A new study on use of force by US police has found that, while black, Native and Latino Americans are more likely to be stopped, searched and arrested by police, race does not affect the risk of injury or death in a police encounter.
Study author Ted Miller said he found the results “surprising”, but hopes that the paper will “improve the quality of discussion” around race in US policing.
“If we want to intervene, it’s the excessive arrest of minorities that’s the problem. We need to bring down the numbers of them being arrested in the first place,” said Miller, a research scientist at the Pacific Institute for Research and Evaluation in Maryland.
The study, published in the online journal Injury Prevention, culled police fatality data from the Guardian’s The Counted project, along with national hospital discharge data, FBI data on police stops and the Police Public Contact Survey to arrive at its findings. The different data sources are drawn from different years spanning from 2011 to 2015, but Miller says the numbers across years are extremely stable and appropriate for comparison, adding: “This is not a problem that’s suddenly exploded.”
Miller discovered that, in line with the Guardian’s own findings, black men are disproportionately killed by police. But his research suggests that the reason for this increased risk is the greater number of stops and arrests by police, what is known as “excess exposure”.
Miller looked at the number of injuries and fatalities relative to the number of police interactions (arrests and stops), and found that the rate was the same across races.
Miller and his team calculated that during one in 291 stops or arrests by police in 2012, a bystander was either hospitalized or killed. They found that both the hospital-admitted and fatal injury ratios did not differ significantly between racial and ethnic groups. Although the report suggests that its estimates are “far from exact”, Miller said he had little doubt about the central conclusion. Given the confidence rates, “we can take that to the bank and say ‘that’s what’s there in this country’.
“What this study says is that it doesn’t matter what your race is when you’re in a stop and frisk situation or arrest situation with a police officer. Your chance of being injured or killed is the same regardless of race – it’s equally dangerous for everyone,” Miller said.
The details of people who had been stopped in traffic and searched, stopped on the street, or otherwise arrested were looked at in terms of race, gender, age and location (rural versus urban). The highest-risk groups were found to be males between 30 and 44 years old. Miller believes that the riskiest factor for someone in an “adversarial” situation with police is whether they are carrying a gun themselves.
According to the Guardian’s Counted data, in 2016, 58% of black Americans who have been shot and killed by police were carrying a gun at the time, compared with 50% for whites. Black Americans are twice as likely to be completely unarmed when shot and killed by police in 2016. It is possible for both to be true in large part because white Americans have been about four times more likely to have a knife when shot than blacks.
Miller said that his findings were consistent with simulation studies that have found “police are no more likely to fire on unarmed blacks than unarmed whites”, citing a 2007 study from the University of Colorado. Researchers there found that, while the general public is more likely to erroneously shoot an unarmed black person in a computer simulation, police officers in the same game did not make the same errors. “Officers showed greater sensitivity,” that study reads, and “this tended to be particularly true with Black targets”. The same study found that both police and the general public were susceptible to anti-black bias in their reaction time.
Sam Sinyangwe, an organizer with the Campaign Zero campaign to reduce police violence, said that the numbers “likely systematically underreported the number of officer caused injuries to black people and especially to Latinos/Asians and undocumented folks” because “marginalized populations are less likely to have health insurance, less likely to be located near a hospital, and less likely to utilize those resources when available
”.
Miller countered that, “if you’re injured seriously enough that you’re hospital admitted, you’re going to be admitted regardless of whether you can afford it because the hospitals are required to provide care.” He added that, in those situations where a person is injured badly in an interaction with police, emergency medical services will typically be requested to the scene by officers, making it an involuntary decision, “so i don’t really think that’s a factor,” Miller said.
The study becomes the second in as many weeks to cast doubt on what has become prevailing wisdom over the past two years of robust nationwide attention on race and policing: that black Americans face a uniquely acute threat of death in interactions with police. In a controversial paper released last week, the Harvard researcher Roland Fryer similarly concluded that in the case of so-called “officer-involved shootings – we find no racialized differences in either the raw data or when contextual factors are taken into account.”
Fryer’s study was widely critiqued for its lack of peer review, its small geographic sample and its reliance on police narrative for characterizing encounters.
A third study, also released this month, from the Center for Policing Equity (PE), a criminal justice thinktank, reached a conclusion similar to Fryer’s. Its data suggests that in most use-of-force situations, black Americans were “more likely than Whites to be targeted for force”, including the drawing and pointing of weapons, use of pepper spray, Tasers and hands. But when it came to lethal force, the mean use of force rate was nearly double for whites versus blacks. The PE data was pulled from Bureau of Justice Statistics numbers for eight locales, and calculated use of force compared against arrests.
Miller said the first time he did his calculations, like PE, he used arrest data but decided those numbers might not be reliable without also looking at stops, as they assume officers use their discretion for arrests evenly among all arrests. “Maybe the police are biased in who they arrest, and who they stop,” Miller said, “so you need to look at who the police are stopping as well.”
Mona Chalabi also contributed to this report.