New evidence that what we think about cops and race is far too simplistic

A new study undermines the narrative of racial bias in police killings. The study is flawed, experts say, but still a useful addition to the debate.

There is no racial bias when it comes to fatal police shootings.

That’s the startling conclusion of a new study by economist Roland Fryer Jr., the youngest African-American to receive tenure at Harvard. 

The data don’t completely undermine the fundamental premise of the police-and-race debate now shaking the United States. Dr. Fryer’s research suggests that police treat black and white suspects differently in other ways – blacks are 18 percent more likely to be pushed to the ground, 16 percent more likely to be handcuffed without arrest, and 24 percent more likely to have a gun pointed at them, for example. 

But the study challenges broad assumptions that racism is a key driver in the killing of black suspects by police. 

Among those who found the results surprising was Fryer himself, who began researching police shootings because of "his anger after the deaths of Michael Brown and Freddie Gray,” he told The New York Times. And for families who have lost loved ones after police encounters that began with a broken taillight, a pocketknife, loose cigarettes, or CDs, the new study is unlikely to ameliorate the sense that their family suffered a massive injustice.

But Fryer's examination of the patterns lying behind the tragedies that have driven the national conversation on race for the past two years both adds detail and understanding to the circumstances leading up to a police shooting, and shows the paramount need for more research.

Criminologists have qualms with the study, saying it has not yet been fully vetted by other experts and has flaws in its execution. But they suggest the findings could be a valuable attempt to debunk what they see as the current, simplistic narrative about police violence and race. The statistics that exist indicate the story is far more nuanced than most media accounts suggest, they say.

Eugene O’Donnell, a criminologist at John Jay College of Criminal Justice and a former New York police officer, says he is not surprised by Fryer's findings. 

“You’re talking about a conversation with 10,000 nuances, but it’s revolving over the same visceral video,” he says. “We’re not having a national conversation on the realities of policing because that’s boring to people,” so instead the issue becomes black versus white, both literally and figuratively. 

“Ninety percent of the narrative of is a repeat of the same conversation,” he adds. “We’re just on a loop.”

Hints of a counternarrative

Some data support the general thesis of Fryer’s study. Crime rates have gone down since 2000, but lethal police shootings of both white and black Americans have increased, says James Alan Fox, a criminologist at Northeastern University in Boston.  

In fact, the increase among white Americans has been steeper – “certainly not what one might expect if you believe that blacks are treated differently by cops,” says Professor Fox. 

“From 2000 to 2013, the number of whites killed by the police increased 57 percent, while the number of blacks killed by the police increased 42 percent,” Fox wrote in a 2015 opinion article for USA Today. “Are we more critical of certain white-on-black police encounters by presuming racism as a contributing factor rather than seeing the matter as a few police officers doing their job recklessly?”

In a sign of how different statistics show different nuances, a Washington Post database of the 990 fatal police shootings in 2015 shows 90 unarmed males killed: 36 were black, 31 were white, 18 were Hispanic, and five were “other.” 

African Americans make up 13 percent of the population but 40 percent of the police shootings, seemingly supporting the general narrative. 

But the “unarmed black man” description doesn’t tell the whole story, writes Heather Mac Donald in a commentary for the Marshall Project, a nonprofit news organization focused on US criminal justice. In some of these 36 cases, unarmed suspects physically beat up officers to the point where they feared for their lives or reached for officers’ guns and equipment.   

“One can debate the tactics used and the moment when an officer would have been justified in opening fire,” writes Ms. Mac Donald, “but these cases are more complicated and morally ambiguous than a simple ‘unarmed’ classification would leave a reader to believe.”

Experts warn that race could play a subconscious factor in how police officers perceive a personal threat – something that might not show in the data.

But “the more you look at the facts, the picture will get murkier not clearer. I haven’t seen a clear pattern of widespread, demonstrable abuse with unjustified shootings,” says Professor O’Donnell of John Jay College. “It’s quite clear we’re trapped in a narrative – it’s understandable, but it’s afactual.”

Fryer’s study, which examined 1,332 shootings since 2000 in 10 cities – including Houston, Los Angeles, and Orlando – in three states, is seen as an intriguing contribution to a debate that Fryer calls “virtually data free.” But criminologists are hesitant to take it at face value for several reasons. 

• Some police departments, such as New York City, refused to give Fryer’s team their data. 
• Fryer’s team also trusted police officers’ classification of “justifiable lethal force” – which the recent videos of the deaths of Walter Scott, Eric Garner, Alton Sterling, and others have called into question. 
• The study is classified as a “working paper” without peer review. 
• The study ignores the racial bias that causes officers to stop blacks vs. whites in the first place. 

What we don't know

Many of the caveats could be eliminated by better data systems in the US justice system, says Phil Stinson, a criminologist at Bowling Green State University.

“That, to me, is the biggest takeaway from research like this – that our recording system for police data is bad,” says Professor Stinson. “Not only is it bad, but it is different in different places.”

For example, the race of a victim only comes up in some types of court records, officer training differs, and definitions of force vary.

“It’s just really, really difficult to look at this issue. We are dealing with a decentralized system in a large country,” adds Stinson. “And up until a few years ago, researchers didn’t even care about looking into police shootings.”

However, studies like Fryer’s can help Americans break away from an exaggerated narrative and realize our broken data system, says Tod Burke, a criminologist at Radford University in Virginia and a former police officer in Maryland. 

“This isn’t the end, it’s a continuing process – let’s continue the discussion but use this study and its results so we have data and not just ‘I see’ or ‘I believe,’ ” he says. “The results may show it’s not as bad as we think, but the fact that we’re having this conversation shows there’s an issue that needs to be addressed in police-community relations.”