As high-profile police shootings continue to hold the nation’s attention, a new study on the extent of racial bias in police officers’ use of force is making some big waves. In fact, this research comes to a particularly surprising conclusion: the police are actually less likely to shoot black civilians. But there are some important details to work through before we jump to any conclusions.
The study, conducted by renowned Harvard economist Roland Fryer, did identify some significant bias in the way police officers use force. This is a bias that exists at nearly all levels–like putting hands on a civilian, pushing a person up against a wall, using handcuffs on someone without arresting them, and even using pepper spray or a baton–but when it comes to lethal force, the most severe of all, police may actually be slightly less likely to kill black civilians.
But before we jump to conclusions or even accept that conclusion on its face, it’s important to sort through a large number of methodological nuance to understand what we can take away from this research. Fair warning, if you came here looking for a clear-cut conclusion, you’ll have to read the rest of this article to get the full picture.
Some Background
First, it’s important to point out that Roland Fryer’s study involved a significant undertaking by a team of researchers. And Fryer himself certainly wasn’t expecting the final conclusion. “It is the most surprising result of my career,” Fryer told the New York Times. The research is also still a working paper, which means that it hasn’t yet been peer-reviewed and published in a formal journal. It was put out by the National Bureau of Economic Research for experts in the field to look at and discuss its findings prior to a more formal release.
The debate about police shootings is one that has notoriously had a lack of hard data to discuss. We simply don’t have good statistics on the number and characteristics of incidents when police officers used lethal force. Independent counts have started to fill this gap–of note are the Washington Post’s database of police shootings and the Guardian’s “The Counted” project–but there is little national data with the depth necessary to identify bias. With this research, Fryer and his team provide some important analysis to a discussion that has few statistics to draw from.
A look at the methods
So let’s take a closer look at the data and how it was used. To measure racial bias when it comes to all levels of force the researchers looked at statistics provided by several police departments, notably New York City’s “stop, question and frisk” program, as well as nationally representative survey data that measures interactions with the police. To look at the use of lethal force–officer-involved shootings–the researchers had to assemble their own dataset from 10 police departments in three states. The researchers managed to get a particularly interesting dataset from the Houston police department, which provided a large number of reports on interactions between police and civilians.
Of that Houston dataset, the researchers took a random sample of files with “arrests codes in which lethal force is more likely to be justified: attempted capital murder of a public safety officer, aggravated assault on a public safety officer, resisting arrest, evading arrest, and interfering in arrest.” However, this data was unique to the Houston police department. The other police departments could only give details for instances when lethal force was used, but we would need data about when officers decided not to use force in order to properly identify the effects of bias.
The conclusions and what they mean
First, it’s important to note that the researchers only look at actual interactions between officers and civilians. This means that the study does not engage with one important racial bias in policing–that officers are more likely to stop and interact with black civilians than white civilians.
That caveat aside, when it comes to the lethal use of force, here’s how Fryer summarizes his findings:
Using data from Houston, Texas–where we have both officer-involved shootings and a randomly chosen set of potential interactions with police where lethal force may have been justified–we find, in the raw data, that blacks are 23.8 percent less likely to be shot at by police relative to whites.
The researchers used the raw data referenced in the quote and then controlled for a range of factors–civilian behavior, possession of a weapon, the situational context, and much more–but still found no presence of bias in the use of lethal force by the Houston police department.
Importantly, this is only based on the Houston data, so while we may be able to conclude that officer-involved shootings in Houston are not subject to racial bias, we cannot really take that to mean that the same holds true nationally or for any other police department. Even then, this conclusion rests on the researchers’ ability to control for several important variables like behavior and context.
It is also important to note that this data came from police officer summaries of their interactions with civilians. While we cannot cast doubt on an entire police department, there have been several cases where the story provided by police officers has been refuted with video evidence.
Police officers’ bias in nonlethal force
Looking beyond lethal force, Fryer and his team found persistent racial bias at all other levels of force. Using the New York City Police Department’s data on its stop and frisk program between 2003 and 2013, the researchers found a notable relationship between race and the use of force. After controlling for as many variables as possible, the researchers note that black civilians are 17 percent more likely to be subject to the use of force than white civilians. For Hispanics, the rate was 12 percent higher than for white people.
Interestingly, this pattern largely remained consistent at a range of different levels. The lowest level of force measured–officers laying their hands on a subject–occurred much more frequently than the highest–using pepper spray or a baton. But the rate at which minority civilians incur the use of force largely remained consistent at all levels short of lethal force. As Fryer puts it, “The use of high levels of force in these data are rare. Yet, it is consistently rarer for whites relative to blacks.” What’s particularly interesting about this is that the researchers managed to find bias in data collected and provided by the NYPD itself.
The researchers also looked at a nationally representative survey to identify the extent to which racial bias exists in these interactions. Using data from the Police-Public Contact Survey conducted by the Bureau of Justice Statistics, they found even larger differences in the use of force by race. While the researchers note that the rate at which officers used force was considerably lower in the survey data–about 15 percent of white civilians experienced the use of force in the stop and frisk data while only 1 percent reported experiencing force in the survey data–a pattern of bias remains for different racial groups. They conclude:
Differences in quantitative magnitudes aside, the PPCS paints a similar portrait–large racial differences in police use of force that cannot be explained using a large and varied set of controls.
After looking over his research, Fryer argues that police may act according to perceived costs. He suggests that there may not be racial bias in the use of lethal force because doing so is particularly costly–there is often internal reviews and the decision to shoot someone can have profound life consequences for the police officer as well as the victim. However, the same costs may not exist when using nonlethal force. Fryer argues that if we wish to reduce racial bias for lower levels of force, we should increase the costs associated with using them. Put simply, if we want to reduce this bias, police need to feel that they will be held accountable for unnecessarily using force.