February 2, 2015
Force Science Institute Executive Director testifies before U.S. Presidential Task Force on Policing
MANKATO, MN—Dr. Bill Lewinski, executive director of the Force Science Institute, recently testified before the Task Force on 21st Century Policing, which was created by executive order of President Obama and charged with writing recommendations for "how policing practices can promote effective crime reduction while building public trust."
Lewinski’s January 31, 2015 appearance came after personal invitation from the governing body to share FSI’s insights into issues related to body worn cameras during a “listening session” conducted at the University of Cincinnati. Following his presentation and those of other witnesses, Lewinski fielded direct questions from a diversified group of Task Force Members including Task Force Chair Charles Ramsey, Commissioner of the Philadelphia Police Department; Laurie Robinson, the Clarence J. Robinson Professor of Criminology, Law and Society at George Mason University; Susan Rahr, Executive Director of the Washington State Criminal Justice Training Commission; and Tracey Meares, the Walton Hale Hamilton Professor of Law at Yale Law School, among numerous others.
Lewinski’s testimony focused on the widespread need for training and education on the scientifically founded limitations of body cameras and the application of that knowledge during evaluations of footage of force encounters involving police.
“In my testimony I made clear that the Force Science Institute is in favor of the use of cameras in law enforcement,” Lewinski said, “but that support comes with a caveat. Being unaware of, or ignoring, the limitations of body camera footage can have a profoundly damaging impact on the true goal of the use of that technology—which is surfacing the truth.
“Cameras are a powerful tool that can help investigators, supervisors and the community gain a more complete understanding of what happened during an encounter,” Lewinski continued, “but they are not to be considered the only source of detail.”
In the interest of sharing the details of the limitations of body cameras, the Force Science Institute released an article titled, “Police and body cameras: 10 limitations that must be considered.” A copy of the article appears below and a printer-friendly version can be found online at: www.forcescience.org/bodycams.pdf
Police and body cameras: 10 limitations that must be considered
A special report from the Force Science Institute
The idea is building that once every cop is equipped with a body camera, the controversy will be taken out of police shootings and other uses of force because “what really happened” will be captured on video for all to see.
Well, to borrow the title from an old Gershwin tune, “It Ain’t Necessarily So.”
There is no doubt that body cameras—like dash cams, cell phone cams, and surveillance cams—can provide a unique perspective on police encounters. But like those other devices, a camera mounted on an officer’s uniform or head has limitations that need to be understood and considered when evaluating the images they record.
“Rushing to condemn an officer for inappropriate behavior based solely on body-camera evidence can be a dicey proposition,” cautions Dr. Bill Lewinski, executive director of the Force Science Institute. “Certainly, a camera can provide more information about what happened on the street. But it can’t necessarily provide all the information needed to make a fair and impartial final judgment. There still may be influential human factors involved, apart from what the camera sees.”
Following are 10 limitations, as enumerated by Dr. Lewinski, that are important to keep in mind regarding body-camera evidence and, for the most part, recordings from other cameras as well.
1. A camera doesn’t follow officers’ eyes or see as they see.
At the current level of development, a body camera is not an eye-tracker like the Force Science Institute has used in some of its studies of officer attention. That complex apparatus can follow the movement of eyes and superimpose on video small red circles that mark precisely where someone is looking from one microsecond to the next.
“A body camera photographs a broad scene but it can’t document where within that scene an officer is looking at any given instant,” Lewinski says. “If the officer glances away from where the camera is concentrating, he may not see action within the camera frame that appears to be occurring ‘right before their eyes.’
“Likewise, the camera can’t acknowledge physiological and psychological phenomena that an officer may experience under high stress. As a survival mechanism, your brain may suppress some incoming visual images that seem unimportant in a life-threatening situation so you can completely focus very narrowly on the threat. You won’t be aware of what your brain is screening out.
