Kudos to the Washington Post

There has been an immense amount of coverage of problematic police shootings. Much of this coverage has been emotional and lacking in any sort of empirical grounding. Doing what great newspapers can and should do, the Washington Post has undertaken a project for 2015 to address the lack of transparency and accuracy of data. They are collecting data on all police shootings that result in death and report in Fatal police shootings in 2015 approaching 400 nationwide by Julie Tate, Jennifer Jenkins, Steven Rich, Keith L. Alexander and Wesley Lowery that so far this year there have been 385 deaths as a result of police shootings with a projected total for the year of around 1,000, a number nearly twice what the official data indicates.

There has been an attempt in many quarters to make this all an issue of police brutality, insensitivity, racism, etc. But the facts don't support such storytelling. Yes, there are tragics events, too many to be blithely dismissed. But the job that police face is far more complex than is readily acknowledged and they are the frontline for many deeply rooted and horribly complex issues (substance abuse and dependency, mental illness, etc.) which are not subject to easy resolution.

In the most recent reporting on their project, Fatal police shootings in 2015 approaching 400 nationwide by Julie Tate, Jennifer Jenkins, Steven Rich, Keith L. Alexander and Wesley Lowery, there is some interesting information that shows how complicated the issue is.

The very opening lines illustrate how difficult it is to stick with the facts. The authors open with:

In an alley in Denver, police gunned down a 17-year-old girl joyriding in a stolen car.

With that framing, it immediately suggests some negligence and culpability on the part of the police. A youngster joy riding. Death should not be an outcome in such a situation. But elsewhere in the reporting, the context is different. It indicates that the teen was:

killed by Denver police officers in January as she and friends allegedly tried to run them down in a stolen car.

That's quite a different frame.

While all the news reports tend to focus on shootings of black males, that is an unrepresentative picture. 25% of the deaths this year are black males. That is disproportionate to their numbers in the general population (13%) but is actually well below their numbers in terms of murders (50%) and more generally all violent crime (56% per the FBI). So while there is a common leit motif that police are disproportionately shooting young black men, the numbers indicate otherwise.

What about the story that police often shoot unarmed victims. 87% of those killed were armed (guns, cars, knives, etc. as well as with toys or facsimiles that appeared to be weapons). All deaths should be reduced but when nearly 90% involve weapons, the police response begins to make more sense. Especially when you recognize that police, in having to deal with uncertain situations every day, have to always anticipate the worst scenario first.

What else does the reporting tell us that is broadly inconsistent with the standard storyline? It confirms what was already known, which is that police shooting deaths are materially under reported in the official statistics. Of particular value is that the reporting allows us to estimate that underreporting. We now know that the official figures only capture about half of the police shooting deaths that actually occur.

Half the police shooting victims are white, somewhat underrepresented in terms of their proportion of the population (67%) but significantly overrepresented in terms of their involvement in violent crime.

5% of police shooting victims are women.

Only 8 of the 385 were under 18 years old. You'd want that to be zero but I would have thought that there might be a lot more 17 year-olds on a tear.

50% of all the killings resulted from domestic disturbances, often involving people with mental illnesses and off of their medications. 24% of all the victims were mentally ill.

Most of the victims (the Post does not specify how many) had prior criminal records.

From the examples in the article and from the numbers shared, it is clear that police shootings are rare events and almost always occur in fraught and dangerous circumstances. It is also clear that improved training might have some marginal impact on overall numbers. What is the minimum that might reasonably be expected to be accomplished? You'd like to say 0 but the mix of weapons (87%), demonstrated violence, mental health, etc. all call into question what might realistically be expected.

As both the Post and many in the article point out, you won't know, in fact, can't know, until you start actually measuring the magnitude of the problem and the details surrounding every one of the incidents. The Washington Post is doing the right thing. Instead of complaining about the absence of valid numbers and/or fanning false perceptions, they are actually investing heavily in getting the facts. This is not about race, no matter how much social justice warriors might want to make it so. It is about mental health and weaponized criminals, and domestic violence. The police face an unenviable job. With good numbers and details behind the incidents, we ought to bring the overall numbers down. If there are 1,000 in a year, perhaps we can bring it down 100-200. 200 lives would be a tremendous accomplishment. But until something can be done about the actions and behaviors of criminals themselves and about the care extended to those who are mentally ill, it seems challenging to believe that we can materially reduce the numbers beyond that improvement.