About . . . . . . Classes . . . . . . Books . . . . . . Vita . . . . . . . Links. . . . . . Blog

by Peter Moskos

October 15, 2017

Cops in Conservative Cities Shoot & Kill More Often

Forbes came out with a list of the 10 most conservative and liberal cities in America.

Top ten conservative, in rank order:
Oklahoma City
Virginia beach
colorado springs
Arlington, TX
Top ten liberal, in rank order:
San Fran
I'm not going to argue with the rankings. I don't really care. But here's what I thought: I bet police shoot a lot more people in the conservative cities. Related to and perhaps correlated with the fact police shoot more people, per capita, in states that are more white.

How's this for a working hypothesis? Other things being constant (they rarely are), police shoot more people when nobody cares about police-involved shootings. And white people -- particularly conservative white people -- don't really care about police-involved shootings. Period. No matter the race of those shot. And when there's never any pushback or criticism of police, laws and training and culture do not change.

Based on Washington Post data from January 2014 through September 2016, the annual rate (per 100,000) of police-involved homicides in the top 10 conservatives cities (n = 82) is 0.61. The annual rate of police-involved homicide in the top 10 liberal cities (n=78) is 0.20.

Now New York City accounts for a lot of that, in terms of population. But even were one to remove NYC for simply being too big, the rate in the liberal cities is 0.39, or 64 percent of the conservative city average. Even without New York, cops in the most liberal cities are more than a third less likely to shoot and kill people. Are other factors involved? Sure. And they might be correlated to political ideology. Go figure them out, if you wish.

Also of note, and I'm just looking at 2016 murder numbers, the murder rate in the top ten liberal cities in 9.96, which isn't that much higher than the homicide rate of 8.01 in the top 10 conservative cities. If you take NYC out of the equation, the homicide rate for the other 9 liberal cities goes way up to 20. But if you consider that murder is higher in the top-10 liberal cities, the lower rate of police-involved homicides is all the more impressive.

I mean, think of it this way: community violence and police-involved violence are very related. A lot of the people police shoot are violent criminals with guns, some in the process of using them. The more violent criminals there are running around shooting people, the more people police will shoot. Always has been, always will.

That said....

There were 138 murders in DC last year and every year (for the past 2.75 years) police shoot and kill 4 people. In Tulsa and Oklahoma City (which combined have 1 million people) there were 142 murdered last year and police shoot and kill 10 people. That's a big difference. Police do shoot a lot more people out west. And it's not just in conservative cities. In fact, given the low levels of murder in Seattle and San Francisco, the high number of people killed by police stand out.

Anaheim had but 7 murders last year and police shot and killed 5 people since 2015. In Boston, Arlington and Detroit, police also shot and killed 5 people in the past 2.75 years, but there were 49, 21, and 303 murders, respectively, in these cities. Why? My guess: a combination of cops being better trained, less afraid, and less trigger happy in these cities combined with cops also being less proactive.

Here's the raw data I used. (Rate modifier is used in column G, (population/100,000)/2.75, because I'm using 2.75 years of police-involved homicide data.)


David Thacher said...

Interesting. Maybe the network of liberal-minded police leadership knows something about controlling use of force after all.

It would be a lot of work, but illuminating, to look at the history of use of force controls in a few cities on each side of the liberal/conservative divide. New York's story is certainly consistent with your theory: It was specific widespread public outcry about a police shooting of an 11-yo boy that led the department to establish serious use of force standards that have made them a leader to this day. I don't know exactly what got the ball rolling in DC, but it seems to have been a series of shootings in the 1990s and they pretty quickly became a national model for use of force controls (aided by some kicks in the pants from the DOJ). I imagine you could tell this story for every agency in a liberal city.

The interesting question would be other side: The "conservative" cities clearly *had* lots of bad shootings, was there really no comparably meaningful attempt to improve use of force controls in response? Why? Would be especially interesting to look at a city like Colorado Springs, which I think had some capable leadership at one point; why wasn't that energy and talent devoted to use of force controls (if indeed it wasn't; I don't actually know, but if the shooting #s are high apparently not...)?

Gary Cordner said...

I agree, very interesting, thanks Peter. I assume a conservative interpretation would be deterrence-related -- weak policing (not shooting enough bad guys) leads to more murders. Or, police in liberal cities are afraid to do their jobs.

I do see several cities on the conservative list, besides Colorado Springs, known for having had progressive leadership during some or all of the last decade (e.g. Arlington TX), as well as several on the liberal list that have been a total mess (in terms of leadership). David's suggestion of looking into the local dynamics of a few cities on each list is an excellent thought. My hunch would be that there are some underlying national & regional factors but also each city plays out its own story in its own way.

Mike Maltz said...

Peter, take a look at Randy Roth's op-ed in the Washington Post this past weekend. It contains a historical parallel to your observation.



Jay said...

The obvious explanation is that it's far more about politics than murder. Conservatives are reluctant to place restraints on their police. Trump and Sessions are pretty clear that this is their position.

Gary Cordner said...

Related topic, saw that the 2016 LEOKA came out. I looked at the last 10 years and calculated rates per FT officer for the FBI's UCR population & agency type categories. Highest risk for an officer being feloniously killed was in cities 100,000-249,999 (about 45% higher than the average for all groups), lowest was state LEOs. For accidentally killed, state LEO risk was highest by far, more than twice as high as for any other group. That pushed state LEOs into the highest overall fatality rate (felonious + accidental), about 50% higher than the average for all groups. For overall fatality rate, cities 250,000+ were second safest. After state LEOs, cities 100,000-249,999 and cities <10,000 were next riskiest.

Peter Moskos said...

I assume state police accidental death means car crashes on highways, right?
I bet that city populations are less useful categories than geography, in as much as geography is associated with police per capita. Fewer cops out west. My theory: Backup is good. More cops are safer for public and police alike. Shouldn't be too hard to figure out that out. Though I'm not going to any time soon. But I want to put that out there.

Gary Cordner said...

Yes, I assume the state police accidental fatalities are mainly car crashes, and that they drive more miles, respond further to emergency calls, drive on more windy country roads in bad weather, spend more time standing along the side of interstates, maybe are farther from a trauma center when they do wreck, and probably also just tend to drive too fast.

The lower felonious fatalities rate in biggest cities could be due to more backup, as you say, or could also be a bit deceiving due to bigger agencies having more sworn personnel in specialized and administrative positions, i.e. their true denominator of "cops in the field doing real police work" might be disproportionately smaller, and thus their fatality rate of cops who actually make stops and respond to calls higher, just due to their size. But why the fatality risk in the next biggest size category, 100-249,999, would be a lot higher is a mystery to me, as they should be affected by many of the same size-related factors, including backup availability.

Thos Wallace said...

I don't see much in common between the various 'Liberal' cities. Half of them I couldn't afford to live in and the other half I would live in under any circumstances.

One theory is that policing is simply more dangerous when criminals have cars. An armed criminal in a car makes any confrontation much more dangerous.

Peter Moskos said...

Criminals in cars? That's a very interesting concept.