top of page

Research - Full

Reducing Intergroup Conflict and Inequalities


Political Polarization and Democracy

Scholars have argued that democratic societies benefit from polarization in the sense that political parties with clearly distinguishable policies make it easier for voters to choose the parties that represent their interests. However, over the last few decades, a different, affective form of polarization has emerged in the U.S. and other countries: a growing tendency for partisans to strongly dislike the other side. Many scholars and practitioners alike have warned about the disastrous consequences of this affective polarization for democratic societies. The concern is that people are so appalled or threatened by their political opponents that they engage themselves in, or support copartisans’, social and economic discrimination and/or undemocratic practices against outpartisans even though such behaviors are detrimental to society as a whole. I am particularly concerned that political elites can use polarizing and anti-democratic tools to keep traditionally disadvantaged groups from reducing long standing inequalities.


My early graduate work on affective polarizations experimentally tested several strategies to reduce partisan animosity. This work finds that including members of a political outgroup in a discussion (Voelkel et al., 2021, Journal of Experimental Social Psychology) and increasing belief in the utility of cross-partisan empathy (Santos, Voelkel et al., 2022, Psychological Science, Washington Post) are two promising strategies for reducing ideological and partisan animosity. In contrast, highlighting people’s overconfidence in their own understanding of policies is not a promising strategy for reducing animosity based on worldview differences (Voelkel et al., 2018, Comprehensive Results in Social Psychology).


In more recent work, I have started to empirically question the assumption that reducing affective polarization has positive implications for societies. The results are more nuanced than what one might expect. On the one hand, my work shows that people judge harmful and unfair actions against others as significantly less immoral when the others are political outgroup members (Voelkel & Brandt, 2019, Personality and Social Psychology Bulletin, New York Times). Furthermore, reducing partisan animosity makes people more persuasive in interactions with outpartisans (Santos, Voelkel et al., 2022, Psychological Science, Washington Post), enabling consensus and coalition formation. On the other hand, strategies that directly target partisan animosity are not necessarily effective for reducing anti-democratic attitudes, such as support for undemocratic politicians or political violence (Voelkel et al., 2023, Nature Human Behaviour, New York Times). The potential lack of a causal link between affective polarization and anti-democratic attitudes exposed that while many papers have demonstrated how to reduce affective polarization, knowledge on how to reduce anti-democratic attitudes is scarce.


To fill this gap, the flagship project of my dissertation (Voelkel et al., 2023, reject and resubmit at Science, Washington Post) used a novel methodological approach to accelerate the social psychology of anti-democratic attitudes. I organized a mass collaboration project, called the Strengthening Democracy Challenge, calling for members from across the social sciences as well as practitioners to submit their ideas for reducing support for undemocratic practices, support for partisan violence, and partisan animosity. The project received more than 250 submissions from more than 400 researchers and practitioners from 17 countries. The 25 most promising interventions were selected based on expert reviews. In this project, we identified that the most promising strategies for reducing support for undemocratic practices and candidates among U.S. partisans are to (a) correct exaggerated misperceptions about outpartisans’ support for undemocratic practices and (b) highlight the disastrous consequences of democratic collapse. In contrast, the most promising strategies for reducing partisan animosity were quite different: (a) showcasing sympathetic exemplars with different political beliefs and (b) highlighting common identities that unite Democrats and Republicans.


In follow-up work, I have contributed to make the insights gained from the Strengthening Democracy Challenge more impactful. First, we have applied the strategy of correcting exaggerated stereotypes of outpartisans to support for undemocratic practices among a key group of political decision makers, state legislators. We found that a short intervention is successful in correcting misperceptions among state legislators and reducing their support for undemocratic practices (Druckman, Kang, Chu, Stagnaro, Voelkel et al., 2023, PNAS). Second, a challenge of applying the identified strategies outside of survey experiments is that people may not be motivated to complete interventions. Therefore, we have developed and tested a multi-stage approach: first, use the strategy of increasing curiosity about outpartisans, which then opens up people to participating in a task that corrects their misperceptions about outpartisans (Santos, Voelkel et al., 2023, under review at Nature Communications).