“Your brain may also play visual tricks on you that the camera can’t match. If someone is driving a vehicle toward an officer, for example, it will seem to be closer, larger, and faster than it really is because of a phenomenon called ‘looming.’ Camera footage may not convey the same sense of threat that the officer experienced.
“In short, there can be a huge disconnect between an officer’s field of view and her visual perception and the camera’s. Later, someone reviewing what’s caught on camera and judging the officer’s actions could have a profoundly different sense of what happened than the officer had at the time it was occurring.”
2. Some important danger cues can’t be recorded.
“Tactile cues that are often important to officers in deciding to use force are difficult for cameras to capture,” Lewinski says. “‘Resistive tension’ is a prime example.
“An officer can usually tell when they touch a suspect whether he or she is going to resist. The officer may quickly apply force as a preemptive measure, but on camera it may look like that officer made an unprovoked attack, because the sensory cue he felt doesn’t record visually.”
And, of course, the camera can’t record the history and experience an officer brings to an encounter. “Suspect behavior that may appear innocuous on film to a non-law enforcement civilian can convey the risk of mortal danger to a streetwise officer,” Lewinski says. “For instance, an assaultive subject who brings his hands up may look to a civilian like he’s surrendering, but to an officer, based on past experience, that can be a very intimidating and combative movement, signaling his preparation for a fighting attack. The camera just captures the action, not an officer’s interpretation.”
3. Camera speed differs from the speed of life.
Because body cameras record at much higher speeds than typical convenience store or correctional facility security cameras, it’s less likely that important details will be lost in the millisecond gaps between frames, as sometimes happens with those cruder devices.
“But it’s still theoretically possible that something as brief as a muzzle flash or the glint of a knife blade that may become a factor in a use-of-force case could still fail to be recorded,” Lewinski says.
Of greater consequence, he believes, is the body camera’s depiction of action and reaction times.
“Because of the reactionary curve, an officer can be half a second or more behind the action as it unfolds on the screen,” Lewinski explains. “Whether he’s shooting or stopping shooting, his recognition, decision-making, and physical activation all take time—but obviously can’t be shown on camera.
“People who don’t understand this reactionary process won’t factor it in when viewing the footage. They’ll think the officer is keeping pace with the speed of the action as the camera records it. So without knowledgeable input, they aren’t likely to understand how an officer can unintentionally end up placing rounds in a suspect’s back or firing additional shots after a threat has ended.”
4. A camera may see better than a human does in low light.
“The high-tech imaging of body cameras allows them to record with clarity in many low-light settings,” Lewinski says. “When footage is screened later, it may actually be possible to see elements of the scene in sharper detail than an officer could at the time the camera was activated.
“If you are receiving less visual information than the camera is recording under time-pressured circumstances, you are going to be more dependent on context and movement in assessing and reacting to potential threats. In dim light, a suspect’s posturing will likely mean more to an officer immediately than some object he’s holding. When footage is reviewed later, it may be evident that the object in his hand was a cell phone, say, rather than a gun. If an officer is expected to have seen that as clearly as the camera did, his reaction might seem highly inappropriate.”
On the other hand, he notes, cameras do not always deal well with lighting transitions. “Going suddenly from bright to dim light or vice versa, a camera may briefly blank out images altogether,” he says.
5. An officer’s body may block the view.
“How much of a scene a camera captures is highly dependent on where it’s positioned and where the action takes place,” Lewinski notes. “Depending on location and angle, a picture may be blocked by an officer’s own body parts, from his nose to his hands.
“If an officer is firing a gun or a Taser, for example, a camera on his chest may not record much more than his extended arms and hands. Or just blading his stance may obscure the camera’s view. Critical moments within a scenario that the officer can see may be missed entirely by his body cam because of these dynamics, ultimately masking what a reviewer may need to see to make a fair judgment.”
6. A camera only records in 2-D.
Because cameras don’t record depth of field—the third dimension that’s perceived by the human eye—accurately judging distances on their footage can be difficult.
“Depending on the lens involved, cameras may compress distances between objects or make them appear closer than they really are,” Lewinski says. “Without a proper sense of distance, a reviewer may misinterpret the level of threat an officer was facing.”
