Do police body cameras work?

THERE IS GROWING pressure across the country to equip police officers with body-worn cameras, driven by a belief that the technology will reduce unwarranted use of force by police and is also likely to lower the incidence of assaults on police.

In her 2013 ruling against the New York City police department’s “stop and frisk” policy, a federal judge ordered a body camera pilot study in one New York precinct. Since then, calls for body cameras have intensified with each high-profile case of police using lethal force in an encounter with an unarmed black man. Hundreds of departments are now using the technology.

But what does the evidence on body cameras show? The weight of the evidence supports the value of body cameras, but the evidence is not clear-cut, research on the topic is still in its infancy, and the widespread adoption of the technology by departments will come at significant cost — money that won’t be available for other uses.

The main idea behind the use of body cameras is based on a concept known as deterrence theory — that the act of being observed or monitored changes behavior and will induce people to act more responsibly and with greater restraint during what are often very charged situations where emotions run high. This ought to be true for both police officers and members of the public they interact with.

The argument for body cameras also rests on the broader idea that they will yield greater accountability for police behavior and increased levels of trust in police by residents. That includes their potential to aid in the fair and speedy resolution of citizen complaints about police behavior. Body cameras are also being viewed by prosecutors as providing potentially valuable evidence for use in criminal proceedings.

The deterrence theory apparently did not work in a recent high-profile case in which body cameras worn by police in Chicago show officers repeatedly firing — in violation of department policy — at a car they are pursuing that had been reported stolen. The body camera video is, however, likely to be crucial to any proceedings that hold officers accountable for their actions. The case also highlighted the importance of following body camera protocols, as the unarmed suspect was shot to death in a foot chase after the car he was driving crashed into a police cruiser, an encounter that was not captured on video because the officer had not activated his body camera.

Research on body cameras is still in its early stages, with only a handful of studies completed to date, not all of them showing clear benefits. A jolt went through the criminal justice research field this spring when a report by a British group at the forefront of body camera research suggested that body cameras were not only ineffective in reducing the use of force by police but were actually associated with a higher rate of assaults of against police officers.

The research, led by Barak Ariel of the University of Cambridge and Alex Sutherland of the Europe office of the Rand Corporation, aggregated the results of 10 trials of body cameras in an effort to provide a more reliable estimate of their effect based on a larger sample size. The overall finding was that body-worn cameras (BWCs) did not lead to a lower incidence of use of force by police. Not only did that finding run counter to the presumed effect of body cameras, the study reported that camera use appeared to increase by 15 percent  “the likelihood of an officer being assaulted during a shift compared to not wearing the cameras.”

In a separate paper, the authors carried out a deeper dive on the data to try to better understand the surprising findings. “Taken at face value,” the overall findings showing no effect of BWCs “provide weak support for the deployment of BWCs in policing,” they wrote. But the lack of any effect seen when study results were averaged together “masks instances [in the individual studies] where use of force decreased and others where it increased.”

When the studies were grouped together based on whether police officers followed the protocol to have the cameras record every encounter with the public, the three studies with high compliance showed a 37 percent decrease in officer use of force associated with BWCs, a finding consistent with the theory behind body cameras.

In four studies, the researchers found that the protocol assigning officers to the treatment group (use of cameras) and control group (no cameras) broke down completely: Officers assigned to camera use often used their own discretion in deciding when to record encounters, while some of those assigned to be controls wound up using cameras, thereby “contaminating” the control group. In these studies, there was no difference in the use of force between the treatment and control groups. But these were more like Wild West free-for-alls where no rules applied than rigorously controlled experiments. Camera use may, in fact, have been fairly similar in the two groups, making the the lack of any differences in use of force between them not surprising.

The surprising finding was in a third group of studies where officers had discretion in the treatment group (meaning they could activate or not activate cameras as they wished) but the control group stuck to the protocol by not using cameras. In those studies, use of force increased by 71 percent in the camera-assigned group.

