The Snopes Clarification

Having read the Snopes response to the claim “Black police officers are more likely to kill black people,” I was surprised to find that there were references in there to the “unarmed black teenager” Michael Brown and the assertion that “National statistics show that black people are statistically three times more likely to die at the hands of police officers than white people.”

Neither of those two statements are related in any way to the question at hand, both are editorialization about orthogonal issues - it’s almost as if the author didn’t want to be able to prove something positive (or disprove something negative) and so took the time to remind us how there are so many other problems, lest we forget.

Since I am quoted as one of the experts relied upon by the journalist, and since the journalist used one aspect but not the central focus of my response, I thought I would post her original question and my original answer, both unedited (except for one capitalization, in the title of my book).

I’m pretty angry. I answered carefully the writer’s question as to whether police target black men by race with detailed analysis and links to open, available, published, and reviewed data. She ignored it entirely in the final story. And our data jibed generally with that of Lott and Moody.

UPDATE: The post has been updated. Read this and I will update at the bottom.

Here’s the exchange (we spoke on the phone as well, during which I gave color, but the written quotes are what I provided):

Dear Mr Selby,

I am a fact checker for the web site Snopes.com. I have been in consistent contact with D. Brian Burghart who has a police shooting database, but for this question, he referred me to you.

A reader sent us an inquiry about a study that seems to indicate that black people are more likely to be killed by black police officers. From the information I am aware of, there isn’t enough data about the race of police officers involved in shootings, because there isn’t a lot of data available in the absence of comprehensive federal data. I could be wrong of course, but I was curious whether this is a topic you could help with. The study abstract is here, and you can get a PDF of it here as well: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2870189&download=yes

Thank you in advance for your help.

Sincerely,

Bethania Palma Snopes.com

Hi, Bethania,

Thanks for fact checking this. It’s important work, and it is not easy. I’m familiar with the Lott and Moody paper and we have spoken about it. Before I answer, let me state something, because our findings are similar to Lott and Moody’s and don’t necessarily mean what people seem to think it means.

There are in fact two distinct nationally important questions related to policing and race that must be addressed, and the general public seems to have conflated or at least blurred the lines between them: 

1. Are police targeting minorities for special attention?
2. Do officers behave differently or use greater force when confronting a black suspect than a white one? 

To illustrate the difference, consider this: if a computer selected at random cars to pull over, officers would not be able to select drivers to stop based on race. But if an officer gave more tickets to black drivers than to white drivers after the computer had selected the car, the officers would be treating black drivers differently.

To address both these questions, we must separate them and study the available data. Our study, and some others including that by Lott and Moody, have tried to start this, but the data are very difficult to come by ON THE SECOND question. I note that independently, Lott and Moody and my group came to the same conclusion, that to accomplish this, we must clean and code a large range of data from a large range of sources, and make it publicly available. Ours is available here:

https://github.com/StreetCredSoftware/PKIC

Lott and Moody’s is available here:

http://crimeresearch.org/wp-content/uploads/2013/06/police.shoot_.zip

Quoting from my book, In Context: Understanding Police Killings of Unarmed Civilians:

First, in initial contacts of those unarmed civilians who ultimately died after a confrontation with police in 2015, [like Lott and Moody] we found that police did not selectively target by race. 

Second, based on the data we collected, it is not possible to determine whether, once contact is made, police treated black or Hispanic people differently from white people. The distinction is nuanced, and it is critical as America considers ways to consider whether there is unfair policing and biased behavior by officers.

What we found in our study was that of people stopped by police in encounters that ultimately resulted in the death of an unarmed person in 2015, the ratio of blacks to whites to Hispanics was statistically the same whether police initiated the encounter or whether the encounter was initiated by a member of the community. From this we conclude that the police are not targeting for any special attention minorities. 

We are unable to answer the second question based on the data at hand. 

We DO note, however, that part of the reason for the spotty coverage was extreme media bias towards more thorough and extensive coverage of incidents in which the decedent was black. In our analysis of 679 media reports on the killings of unarmed civilians, in 2015, the American news media was more than three times more likely to name the race of the officer involved in a police killing of an unarmed person if the decedent was black, than if the decedent was White or Hispanic. 

From my reading of the Lott and Moody report, they encountered similar inconsistencies in media coverage and data availability, but I do not presume to speak for them.

I hope this answers your question. 

UPDATE: On 29 December the post was updated. It removed the “teenager” comment, added much more of my original quote from this email, and then added, baffingly, a USAToday story about how black people are more likely to be killed in car chases than white people. Why this was added I have no idea. It then repeated the “black people are three times more likely to die than white people” staistic, again, for what reason I cannot imagine.

Snopes had one job: to answer the question, “are Black police officers more likely to kill black people?” In doing so, it is considering questions of whether black officers are more likely to kill black people, and then whether black people are more likely to be killed in car chases and crashes regardless of the race of the police officer (no one asked about this?) and then implying that black people are likely to be shot just, say, walking down the street by police. This is the site people are turning to to halt the spread of fake news?

Because to me, from a fact-checking site, anything that answers anything other than the question asked, seems…Well, it just seems strange and out of place.