Article Summary
Introduction
Concerns about rising political violence have sparked renewed interest in understanding the psychological factors which drive support for political violence against one’s opposing political party. Drawing from a growing body of psychological research, this paper seeks to answer three interrelated questions, all of which seek to help us understand the consequences of misperceptions. First, do partisans on average overestimate how much members of the opposing side support political violence, which social scientists call “metaperceptions,” and if so by what magnitude? Second, is there a relationship between overestimating how much the opposing side supports political violence and supporting political violence amongst your own side? And three, can an intervention designed to correct these inaccurate metaperceptions of the other side actually decrease partisans’ support for political violence amongst their own side? In other words, do people correctly guess how members of the opposing party feel about political violence, and if they do not, is it because they themselves support political violence? If so, does correcting the information about the other side temper support for violence?
The authors do find that partisans wildly overestimate how much members of the opposing side support political violence, and that this bias is related to partisans being more supportive of political violence themselves. Encouragingly though, the authors provide experimental evidence that these inaccurate metaperceptions can be corrected by providing partisans with accurate information about what the other side actually believes, namely that almost all partisans reject political violence, and that doing so durably reduces people’s own support for political violence.
Analytical Approach
This paper uses four studies to come to its conclusions. Studies 1 and 2 were correlational and measured partisans’ levels of support for and willingness to engage in political violence, along with their estimates of how the average member of the other side responded to those questions. This allowed the authors to measure participants (in)accuracy in these metaperceptions of out-party beliefs about political violence. Studies 3 and 4a were experimental, where participants were assigned to either a control condition or informational intervention. The informational intervention informed partisans of the actual beliefs of the other side regarding political violence (i.e., provided these partisans with the values collected in Studies 1 and 2), then afterwards asked participants their own support for and willingness to engage in political violence. Their hypothesis was that these informational interventions would reduce partisans’ support for and willingness to engage in political violence. Study 4b followed up a month later with participants from Study 4a to examine whether their attitudes about political violence had reverted back to baseline, or remained lowered as a result of the informational intervention.
Experimental Manipulations
| Study | Manipulation |
|---|---|
| Study 3 | Random assignment to control condition (no new information) or informational intervention, where they were told the other side’s true level of support for political violence (based on data from Study 1). These true values were clearly juxtaposed against the participant’s previous estimate, making it easy for them to assess their (in)accuracy. |
| Study 4a | Random assignment to control condition (no new information) or informational intervention, where they were told the other side’s true level of willingness to engage in political violence (based on data from Study 2). These true values were clearly juxtaposed against the participant’s previous estimate, making it easy for them to assess their (in)accuracy. |
Findings
Studies 1 and 2 found that partisans vastly overestimated the other party’s support for political violence (Study 1) and willingness to engage in political violence (Study 2). Estimates ranged from 245 to 442% higher than actual levels. These misperceptions were of equal magnitude for Democrats and Republicans. In Study 2, they measured support for both offensive and defensive political violence. While support for defensive political violence was higher than offensive violence, partisans still meaningfully overestimated the actual values. Studies 3 and 4a found that participants presented with the actual values of the other side’s support for political violence (Study 3) and willingness to engage in political violence (Study 4b) were less supportive of/willing to engage in political violence, compared to participants in the control condition who were not supplied the actual values for the other side. Critically, they found that the informational intervention was most effective on partisans who were more inaccurate at baseline, suggesting that inaccurate beliefs about what the other side thinks about political violence are causing partisans themselves to be more accepting of political violence. Study 4b followed up with the participants from Study 4a a month later and found that the effect of the intervention persisted. Even a month later, participants assigned to the informational intervention were less willing to engage in political violence compared to those in the control condition.
Implications
This research suggests that perceptions of social norms related to violence can actually cause support for political violence. If a partisan (incorrectly) thinks the other side is okay with political violence, this may actually lead that partisan to themselves view political violence as okay. However, this research finds that partisans are thankfully sensitive to the truth. When they learn that they vastly overestimate the other side’s support for political violence, they themselves become less supportive of it. This demonstrates the power of social norms: people won’t support something if they believe no one else supports it. And thankfully, this work demonstrates that almost all partisans reject political violence.
These findings suggest that public awareness campaigns that inform people of just how rare political violence is, and that almost everyone on both sides of the aisle rejects political violence, may help to further reduce the rare instances of political violence.
Questions left unanswered
An interesting question that was outside the scope of this paper is the source of these exaggerated beliefs. Could it perhaps be media coverage of political violence, or partisan media’s demonization of the other side, that causes people to hold inaccurate beliefs, in turn potentially fueling support for political violence amongst viewers? Hopefully future research will provide some answers.
Another provocative insight relates to affective polarization. While not the primary focus of this research, the authors found that affective polarization was unrelated to support for and willingness to engage in political violence. Moreover, they found that the interventions had no effect on affective polarization, relative to the control condition. This suggests that affective polarization and support for/willingness to engage in political violence are separate phenomena. While it may be encouraging to know that increasing polarization perhaps does not contribute to an increase in political violence, this leaves many open scientific questions about the nature of polarization and its relationship to political violence.
Methods and Analysis
Was the study and its analyses pre-registered?: Yes
Did the study rely on proxy variables to measure polarization?: No
Were standard p-value thresholds used (p<.05 or 95% Confidence Intervals that don’t overlap zero)?: Yes
- Largest p-value presented as significant: 0.05
Were correlational results interpreted with causal language?: No
Limitations / Weaknesses
First, the sample in Study 3 included only strong partisans, not a representative sample of the general population. This limits the extent to which we can be confident the intervention tested in Study 3 will be effective if deployed in other contexts or with other populations. Second, as is the case with all self-reported behavioral measures, there is always a concern that expressed willingness to engage in a particular behavior is not actually predictive of real-world behavior. In this instance where the behavior is political violence, which is both very rare and difficult to study directly, caution is needed when extrapolating from these experiments to instances of real-world political violence. Third, there were no financial incentives for participants to make accurate judgments at baseline. Past research finds that incentivizing accurate judgments about out-partisans increases accuracy, which suggests that some of the inaccuracy observed in this paper’s studies are the result of “heat of the moment” responding and not reflective of what participants actually think if forced to be a little more reflective. Lastly, because this research focuses on what partisans think the other party believes, it cannot speak directly to what might motivate political independents to support political violence.
Open Data & Analyses
Does the article make the replication data publicly available?: Yes
Does the article make the replication analysis scripts publicly available?: Yes
Article Citation
Mernyk, J. S., Pink, S. L., Druckman, J. N., & Willer, R. (2022). Correcting inaccurate metaperceptions reduces Americans’ support for partisan violence. Proceedings of the National Academy of Sciences, 119(16), e2116851119. https://doi.org/10.1073/pnas.2116851119
Bibtex
1
2
3
4
5
6
7
8
9
10
@article{mernyk_correcting_2022,
author = {Mernyk, Joseph S. and Pink, Sophia L. and Druckman, James N. and Willer, Robb},
journal = {Proceedings of the National Academy of Sciences},
month = apr,
number = {16},
pages = {e2116851119},
title = {Correcting inaccurate metaperceptions reduces {Americans}' support for partisan violence},
volume = {119},
year = {2022}}