Policy

Defunders Call for Alternatives to Cops—But Do They Work?

Analysis: The evidence on 'violence interrupters' and de-escalation is mixed at best

defund the police
Getty Images

Progressive protesters envision a future where unarmed "de-escalators" and "violence interrupters" replace police—but the evidence indicates those practices would be no match for violent crime.

The burgeoning movement to defund police has in recent days produced a bevy of proposals for replacing cops, from hiring more drug counselors to creating unarmed traffic officers. Asked how to address the 1.2 million violent crimes that occur every year, police critics have argued that funds should be redirected to unarmed agents trained to stop violence with words: so-called interrupters or de-escalators whose job is to peacefully defuse potentially violent situations.

These ideas would expand on policies already adopted in part by police forces in New York, Chicago, Baltimore, and elsewhere. But, as local lawmakers consider replacing cops wholesale with unarmed counterparts, a closer look at the evidence indicates that such alternatives are erratically and questionably effective at stopping violent crime.

Although it is widely used, there is surprisingly little high-quality evidence to support de-escalation as a tactic. Violence interrupters are more widely analyzed, but the results of numerous pilot programs are mixed at best, with some neighborhoods actually seeing increases in crime under their auspices.

Though they may sound like a good idea to those who fear police violence, it is hard to conclude that the nonviolent alternatives currently being trumpeted have much to back them up.

Calls for replacing police with unarmed "violence interrupters" are growing louder. Sociologist Patrick Sharkey, writing in the Washington Post, mentioned "conflict mediators" and "violence interrupters" as part of a hypothetical nonviolent policing alternative. Nashville-based protest group Gideon's Army has argued interrupters are an effective alternative to police.

Some local governments are listening. The mayor of Harrisburg, Pa., plans to move police budgets toward interrupters. New York City comptroller Scott Stringer has called for moving $1.1 billion out of the NYPD's budget toward "social workers, counselors, community-based violence interrupters, and other trained professionals."

Just because a policy has support among activists or lawmakers, however, does not mean it works.

Police-worn body cameras, for example, are wildly popular; both Democrats and Republicans are expected to include more funding for them in reform proposals. But study after study after study has found that body cameras don't actually reduce police misconduct. One particularly large study found they actually increase the risk of assault against officers.

Such findings seem counterintuitive, but naturally raise the question: Does research support violence interruptors and de-escalators?

De-escalation—using verbal tactics, rather than force, to calm violent or angry people—began to see widespread use after the 2014 protests in Ferguson, Mo. In a 2015 report, the Obama administration explicitly called for de-escalation training to be taught to police to minimize the use of potentially lethal force. Defunding advocates argue that such tactics, in the hands of unarmed police alternatives, could mitigate violence.

But there's little robust evidence about the effectiveness of de-escalation training. Natalie Todak, a professor at the University of Alabama at Birmingham, wrote in her 2017 dissertation that "de-escalation has not received specific empirical attention." Todak told the Washington Free Beacon by email that to the best of her knowledge, there remains "a huge gap in the research" on its effects.

A comprehensive review released in January analyzed 64 studies conducted over 40 years, finding that while there were few downsides and "slight to moderate individual and organizational improvements" from de-escalation training, any conclusion "is limited by the questionable quality of almost all evaluation research designs." In other words, none of the studies allow us to draw strong conclusions about what, if anything, de-escalation training accomplishes.

"As it stands, de-escalation training is a promising practice; that is, a well-intended police reform whose consequences are largely unknown," the study's authors conclude. "[R]ecommendations that de-escalation must be used as a primary tool should await additional evidence regarding its effectiveness and any unintended consequences that may impact officer and public safety."

The story on interrupters is a little more complicated. Many cities have implemented violence interrupter programs, most following a model first developed to fight crime in Chicago. Initially called "CeaseFire," and later relabeled "Cure Violence," the program uses what creator professor Gary Slutkin describes as a "public health" approach to violence.

Under the Cure Violence framework, "violence interrupters," usually residents of a particular violence-plagued community, establish rapport with young men who are in gangs or otherwise at risk for violence. Interrupters often "patrol" neighborhoods, defusing tensions through discussion. They also discourage retaliation after shootings and sometimes offer would-be shooters access to social services.

How well do these programs work to reduce violence? As many cities have implemented Cure Violence programs, there have been lots of opportunities to assess this question.

A pioneering evaluation of Chicago's program, authored by a research team funded by the National Institute of Justice, offers characteristic results. The study compared seven Cure Violence "treated" neighborhoods in Chicago to comparison neighborhoods. In three of the treatment areas, there was a significant decline in shootings, but in the others there were not; gun homicides fell by a statistically significant amount in just one of the neighborhoods, while gang homicides changed in none of them.

That's not nothing, but it is less decisive than one might hope. The NIJ evaluated the practice as "promising," but declined to give it the "effective" rating.

Other cities' programs follow a similar pattern:

  • Baltimore's "safe streets" program cut homicides and nonfatal shootings in one neighborhood, but another saw a drop in homicides and an increase in shootings, while a third saw no effect on homicides and a drop in shootings.
  • In two New York City neighborhoods, Cure Violence significantly cut gun injuries, but only significantly reduced shooting victimizations (being shot at) in one neighborhood.
  • In another Chicago study, Cure Violence neighborhoods saw homicides fall by more than the city as a whole, but shootings by the same amount and violent crime by less.
  • A Phoenix pilot saw violence and assault decline significantly in treatment areas, but shootings increase significantly.
  • Newark's implementation had no significant effect on gunshot wounds in the three years following its debut.
  • A Pittsburgh pilot saw rates of homicide, aggravated assault, and gun assault rise following implementation.

There are more successful outcomes, though they sometimes rely on questionable research designs. Studies in Boston and Las Vegas find big effects, but do not use other parts of the city as controls. Other successes are questionably analogous: Cure Violence helped a lot in Trinidad and Tobago, where the murder rate is six times higher than America's.

There's one other problem with the idea that violence interrupters can replace cops: Many of the most successful variations rely on the police to discourage violence.

The Group Violence Reduction Strategy, a variant on Cure Violence which cut shootings and homicides in New Orleans, offers potential offenders both access to social services and the threat of imprisonment if they commit acts of violence. Similarly, Los Angeles's implementation of Cure Violence had a robust effect on gang violence; a National Institute of Justice analysis found that that outcome was driven not by increased access to social services, but by "intensive law enforcement."

What does all this evidence add up to? Neither de-escalation nor violence interruption are obviously bad ideas, and there's even evidence that the latter is actively helpful.

It is hard to conclude from the available research, however, that either is anything like a substitute for the police, who remain one of the most proven violence-reducing forces available. When it comes to stopping violent crime, words are not as effective as we might like them to be—sometimes, only force will do.