EDITOR’S NOTE: May 20, 2015: On Oct. 6, 2014, Bloomberg Politics published the below article about a canvassing tactic developed by the Los Angeles LGBT Center’s Leadership Lab in the wake of the passage of Proposition 8 in 2008, in which volunteers were invited to discuss their own sexual identity to voters at the doorstep. Many of the claims in the article about the effectiveness of the tactic in changing opinions on gay marriage were based on the experimental research performed by UCLA graduate student Michael LaCour, whose partnership with the Leadership Lab was described in the article. Two months after Bloomberg Politics first reported on LaCour’s research, his findings were published in the journal Science in an article co-authored with Columbia University political scientist Don Green.
In May 2015, after attempting and failing to replicate the design of the surveys that LaCour used to measure opinion change among those whom the LGBT Center’s volunteers has canvassed, graduate students David Broockman (of Stanford) and Joshua Kalla (of UC-Berkeley) concluded that the data used by LaCour and Green suffered from “irregularities.” Subsequently, a variety of LaCour’s collaborators and colleagues have raised doubts about whether he ever collected any of the data supposedly analyzed to demonstrate the effectiveness of the Leadership Lab’s tactics. On May 19, persuaded that the data LaCour had presented to him had been likely faked, Green requested that Science retract the article they had written on the subject. “Michael LaCour’s failure to produce the raw data,” Green wrote to the journal’s editors, “undermines the credibility of the findings.” Bloomberg Politics contacted LaCour, but he had not completed his response.
To those who toil on the field side of campaigns, LaCour and Green’s now-discredited finding about the value of meaningful face-to-face encounter had offered validation. The hard work of knocking on unfamiliar doors actually could change minds. To everyone else, the LaCour/Green paper offered the shape of an answer to the vexing public-opinion question of our time: how did so many Americans change their views on gay marriage so quickly?
The Leadership Lab was conducting these canvasses well before its executive director, Dave Fleischer, met Green or was introduced to LaCour—11,000 conversations by volunteers with voters in the Los Angeles area over four years. Many of these were captured on video and are themselves inspiring, about the power of a candid conversation with a stranger to change minds. In some cases they offer support for the animating thesis of Fleischer’s efforts to transform campaigns in the wake of Proposition 8’s failure: that interactions with a gay or lesbian person can play an important role in softening opposition to gay rights. But those chats, are anecdotes about this grand social change we have witnessed but yet to fully understand. They are not evidence.
When Gallup first asked about legal recognition for same-sex unions, in 1996, only 27 percent of Americans said they were supportive of gay marriage. This May, that number was 55 percent. The stories we tell about how this happened usually minimize politicians and campaigns, preferring the logic of the natural evolution of public opinion: People realize they know gays and lesbians, grow comfortable around them, and come to accept their political demands in human terms.
That view is supported by the vast body of social-science research showing that personal interactions can change views rooted in prejudice. What psychologist Gordon Allport hypothesized in the 1950s as “contact theory” now tints many conversion narratives on gay marriage. In 2012, Barack Obama pointed to his children’s friends raised by same-sex couples to explain why he changed his mind; the next year, Ohio Republican Senator Rob Portman announced he’d been influenced by his gay son.
Recently, a series of unusual field experiments in California have tested whether contact even with strangers can accelerate the process of changing individual opinions on controversial issues. The architect of this approach is a political activist named Dave Fleischer, who moved to California after the passage in 2008 of Proposition 8, a ballot measure that overturned same-sex marriage in the state. The initiative prevailed on what gay activists had considered friendly turf, winning a majority in heavily Democratic Los Angeles County.
Fleischer and his team at Vote For Equality, a program run by the Los Angeles LGBT Center, identified precincts in the L.A. suburbs where Prop 8 had received support and sent volunteers to knock on doors to ask voters why they opposed same-sex marriage. “It took me over 40 years of canvassing before it occurred to me to do that,” Fleischer says. The first script he drafted for these visits, in January 2009, pivoted around two simple questions: “How do you feel about marriage for gay and lesbian couples?” Then: “Can I ask you why you feel that way?” At that point, a stage direction advised: “Optional: Insert Brief Personal Statement.” Many of the volunteers were gay. The instructions invited them to out themselves to someone they had just met.
