Political violence and inaccurate metaperceptions

The January 6, 2021 attack on the United States Capitol, which left five dead and hundreds injured, was a stark demonstration that political violence remains a clear and present threat to democratic institutions. Perhaps because of the rarity of such events in the United States’ near past, recent scholarship on political and intergroup conflict has focused largely on topics of polarization and attitude extremism rather than political violence. It is in this social and scientific context that Mernyk et al. (1) provide timely and compelling evidence that inaccurate metaperceptions serve as a psychological driver of support for and willingness to engage in political violence, and that correcting such inaccurate metaperceptions can durably attenuate partisans’ positive attitudes toward political violence.

Mernyk et al. (1) begin by documenting inaccurate partisan metaperceptions related to political violence. Both Democrats and Republicans substantially overestimate the extent to which outpartisans support and are willing to engage in political violence. Next, they introduce a corrective intervention which informs partisans of outpartisans’ true and low levels of support for and willingness to engage in political violence. This intervention reduces partisans’ own support for and willingness to engage in political violence, and they find that this effect lasts for at least a month.

The success of their intervention provides two critical takeaways. First, attitudes about political violence can be changed and increasing levels of polarization have not cemented partisans’ tolerance for political violence. Second, these interventions provide experimental evidence that attitudes about political violence are caused in part by judgments of how much others support and are willing to engage in political violence. Put simply, if partisans (inaccurately) think other partisans are tolerant of political violence, they will themselves become more tolerant of political violence. This dynamic parallels similar findings across social domains, where (mis)perceptions of social norms and others’ attitudes can lead people to shift their own opinions closer to the (mis)perceived values (2–4).

Metaperceptions and False Polarization

Mernyk et al.’s (1) findings also echo a growing body of work on misperceptions of political polarization and how this “false polarization” contributes to actual polarization (5–8). Broadly defined, false polarization is the phenomenon where people believe politics is more polarized than it is in truth. Individuals overestimate the levels of disagreement between partisans (9–11), and perceived polarization is a stronger predictor of negative outgroup evaluations than actual polarization (7). However, while congruent with work on false polarization, considering how Mernyk et al.’s (1) findings do and do not differ from past work on false polarization can help inform a broader understanding of how misperceptions contribute to political conflict.

Mernyk et al. (1) build upon similar work on intergroup metaperceptions and in some respects their findings are directly analogous to past findings. For example, inaccurate perceptions of how much the outgroup dehumanizes one’s ingroup (metadehumanization) predicts support for breaking democratic norms (12), and correcting inaccuracies in metadehumanization can reduce reciprocal dehumanization toward the outgroup (13). Similarly, inaccurate perceptions of how the outgroup will react to collective intergroup behaviors (group metaperception) predicts polarized motive attributions, and correcting inaccurate group metaperceptions has successfully reduced polarized attributions in nine countries (14, 15). Like Mernyk et al.’s (1) findings, group metaperception interventions are more effective on those who are more inaccurate at baseline, providing further evidence for a causal relationship between inaccuracies and intergroup attitudes. Inaccuracy also persists for metaperceptions of ingroup members in addition to metaperceptions of outgroup members, although ingroup metaperceptions are less inaccurate (6).

Disentangling the Roots of Political Violence and Polarization

Despite the similarities with past findings, Mernyk et al. (1) begin to illuminate the ways in which the umbrella of false polarization may miss critical distinctions in the psychological processes driving negative political attitudes and outcomes. They find that while their corrective interventions have durable impacts on attitudes toward political violence, attitudes toward political violence are only weakly related to affective polarization, and the interventions have no observable effect on polarization levels. I would caution against interpreting this to suggest that the interventions’ effectiveness is narrow or limited. Rather, it suggests that the causal relationship between inaccurate metaperceptions and attitudes is highly domain-specific, an inference supported by a growing body of work. For example, while metadehumanization is associated with outgroup hostility and reciprocal dehumanization, metaprejudice is not (16), suggesting that (meta) dehumanization and (meta) prejudice have differing associated outcomes and potentially distinct antecedents. Similarly, the group metaperception interventions which have been found to reduce polarization in nine different countries (14) have no effect on antidemocratic attitudes (17), suggesting affective polarization and antidemocratic attitudes are not as psychologically linked as many scholars have assumed.

The evidence that intergroup metaperceptions in one content domain are directly linked to attitudes in that domain, but rarely tied to metaperceptions or attitudes in other domains, buttresses a central argument made by Lees and Cikara (6). They argue that the false polarization hypothesis, which states that inaccurate (meta) beliefs are contributing to negative intergroup outcomes, is no doubt true. However, they also argue greater attention is needed to the nature of those specific beliefs and how they are psychologically distinct from one another. Support for political violence is distinct from outgroup prejudice, which is distinct from antidemocratic attitudes, which is distinct from dehumanization, which is distinct from extreme policy positions, et cetera. Yet, lumping them all under the category of “polarization” obscures meaningful differences that have implications both for scientific understanding and policy prescriptions. By measuring both affective polarization and attitudes toward political violence, Mernyk et al. (1) convincingly demonstrate that support for and willingness to engage in political violence can be attenuated through corrective metaperception interventions related to political violence, but that those interventions will not move the needle on affective polarization.

A question left open by Mernyk et al. (1), and indeed by most research on false polarization, is the source of these inaccurate perceptions. Wilson et al. (18) suggest that political elites, the news media, and social media all play a role in causing political misperceptions. These factors are likely strong drivers of inaccurate metaperceptions about political violence. Political violence is rare, and a large majority of voters reject the use of violence for political ends (19), which means that most people’s experience of political violence is through the media. As such, the role of the media in amplifying messages which may lead to inaccurate metaperceptions is a fruitful avenue of future research. More worryingly, if excessive coverage of political violence by the news media is causing inaccurate metaperceptions, then Mernyk et al.’s (1) findings suggest such coverage may even be indirectly contributing to support for political violence.

In the wake of the January 6, 2021 attack on the United States Capitol, it is reassuring to see Mernyk et al. (1) demonstrate that despite the widespread concerns of intractable polarization and political conflict, partisans are sensitive to the truth and will update their beliefs accordingly. The real-world value of reducing support for and willingness to engage in political violence is apparent and immediate, and Mernyk et al.’s (1) work deftly builds upon past work related to false polarization and generalizes it to this timely domain in need of further scholarship.