For one recent study, which is currently under review, the authors examined what happened when three different types of protesters—animal rights, Black Lives Matter, and anti-Trump—used either moderate or extreme protest tactics.
“A prototypical extreme protest is something where vandalism occurs or violence is threatened, or protesters behave in a violent way, or an interstate highway gets shut down,” said Robb Willer, a Stanford University sociologist and co-author of the study. The Women’s March, Willer says, is a prototype of a moderate protest—one without hateful rhetoric or violence.
First, the researchers found that the study’s participants identified less with and were less willing to support a fictional group of animal-rights protesters who broke into an animal-testing facility than with those who marched peacefully. Then, both African-American and white participants felt more support toward Black Lives Matter protesters if they read that the protesters chanted anti-racist slogans than if they encouraged violence against police officers.
Finally, and perhaps most topically: The researchers showed people a video of a “moderate” anti-Trump protest, in which protesters held signs and chanted, as well as a news report about an “extreme” protest, in which protestors caused a traffic jam and blocked Trump supporters from reaching a Trump rally. People shown the extreme anti-Trump protests actually supported Trump more—an effect that occurred, to varying degrees, among liberals and conservatives alike.
Over and over, the researchers found the reason the extreme protesters were dissuasive is that less-radical bystanders couldn’t identify with them. People generally don’t see themselves as disruptors of the social order, Willer told me, even for causes they believe in. Ultimately, our belief in something is surpassed by our desire to conform.
“When the social order is being greatly disrupted, when property is being destroyed, when there’s some risk of harm to people, that leads to a dis-identification effect, where people say ‘I’m not like those people,’” Willer said.
The problem is, the extreme protesters didn’t realize this would happen. When Willer and his co-authors surveyed people about the causes they believe in and what they would be willing to do for the cause, the truest believers were willing to go to the most extreme lengths—and they thought the tactics would help gin up support.
“It can be really difficult to take the perspective of a bystander who has not yet joined a movement, when you’re interacting mostly with other activists,” Willer said. “Bystanders are asking, ‘Am I like them? Can I see this issue the way they see it?’”
The findings echo the results of another recent study by Princeton University’s Omar Wasow, which found that nonviolent, civil-rights protests of the 1960s boosted votes among whites for Democratic candidates, who supported civil rights, while violent protests increased support for Republicans, and might have even tipped the the 1968 election for Richard Nixon.