After more than 11,000 visits, it became clear that just talking to a gay person was softening opponents’ resistance to same-sex marriage. The conversations often lasted 20 minutes or more. “From the point of view of the voter, we’re having a very low-stakes conversation,” Fleischer says. “We’re talking with them privately, one-on-one, out of earshot of anyone else. And this person is listening and is not judging them. And is allowing them, and in fact encouraging them, to talk about things that really matter to them. Well, how often does that happen in life? Almost never.”
In 2013, Fleischer met with Donald Green, a Columbia University political scientist who specializes in voter behavior. “I was of the opinion that long-standing attitudes are very slow to change,” says Green, “and that even if you do change them they’ll snap back to the old opinion not long after the contact ends.” Green offered to help test the hypothesis and introduced Fleischer to Michael LaCour, a University of California at Los Angeles graduate student.
As an undergraduate at the University of Texas at Austin, LaCour designed an experiment to test how people’s opinions about gays changed after seeing Facebook posts from their friends featuring an “I’m Coming Out” badge. (Facebook refused to cooperate with the experiment because it pushed a political agenda.) With Green, LaCour designed an experiment to measure the effect Fleischer’s volunteers were having on people’s views in the real world.
LaCour sent volunteers from Vote For Equality to visit about 1,000 households in precincts where Proposition 8 fared well. Some households were randomly assigned to hear a version of Fleischer’s script on same-sex marriage and others got a placebo in which the canvasser promoted recycling instead. LaCour also kept track of which voters who heard the same-sex marriage script had been visited by gay volunteers. In follow-up surveys with those households, average opinion in favor of gay marriage—now measured on a 100-point “feeling thermometer”—jumped by double digits among those who talked about same-sex marriage with gay canvassers, nearly twice the change among those who talked about gay marriage with straight volunteers.
It was contact theory in practice. The experience of discussing same-sex marriage appeared to jolt people into an instant reconsideration of attitudes they might have characterized, just before the doorbell rang, as deeply held. “You may have a co-worker who’s gay, but are they directly discussing the issue at hand?” asks LaCour. “It’s fairly uncomfortable and truly weird to have a complete stranger come out at the door.” In follow-up calls, LaCour also surveyed people living with voters who met volunteers. The results were astonishing: Those people’s opinions changed, too. The conversations that took place after the canvassers left seemed to be almost as powerfully mind-changing as the originals. The effects were still measurable a full year later.
For those looking to change voters’ minds on contentious social issues, the findings offer a tantalizing promise. Green is already working on an experiment in Texas where Latino canvassers out themselves as Democrats to Latino voters. He’s also seeking funding for one where volunteers pushing for gun control disclose they are victims of gun violence. (Such a design could easily be flipped to advance an opposing agenda, by having canvassers reveal they are gun owners.) “The question for those who want to bring this to practice is, how much can you water it down and still have some of the impact?” Green says. “If you have something less than a 22-minute conversation, can you still change someone’s mind?”
Fleischer is now helping Planned Parenthood Los Angeles overhaul its canvassing program. After failing to stop several statewide ballot initiatives that added parental notification requirements for minors seeking abortions, Planned Parenthood’s leadership concluded that its traditional messaging strategies didn’t work, says Sue Dunlap, the chief executive officer of the Los Angeles chapter. Now they’re trying Fleischer’s approach. “We’re making space for people to have complicated, complex feelings and still move forward,” Dunlap says.
On a recent Saturday morning, several dozen volunteers gathered in Whittier, Calif. Trainers told them to start by asking voters to measure their support for abortion rights on a scale of 1 to 10. Then, the reveal: The volunteers would talk about their own experiences with abortion. “Build rapport with the voter by being yourself,” advised a trainer named Virginia Millacci. “When we’re vulnerable with the voter, it really pays off.”