ACCORDING TO INITIAL reports from the Louisiana State Police, Ronald Greene died in a car crash. This was in May 2019, one year before the murder of George Floyd. Mr Greene’s family was suspicious, so they pressed the police to release footage from the cameras worn by six of the officers on the scene. Not all of the devices were switched on, and it took a 17-month fight before they could watch the footage (the public did not see it until the Associated Press leaked it, seven months after that). The footage showed the troopers repeatedly stunning Mr Greene with a taser, punching him, and leaving him moaning face-down on the road for nine minutes; he died on the way to hospital. Body-worn cameras were present. Transparency was not.
After the murder of George Floyd on May 25th 2020, a law named after him was introduced into Congress the following month with the aim of increasing police accountability. It was passed by the House of Representatives but is pending in the Senate. The House version encompasses a range of policies, including restrictions on the use of force (any application of physical restraint other than handcuffing), creating a national database of officer misconduct, requiring federal officers to wear cameras and limiting qualified immunity (a judicial doctrine that shields officers from civil liability). As the act makes its way through Congress, much attention has focused on whether it will gain bipartisan support. But the more important and equally uncertain question is whether it will actually make policing more transparent and accountable. If and when the law is adopted, the details will be important.
Much of the Floyd Act applies only to federal law-enforcement officials: the Federal Bureau of Investigation, United States Park Police, officers in the various federal departments, and others. They account for only about one-fifth of America’s 700,000 law-enforcement officers. But the bill can nevertheless serve as a standard for other departments, explains Marc Levin, a lawyer at the Council on Criminal Justice, a think tank. “It’s possible that just having this act go through, could have…a cascading effect and encourage more state and local jurisdictions to act.” But to achieve the best outcome, says Mr Levin, policies should be based on evidence, not just sound promising.
A large-scale study by the Becker Friedman Institute at the University of Chicago found that body-worn cameras reduce the number of police misconduct complaints by 17%. They also probably reduce the use of force. Evidence suggests a six-in-seven chance that the cameras reduce the number of use-of-force incidents by 10%. They also find that body-worn cameras are worth the cost. For every dollar spent, society gets about five dollars back in benefit, mostly from averting uses of force and the associated costs (investigation time, compensation to community members and administrative costs). But as of 2020, only about 8,000 departments (about 45%) used body-worn cameras and just seven states required them. The Floyd Act, as currently written, requires federal uniformed officers to wear body cameras, and requires state and local law enforcement to buy them with federal funds.
But just putting cameras on officers is not enough, as shown by Mr Greene’s tragedy, and also by that of Andrew Brown, Jr, another black man, killed by police on April 21st in North Carolina. Despite the presence of seven officers with body-worn cameras and dashboard-camera video, it took four weeks for the public to see any footage. North Carolina requires a judge’s approval to release video, as does Louisiana, where Mr Greene was killed. Eleven other states also restrict public access to body-worn camera footage. By contrast, when Ma’Khia Bryant, a black teenager, was shot and killed by an officer in Ohio, the video footage was released publicly in a few hours. Ohio’s body-camera policy, signed by Republican governor John Kasich in 2019, makes the footage part of the public record.
“A bad result was kind of written in stone when [North Carolina] passed their [body-worn camera] law,” explains Chad Marlow of the American Civil Liberties Union. “It was extraordinarily foreseeable…You cannot have officers wear body cameras, and suggest that that’s a value to the public if the public doesn’t have access to the footage.” As currently written, the Floyd Act requires that federal agencies, in most cases, release any footage showing a death or serious injury in police custody within five days of being requested to do so.
Beyond releasing footage, departments ought to specify how officers should use their cameras. Some police departments allow officers to decide when to switch them on. Others outline strict protocols for when a camera must be activated (at the start of any call, for example). Still others use cameras that automatically turn on with sirens or the removal of a gun from its holster. Such details matter. One study found that among officers systematically recording incidents on arrival, rather than using discretion, use-of-force incidents decreased by 37% in comparison to officers not using body-worn cameras. Among those who chose when to turn their cameras on and off, use-of-force incidents increased by 71% against this control group.
In Mr Greene’s situation, several moments of footage are captured in silence because officers may have deactivated the microphone. One officer was suspended without pay for 50 hours for dragging Mr Greene and deactivating the camera. But the discretion may have resulted in the loss of valuable information. Limiting choice protects officers, too, explains Mr Marlow. “In the climate that we live nowadays, if police officers have discretion and make the wrong call, a malevolent motive is going to be assigned to that officer.” The Floyd Act currently requires a federal law-enforcement officer to activate cameras when “responding to a call for service or at the initiation of any other law enforcement or investigative stop” and to deactivate only when the officer leaves the scene. This seems the best sort of regulation to ensure transparency.
Overall the George Floyd Act is fairly comprehensive, says Nancy La Vigne of the Council on Criminal Justice. But it risks being ineffective if legislators amend the act without considering how each policy performs in practice. Some other parts of the bill are strongly supported by the evidence. De-escalation training, which involves teaching officers techniques to require less frequent and severe uses of force, was found to reduce use-of-force incidents by 28%, decrease citizen injuries by 26% and lower officer injuries by 36%. This sort of training is effective, but it is only mandated in 16 states.
But not all aspects of the Floyd Act are backed by evidence, though. Implicit bias training, teaching officers about their unconscious biases, does not always work as advertised, yet it has been implemented in 69% of police departments. Former president Donald Trump signed an executive order against unconscious-bias training that involved “race or sex stereotyping”, which was reversed by President Joe Biden. The Trump order only banned training focused on “scapegoating” one race, but the message is clear: implicit-bias training is politically unpopular on the right. And as implemented currently, it is ineffective, according to the University of Chicago Crime Lab. It is possible that implicit-bias training could be effective in the future, says the Crime Lab report. But Ms La Vigne argues that with limited time and resources, policies should focus on training programmes most likely to make a positive change. With so much on its plate, the Floyd Act should stick to the evidence.