It is hard to follow current events without noticing that we are living in an era of extreme political polarization. Yet, there may be some good news: as polarized as politics has become, it is probably not as bad as most people think. Research in social psychology has demonstrated that people on each side of controversial political issues, such as same-sex marriage, capital punishment, and affirmative action, tend to believe their opponents hold a more extreme position on the issue than these opponents actually do. This “False Polarization Effect” has been explained using the theory of naive realism.

According to Naïve Realism, people believe they see a direct, unfiltered view of objective reality, and think their opinions are based on a rational interpretation of this reality. As a consequence of this, people believe that those who disagree with them either lack the true information about an issue, or are motivated to reach the wrong conclusion because of self-interest or ideological bias. Political opponents (and to some extent even political compatriots) are seen as biased and misinformed ideologues who fail to correctly perceive reality. This leads people to perceive that their opponents will be more extreme and ideological than they really are.

While this suggests that the current polarization may not be as bad as it appears, it also raises an interesting question. If opponents are more moderate than most people think, why is it so difficult for people with opposing views to have discussions about political issues? Shouldn’t political discussions proceed amicably, with both sides happy to discover they share more common ground with their opponents than expected?

We propose that political discussions often proceed poorly because when people describe their views, they describe not only their position on the issue, but also the certainty with which they hold this position.

We argue that people tend to be quite certain that their position is correct, even if this position is moderate. For example, on the issue of marijuana legalization, one might be very certain about the moderate position that marijuana should be available for medical purposes, but not for recreation. After all, if people believe that they see an unbiased version of reality, they should be certain that their views are correct.

In addition, we think people perceive their opponents as less certain than themselves. Although people are likely aware that their opponents are certain, they may assume that the extreme, biased opinions of their opponents must conflict with true reality at some point, leaving opponents some lingering doubts about their position. Thus, people might falsely moderate the certainty of their opponents, with each side certain their position must be correct, and expecting their opponents to harbor doubts. This could help explain why political discussions can go poorly: each side underestimates the certainty of their opponents, and is frustrated by their opponent’s unexpected stubbornness and refusal to change their distorted position.

In a recent paper, we tested whether people do in fact, falsely moderate the certainty of others. We asked undergraduate students to indicate which side of several political issues they supported, and how certain they were that their position on the issue was correct. After this, we asked participants to estimate how certain the typical student at their university on each side of the issue was. Consistent with our predictions, people were very certain in their views: on a scale from 1 (“not very certain”) to 9 (“very certain”), the average reported certainty was 7.61, and the most frequent response was 9. Moreover, for the majority of issues, people believed that they were more certain than their opponents were (average estimated certainty for opponents was 6.19), demonstrating false moderation.

Thus, our research indicates that people underestimate the certainty of those they disagree with. This suggests that correcting falsely moderate perceptions of opponents’ certainty may improve political discourse. If each side anticipates the strong conviction their opponents are likely to hold, they may be better prepared for a political discussion, which may prevent some of the frustration that typically ensues. However, this approach may have downsides as well. Informing partisans that their opponents are likely quite certain about their position may increase pessimism about discussion and forestall attempts to resolve differences.

Another method of improving discourse may be to focus on reducing the high levels of certainty each side has that they are correct. Challenging the certainty with which individuals hold their convictions may be difficult, especially since partisan media and the echo chambers of the internet have created a “golden age of confirmation bias“. However, a few recent studies have shown promise in this area (Fernbach et al., 2013; Hart et al., 2015). For example, Philip Fernbach and his colleagues found that peoples’ high level of confidence in political policies can be reduced by asking them to provide in-depth, mechanistic explanations for how these policies work.

Thus, changing perceptions of others’ certainty or reducing one’s own certainty might reduce political conflict. There’s a lot still to be learned in this area, but we believe that certainty and perceptions of certainty play an important, yet understudied role in conflict and polarization.