The Polarization Myth

I n January 2021, in the turbulent wake of the last presidential contest, a former professor named Todd Rose asked some 2,000 people a question. The survey was, at least on the surface, designed to deduce what kind of country Americans would like future generations to inherit.

Each person was presented with 55 separate goal statements for the ­nation—“People have individual rights” was one; “People have high-quality health care” was another—and asked to rank them in order of importance. Each person was also asked how each goal would be ranked by “other people.”

When the results were tallied, the surprise was not that “People have individual rights” came in first, or that “People have high-quality health care” finished second. The surprise was the third highest priority: “Successfully address climate change.” We know that’s a surprise because, on the list of what “other people” considered important, climate came in 33rd. In other words, no one thought their fellow Americans saw climate as the high-priority item nearly everyone actually considered it to be.

That gap—between what we ourselves think and what we reckon ­others must be thinking—may hold the power to upend a great deal of what we believe we know about American civic life.

“People are lousy at figuring out what the group thinks,” Rose says. This collective blind spot is a quirk he would underline to students when he was teaching the neuroscience of learning at Harvard. At Populace, the think tank he co-founded to put such knowledge to practical use, the foible plays a prominent role in efforts to undo what Rose calls the “shared illusion” that Americans are hopelessly divided.

And divided we certainly think we are. The only thing Americans seem to agree on is that Americans cannot agree on anything. It’s hardly worth summarizing the headlines about doom and radicalization. In the prelude to a November ballot featuring the candidate synonymous with polarization, all the dapple and nuance of life is once again being reduced to a binary. Choose a side: red or blue.

Yet in the wintry interval between Jan. 6 and Inauguration Day 2021, that Populace survey , dubbed the American Aspirations Index, found “stunning agreement” on national goals across every segment of the U.S. population, including, to a significant extent, among those who voted for Donald Trump and those who voted for Joe Biden. On the few points where the survey registered disagreement (notably, on immigration and borders), the dissent was intense. But intense disagreement was the exception, not the rule.

Much of what news reports, politicians, and pollsters call polarization, Rose understands as “learned ­divisiveness”—­division propagated by the assumption that it exists even where it does not.

It’s a bold, and boldly optimistic, notion, but a notion supported by more than just one survey. At universities across the U.S., researchers have been looking hard at the mechanics of polarization. Picture them under the hood, bent over the engine that’s supposed to be driving us, possibly over a cliff. Every now and then, one reaches back with something they’ve managed to pry loose, sets it on the fender. These studies, hiding under titles like “Reducing Explicit Blatant Dehumanization by Correcting Exaggerated Meta-­Perceptions,” together make up a growing body of evidence that challenges the received wisdom about this political moment.

Maybe, they suggest, America has the wrong idea about polarization. It may not be nearly the engine we thought. It’s possible that what it produces, as much as anything, is noise.

Consider: Ordinary people in both parties turn out to like ordinary people in the other party well enough. In a 2021 study in the Journal of Politics , researchers found that when a person in one political party was asked what they think of someone in the other party, their answer was pretty negative. That certainly sounds like polarization. But it turns out the “someones” respondents had in mind were partisans holding forth on cable news.

If told the truth—that a typical member of the opposite party actually holds moderate views and talks about politics only occasionally—the animus dissolved into indifference. And if told that the same moderate person only rarely discusses politics, the sentiment edged into the positive zone. These folks might actually get along.

“There are people who are certainly polarized,” says Yanna Krupnikov, a study co-author now at the University of Michigan. “They are 100% polarized. They deeply hate the other side. They are extraordinarily loud. They are extraordinarily important in American politics.” But those people, she adds, are not typical Americans. They are people who live and breathe politics—the partisans and activists whom academics refer to in this context as elites.

“Elite politics is quite polarized,” Krupnikov says. “So the question is, does that mean everyone else is?”

Why not ask “everyone else” whether America is really that divided? Pollsters do, all the time. But there’s a problem. Ordinary folks think Americans are much more partisan than they are. In the same study, people grossly overestimated (by 78%) the size of the most polarized group within each party—that is, Democrats who call themselves liberal and Republicans who call themselves conservative. At the same time, ordinary Americans grossly under estimated (by 77%) the share of the other party who are moderate. That share is, in fact, at least half of either party. “People probably are exactly right about how polarized their leaders are,” says Robb Willer, a sociologist at Stanford. “They get it very wrong for the general public.”

It gets worse: the more involved in politics a person is, the more distorted their view of the other side, a 2019 YouGov survey found . In other words, engagement in civic life actually serves to narrow one’s perspective on the world.

