Political Polarization May Be Driven by Real Disagreements, Not Misperceptions

Rick Harrison
An American flag made out of rock with a break in the middle

A popular theory among social scientists is that Democrats and Republicans greatly exaggerate the other side’s unity and extremeness, intensifying animosity. Some efforts to reduce hostility have often centered on correcting those misconceptions.

But a new study by faculty from the Institution for Social and Policy Studies’ Center for the Study of American Politics (CSAP) found that Democrats and Republicans often make the same mistakes about their own parties, suggesting that partisan hostility may be rooted less in caricatures of the opposition than in genuine disagreements over what the parties stand for and do.

“It could just be properly understanding the differences between the parties,” said Gregory Huber, the study’s co-author and director of CSAP. “You disagree. That’s why you are in your party and not the other one.”

Published last month in Frontiers in Political Science and co-authored by Lilla Orr, Yale Ph.D. ’21 and assistant professor of data science and statistics at the University of Richmond, the new study drew on three surveys, including national YouGov surveys of more than 5,000 U.S. adults before each of the 2020 and 2024 elections and compared respondents’ estimates with answers to identically worded policy questions from the Cooperative Election Study, a large national survey of American voters.

They included questions such as: What percentage of Democrats are registered Democrats? What percentage voted for their party’s presidential nominee? What percentage support their party’s entire platform? 

Huber and Orr found little evidence that people see their own party as diverse and nuanced and the other party as a group in which everyone thinks alike. They found that Democrats and Republicans gave remarkably similar estimates about the loyalty of members, regardless of whether they were evaluating their own party or the opposition.

In addition, the researchers asked thousands of Americans to estimate how many Democrats and Republicans support certain policies, what kinds of people make up each party, and how party positions were changing over time.

Gregory Huber

Gregory Huber

The researchers found that when survey respondents estimated support for issues such as abortion, immigration, environmental policies, and foreign policy, they were not always accurate. But Democrats and Republicans often made similar errors about both parties. People were slightly more accurate when estimating the views of their own party than those of the opposing party.

“The problem does not seem to be our artificial view that the other party is more extreme,” Huber said. “Nor does that correlate much with whether we dislike them.”

The researchers examined both the accuracy of perceptions at a particular moment and whether those perceptions changed along with the parties between elections. When comparing perceptions before the last two presidential elections, the researchers found that public perceptions moved in the same direction as actual party opinion on seven of the eight policy measures they examined. But survey respondents struggled to accurately understand changes in racial, religious, income, and social makeup of the parties.

“People have their ear to the ground and understand the broad winds that are blowing,” Huber said. “They’re really bad at judging demographics. And everybody’s bad at demographics.”

The researchers were particularly interested in what they call “differential stereotyping,” or the tendency to see the opposing party as more uniformly extreme than members of that party see themselves. They found that people who expressed greater dislike of the opposing party were no more likely to depict it as uniformly extreme, relative to their perceptions of their own party.

“It’s interesting that when there are misperceptions, they are often shared,” Huber said. “Democrats stereotype Democrats about the same amount that Republicans stereotype Democrats. It’s not about the other party. It’s just about misperceptions.”

He suggested researchers and political actors should rethink efforts to reduce polarization by correcting misconceptions.

“By correcting misconceptions, you’re never telling them, ‘Oh, that other party actually agrees with you,’” Huber said. “You’re actually telling them they disagree with you, only maybe not as much as you said they did when you were asked on a survey.”

Huber cautioned that broad questions about “Democrats” or “Republicans” may prompt people to think about national political parties, their policies, or prominent figures in Washington, D.C., rather than individual voters in their communities.

“When you say you don’t like the other party, you’re not thinking of Bob down the street,” Huber said. “You’re thinking about the Republican or Democratic parties, often the most stereotypical Democrat or Republican in Washington, the policies they pursue, and if you disagree with them.”

The findings do not show that Americans’ perceptions are always accurate. Rather, they suggest that mistaken beliefs about the opposing party may not be the principal source of partisan hostility.

“Real disagreements,” Huber said. “Not necessarily misperceptions.”