The murder of George Floyd by Minneapolis police officers was captured on video, not once but half a dozen times. As we try to understand why a police officer continued compressing a man’s neck and spine for minutes after he’d lost consciousness, we have footage from security cameras at Cup Foods, where Floyd allegedly paid for cigarettes with a counterfeit $20 bill. As we wrestle with the sight of three officers standing by as their colleague killed Floyd, we have footage from the cell phones of witnesses who begged the officers to let Floyd off the ground. In the murder trial of Officer Derek Chauvin, who was patrolling despite 17 civilian complaints against him and previous involvement in two shootings of suspects, his defense may hinge on video from the body cameras he and other officers were wearing.
None of these videos saved George Floyd’s life, and it is possible that none of them will convict his murderer.
Officer Chauvin knew this. In the video shot by 17-year-old Darnella Frazier, you can see him lock eyes with the teenager. He knows she’s filming, and knows that the video is likely being streamed to Facebook, to the horror of those watching it. After all, in a suburb of nearby St. Paul four years earlier, Officer Jeronimo Yanez shot and killed Philando Castile while Castile’s partner streamed the video to Facebook. Yanez’s police car dashcam also recorded the seven shots he pumped into Castile’s body. He was charged and acquitted.
After Castile’s death, I wrote a piece for MIT Technology Review about “sousveillance,” the idea posited by the inventor Steve Mann, the “father of wearable computing,” that connected cameras controlled by citizens could be used to hold power accountable. Even though bystander video of Eric Garner being choked to death by New York police officer Daniel Pantaleo in 2014 had led not to Pantaleo’s indictment but to the arrest of Ramsey Orta, the man who filmed the murder, I offered my hope that “the ubiquity of cell-phone cameras combined with video streaming services like Periscope, YouTube, and Facebook Live has set the stage for citizens to hold the police responsible for excessive use of force.”
I was wrong.
Much of what we think about surveillance comes from the French philosopher Michel Foucault. Foucault examined the ideas of the English reformer Jeremy Bentham, who proposed a prison—the panopticon or Inspection-House—in which every cell was observable from a central watchtower. The possibility that someone might be watching, Bentham believed, would be enough to prevent bad behavior by prisoners. Foucault observed that this knowledge of being watched forces us to police ourselves; our act of disciplining ourselves as if we were always under observation, more than the threat of corporal punishment, is the primary mechanism of “political technology” and power in modern society.
The hope for sousveillance comes from the same logic. If police officers know they’re being watched both by their body cameras and by civilians with cell phones, they will discipline themselves and refrain from engaging in unnecessary violence. It’s a good theory, but in practice, it hasn’t worked. A large study in 2017 by the Washington, DC, mayor’s office assigned more than a thousand police officers in the District to wear body cameras and more than a thousand to go camera-free. The researchers hoped to find evidence that wearing cameras correlated with better policing, less use of force, and fewer civilian complaints. They found none: the difference in behavior between the officers who knew they were being watched and the officers who knew they were not was statistically insignificant. Another study, which analyzed the results of 10 randomized controlled trials of body camera use in different nations, was helpfully titled “Wearing body cameras increases assaults against officers and does not reduce police use of force.”
Reacting to the DC study, some scholars have hoped that if cameras don’t deter officers from violent behavior, at least the film can hold them accountable afterwards. There, too, body cameras rarely work the way we hope. While careful, frame-by-frame analysis of video often shows that victims of police shootings were unarmed and that officers mistook innocuous objects for weapons, attorneys for the defense screen the videos at normal speed to show how tense, fast, and scary confrontations between police and suspects can be. A 1989 Supreme Court decision means that if police officers have an “objectively reasonable” fear that their lives or safety are in danger, they are justified in using deadly force. Videos from body cameras and bystander cell phones have worked to bolster “reasonable fear” defense claims as much as they have demonstrated the culpability of police officers.
It turns out that images matter, but so does power. Bentham’s panopticon works because the warden of the prison has the power to punish you if he witnesses your misbehavior. But Bentham’s other hope for the panopticon—that the behavior of the warden would be transparent and evaluated by all who saw him—has never come to pass. Over 10 years, from 2005 to 2014, only 48 officers were charged with murder or manslaughter for use of lethal force, though more than 1,000 people a year are killed by police in the United States.
As he stared at Darnella Frazier, Officer Chauvin knew this, because it’s impossible to work in law enforcement in the US and not know this. The institutions that protect police officers from facing legal consequences for their actions—internal affairs divisions, civil service job protections, police unions, “reasonable fear”—work far better than the institutions that hold them responsible for abuses.
The hope that pervasive cameras by themselves would counterbalance the systemic racism that leads to the overpolicing of communities of color and the disproportionate use of force against black men was simply a techno-utopian fantasy. It was a hope that police violence could be an information problem like Uber rides or Amazon recommendations, solvable by increasing the flows of data. But after years of increasingly widespread bodycam use and ever more pervasive social media, it’s clear that information can work only when it’s harnessed to power. If there’s one thing that Americans—particularly people of color in America—have learned from George Floyd, Philando Castile, and Eric Garner, it’s that individuals armed with images are largely powerless to make systemic change.
That’s the reason people have taken to the streets in Minneapolis, DC, New York, and so many other cities. There’s one thing images of police brutality seem to have the power to do: shock, outrage, and mobilize people to demand systemic change. That alone is the reason to keep filming.
What happened to the microfinance organization Kiva?
A group of strikers argue that the organization seems more focused on making money than creating change. Are they right?
How one elite university is approaching ChatGPT this school year
Why Yale never considered banning the technology.
Six ways that AI could change politics
A new era of AI-powered domestic politics may be coming. Watch for these milestones to know when it’s arrived.
Cryptography may offer a solution to the massive AI-labeling problem
An internet protocol called C2PA adds a “nutrition label” to images, video, and audio.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.