That hardly recommends today’s politics, and goes a long way toward explaining why many people avoid partisans. “They dislike people who are ­really ideologically extreme, who are very politically invested, who want to come and talk to them about politics,” says Matthew Levendusky, a University of Pennsylvania professor of political science. And it’s not as if they’re trying to avoid confrontation, he adds: “It’s also the case that people aren’t ­really that fond of people from their own side who want to talk to them about politics.”

So people who do like to talk about politics talk to each other instead, and a striking social dynamic plays out: ­political enthusiasts will pretend to be even more polarized than they are. For a 2023 study published in Public Opinion Quarterly , people who described themselves as heavily invested in politics admitted that they would dial up their anger to impress fellow partisans. According to Elizabeth C. Connors, the University of South Carolina professor who conducted the study, the falseness partisans described about their own behavior reached levels “rarely seen in social sciences.”

Her takeaway: “If you’re a partisan and you’re going to say you’re a Republican or you’re going to say you’re a Democrat, you need to be a polarized one. Or else you’re not a good one.”

Such performative behavior of course complicates efforts to gauge how divided Americans have become. “If you ask a true racist their views, they’re going to lower the temperature, and report that they’re less racist than they actually are,” says Sean J. Westwood, who studies polarization at Dartmouth. “If you ask someone about partisanship and partisan hatred, they tend to do the reverse.”

So, yes, American politics has grown more divided—but largely among people who live and breathe politics. And these people exaggerate their own polarity to win the approval of other people who also live and breathe politics. It’s also true that the number of these people has grown over the past 40 years, as more Republicans identified as conservative and more Democrats as liberal.

That growth is a big reason that, for example, the U.S. House of Representatives is no longer actually representative. Most House seats—often by design—are for districts dominated by one party, so the decisive election is the primary, a low-­turnout affair in which the enthusiasm of activists has outsize impact. And, once in Washington, studies show, the Congress­person routinely cast votes more ideological than their ­typical constituents. But, still, in ­neither party do the ideologues make up the majority, even if it sure can feel that way. In truth, most Americans agree on most things.

“That’s kind of surprising to a lot of people,” says James Druckman, a political scientist at the University of Rochester. “But it’s pretty well documented that the typical voter of each party is not that far from the typical voter of the other party on most issues. If you look at other countries, the distance is a lot greater.”

Read more: Americans Say They Feel Pretty Good—at Least Compared to the Country as a Whole

Yet that relatively modest distance seems like a chasm, in no small part because of what’s called “conformity bias.” Researchers have long known that when asked a question by a pollster, people tend to color their reply by what they think they’re expected to say. This idea can make it easier to understand why, when the national narrative is about extremes, as it is now, moderate people self-report as being less moderate than they really are.

“This tug toward the fringes,” as Populace’s Rose calls it, threatens to empty out the middle ground where many Americans might prefer to stay, but fear they’ll be alone there. Their isolation may be an illusion—like the idea that no one but you cares about climate change—but it can feel real enough.

Remember how bad humans are at figuring out what other people are thinking, at least as a group? It’s reinforced by another bug in our mental software. Our brains mistake repetition for majority opinion.

As the delightful subtitle on a 2007 study in the Journal of Personality and Social Psychology put it: “A repetitive voice can sound like a chorus.” The study gathered people in a group to discuss something, then asked individuals to state the majority opinion of the group. What people offered up was the opinion they had heard several times—even when it had been voiced by just one person, saying the same thing over and over. Other studies have documented the same phenomenon.

“Your brain has this stupid shortcut for how it estimates the majority,” says Rose. The shortcut sheds light on why people frequently mistake the views of political activists, such as those on Fox News Channel and MSNBC, for the views of most Americans. Regular viewers do appear to be genuinely polarized. But in a 2022 study , Fox News viewers who were paid to watch CNN registered a significant moderation in their views after just a few weeks. “You change their media environment and their attitudes change pretty meaningfully,” says University of California, Berkeley, political scientist David Broockman.

But for those who don’t embrace an ideology, the “tug toward the fringes” can be a source of stress. Populace figured out a way to measure this unease in another of its surveys—one that helps explain how we know moderates are inhibited about revealing their views to pollsters.

This survey, in 2022, aimed to avoid the distortions of conformity bias by masking both the respondent’s identity and, more subtly, the question being asked, by hiding the “target” among a series of multiple-­choice questions. Because this method requires several rounds of polling to see which results are significant, it’s expensive and time-consuming—but it’s thought to reliably reveal information people might not consciously choose to share. (“The IRS uses this,” Rose says.)

