The Surprising Power of Conversations to Change People’s Minds—Including Our Own

Many of us avoid talking politics with people who disagree with us. After all, what’s the point? No one’s minds will change, and it might only hurt the relationship.

But new research suggests that’s not necessarily the case. Instead, we tend to overestimate how badly these kinds of conversations will go and underestimate how much people’s views can change within a conversation, including our own viewpoint.

“Entering into a conversation is perhaps one surprising way of acquiring a broader view of an issue that you might care about,” says lead researcher Michael Kardas of the University of Wisconsin, Madison. “Having conversations, despite people’s predictions, [can] decrease polarization.”

Negative predictions can be overblown

In the study, diverse groups of adults from the United States (and, in some cases, from Britain) were recruited to discuss conflicting viewpoints on different issues, including a non-political one (whether dogs or cats make better pets) and two sociopolitical ones (whether cancel culture is good or bad, or whether Joe Biden was a good president).

Before having these 10-minute conversations, participants first reported on the strength of their views, their reasons why they held these beliefs, and their expectations for a discussion involving someone with an opposing view—in particular, whether they anticipated they or their partner’s views might shift and become more similar to each other.

Researchers gave participants prompts to guide their 10-minute conversation, including sharing their reasoning for their views and asking about their partner’s reasoning. After the discussion, they reported again on the strength of their views. Kardas and his team then compared their current ratings to their earlier predictions.

In all cases, people’s views changed more than they’d anticipated—and so did their partner’s views. Contrary to most expectations, the conversations ended up depolarizing their opinions, to some extent. For Kardas, this was a surprising result.

“People went into the conversations anticipating that no one was going to change their mind,” he says. “But it turned out that even on a conversation about a topic related to personal preferences, like cats versus dogs, people became less extreme in their views to a greater extent than they themselves expected.”

When the researchers surveyed the participants a week later, they found that people’s views on the issues were still less polarized than they were initially, though not as much as they were immediately after the conversation ended. Interestingly, though, the people whose attitudes changed the most, post-conversation, were also most likely to stay that way.

“This suggests that conversations offer, at least to some extent, a somewhat durable route to informing people about issues that carries forward into their later attitudes,” says Kardas.

Why discussion may decrease polarization

To get at why conversations across difference can be depolarizing, the researchers did several sub-experiments and analyses. One thing they discovered is that having these cross-group conversations widens your perspective, because you better understand what information someone draws upon to form their opinion.

For example, says Kardas, let’s say someone believes cancel culture is a good thing. They may have come to that opinion because they’re focused on an example like Harvey Weinstein, the movie producer who was exposed as a sexual predator and then “cancelled.” Someone else may believe cancel culture is bad, because they’ve seen it tar the reputation of someone whose behavior wasn’t criminal, but merely provocative—for example, Colin Kaepernick, who was “cancelled” for kneeling during the playing of the national anthem to protest racial injustice. Hearing how people came by their views for and against something can help us find points of understanding or even agreement, softening our own stance.

• 

  Bridging Differences in Higher Education Playbook

  16 research-based bridging practices each grounded in science and real campus case studies

  Download

“When people formulate their attitudes, they’re often drawing on only a small subset, a small sliver of information that comes to mind when they think of the issue,” says Kardas. “[But] when you actually have a conversation, you recognize that the basis of your attitude and the basis of another person’s attitude are actually drawing on different facets of a multifaceted issue.”

That’s why it’s important to recognize that these conversations across difference have potential, he says. Otherwise, we just get our information only from sources we already agree with, like friends, family, or, often, social media. That might keep us stuck in our shared fears.

“If people expect that a conversation will be unproductive, unfruitful, unnecessary, because no one’s going to learn anything, no one’s going to change their minds anyway, then they’ll stay in the areas of social media and with people that share their views,” he says.

Knowing can make a difference

Can these findings actually change anything, though? Won’t we still be reluctant to talk to people who don’t agree with us?

To find out, Kardas and his team did another experiment. They gave some participants information they learned from their research showing that conversations across difference can reduce polarization, because a conversation partner’s views will become more moderate, their own views would become more moderate, or both views would move closer together. Other participants weren’t told anything about these results.

Then they compared how willing participants were to have conversations across difference (in this case, different views of Joe Biden’s performance). In all cases, those who were told about the research were more willing to discuss the polarizing issue than those who weren’t told.

The study adds to other research showing that our expectations for engagement across political and other differences might be overly negative, keeping us from hearing each another.

For example, Kardas says, a 2019 study found that people overestimated how badly they would feel listening to speeches by politicians on the other side of the political spectrum from themselves. Once informed that their fears about engaging across differences were likely exaggerated, people became 24–34% more engaged with diverse sources of information.

“It tends to be less emotionally jarring than people think it will be when you’re exposed to information that doesn’t align with your own beliefs,” says Kardas.

Similarly, research has shown that people overestimate how polarized we are politically and that correcting that view can help people become less extreme around their own positions. In another study, giving people more accurate information about how those in an opposing political party viewed members of their own party (less negatively than anticipated) helped people attribute more positive qualities to political opponents.

This is good to keep in mind, because many of us already have overblown fears around having conversations with people we don’t know, let alone someone we disagree with. If you expect people to be angry or upset with you based on your views, it could keep you from talking to them and, potentially, increasing civility.

“The same tendency for conversations to reveal unexpected areas of agreement that I’ve detected . . . also [seems to] produce a less hostile and more congenial conversation than people expect,” says Kardas.

This is good news for organizations who work to bring people on opposite sides of political issues together to have conversations, says Kardas. Though he doesn’t necessarily think depolarization is the goal of having these conversations, encouraging people to have them could result in more warmth and shared knowledge, making it easier to find consensus around solving our social problems.

“To the extent that people can acquire a complete understanding of an issue, and then arrive at what they themselves believe is the appropriate attitude, we can collectively be better informed citizens about the issues we consider socially and politically important,” he says.