When We Think The Other Side is Bad, We Act Badly – BCB #115

When we overestimate the opposition’s inclination to play foul, we risk delivering a self-fulfilling prophecy.

Right now, both Red and Blue are convinced that the other side is poised to steal the election. Red is preparing measures to contest an election that hasn’t taken place yet—insisting that Blue will cheat, claiming the electoral system is vulnerable, and challenging swing states’ election procedures in court. Blue is also preparing a game plan, developing legal strategies in advance of close races which they expect to be contested through lawsuits, demands for recounts, and refusals to certify.

Meanwhile, voter confidence in election integrity is dire. Less than half of Americans have faith that votes in the presidential election will be counted accurately, with Red much more pessimistic than Blue.

The ease with which both Red and Blue are able to imagine the other side cheating in the election tracks with research showing that people overestimate the willingness of their opponents to break democratic norms. Such assumptions make people more likely to support antidemocratic behavior on their own side, which, in turn, creates a race to the bottom as each side sets out to preempt anticipated betrayal. Our metaperceptions—what we believe about what others believe—are deeply distorted. 

In a 2023 study by researchers from the University of California and MIT, participants were asked about democracy-subverting scenarios—like reducing the number of polling stations in areas likely to vote for the opposition, or ignoring controversial rulings from partisan judges. They were asked whether they would support such measures, and whether they believed the opposition would support them. The results show that both Red and Blue participants hugely overestimate the willingness of their opposition to subvert democracy.

(Source)

The effect was stronger for the more partisan participants. That is, the more strongly someone identifies with their own political tribe, the more they overestimate the other tribe’s willingness to subvert democracy.

Patterns like this are familiar: More In Common’s perception gap research shows the extent to which both Red and Blue misperceive one another’s support for extreme  policies (such as totally open borders or complete abortion bans). But the “subversion dilemma,” as it’s described by the study’s authors, is much greater than the policy perception gap:

On a 0 to 1 scale, we found that partisans inaccurately perceived opposing partisans by an average of 0.09 on policy views and 0.14 on dehumanization of opposing partisans, while they inaccurately perceived the other side’s willingness to subvert democracy by 0.40.

Worse, the more people believe that their opponents are willing to break democratic norms, the more willing they are to subvert democracy themselves. In highly polarized societies, voters are less likely to switch to voting for the other side, which can make them more tolerant of undemocratic behavior on the part of their own candidate. When those in positions of power are able to count on the support of their voters even while they violate democratic norms, it undermines democracy as a whole. When we overestimate the opposition’s inclination to play foul, we risk delivering a self-fulfilling prophecy.

Almost no one supports violence, but each side thinks the other does 

The discrepancy between perception and reality persists when it comes to endorsing political violence. Only a few percent of Americans say that political violence is justified even in extreme cases—there have been a spate of headlines reporting much higher numbers in the last few years, but these seem to have been largely mistaken.

A 2020 study from the University of Pennsylvania found that people massively overestimate their opponents’ actual support for political violence—and this increases their own propensity to consider violence an acceptable option. Each side greatly overestimates the other’s willingness to use violence for political purposes (like contesting an election they think is stolen), regardless of exactly how the question is asked.

Self-reports vs. out-party perceptions of support for political violence. SPV is “support for partisan violence,” where participants were asked “How much do you feel it is justified for [your own party] to use violence in advancing their political goals these days?” WEV is “willingness to engage in violence,” where participants were asked “If [an out-partisan] is declared the winner of a contested election, how likely would you be to engage in violence?”(Source)

The effect of these metaperceptions is clear, both in campaign discourse and between voters on social media.

In an environment increasingly haunted by the threat of political violence, the rhetoric between hostile Red and Blue camps is escalating to the point where each portrays the other as an existential threat—ripe conditions for democratic backsliding through an endless downward spiral. Blue cites January 6 and the controversial Project 2025 agenda to stoke fear about Red's political objectives. Red says this sort of language is responsible for the recent attempt on Trump’s life, is tantamount to incitement, and similarly claims that Democratic policy is a threat to democracy.

Despite the FBI's assurance that the Trump rally gunman acted alone and lacked a clear political alignment, both the Red and Blue media bubbles have erupted with conspiracy theories, each imagining the shooter as a partisan actor despite the lack of evidence, or claiming that the incident was staged. Unfounded accusations that the Trump assassination attempt was a Democrat plot or a Republican false-flag abound. There’s no evidence to support either of these claims, but plenty of willingness to believe.

Correcting misperceptions reduces support for political violence

Regardless of what they may believe about one another, Americans do value their democratic institutions, and the role of those who oversee them. In the midterms, voters withheld support from candidates who based their campaigns on the idea that the 2020 election was stolen or who had filed bogus lawsuits, showing that undermining the election process is not a vote winner. After the chaos that followed the 2020 election, some Republican leaders organized to affirm the integrity of an election they had lost. 

The good news is that, while people do hugely overestimate political opponents’ willingness to subvert democratic norms and engage in violence, metaperceptions can be corrected—with rapid and long-lasting effects.

In the 2023 Subversion Dilemma study discussed above, when participants learned that many fewer of their opponents actually supported measures to subvert democracy than they thought, it had a meaningful effect on their own responses. When people knew how the opposing tribe had responded, their own support for democratic norms increased. The same is true for people’s tolerance for political violence: when people’s overestimation of the other side’s willingness to use violence is corrected, they also become less inclined to say that violence is acceptable. The University of Pennsylvania study found that correcting people’s misperceptions of their opponents reduced their own support for violence by 34%. And the effect is greatest for those with the most distorted metaperceptions.

The trap of democratic backsliding is a prisoner’s dilemma. It’s in everybody’s interests to stick to their values, but the fear that the other side will betray the social contract creates an incentive for a preemptive strike. The good news is that despite the challenges, a little bit of accurate information goes a long way.

Quote of the Week

If citizens do not believe that opposing partisans will hold their representatives in check, they have a powerful incentive to give their own politicians broad leeway to do whatever it takes to save democracy from their opponents.

— Alia Braley, Gabriel S Lenz, Dhaval Adjodah, Hossein Rahnama, and Alex Pentland