Americans often fail to notice when Democrats and Republicans are moving in the same direction

Informing people about how the political attitudes of opposing parties are moving in the same direction over time might help reduce partisan hostility. New research published in Political Psychology suggests that people frequently fail to notice how much support for certain social issues has grown within both parties. When people read about these shared trajectories of opinion change without being reminded of the current gap between parties, they tend to expect more future common ground and report slightly warmer feelings toward their political opponents.

To understand how political divides operate, psychologists often look at descriptive social norms, which are individuals’ perceptions of what other people believe or do. In politics, perceptions of these norms are frequently distorted. For instance, a study covered by PsyPost in 2021 indicated that both Democrats and Republicans overestimate how much the opposing party dislikes them. When people assume their political opponents hold extreme or hostile views, it often drives further polarization.

Researchers often try to correct these false assumptions by presenting static norms, which describe the current level of support for a specific issue within a group. However, static norms only capture a single moment in time. Dynamic norms, on the other hand, describe how a group’s beliefs or behaviors are changing over time. A 2017 study found that showing people a trend is growing in popularity can persuade them to adopt new behaviors, even if current support is relatively low.

Applying this idea to politics requires overcoming another layer of misperception. A 2022 study indicated that Americans often misjudge how public opinion has shifted over decades, usually overestimating how much attitudes have changed, especially in a liberal direction. The authors of the new study set out to test whether correcting misperceptions about these trends could soften partisan hostility. How such a message is framed may matter: a 2021 paper argued that describing partisan divisions can make division seem like the norm and inadvertently deepen polarization.

The research was led by Trystan Loustau, a doctoral candidate in the Department of Psychology and Neuroscience at Boston College. “A lot of research shows that Democrats and Republicans have distorted perceptions of one another, and that these misperceptions can make political divisions seem larger than they really are,” Loustau told PsyPost. “But most of that work focuses on a snapshot of the present: what Democrats or Republicans believe right now. We were interested in whether people also misperceive change: even when members of both parties are moving in the same direction on an issue, do people recognize that convergence?”

“That seemed especially important because the direction people think society is moving may shape how they imagine its future,” Loustau explained. “If you believe the two parties are becoming increasingly divided, you might respond very differently to political opponents than if you believe there are areas where they are gradually finding more common ground.”

In the first study, 368 Democrats and Republicans estimated current support and changes in support for 12 social issues on which support had actually risen in both parties. The researchers found that participants systematically underestimated attitude change. Only about half of the estimates accurately recognized that support for these issues was increasing. They also underestimated current levels of support by about 12 points for Democrats and 20 points for Republicans.

Misperceptions were especially pronounced when evaluating Republicans. Only about 25 percent of Democrats’ estimates and 44 percent of Republicans’ estimates regarding Republican attitude changes were accurate, compared to 77 percent and 66 percent accuracy for Democratic attitude changes. This finding is in tension with the 2022 study: here, participants often missed real increases in support for progressive positions, especially among Republicans. The authors suggest people may expect Democrats, but not Republicans, to move left.

In addition, the researchers found a correlation between recognizing these trends and intergroup attitudes. For every additional 10 percent of estimates that correctly identified change in the opposing party, participants felt about 3 points warmer toward that party on a 0-to-100 scale. This association held after accounting for accuracy about current support, which itself was unrelated to warmth.

“Political disagreement is real, but our perceptions of it are not always accurate,” Loustau pointed out. “We may be particularly bad at noticing when attitudes are changing.”

To see if correcting these misperceptions could help, the researchers randomly assigned 1,187 partisans to a dynamic norm condition, a static norm condition, or a control condition featuring apolitical topics. The dynamic norm message provided both the past and current levels of partisan support alongside the percentage-point increases over time. The static norm message only provided current support levels.

The results indicated that the dynamic norm group expected more future common ground between parties than the control group. Combining the current support levels with the trend data did not produce broader positive changes in attitudes toward the outparty, meaning the opposing party, such as warmth or expectations of belonging. The authors suggest that presenting the current levels of support alongside the trend may have inadvertently reminded participants of the existing gaps between the parties.

For the third study, the team refined the intervention using a new sample of 1,154 participants. This time, the dynamic norm condition isolated the trend information by completely removing references to current levels of partisan support. The message focused solely on how support had grown within each party over time. Participants were randomly assigned to read the refined message, the original combined message from the second study, the static message, or the control material.

The refined message increased expected common ground and expected future feelings of belonging with the outparty. People who read it reported outparty warmth of about 32.5 on a 100-point scale, compared to 25.9 in the control group, and also rated members of the other party more positively. The researchers noted that a recent review of 77 depolarization interventions found an average gain of about 5.4 points in outparty warmth, making the refined message’s 6.6-point difference comparable to that average.

The refined message was the only one to outperform the control condition on all four of these measures. On warmth and positive evaluations, however, it did not statistically outperform the original combined message or the static message, and the static message also raised warmth relative to control. No version of the message changed how much people saw the other party as overlapping with themselves.

“What surprised me most was how much the way we presented the same information mattered,” Loustau said. “It was a useful reminder that more information is not necessarily better: emphasizing how far apart the parties remain may overshadow evidence that they are moving in the same direction.”

He added, “The broader takeaway is that focusing exclusively on today’s partisan divisions can obscure meaningful ways in which public opinion is changing.”

The findings are in line with research covered by PsyPost in 2024, which found that showing partisans how much their core personal values overlap increased perceived similarity, strengthened beliefs that compromise is possible, and made people more hopeful. That study focused on present-day overlap in values, whereas the new study focused on how opinions on specific issues have shifted over time.

As with all research, there are a few things to keep in mind. The sample was an online group of participants rather than a nationally representative one, and outcomes were measured right after reading the messages, with no follow-up.

The study also relied on real statistics for issues that were primarily progressive in nature, so it is not known whether these effects would hold for conservative policies or in different political contexts. Using real polling data also meant the researchers could not test larger shifts in opinion, and many of the real changes they presented were modest.

“The effects on attitudes toward the opposing party were small, and I think it is important to be clear about that,” Loustau noted. “A brief informational message is not going to eliminate partisan polarization. At the same time, polarization is a large, persistent social problem, so even small changes produced by a relatively minimal intervention can be informative, particularly because the intervention consisted simply of changing which accurate information about public opinion people saw.”

“I would view the findings less as evidence for a single solution to polarization and more as evidence that people’s expectations about where political groups are headed can influence how they think about those groups in the present,” he observed.

It is also possible that emphasizing negative trajectories could have the opposite effect. Messages highlighting increasing disagreement or polarization might inadvertently increase hostility. Future research could explore whether this kind of dynamic messaging works for other intergroup conflicts, such as those involving race or religion, and whether the observed changes in attitude persist over longer periods of time.

“Long term, I’m interested in what happens when we encourage people to think about political differences in a more future-oriented way,” Loustau said. “Rather than focusing only on how different we are right now, I want to understand what changes when people recognize both their own and others’ capacity to change. That shared capacity may itself be a form of common ground, opening space to imagine not just what society currently looks like, but what we want our relationships and world to look like in the future.”

The study, “Partisans misperceive and are influenced by dynamic norms,” was authored by Trystan Loustau, Ashley Fabrizio, Liane Young, Jordan Wylie, and Stylianos Syropoulos.