Among the revelations of “ Private Opinion in America ” is that men are less supportive of abortion being a matter between a woman and her doctor than public surveys suggest, but also that people are less concerned than other polls suggest about the amount of time public schools spend talking about race.

On many topics, the gap was fairly small—a few percentage points—­between the opinion someone held privately and the one publicly expressed. And the results varied by demographic and political party. Yet every group polled registered double-­digit gaps on at least one issue.

Read more: The Split in How Americans Think About Our Collective Past Is Real—But There’s a Way Out of the ‘History Wars’

One group in particular was revealed to have struggled mightily to be candid with ordinary pollsters. For political independents, people without a party, the gap between private thought and public expression ran to double digits on more than half the issues—a striking amount of dissonance. This discrepancy ought to seem odd. After all, political moderates still constitute the majority in the U.S. electorate. But in a public sphere dominated by extremes, independents are made to feel that they have no place.

A more striking measure of that distress popped up in Gallup’s annual poll asking Americans, “What one country anywhere in the world do you consider to be the United States’ worst enemy today?” One of the options is “the United States itself.” This year, that was the choice of 2% of Democrats and 1% of Republicans. But 11% of independents judged the U.S. as its own worst enemy—more than selected ­either North Korea or Iran.

People do, of course, disagree. If they didn’t, there wouldn’t be much need for democracy. There are real differences in opinion on topics that are, to many Americans, a matter of life and death. It matters that you vote. And there’s a reason the past decade or so has been a time in which friendships, families, and civic life have been riven by politics. Which is to say, no discussion of polarization can ignore Donald Trump. ­Division is kind of his brand. Whether or not Trump deliberately exploited the national tug toward the extremes to get elected in 2016, the trend accelerated during his time in office.

When it comes to measuring perceived polarization, political scientists regard the quadrennial surveys by American National Election Studies as the gold standard. Every four years, it asks members of one party how warmly or coolly they feel toward the other party. During Trump’s term, the temperature dropped a record amount. Studies of presidential rhetoric note that he stood out among modern Presidents for seldom using language intended to unite the country.

And yet, at the end of those four years, moderates remained the majority, even as politics grew nastier. “National unity” actually turns out to be of scant interest to most people, finishing 50th in the American Aspirations survey. “Treating one another with respect,” however, ranked 14th. In a country where most people agree on most things, the acid tone of public debate amounts to a paradox that Lilliana Mason, a political psychologist at Johns Hopkins, captured in the title of a 2014 paper, “I Disrespectfully Agree.”

Mason says insult politics masks the underlying congruity on most issues by stirring emotions attached to differences in sensibility or social identity—the “culture war” topics that animate activists on both sides. “Americans are, on average, moderate on most policy preferences,” she says. “But one of the things that our current politics does is it makes us think the most about the policies that we get the most mad about.”

Fortunately, when people learn the truth about the other side, they feel better.

“Polarization appears to be largely driven by misperceptions,” Rachel Kleinfeld concluded in a sweeping survey of the topic for the Carnegie Endowment for International Peace. Even the worst reports going around—that, for instance, significant numbers of Americans supposedly favor armed ­revolt—turn out to be misleading. Those polls reflect a perception that the “other side” is ­already ­planning ­violence. ­Informed of the actual ­situation, the reaction recedes . Stanford’s Willer says the propensity for political violence is overreported by 300% to 400%.

Mason agrees. In so many studies, people register surprise that their assumptions about their rivals are wrong. “They’re like, ‘Oh, I didn’t know that,’ and then they feel better about the other side,” she says. “And then they go out into the real world and everything around them is like no, no, no, they’re demons. And so the effect doesn’t last, right? It has to be everywhere.”

What looks like a gulf may be more like a flooded sidewalk—shared space that’s still there, just really hard to see. In American Aspirations, more respondents said politicians should focus on finding common ground than said politicians should be fighting for them. But—sure enough—they also thought “other people” felt the opposite.

And of course November looms, with its promise of cleaving the nation down the middle with a this-or-that choice. Yet face to face, most people still get along, especially if they’re polite enough not to talk only about politics all the time.

But even if they do, look: In 2022, a Berkeley study followed what scholars have determined are the most insular partisans of all—liberal Democrats—as they knocked on doors in conservative neighborhoods, canvassing for votes. The activists didn’t change many minds. But afterward, many reported a new respect for people who saw things differently. — With reporting by Julia Zorthian

Contact us at letters@time.com .