New Study Finds Body Cameras Aren’t Objective Witnesses

There were multiple cameras watching when a Charlotte, North Carolina, police officer fatally shot Keith Lamont Scott in September 2016. One, a body camera, is shaky and dynamic, moving along with an officer and seeing what the officer saw. The other, a dashboard camera inside a police car, provides a static and removed view of the whole scene.

The two leave viewers with very different impressions of the incident—and, research shows, different takeaways about the amount of blame that could be attributed to the officers. People are more likely to absolve officers after viewing footage from body cameras than after seeing the same incident captured on a dashboard camera, according to a new study.

That’s because viewers usually don’t see the person wearing the body camera, says lead study author Broderick Turner, a doctoral student in the Kellogg School of Management at Northwestern University. “We need to see a person to attach judgments of attribution. There’s less of a person available in the body camera than the dash cam,” he said. “We need to see people doing things in order to think that they did them.” The study analyzed nearly 400 police videos, and found that officers spent far less time on screen in body camera videos than dashboard camera videos.

Cameras are inexpensive ways for a city to signal efforts toward improved accountability for police officers, and they have been advocated by activists and law enforcement leadership alike. However, though dashboard and body cameras can be valuable tools to monitor police activity, research shows that the footage they collect isn’t impartial. The perspective of the camera can skew the way viewers interpret what they see.

The thought has been that video footage is objective and will always show exactly what happens. We need to remember that’s not always the case.

Kristyn Jones John Jay College of Criminal Justice

The video from North Carolina was used as part of Turner’s research comparing viewers’ perceptions of footage from body cameras to their perceptions of footage from dashboard cameras. The findings, published in the journal PNAS, also found that participants in the study were less likely to say that they would indict an officer wearing a body camera than one recorded on a dashboard camera.

“The thought has been that video footage is objective and will always show exactly what happens. We need to remember that’s not always the case,” said Kristyn Jones, a doctoral student in the psychology and law program at John Jay College of Criminal Justice.

In addition to keeping the wearers off-screen, body cameras keep suspects or civilians engaging with officers as the dominant actors in videos. Researchers have speculated that because suspects or civilians are the focus, viewers might see them as having more responsibility, due to a cognitive effect called illusory causation that makes people more likely to ascribe intent to the dominant thing in their view. Body cameras placed at shoulder or chest level will also naturally point at video subjects from below. That perspective makes them look taller and larger than they actually are, which might also make them seem more threatening.

The type of information presented alongside body camera footage can also influence how viewers interpret it, Jones said. She has studied the effects of officer-generated reports on viewer impressions of incidents recorded by body cameras.

“We used an example of an officer who uses a baton to strike a civilian repeatedly. It doesn’t appear [the civilian] is doing anything violent, but the report says the officer behaved that way because the civilian had hit him with his elbow, and was carrying a knife, even though that wasn’t shown in the video,” Jones explained. “People then relied on the officer’s report to make sense of what they were seeing.” Participants in the study who said they identified with police officers were also more likely to use the information in the report, even though it didn’t match what they saw in the video.

Body cameras are great. They just can’t be the only solution.

Broderick Turner Kellogg School of Management at Northwestern University

Despite the results, Jones notes that none of these findings discount the role of body cameras in policing and police accountability. “I think body camera footage might be part of the answer, but it’s only one aspect,” she said.

Turner agreed. “Body cameras are great. They just can’t be the only solution.”

They might, for example, require additional context when they are presented in court to juries or judges. One strategy, Jones suggests, would be to provide instruction to jurors on how to view videos, similar to the way they are given instructions on how to evaluate eyewitness testimony. It’s probably not enough, she says, to just set up a television and let the tape roll.

“We still don’t know how jurors might view these videos, even though we’re starting to understand it better,” she said. Research into body and dashboard cameras is still in early stages, and there are dozens more questions to answer: Would seeing more video angles cut down on some of the observed biases? How does the order in which written and video reports are presented in court change viewers’ interpretations?

The officer who shot Keith Lamont Scott was not charged by prosecutors, who determined that the shooting was justified. As Turner noted, police officers are hardly ever indicted. He’s hopeful improved use and understanding of videos might help change things. “Thinking optimistically, indictments could go up if there was more accountability and it was clear what happened,” he said. “Realistically, I don’t know if it’s a problem with evidence or the structural nature of policing.”

But in the meantime, Turner says he will continue pressing forward with his research.

“We have all this video now, not just on police, but on everyone,” he said. “We still don’t really know what it means to have it and how it changes the way we understand things.”