are directly shaped by the theming but also for those that should be unrelated to it, suggesting that outgroup bias plays a role. However, this political branding does not proportionally attract Peronists: only two out of seven dependent variables indicate a positive effect, and willingness to visit is also negatively impacted. This latter effect is driven by moderate Peronists, while extreme Peronists show no significant response. These results contribute to political science literature on affective polarization in daily life and offer insights for hospitality management and marketing regarding the risks of politically themed branding.

Rethinking Conservatism in Comparative Perspective 

Coauthored with Marc Hetherington (UNC Chapel Hill), Andrew Engelhardt (Stony Brook University) and Martín Opertti (Duke University)

Manuscript in progress. 

Inspired by Huntington (1957) and McClosky (1958), we argue for the importance of psychological ly informed conservatism to understand political conflict across time and space. Recognizing that conservative writers approached politics with a quite similar perspective despite reacting to markedly different realities over many centuries, McClosky (1958, 28) argued that “both conservatism and liberalism may be ‘natural’ or polar positions around which individuals of certain habitual outlooks, temperaments, and sensibilities can comfortably come to rest and be united with others of like disposition.” That approach runs counter to how people have considered the concept since Converse (1964). Commentators and political scientists tend to ignore these outlooks, temperaments, and dispositions in favor of economic or governing philosophies. Not only does this approach cause problems for labeling an actor or policy as conservative, it also contributes to misunderstandings or incomplete answers to many questions around the more polarized politics many countries face today. In this paper, we show how a conservative worldview that unites insights about ideology from Huntington and McClosky with those from psychology helps us understand political divides. Across three country contexts (U.S., Argentina, and Spain), we demonstrate how a novel measure of different facets of conservative worldview maps to political opinions in theoretically predictable ways. 

How Partisan Can Media Get? Testing the Limits of the Hostile Media Effect

Coauthored with Paloma Alonso (Universidad Torcuato Di Tella)

Manuscript in progress. 

This paper addresses the effect of partisan media brands on citizens' perception of non-political content. The level of bias that people perceive on news and hosts is influenced by media outlet brands. That phenomenon is known as the hostile media effect. Even though the literature has extensively documented this for political news, little is known about the way in which media brands (and specifically partisan media brands) can impact citizens' reactions to content for which bias is inconsequential. In other words, there is not much research testing the limits of the hostile media effect. In this project, we intend to fill that gap with a series of survey experiments that analyze how the presence of a partisan media brand can affect people's reactions to different non-political situations, such as guests, products that are advertised and victims of crimes.