The Fallout from Black Lives Matter Protests

L ast week, Vox spotlighted some results from an important new study by Travis Campbell.

The study looks at the effects of Black Lives Matter protests, using data that end in 2019 (and thus do not include the George Floyd protests last year). The good news is that the protests did seem to reduce police-involved homicides in the places where they occurred — by maybe 15 to 20 percent, relative to trends elsewhere — and also led departments to adopt body cameras and community policing. The bad news is that the protests also coincided with an increase in murder, by somewhere around 10 percent, and reduced the employment of black officers.

Since everyday murders are so much more common than police killings, these effects produce a lopsided death toll: perhaps 300 police homicides prevented, vs. 1,000–6,000 more murders.

This study isn’t gospel. It’s the work of a Ph.D. student and has not yet been peer reviewed and formally published, and of course even published studies are hardly infallible. It’s also limited in various ways, including the time period and the fact that, by comparing individual areas with each other, it sidelines the question of what nationwide impact the protests had.

But at this point, more than half a decade out from Ferguson, the evidence is strong that the cycle we’re trapped in — police killing, followed by intense and sometimes violent protests, followed by crime spikes as cops back off — is not serving anyone well. We need both careful police reform, and an activist class that’s pickier in choosing its battles.

In city after city, violence has exploded after policing incidents went viral. Baltimore broke its murder record following the Freddie Gray incident, and its murder rate has yet to return to normal. Chicago had a surge in gun violence upon the release of the Laquan McDonald video. Minneapolis experienced what Charles Fain Lehman called an “unprecedented spike in gun assaults” following the George Floyd protests. At the national level, murder rose in 2015 and 2016 after decades of declines, and shot up again in 2020, especially over the summer.

These patterns are backed up not only by the new study, but also by a paper last year from Tanaya Devi and Roland Fryer. Rather than focusing on BLM protests per se, Devi and Fryer studied the effect of “pattern or practice” investigations of police departments. These investigations usually reduce crime, they found. But in the five cases where the investigations were set off by viral incidents, crime rose instead, resulting in 900 excess homicides.

There is also evidence that depolicing is part of what connects these incidents and protests to increases in violence. The new study, for example, finds that property crimes reported to the police fall after BLM protests — which likely happens because the community trusts police less and fails to report crimes, not because actual property crime is down. The police themselves make arrests in a lower share of reported property-crime cases.

Devi and Fryer similarly found decreased police activity following many of the viral incidents they studied: “In Chicago, the number of police-civilian interactions decreased by almost 90% in the month after the investigation was announced. In Riverside CA, interactions decreased 54%. In St. Louis, self-initiated police activities declined by 46%.”

So where does that leave all of us? I would suggest three changes can help: Activists should choose their cases carefully, legislators should address activists’ legitimate concerns, and police departments should find ways to keep police activity up even when officers are under scrutiny.

Americans have a First Amendment right to protest the government’s actions, and police who use force without justification need to be held to account. Further, I think it’s clear, both from the new study and just from watching the police-reform debate over the past decade, that protests can change policy. But when protests can themselves turn violent, and when they can set off crime increases that cause far more deaths than bad policing does, activists might want to choose their cases and their tactics more carefully. After all, some of these protests, from Ferguson, Mo., to Kenosha, Wis., ended in violence and were based on misleading-at-best accounts of the incidents that triggered them.

Meanwhile, as I’ve detailed in the past, there are ways we can improve policing and better hold police accountable without demonizing and demoralizing cops across the board — and without reducing legitimate police activity (which increases crime, and which no demographic of Americans wants). We can curtail “qualified immunity,” limit the power of police unions, require body cameras, and more aggressively weed out bad cops. We can improve training, restrict neck holds, and expand community violence-prevention programs with strong evidence behind them. If we bring down overall crime rates and reduce unjustified police shootings, we take away some of the impetus for these protests.

And since reduced policing in the wake of viral incidents seems to be part of the puzzle here, departments should explore ways of combating this pattern, including disciplining cops who refuse to do their jobs. Yes, it’s human nature to slack off when you feel targeted and unappreciated. No, that’s not a good reason to let criminals run wild killing people.

Right now we seem caught in a doom loop, where police violence — some of it justified, some of it not — sparks protests, depolicing, and crime, until the next viral incident comes along. The bodies only keep piling up as we struggle to find a way out.