The authors say a “causal mechanism” explaining that outcome is unclear. They speculate that the selective activation of cameras by officers may be closely correlated with situations that are already escalating toward potential conflict. What’s more, they think that turning cameras on during such situations may prompt an even more aggressive response from the citizen in the encounter, which in turn leads to more use of force by the officer. A key conclusion the authors draw is that procedures for consistent use of BWCs in all encounters must be followed rigorously, and accompanied by verbal reminders by officers at the start of such encounters that a body camera is recording the interaction.

Sutherland, the Rand Corporation coauthor, says the results underscore the fact that research findings are often far from black or white. It is a “classic researcher response,” he says in an interview, but “the results aren’t as clear-cut as people might have assumed.”

Meanwhile, the finding that wearing body cameras led to a greater incidence of assaults against police officers flies entirely in the face of expectations based on the deterrence theory of body cameras. The researchers laid out several possible — but unproven — alternate explanations. BWCs, they say, may make officers less assertive in their initial exchanges with the public and that somehow makes them more vulnerable to assaults. Or, they say, perhaps the knowledge that there is supporting video evidence made officers in the camera groups more likely to report instances of assault against them.

The study may have captured “a change in reporting rather than a change in citizen behavior,” says Michael White, a professor in the School of Criminology and Criminal Justice at Arizona State University who is co-directing a US Department of Justice training and technical assistance program for police departments in use of body cameras.

Further muddling the picture is a report released in mid-July by researchers from Temple University suggesting police wearing body cameras are more, not less, likely to deploy lethal force. The study examined police killings of civilians throughout the US in 2015 and found officers wearing cameras were slightly more likely to a kill a suspect. The authors theorize that knowledge that there will be clear video evidence to back them makes officers more apt to use deadly force in situations where they might otherwise exercise restraint because of concern about the uncertainty of what a follow-up inquest would yield.

Such unanticipated negative effects are precisely why more research is needed before drawing definitive conclusions on the use of body cameras.

The entire field of body camera research is only few years old. The first randomized, controlled trial of body cameras was completed just three years ago. Such studies — in which subjects are allocated at random to receive the treatment being studied (in this case the use of body cameras) or to serve as controls (by not wearing cameras) — are considered the gold standard in research. In this initial trial, a 12-month study of officers in the Rialto, California, police department, there was a 60 percent reduction in incidents of police use of force among those wearing cameras compared with those not wearing body cameras. There was an even more striking 88 percent decrease in citizen complaints about police conduct correlated with camera use.

In December 2014, the White House announced a proposal to provide $75 million in matching funds to help local police departments institute body camera programs. The cost of body cameras themselves can run as high as $1,000 each. But departments will face far greater costs associated with storage and retrieval of tens of thousands of hours of video. Denver, for example, has committed $6.1 million over five years for a body camera program, but cameras themselves only account for 8 percent of the total contract.

Dominating the body-worn camera market is Taser International, the company that is best known for its electronic stun gun. Taser, which aggressively markets its Axon body cameras and cloud-based video storage on its site, Evidence.com, controls about three-quarters of the body-worn camera business in the US.

In Massachusetts, the Methuen police began the routine use of body cameras on the city’s 47-officer force in June. Other departments, including Boston, are poised to join the camera wave.

Boston officials have agreed to launch a six-month pilot study of BWCs. The start date has been pushed back several times. At a community meeting on body cameras last week in Mattapan, Police Commissioner William Evans said he hopes to have the pilot study underway by September 1.

Evans, who boasts about the restraint his officers show and their good relations with residents, has expressed mixed feelings about body cameras, suggesting at one point that he thought Boston may not need the technology. More recently he has said he’s on board with the pilot program and is prepared to move to broader deployment of body cameras afterwards, if that’s the decision that is made. But he still seems ambivalent.

“I’ve all along said I don’t want to rush into it because I want to get it right,” Evans said at the meeting about a instituting a body camera program in Boston. “Some cities are abandoning it because of costs and they saw it didn’t increase trust,” he said, citing Arlington, Texas, as an example.

