This article was published in partnership with The Marshall Project, a nonprofit news organization covering the U.S. criminal justice system.
MINNEAPOLIS — Last Wednesday, Marcell Harris was hit by a rubber bullet. He had joined the second day of protests in this city over the killing of George Floyd, a black man who died after a police officer kneeled on his neck for more than eight minutes while bystanders filmed. Though these protests began with peaceful demonstrations outside the city’s 3rd Precinct, interactions between police and protesters had escalated. Police unleashed pepper spray, projectiles and tear gas. Protesters threw water bottles, built barricades and destroyed nearby property.
Harris said he had used his backpack as a shield and maneuvered close enough to take the baton of the officer who shot him. On Thursday night, he returned to the same spot to watch the precinct burn. With no police presence to be seen, he and other protesters were celebrating a victory. “I’m nonviolent,” he said. “But this feels emotional. George Floyd popped the bubble. It feels like the beginning of the end.” The end of what? “What we’ve been going through,” he said, referring to heavy-handed and often deadly policing of African Americans. “All the bullshit.”
Watching a peaceful protest turn into something much less palatable is hard. There has been a lot of hard the past few days, as people in dozens of cities have released pent-up anger against discriminatory police tactics. Cars and buildings have burned. Store windows have been smashed. Protesters and police have been hurt. When protests take a turn like this we naturally wonder … why? Was this preventable? Does anyone know how to stop it from happening?
Turns out, we do know some of these answers. Researchers have spent 50 years studying the way crowds of protesters and crowds of police behave — and what happens when the two interact. One thing they will tell you is that when the police respond by escalating force — wearing riot gear from the start, or using tear gas on protesters — it doesn’t work. In fact, disproportionate police force is one of the things that can make a peaceful protest not so peaceful. But if we know that (and have known that for decades), why are police still doing it?
“There’s this failed mindset of ‘if we show force, immediately we will deter criminal activity or unruly activity’ and show me where that has worked,” said Scott Thomson, the former chief of police in Camden, New Jersey.
“That’s the primal response,” he said. “The adrenaline starts to pump, the temperature in the room is rising, and you want to go one step higher. But what we need to know as professionals is that there are times, if we go one step higher, we are forcing them to go one step higher.”
Interactions between police and protesters are, by their very nature, tough to study. Even when researchers get a good vantage point to observe protests in the real world — for example, by embedding within a crowd — the data that comes out is more descriptive and narrative as opposed to quantitative. Some kinds of protests are highly organized with top-down plans that are months in the making. Others, like many of the events across America this past week, are spontaneous outpourings of grief and anger. The social and political context of the time and place also affect what happens. Even a single protest isn’t really a single protest. “You have lots of mini protests happening in many places,” said Edward Maguire, a professor of criminology and criminal justice at Arizona State University. “There’s different dynamics. Some peaceful. Some not. And different police tactics.” In Baltimore on Saturday, for example, a police lieutenant mollified a crowd by reading out loud the names of victims of police brutality, while protesters outside City Hall threw bottles at police in riot gear and police used tear gas on the crowd, WBFF-TV reported.
But just because there’s no data about protests that can be easily compared in a chart doesn’t mean we’re bereft of information, said Pat Gillham, a professor of sociology at Western Washington University. There’s 50 years of research on violence at protests, dating back to the three federal commissions formed between 1967 and 1970. All three concluded that when police escalate force — using weapons, tear gas, mass arrests and other tools to make protesters do what the police want — those efforts can often go wrong, creating the very violence that force was meant to prevent. For example, the Kerner Commission, which was formed in 1967 to specifically investigate urban riots, found that police action was pivotal in starting half of the 24 riots the commission studied in detail. It recommended that police eliminate “abrasive policing tactics” and that cities establish fair ways to address complaints against police.
Experts say the following decades of research have turned up similar findings. Escalating force by police leads to more violence, not less. It tends to create feedback loops, where protesters escalate against police, police escalate even further, and both sides become increasingly angry and afraid.
“Do we know [this] in the way that you know if you put two chemicals together things explode?” said John Noakes, professor of sociology, anthropology and criminal justice at Arcadia University. “No. But there is a general consensus.”
De-escalation, of course, does not guarantee that a protest will remain peaceful, and when protests take an unpredictable turn, it can be challenging for police to estimate the appropriate level of force.
Former law-enforcement officials also said good policing of demonstrations isn’t as simple as just showing up with an approachable demeanor. “The time to make friends isn’t when you need them,” Thomson said. “You have to be in front of it.”
James Ginger, a veteran police monitor who is now overseeing the Albuquerque Police’s settlement with the U.S. Department of Justice, agreed that only this longer-term trust-building exercise works. “Trying to find folks at the last minute that you can put out there in soft clothes and talk to people, frankly and in my opinion, wouldn’t work that well,” Ginger said. “You’ve got to till the soil before you can grow the beans.”
Still, if researchers know it’s not a good idea for police to use force against protests and demonstrations, and that information has been available for decades, why do we still see situations like this happening all over the country?