Gender Inequality

Another widespread problem in society is the persistence of gender inequality. Scholars have argued that one pathway towards gender equality is through political representation. However, many countries are still mostly represented by men. In the U.S., all 46 presidents have been men. Although traditional explanations focus on taste-based preferences for men, one of my papers that I co-led with another graduate student (Corbett*, Voelkel* et al., 2022, PNAS, New York Times; *: equal authorship) demonstrates that pragmatic bias can undermine women’s access to political leadership positions as well. That is, even voters who do not hold preferences for men themselves may fail to choose women candidates in primaries because they are concerned about their electability. In this paper, we demonstrate that voters in the 2020 Democratic presidential primaries perceived women candidates as less electable than men candidates although other research suggests that voters do not prefer men candidates over women candidates in general elections. We experimentally tested an intervention citing this research, finding that informing Democratic primary voters about the similar electability of men and women candidates significantly increases voters’ perceived electability of, and intention to vote for, women candidates. This research suggests that correcting misperceptions about others’ prejudice is crucial for facilitating equal political representation.


Economic Inequality

Majorities of Americans hold the belief for decades that income inequality is too large. Yet economic inequality in the United States has surged since the 1970s. One pathway towards reducing economic inequality is through electing politicians passing progressive economic policies. One of my papers (Voelkel et al., 2023, PNAS Nexus, New York Times) suggests that Americans want to reduce inequality in the U.S. but often vote against progressive politicians because progressive politicians argue for progressive policies in ways that do not resonate with the large group of conservative Americans. We demonstrate in a text analysis of all presidential debates since 2020 that progressive politicians are less likely to refer to values such as patriotism and respect for authority than conservative politicians. In two experiments, we find that a progressive politician who frames progressive policies as consistent with such conservative values broadens their base of support among moderate and conservative voters. This research suggests that recognizing that the alignment between values and candidate support is malleable, enabling broader coalitions for reducing economic inequality by reframing the values voters associate with progressive platforms. This strategy can be applied to other domains as well. For example, we find that reframing arguments in favor of providing opportunities for immigrants as being consistent with patriotism and American traditions increases Americans’s support for increasing immigration (Voelkel et al., 2022, The ANNALS of the American Academy of Political and Social Science).



The societal problems I study are impossible to address with a single study or even within a single career. Therefore, a central question for my research is how to design my studies to contribute to fast scientific progress.


A key challenge for scientific progress is making results more comparable across different studies. Megastudies are a rising method for accelerating scientific progress. Megastudies are experimental tests of many different treatments using the same control condition, measures, research design, and sampled population. Such megastudies facilitate collaborations across different scientific disciplines as well as with practitioners. Further, megastudies enable researchers to answer previously neglected research questions about the relative effectiveness of different treatments. Such studies are particularly useful for urgent problems, such as the rise of undemocratic leaders or climate change, because they offer insights into the most promising directions for finding effective treatments. Therefore, I have used a new megastudy paradigm in my dissertation (Voelkel et al., 2023, reject and resubmit at Science, Washington Post; Voelkel et al., 2023c, Nature Climate Change, acceptance at Stage 1 of registered report). I have also contributed to other megastudies (Grossmann, …, Voelkel et al., 2023, Nature Human Behaviour; Roy, …, Voelkel et al., 2023, under review at Journal of Personality and Social Psychology) and co-written an article on how to organize such big team science collaborations (Baumgartner, …, Voelkel et al., 2023, Royal Society Open Science).


Another key ingredient of contributing to scientific progress is producing reliable and transparent results. Therefore, all of my papers are open access (typically via a preprint survey) and the data, materials, and code are available via my Open Science Framework page so that other researchers can conveniently build on my work. I have (co-)written guides on how to make one’s own research more open and accessible (Voelkel & Freese, 2022, Handbook of Computational Social Science; Freese, Rauf, & Voelkel, 2022, Social Science Research). In my understanding, open science contributes to accelerating scientific progress by making it easier to find errors and to use materials and code from other researchers to build on their work.


Another key ingredient of contributing to scientific progress is making methodological and statistical advances more accessible. For example, the bias of scientific publications to report significant effects more often than non-significant effects is a major issue in knowledge accumulation. One issue with non-significant effects is that it is hard to determine whether the results favor the null hypothesis of no effect or whether the results are actually ambiguous. Bayesian statistics provide a tool to assess the evidence in favor of the null hypothesis, making  the interpretation of non-significant findings less ambiguous and easier to publish. As a software developer for JASP, an open-source statistics program that promotes the usage of Bayesian statistics in the social sciences, I contributed to making Bayesian analyses easier to access (van Doorn, …, Voelkel, & Wagenmakers, 2021, Psychonomic Bulletin & Review; van den Bergh, …, Voelkel et al., 2020, LAnnee Psychologique). I have also used these tools in my own research. For example, I have shown that treatments that reduce affective polarization have null effects on anti-democratic attitudes, questioning the widely held belief that affective polarization causes anti-democratic attitudes (Voelkel et al., 2023, Nature Human Behaviour, New York Times).

Political Polarization and Democracy
Economi Inequaliy
bottom of page