The variable polarization research skipped: not difference, but being wrong

A group of behavioral scientists who study beliefs in politics and everyday life report that people are not chiefly bothered by others holding different beliefs, but by others holding beliefs they judge to be based on incorrect information ^[1][10]. Across four studies with more than 2,000 U.S. adults, perceived incorrectness predicted both distress and stated willingness to avoid, distrust or block another person, while confidence that a belief was merely different did not predict avoidance ^[2][5][6].

That distinction matters because most of the field has been measuring the other thing. The standard construct is belief homophily, the idea that people prefer others who think like them, and researchers have used it to explain friendships, neighborhoods, media habits and political life ^[3]. It has good circumstantial support: people do cluster residentially by politics and do select congenial news and social media, a pattern linked to echo chambers and polarization ^[11]. The authors, who include themselves in that tradition, argue the account is incomplete rather than wrong, because people do not react uniformly to all disagreements ^[1].

The three reported designs separate the variables in different ways. In a recall study, participants asked to describe a disagreement they saw as factually incorrect reported having felt more disturbed, frustrated and upset than participants who described a disagreement they saw as a difference in priorities ^[4]. One participant's example of the latter was whether couples should live together before marriage; an example of the former was a friend who thought soda and fast food were harmless [c4b]. In a second study, participants stated views on climate change, capital punishment and policing, then read a social media post opposing them; the more confident they were that the author was wrong, the more disturbed they felt and the more likely they said they would be to avoid, distrust or block that person ^[5].

The third design is the one with an actual manipulation. Participants read the same fictional facts about a neighbor who believed a politician was corrupt, and the only thing varied was whether the neighbor's belief was described as different or as incorrect; distress was substantially higher under the incorrect framing ^[7]. Holding the information constant while moving the label puts the effect on the framing rather than on the content of the disagreement ^[1].

The operational consequence is unflattering to a common playbook. If distress and avoidance track perceived error, and perceived difference does not predict avoidance ^[5][6], then interventions organized around surfacing shared beliefs are optimizing a variable that did not move the outcome in these studies, while the variable that did move it, whether the other side is seen as factually mistaken, is left untouched ^[2].

Caveats worth holding. The authors offer three candidate explanations for why perceived falsehood stings, all presented as possibilities rather than tested mechanisms: bad decisions follow from false beliefs, false beliefs threaten a sense of shared reality, and false beliefs are read as signals about the person holding them ^[8].

Watch for the published effect sizes, which this account does not give ^[3]. Watch whether the avoidance measure survives contact with behavior rather than stated intention ^[5]. And watch whether the framing manipulation runs in reverse, since a result that says calling someone wrong inflames a dispute ^[7] is only useful if relabeling the same dispute as a difference in values cools it. The sample is U.S. adults, so nothing here is yet a claim about anywhere else ^[2].