After negotiating an agreement with the city’s main police union to carry out the pilot study, the department sought 100 officers to volunteer to take part. Activists from the Boston Police Camera Action Team, a group that has pushed aggressively for the department to begin using body cameras, expressed concern at the community meeting about evaluating cameras only among officers that volunteer to try them out.

However, no officers had stepped forward as of last Friday’s deadline imposed by the department, prompting Evans to say officers would be assigned to the body camera pilot. The initial reticence among officers does not bode well for the effort if that outlook persists. “Internal buy-in within a department is essential for successful implementation of a BWC program,” White and a group of colleagues wrote in a recent paper in the journal Police Quarterly.

But that wariness could make it easier to carry out a stronger evaluation of the pilot project, as it may now be possible to design the pilot study more in line with with the approach of randomized controlled trials, whose findings are more statistically meaningful.

The design and evaluation of the pilot study will be carried out by Anthony Braga, director of the School of Criminology and Criminal Justice at Northeastern University, and Jack McDevitt, a professor at Northeastern.

Braga said at the Mattapan community meeting that the evaluation would look at four main outcomes: 1) measures of “civility” in police-citizen interactions, which includes both officer use of force and assault and battery by citizens on officers; 2) community reaction, which will be gauged through focus groups seeking to establish whether the use of body cameras increases public trust in the police; 3) police activity, measured by numbers of arrests, field interrogations, and other outcomes (some have suggested that body cameras may decrease police encounters with the public); and 4) legal issues in police-citizen interactions, principally centered on constitutional issues related to search and seizure and due process rights.

City Councilor Tito Jackson seems ready to jump to department-wide use of  body cameras. “I don’t think it is a question of if we should implement this across the board, but when we implement this across the board with every officer,” he said at the meeting. Concerns about the costs of body cameras, he said, should be weighed against the $38 million Boston has paid in recent years to settle lawsuits concerning police treatment of civilians.

Jackson also expressed opposition to the idea that the success of the pilot program would be based on an evaluation by the two Northeastern researchers. “I don’t believe the determination should be made by professors,” he said at the meeting. “I think it should be made by the residents of the city of Boston.”

Public concern about policing practices is legitimate and understandable, and Braga said at the meeting that he and McDevitt are open to input from residents as they design the pilot study. But dismissing the idea of having criminal justice research experts direct the evaluation of the project seems like an argument against collecting rigorous evidence on which to base an important and costly decision.

“This is a very, very, very expensive proposition,” says Brian Kyes, the police chief in Chelsea and president of the Massachusetts Major City Chiefs Association. “You balance that against public concerns, public outcry.” He says cities need to have as thorough an understanding of the effect of body cameras as possible, including knowledge of their costs, and “then you move forward based on that full disclosure.”

“It is important to remember, though, that BWCs are not a silver bullet,” White and his colleagues write in their recent paper. “They will not end police brutality or misconduct; rather, they are one more tool that officers can use to efficiently and effectively do their job,”

That said, “for the most part, the effects of body cameras are very positive,” White says in an interview. “Not only in terms of use of force and citizen complaints, but also for their evidentiary value and in investigating complaints against officers.”

Ongoing studies will be helpful to further flesh out the effects of body cameras, says White. And results from some cities may not be overwhelming. “In a high-functioning department you’re not going to see the giant decline in use of force or complaints, because they are not problems there,” White says. But he says cameras can still be very useful even in those departments, helping to quickly resolve whatever citizen complaints do arise.

As for the overall question of whether body cameras will be become a standard component of policing, “I think this train has left the station,” White says. “There’s an expectation that police departments are going to deploy this technology across the country. I equate it with the CSI effect,” he says of the growing use of DNA and other forensic evidence. “I think we probably won’t even be talking about it in three or four or five years. It will be a routine part of policing that citizens will just expect and demand.”