Talking with those who hold differing political views has little effect, study says

People hope that crosspartisan conversations can help ease affective polarization and promote a healthier democracy. However, a new study by the University of California at Berkeley and Stanford scholars has found that simply getting Democrats and Republicans to talk to each other has scant power to bridge political divides.

The new study co-authored by UC Berkeley political scientist David Broockman and Stanford social psychology doctoral student Erik Santoro was released in the journal Science Advances on June 22.

Given the growing political fissures in the U.S., many activists and scholars have started efforts to bring Democrats and Republicans together for discussion, hoping they can get along or be more willing to vote for politicians of the other party. However, Broockman and Santoro wondered how effective these approaches truly would be.

“A lot of these efforts … are premised on the idea that this would then accomplish something for democracy and make people more supportive of the democratic norms,” Broockman said. “And we essentially found none of that.”

About two and a half years ago, the two authors decided to work together and conduct a focused test of the hypothesis. They started by exploring a simple question: What would happen if you did get Democrats and Republicans to talk?

In two experiments, Broockman and Santoro paired up hundreds of Republicans and Democrats across the county and let them discuss a casual topic: their perfect day.

“We were broadly interested in having a topic that wouldn't be politics and sort of allow them to have a pleasant interaction,” Santoro explained from a psychological point of view.

The first experiment examined whether the crosspartisan talk is effective in the short and long term. It turned out those conversations about their perfect day caused very large decreases in polarization at the beginning, Broockman and Santoro found. However, such positive effects were found to have disappeared in a follow-up survey three months later.

In the second experiment, the two authors asked pairs of participants to discuss potentially tense political topics beyond their perfect day, such as what they liked about their party and disliked about each other's party. They found that these conversations had virtually no effect on reducing polarization.

“This kind of study, as simple as it sounds, is actually quite difficult to pull off,” Broockman said. “Think about the logistics of it, especially during a pandemic for any kind of study that you want a reasonably large sample size.”

The most challenging part, Broockman said, was to invite people from all over the country to take part and make those one-on-one conversations happen. According to the study, more than 10,000 people completed a screener for two experiments; 478 successfully began a conversation for the first and 338 for the second experiments.

Broockman and Santoro put each pair of participants in a real-time video call relying on software called AllSides, and it could be a Republican in Nebraska talking to a Democrat in California, for example.

Although the study demonstrates that the effects of cross-partisan conversations would decay over time and don't seem to work when focusing on disagreements, Broockman doesn't think it simply contradicts those efforts.

“Being one of the first studies on this, I think not everyone would have predicted these results,” Broockman said. “It just really underscores how important it is to study these kinds of interventions to try to improve democracy or reduce polarization.”

Broockman hopes this latest research can lead to more studies examining the limitations of such political efforts, especially those in collaboration with practitioners working on reducing polarization.

Santoro said he wants to study further how people can improve these conversations and make them more effective.

“I'm really interested in if we could give people specific strategies of how to listen or respectfully engage that might lead to reductions in affective polarization or even have other downstream consequences,” Santoro said.