The Racial Gap on Policing Protests

Black and white Americans agree that violence is bad. They disagree about who’s doing it.

Person with hands raised and back to a line of militarized police
Police and protesters on June 1 in Los Angeles. Mario Tama/Getty Images

There’s been a lot of debate about violence in the protests over George Floyd’s death. The debate has encompassed rioting, looting, vandalism, and assaults by police. But what do ordinary Americans, black and white, think about these issues? Polls taken since Floyd’s death have found a consensus against violence. But they’ve also found significant racial differences in what kinds of threats we perceive and condemn.

Last week, a Morning Consult poll asked whether protesters or police were “most responsible for inciting violence” during the demonstrations. A slight majority of white respondents blamed the protesters; only a quarter blamed the police. But black respondents overwhelmingly blamed the police, 68 percent to 14 percent. A Yahoo News/YouGov poll taken this week got almost the same result among white respondents. But black respondents, by a ratio of two to one, said police had been more violent than protesters.

On questions about police conduct in the protests, there’s a big racial gap. In the Yahoo News poll, more than 60 percent of white respondents said only a few officers, or almost none, had “responded to peaceful protests with violence.” Fewer than 40 percent said many or most officers had responded that way. Among black respondents, the numbers were reversed. In other polls, white respondents have been slightly more likely to say the police response was too aggressive than to say it wasn’t aggressive enough. But black respondents, by margins of 50 percent to 6 percent (in a Marist poll) and 61 percent to 8 percent (in an Economist/YouGov poll), said police used too much force, not too little.

Americans of all races and ethnicities claim to support peaceful protests, but they disagree as to how peaceful the Floyd protests are. In the Yahoo News poll, two-thirds of white respondents said many or most protesters had broken the law. By contrast, nearly 60 percent of black respondents said only a few protesters, or almost none, had broken the law. In the Marist poll, white respondents said by a margin of 26 percentage points that the demonstrations were “mostly legitimate protests,” not “mostly people acting unlawfully.” Among Latinos, the margin was 48 points. Among black respondents, it was 66. In the Morning Consult poll, about 60 percent of white respondents and Hispanic respondents said most protesters were peaceful, not “trying to incite violence or destroy property.” Among black respondents, the number was 81 percent.

Black Americans, on average, express more tolerance for violence by protesters than white or Hispanic Americans do. In a Reuters/Ipsos poll, 47 percent of black respondents (versus 15 percent of white respondents) affirmed that “more violent protests and unrest are an appropriate response to the killing of an unarmed man by police.” In a CNN/SSRS poll, black respondents were more likely (39 percent vs. 23 percent) to affirm that “violent protests” were justified in response to “incidents where African Americans have been harmed or killed by police.” In an Emerson poll, they were slightly more likely (28 percent vs. 20 percent) to affirm that “the burning down of the Minneapolis police station” in response to Floyd’s death “was a justified form of protest.” On each question, Hispanics were in the middle.

The samples of black respondents in some of these polls were small, so while the pattern is consistent and the collective pool is large, the margins in any given survey might be misleading. But to the extent that the margins are accurate, the trend line suggests that the racial gap in opinions about violence by protesters narrows as specific targets are named and as the violence sounds more like opportunistic destruction. When the Emerson poll asked about “looting and acts of destroying property as a means of protest of the death of George Floyd,” the black-white gap disappeared almost completely. All racial and ethnic groups overwhelmingly rejected that behavior.

In Morning Consult surveys, black Americans have been less likely than white Americans to say that it’s very important to protect “private property such as businesses or retail stores from looting or damage” in the protests. In one sample, the black-white gap was 11 points; in another, it was 29 points. Black and Hispanic respondents were also slightly less likely than white respondents, in the Reuters/Ipsos poll, to agree that “the property damage caused by some protesters undermines the original protest’s case for justice.”

Past surveys have also found racial differences on these issues. Three decades ago, the General Social Survey found that black Americans were slightly more willing than white Americans to tolerate protesters “seriously damaging government buildings.” But their tolerance was very low. A 2018 Washington Post-Kaiser Family Foundation poll showed big racial disagreement about “kneeling during the national anthem.” Nearly 70 percent of black respondents said it was sometimes appropriate; only 38 percent of white respondents agreed. But when the poll asked about aggressive or destructive forms of protest—blocking traffic, “burning the American flag,” or “disrupting another group’s rally”—the gap vanished. Black and white respondents decisively agreed that these protests were never appropriate.

Together, these surveys illuminate where we agree and disagree. Black Americans are more likely than white Americans to worry about police violence, as opposed to destructive acts by rioters or looters. They’re less likely to view protesters as violent. They’re more likely to accept violent protest as legitimate in the context of police killings, but their acceptance shrinks as the violence is specified to included looting, burning, or specific destructive acts. Like Latinos and white Americans, black Americans overwhelmingly reject these acts. On that, we’re united.