Tune into American politics today, and you'll hear something far more sinister than simple disagreement. The language has escalated: political parties trash talk each other—blaming rival parties for policy failures or even for causing incidents with national implications.
And reducing polarization and "partisan animosity"—the distrust and hatred of the other party—is remarkably difficult, according to a new study published in the Proceedings of the National Academy of Sciences evaluating past attempts.
The research was led by the Polarization Research Lab , a nonpartisan team of political scientists from Dartmouth and the University of Pennsylvania.
"As long as political and media systems reward outrage with votes and viewership—stoking division—any individual-level effort to depolarize is up against a powerful, unending tide," says senior author Sean Westwood , an associate professor of government at Dartmouth and director of the Lab. "This isn't a problem that can be fixed at the grassroots level alone."
To determine if efforts to reduce partisan animosity have a lasting effect, the researchers conducted a massive meta-analysis of 25 previous studies, encompassing 77 different approaches. These "treatments" included everything from correcting misperceptions about the rival party to encouraging conversations with opponents and calls for civility from party leaders.
The study shows that such superficial interventions are largely ineffective.
On average, treatments improved a person's feelings towards the other political party by a mere 5.3%. The authors note this small gain is dwarfed by the 7% increase in partisan animosity observed between the 2016 and 2020 U.S. elections alone.
The results are not only modest, but fleeting. The researchers found that 75% of the initial reduction in hostility disappears after just one week. Within two weeks, the effects are almost completely gone.
The team also conducted two new large-scale experiments to see if combining or repeating interventions could work better.In one experiment with 3,500 respondents, they tested if "stacking treatments"—by exposing people to multiple interventions at once—would amplify the positive effect.
In another experiment with over 5,000 respondents, they evaluated if providing a "booster shot"—a repeated treatment over time—would make the effects last longer.
The results were clear: neither stacking treatments nor administering them repeatedly produced significantly larger or more durable results. In essence, flooding the airwaves with public service announcements to counter political hatred is not an effective strategy.
"To achieve lasting depolarization in the U.S., a fundamental shift in society is needed," says Westwood. "From the top down, we must address the behavior of political elites and the structural incentives that fuel conflict, and from the bottom up, we need a citizenry with the civic skills to engage constructively across differences."
The co-authors report that while interventions built on genuine dialogue are difficult to scale, they remain the single most effective tool for reducing polarization and require long-term investment.
"Principles of civil discourse and respectful dialogue need to be embedded into the education system in the U.S.," says Westwood. "The future of our democracy depends on it."
Westwood says, "Without more systematic changes, America's divisions will only continue to deepen."
Longtime collaborators Derek Holliday at The George Washington University and Yphtach Lelkes at the University of Pennsylvania also contributed to the study.