That part is harder to answer. At one point, in the 1980s and 1990s, many police departments in the U.S. did try different strategies, Noakes and Maguire said. The “negotiated management” model of protest policing called for officers to meet with protesters in advance to plan events together to specify the times, locations and activities that would happen, even when that included mass arrests.
“There was a time when the playbook was much more straightforward. The police would meet with the organizers of the protest, and they would lay out ground rules together that would provide for an opportunity for protesters to do exactly what they have a right to do,” said Ronal Serpas, a former police chief in New Orleans and Nashville who’s now a professor of criminology at Loyola University in New Orleans.
But the era of negotiated management basically fell apart after the World Trade Organization protests in Seattle in 1999, when protesters blocked streets, broke windows and successfully shut down the WTO meeting and stalled trade talks. When protesters violated the negotiated terms, police responded with tear gas and rubber bullets and took away the wrong lessons, Maguire said. “What a lot of people took from that in policing is, we can’t trust these people. We need to be smarter and overwhelm them to nip these things in the bud,” he said. “We sort of went backwards.”
Of course, as Gillham pointed out, negotiating and managing a protest can’t really work if the protest wasn’t organized ahead of time. That goes double, he said, if the topic of the protest is police brutality. It’s hard to negotiate with someone about the best way to demand they be fired.
Instead, it’s become normal in the U.S. for police departments to revert to tactics that amplify tensions and provoke protesters, Maguire said, including wearing intimidating tactical gear before its use would be warranted. Maguire does training for police officers and has tried, for years, to get buy-in on the idea that there could be a different way. “I have good relationships with police and I’ve been working with them for 25 years, and I’ve never experienced pushback like I do on this,” Maguire said.
De-escalation strategies definitely exist. Anne Nassauer, a professor of sociology at Freie Universität in Berlin, has studied how the Berlin Police Department handles protests and soccer matches. She found that one key element is transparent communication — something Nassauer said helps increase trust and diffuse potentially tense moments. The Berlin police employs people specifically to make announcements in these situations, using different speakers, with local accents or different languages, for things like information about what police are doing, and another speaker for commands. Either way, the messages are delivered in a calm, measured voice.
Communication is also a cornerstone of what police know as “the Madison Model,” created by former Madison, Wisconsin, chief of police David Couper. His strategy for dealing with protesters was to send officers out to talk with demonstrators, engage, ask them why protests are made, listen to their concerns and, above all, empathize.
Not all police officers trust this model, however. “When you have overly aggressive crowds you have to address them,” said Anthony Batts, who led departments in Long Beach and Oakland, California, as well as Baltimore. Batts was police commissioner during the violent clashes between police and protesters that followed the 2015 death of Freddie Gray in Baltimore police custody.
Reached by text, Batts said that certain events, like fires and police retreats, “inspire” crowds. He said from his point of view, methods like the Madison model make crowds “go ballistic.” He said he was speaking generally, and that he does not advocate a harsh police approach to the ongoing demonstrations.
A lot of this pushback from police has to do with some legitimate officer safety concerns related to de-escalation, Maguire said. “But we make the argument that [de-escalation] makes officers more safe, by reducing violent confrontations with protesters. If officers come into a situation already wearing protective body armor and face shields, that can make protesters feel uncomfortable and under attack long before there’s any kind of confrontation,” Maguire said.
It’s also just hard to change police culture. Maguire compared it to trying to change hospital procedures by using evidence-based medicine. Even if the evidence is, “don’t perform this surgery in that way or someone could die,” it can still take 20 years for the new technique to be widely adopted.
The disconnect between rank and file and executive leadership — commonly cited as an impediment to policing reform — also seems to get in the way of improving policing of protests. Take the Atlanta Police Department as an example. On Saturday the city’s chief Erika Shields earned plaudits for meeting face to face with protesters, empathizing with their grief and fear, and even reprimanding some of her own officers: “I’m standing here because what I saw was my people face to face with this crowd and everyone is thinking, ‘How can we use force to diffuse it,’ and I’m not having that.” But mere hours later, her department was trending on social media again — this time because officers had used tasers to force two college students out of their vehicle, even though they did not appear to be posing any threat.
That, experts say, speaks to a cultural attitude that is endemic to the profession, and is hard to change with new chiefs or rules.
Thomson encountered this when he tried to make change in Camden. The police department was so dysfunctional that the city took the unprecedented step of disbanding the force and reconstituting a whole new agency from scratch. “When I had the opportunity to build a new police department, I was able to do in three days what would normally take me three years to do, because of work rules, because of the bureaucracy of collective bargaining agreements — there are a lot of impediments to reform,” Thomson said.
Couper, the creator of the Madison Method, said, “It’s this whole attitude of, ‘We keep order because we kick ass, and it’s us against them.’ (…) We’ve got to root those people out and say, ‘Look, this is the job that we expect. This is how a democracy is policed. If you can’t buy into it. I’m sorry. You just have to find another job.’”
This story was updated to include additional comments from Anthony Batts.