Verified by 
The George Floyd case, consuming not only the U.S. but the world, is sadly yet another one of a progression of black men killed, unarmed, during an interaction with police. As many have written, this is not simply an issue for African Americans, people in the U.S., or any other demographic—it should be a priority for everyone.
Asking questions is one of the ways we, as researchers, can try to make some contribution to this issue that is so painful for so many. We ask two questions: First, beyond media accounts, how common is this problem? Second, how can we begin to understand, in a systematic way, the causes and where we go from here?
The “extent of the problem” question was actually stimulated by reading news accounts and opinion pieces. Some writers have apparently tried to minimize the extent of unjustified deaths in African American encounters with police by claiming the number of deaths is small. In a June 2 article Heather MacDonald writes: “In 2019, police officers fatally shot 1,004 people, most of whom were armed or otherwise dangerous. African Americans were about a quarter of those killed by cops last year (235).”
Apart from questioning the data itself (which some readers did), such contentions are not very reassuring from a practical viewpoint. How many deaths, justified or not, is “small?" The reactions of many protesters and activists to these occurrences make the argument that “the number is smaller than you think” ring hollow. So, what do the numbers really tell us?
To get a better perspective, we wanted to take a look at available data. Joined by our colleagues Alok Baveja of Rutgers and Sayan Mukerjee of T.A. Pai Management Institute, we looked at criminology databases like the Uniform Crime Reporter and census data. Beyond encounters where someone was killed, we considered all arrests across all police departments in the U.S. and the racial makeup of arrests associated with those departments. If there are systematic differences in the proportion of African Americans arrested among different communities, this establishes the problem and its reach in objective numbers.
We looked up the proportion of arrests that involve an African American in each U.S. community compared to the total proportion of African Americans in that community. Because we wanted to see if diverse communities and diverse police departments had any relationship to arrests, we also divided police departments and communities, based on their demographics, into diverse or not diverse units.
So, for example, if a community had a population of 20% African Americans, and 40% of those arrested were African American, the arrest ratio would be 2.0. Actually, that’s pretty close to the ratio we found—in cases where both police departments and the community were diverse.
But what about police departments that were very non-diverse racially, in communities that were also homogeneous? Here that same African American arrest ratio was closer to 4.0—nearly twice as high as in the diverse-diverse situation.
It seems undeniable that the makeup of police and the community they serve, taken together, are associated with real (and pretty dramatic) differences in arrest patterns between cities. One could come up with several explanations for such differences.
For example, some have questioned the amount and type of training needed to become a police officer. A few years ago, a study found that more hours were required to become a barber than a police officer in some states (North Carolina: 1,528 for barbers, 620 for police). In Florida, 1,760 hours are needed to be an interior designer, 770 for police; Michigan demands 4,000 hours for an electrical sign specialist, 594 for police.
To be fair, some of those numbers are minimum requirements and likely some police departments provide more. As the saying goes, the devil is in the details. But even a quick glance into the arrest data shows differences in outcomes for African Americans are real, significant, and not small.
Cases such as George Floyd, regardless of how representative they are of all police interactions, have managed to shine a light on a much larger, foundational problem (we are talking hundreds of thousands of arrests here). We hope this little trip into the evidence on race and arrests will encourage more determined digging and searching for answers. The problem itself makes it clear we all owe this issue that much attention and effort.
Co-authored by Chester Spell
