For anyone who feels that healthy discourse is dead or at least increasingly out of reach in the current American political climate, a Stanford GSB professor says take heart. In a new study published in Science, professor of political economy David Broockman found that a single 10-minute conversation with a stranger could reduce prejudice toward transgender people and increase support for nondiscrimination laws.
In 2015, Broockman and co-author Joshua Kalla, a PhD student at the University of California, Berkeley, ran a field experiment to investigate the effects of a door-to-door canvassing operation conducted by the Los Angeles LGBT Center and SAVE (a South Florida LGBT organization) in Miami. The effort was organized after the Miami-Dade County Commission passed a law protecting transgender people from discrimination in housing, employment, and accommodations. To counter a potential public backlash and an increase in transphobia, the two LGBT advocacy organizations sent dozens of canvassers to knock on doors and talk with voters.
Both before and in several waves following those conversations, the researchers engaged a number of those voters in ostensibly unrelated online surveys to track their feelings toward transgender people. They found not only a dramatic decrease in transphobia and an increase in support for nondiscrimination laws after those conversations, but that the effects persisted for a period of at least three months following the encounters and were resistant to counterarguments.
These findings echo a now retracted paper by Michael LaCour and Donald Green about the effects of canvassing campaigns on support for gay marriage. Here, Broockman discusses what precipitated his research, what he found about changing people’s attitudes, and how a greater understanding of interpersonal interactions might pave a path to help advocates reduce prejudice.
Your study looks at how door-to-door canvassing can reduce prejudice toward transgender people. What made you want to investigate the effects of canvassing to begin with?
I’ve long been interested in the effects of high-quality personal interaction from political campaigns. More than a decade of research has shown that these kinds of high-quality, face-to-face conversations are the best way to turn people out to vote. But there had been much less research on the role of such conversations in changing people’s minds. And likewise, in the literature on prejudice, there had been very few studies that looked at whether there was some way to change somebody’s attitudes about an out group along the basis of a personal conversation.
What motivated this study in particular, and how did you come to work with the Los Angeles LGBT Center?
I wrote this article for Vox with Joshua Kalla in November 2014 about canvassing, citing this Michael LaCour and Donald Green study that I thought was so great, which then came out in Science in December of that year. It also really interested me in the work of the Los Angeles LGBT Center, which was what that study was ostensibly looking at.
I got to know Dave Fleischer, who is the head of the organization’s Leadership LAB, and he asked me to study what they were working on with a Miami organization, SAVE. In planning the study to investigate canvassing’s ability to reduce prejudice toward transgender people, we went back to look at LaCour’s study. We examined the data and procedures much more carefully than one otherwise would because we wanted to follow that recipe. It’s like the difference between eating a meal and hearing 15 seconds about how it was made — how most people were reading that study — versus actually trying to buy the ingredients and follow the recipe step-by-step.
That’s when we realized a lot of things didn’t add up. Ultimately, we brought some things in the data that we thought didn't look right to Green’s attention. We realized that at least some of it was falsely described, and Science retracted it.
Everyone said, “Oh, well, now this hypothesis is disproven.” No, it was just unproven. Despite the retraction, we knew we had to finish the Miami experiment, which had already been months in the planning.
How does your study differ?
The big difference — because I don’t want people to mistakenly think, “Oh, it was actually true all along” — is that with those findings it was all about the identity of the canvasser. It had to be a gay canvasser to change your mind. But we found that the canvasser's identity doesn't really seem to matter; both transgender and non-transgender canvassers were effective. What seems to matter most is that canvassers are experienced. And, of course, we weren’t studying the same thing. These are attitudes toward transgender people, not toward gay people and marriage equality.
In your study, how much did people’s minds change as a result of engaging in these conversations?
We found a pretty large effect — and even more notably, a lasting effect. We saw a decrease in prejudice against transgender people as well as an increase in support for a nondiscrimination law that would protect transgender people from discrimination. The type of laws that, for example, the city of Charlotte passed and that the North Carolina legislature just got rid of. That’s the kind of law that we find that people are more likely to support.
We used a feeling thermometer, which is a way to get a summary sense of how warmly or coolly someone feels toward a group, and saw a shift from those who had the conversation versus those who did not have it. It was similar to the shift you see in public opinion toward gay men and lesbians between 1998 and 2012.
So, when viewed in that way, it’s about a decade and a half of change in 10 minutes.
Conventional wisdom holds that people’s views are deeply ingrained and more or less impervious to persuasion. What do your results have to say about this stability of people’s attitudes?
I do want to caveat very heavily to say this is just one study. We also ran a similar study around the issue of abortion that found no effect. So it’s not like this is some silver bullet, and I think we don't have the full story yet.