In training, he critiques several camera images in which distance distortion became problematic. In one, an officer’s use of force seemed inappropriate because the suspect appears to be too far away to pose an immediate threat. In another, an officer appears to strike a suspect’s head with a flashlight when, in fact, the blow was directed at a hand and never touched the head.
“There are technical means for determining distances on 2-D recordings,” Lewinski says, “but these are not commonly known or accessed by most investigators.”
7. The absence of sophisticated time-stamping may prove critical.
The time-stamping that is automatically imposed on camera footage is a gross number, generally measuring the action minute by minute. “In some high-profile, controversial shooting cases that is not sophisticated enough,” Lewinski says. “To fully analyze and explain an officer’s perceptions, reaction time, judgment, and decision-making it may be critical to break the action down to units of one-hundredths of a second or even less.
“There are post-production computer programs that can electronically encode footage to those specifications, and the Force Science Institute strongly recommends that these be employed. When reviewers see precisely how quickly suspects can move and how fast the various elements of a use-of-force event unfold, it can radically change their perception of what happened and the pressure involved officers were under to act.”
8. One camera may not be enough.
“The more cameras there are recording a force event, the more opportunities there are likely to be to clarify uncertainties,” Lewinski says. “The angle, the ambient lighting, and other elements will almost certainly vary from one officer’s perspective to another’s, and syncing the footage up will provide broader information for understanding the dynamics of what happened. What looks like an egregious action from one angle may seem perfectly justified from another.
“Think of the analysis of plays in a football game. In resolving close calls, referees want to view the action from as many cameras as possible to fully understand what they’re seeing. Ideally, officers deserve the same consideration. The problem is that many times there is only one camera involved, compared to a dozen that may be consulted in a sporting event, and in that case the limitations must be kept even firmer in mind.
9. A camera encourages second-guessing.
“According to the U. S. Supreme Court in Graham v. Connor, an officer’s decisions in tense, uncertain, and rapidly evolving situations are not to be judged with the ‘clear, 20/20 vision of hindsight,’ ” Lewinski notes. “But in the real-world aftermath of a shooting, camera footage provides an almost irresistible temptation for reviewers to play the coulda-shoulda game.
“Under calm and comfortable conditions, they can infinitely replay the action, scrutinize it for hard-to-see detail, slow it down, freeze it. The officer had to assess what he was experiencing while it was happening and under the stress of his life potentially being on the line. That disparity can lead to far different conclusions.
“As part of the incident investigation, we recommend that an officer be permitted to see what his body camera and other cameras recorded. He should be cautioned, however, to regard the footage only as informational. He should not allow it to supplant his first-hand memory of the incident. Justification for a shooting or other use of force will come from what an officer reasonably perceived, not necessarily from what a camera saw.”
10. A camera can never replace a thorough investigation.
When officers oppose wearing cameras, civilians sometimes assume they fear “transparency.” But more often, Lewinski believes, they are concerned that camera recordings will be given undue, if not exclusive, weight in judging their actions.
“A camera’s recording should never be regarded solely as the Truth about a controversial incident,” Lewinski declares. “It needs to be weighed and tested against witness testimony, forensics, the involved officer’s statement, and other elements of a fair, thorough, and impartial investigation that takes human factors into consideration.
“This is in no way intended to belittle the merits of body cameras. Early testing has shown that they tend to reduce the frequency of force encounters as well as complaints against officers.
“But a well-known police defense attorney is not far wrong when he calls cameras ‘the best evidence and the worst evidence.’ The limitations of body cams and others need to be fully understood and evaluated to maximize their effectiveness and to assure that they are not regarded as infallible ‘magic bullets’ by people who do not fully grasp the realities of force dynamics.”
For additional information on the Force Science Institute’s thoughts on video recording and viewing during and after force events visit:
For information on the Force Science Institute visit: www.forcescience.org
Force Science Institute