Would Cameras on Cops Prevent Police Abuse of Force?

In the aftermath of Michael Brown's killing, I'm seeing several editorials and articles arguing that the solution to the tragedy in Ferguson (and elsewhere in the nation) would be to arm police officers with cameras on their badges, that would document their actions. The Washington Post argues:

A study submitted this month by the District’s Police Complaints Board cited the example of the Rialto, Calif., police department, which measured the use of force by officers wearing cameras against a control group of officers who didn’t wear them. The camera-wearing officers were involved in dramatically fewer incidents involving the use of their batons, pepper spray, stun guns or firearms. Behavioral changes were so striking — both in the officers and in citizens they encountered — that complaints against the cops wearing cameras declined by nearly 90 percent.

The study, in full, is here, and the recommendations address a variety of concerns: the quality and durability of the equipment, police officers' control over the initiation of recording, privacy concerns and the need to notify citizens of recording, retention of recorded material, and data access and sharing. There's also the question of cost, compared to the reduction in cost from citizen complaints.

San Francisco implemented a program that would dole out 150 $1,000 cameras to plainclothes officers to document search procedures and minimize concerns about lying and hiding evidence. But in April 2014, these were apparently not in use yet. The argument used against them is that the Rialto experiment took place in a city with a very small police force.

There are probably many benefits to documenting police activity via copcams, but I have a fairly serious misgiving. The first is that video footage is often considered, wrongly, to be objective evidence that cannot be misinterpreted. People tend to reify such evidence as presumably incontrovertible. That is not the case. In 2007, the Supreme Court decided a 1983 suit called Scott v. Harris. Harris sued the police department for giving a dangerous car chase that resulted in an accident, from which he emerged paraplegic. This video, shot from the pursuing car, was admitted in evidence:



Justice Scalia, writing for the majority and denying Harris any compensation, saw the video and concluded that "[r]espondent’s version of events is so utterly discredited by the record that no reasonable jury could have believed him."

But is that really so? Dan Kahan, Dave Hoffman and Dave Braman took Scalia's challenge and showed the video to a diverse sample of 1,350 Americans. They found considerable differences in how their subjects perceived the car chase, its dangerousness, and the justification for police action. Moreover, the subjects' perception of the chase was highly correlated with their general worldviews and personal characteristics. Folks on the conservative side of the political map tended to agree with police actions and argue that Harris was driving in a dangerous way; folks on the progressive side tended to side with Harris and regard police's manner of carrying the chase as unnecessary, excessive, and dangerous.

We don't even have to go that far. Oscar Grant's killing on New Year's Eve, 2009, was documented by cellular phones. The presumably incontrovertible evidence of intentional homicide ended in an involuntary manslaughter conviction for Johannes Mehserle. And lapel cameras did not prevent the Albuquerque shooting of Shaine Sherrill.

This is not to say that copcams are a universally bad idea. But it does mean that the documentation of police action would not put an end to assaults and would not unequivocably settle 1983 lawsuits against the police. As with forensic innovations, anyone hoping for a definitive "cure" for police abuses is going to be disappointed.