With that said, one hypothesis, which would be consistent with some public opinion research, would be that people have absorbed a lot of political messages that they believe on some surface level, but not on a deep level. Maybe the hardest-core partisans do, but a lot of people know what they’re supposed to think, and they think that but not strongly. So when they have a rare moment that someone comes to their door and has a friendly conversation, all of a sudden they feel they’ve really made a decision about it in a lasting way.
How generalizable is that? Can it be generalized to climate change? To gun control? To racial prejudice? It could well be that this is an issue that is unique in some way. This is something that’s animating a lot of politics, so even if this is just a way to reduce prejudice on this issue, I think that’s great.
As evidenced by the recent law passed in North Carolina that you mentioned, and similar laws proposed elsewhere, transphobia seems pervasive even as homophobia seems on the decline. Why do you think that is?
I think this is one of the next frontiers in the battle for LGBT rights. With the same-sex marriage issue largely off the political agenda in the U.S., both advocates and opponents of LGBT equality are turning to this issue more and more.
In some ways, we’re seeing the same patterns that we saw with gay marriage, where 40 years ago in Miami — where this study took place — Anita Bryant had this infamous “Save Our Children” campaign that played on the same ideas about gay people that we're seeing about transgender people now: “You can’t trust them around children. They are all sex offenders. They’re unstable.”
Those same stereotypes are being brought up again. And likewise, just like we saw with gay marriage over the last 10 or 20 years, opponents and supporters are going to use this issue to try to motivate people to vote.
What happened in these canvassing conversations that made them so effective in changing people’s attitudes?
There are two main theoretical paradigms at play here.
The first is the idea of perspective-taking, which is to think about what it’s like to be somebody else in some way. The heart of the intervention is that canvassers build a rapport with people, where their goal is to try to get those people to tell stories of when they were discriminated against in their own life.
The second idea, called active processing, is just a fancy way of saying the difference between quick thinking and slow, effortful thinking. A recurring finding of laboratory studies is that a relatively short intervention can have a pretty lasting effect if people are engaged in this kind of effortful thinking.
Many theories suggest that prejudiced attitudes are resistant to change. Yet you found that this reduction in prejudice was remarkably resilient over time and even held strong against counterarguments. Does this fly in the face of what we thought we knew about attitude durability?
I don’t know if it flies in the face as much as it’s a contrast to what we find when we study more impersonal tactics. The effects of mass media — mail and television ads, for example — typically decay within hours or days. One theory for what happens when people encounter arguments in mass media is that again, they’re not thinking them through. If you ask them about it the next day, they'll remember that they saw it. But they haven't actually taken the time to say, “I should really change how I think about this.”
Our best guess of what’s happening is that it’s not just someone shows up at your door and asks what you think. Instead, someone shows up at your door and asks you questions and gets you to have that kind of effortful thinking. Burning mental calories. That is what is leading people to remember this.
In contrast to the retracted study by LaCour and Green, which suggested that gay canvassers were able to change people’s attitudes about gay marriage, your results show that the identity of the canvasser doesn’t make much of a difference.
Our best guess was that there was a small difference, but not much. Ultimately, the most important thing is what’s happening in the mind of the person they’re talking to. It’s not likely about the message that the canvasser has, and it's not any aspect of them per se. It’s about the work they get the voter to do. More experienced canvassers seem more effective, which points to the idea that knowing how to have these conversations is a skill that can be learned.
What else is particularly exciting about all this for you?
Before, it was basically not feasible to study the effects of these interpersonal interactions on people’s attitudes. We’ve taken a lot of best practices in experimental design and come up with a new method for studying canvassing, which we’re releasing in a companion paper focused on the methodology itself and how to do it.
It opens the door for us to learn a lot more. More than anything, what I’m excited about is to do more work of this type, and to see others do more work of this type, and take advantage of this method.
Your work suggests there’s some optimism for healthy discourse in an increasingly polarized public sphere.
That’s right. There have been two big shifts over the last 50 or 60 years in campaign practice. One is from personal contact to mass media contact. People who are politically ambitious, instead of staying in their communities and trying to build local organization, now move to the coast and send broadcast messages to the interior.
Second, the kinds of messages they're broadcasting are intended to take what people already believe and make them really angry or really excited about those things.
What we found suggests that campaigns might have more success than they expect trying to talk with people that have initially opposing views, even regarding controversial topics and across partisan lines.
So, for example, if you’re a politician, I think our study underscores the following message from other research: It actually might be more effective to go meet voters one-on-one than to focus on raising money to send them mail. That is, it might be in politicians’ and campaigners’ own best interest to hear voters’ concerns one-on-one. Therefore, regardless of the virtues of the causes that embrace this technique, I think our democracy as a whole will be better off as a result.