About . . . . Classes . . . . Books . . . . Vita . . . . Blog. . . . Podcast

by Peter Moskos

April 16, 2015

Well done, NYPD. What's your secret?

The national average, the rate of people killed by police (as they define it, which is pretty loose, but OK) is 0.36 per 100,000. This is over the past 23 months. That's roughly 1,135 killed per year.

This is based on these data from May 2013 to April 2014. I believe it's similar to (but a bit messier than) killedbypolice.net. But it's got city and county data, which isn't at killedbypolice.net.

Now we already know that the rate of being killed by police is a hell of a lot higher in Oklahoma, New Mexico, and Arizona (0.8) -- five times higher -- than it is in New Jersey, Michigan, and New York (0.15).

But states are big and have hundreds of police departments. I want to break it down by city. The rate in California is twice the national average. I don't think San Francisco police are shooting a lot of people. So who is?

Well, Bakersfield (rate = 2.1, which includes killing in the city killings both by the Bakersfield PD and the Kern County Sheriff Dept.), Salinas (2.0), Stockton (1.4), Fresno (1.1), and Santa Ana (0.9) come to mind. These are crazy high rates.

Super high seem to be Kansas City, MO (rate = 2.0), Oklahoma City (1.7), St Louis (1.5), Tulsa (1.4), Phoenix (1.2), and Albuquerque (1.1). Remember all these figures are rough. So I don't mean to rank order, but I do mean to group these cities together.

Bakersfield? Salinas? Maybe it's been a bad two years, but there are only 363,000 people who live in Bakersfield. Between 2012 and 2013, the NYPD killed 21 people. And in the past 23 months 15(!) people have bit the dust in Bakersfield? Do correct me if I'm wrong. The stats may be a fluke. Or maybe it was a bad two years. Or maybe the numbers are wrong. But it's still a hell of a red flag!

The rate in Los Angeles 0.5. That's not quite twice the national average... but it's one-forth of Bakersfield and Salinas. Baltimore's rate is 0.9. Chicago comes in at 0.6.

The NYPD? The big bad NYPD? The killers of Diallo, Gurley, Bell, Garner, and so many other?

Zero-point-one-three. New York City's rate is 0.13. The rate of people killed by police in one-third the national average. This is amazing.

Put another way, Chicagoans are 5 times as likely to be killed by police. Baltimoreans 7 times as likely. And Bakersfield? Lovely Bakersfield? In the streets of Bakersfield you're 16 times more likely to be killed by police than you are in New York City. [Update 2017: This is no longer true. The number of people killed by police in Bakersfield has declined greatly. But the overall numbers for small- to medium-sized cities west of the Mississippi are still very large.]

[Update: See Nick Selby's description of those shot and killed by police in Bakersfield. Maybe the streets just really are meaner.]

Think of this, too, as my NYPD friends do. Shootings by NYPD may be tragic, but compared to the rest of the nation, they really do seem to fall in the category of isolated incidents. Whatever the NYPD is doing to shoot so few people seems to be a case of best practices. Maybe the focus should be not to criticize the NYPD but to learn from it. The systemic problems seem to be out west. And maybe people who want to protest police shootings should protest police who really are shooting too many people.

Go west, young man, go west. There is health in the country, and room away from our crowds of idlers and imbeciles.

[I want to emphasize these results are primarily, not double-checked, and based on unverified data. But the even as just ballpark figures, the differences are too dramatic to ignore.]


Andy D said...

Maybe this isn't the right place to ask this question but I've been curious lately about your opinion, Mr. Moskos, of where (regionally maybe) the cops are the best. I mean "best" in the sense of "best trained" and also "best at handling situations appropriately and with respect for people and their rights." Like, generally are cops "better" in the Northeast Vs the South or in the Pacific Northwest vs the Midwest, etc.

Moskos said...

I just don't know enough to say. I don't get out of NYC much.

But while I can't say who is the best, it certainly seems like the NYPD would be a contender. Not that they couldn't do better (couldn't we all?), but they do seem to do a lot right without doing a whole lot wrong.

What surprises me is related to my stereotype of west-coast police (on states bordering the ocean), which is generally positive. A progressive history and an imagine of professionalism.... But given the rate at which they kill people, I wonder.

Andy D said...

I wonder about the Northwest; none of your high kill rate places were in OR or WA, right? I've had a few interactions with departments out there and they have always SEEMED very progressive and professional but I have no way of knowing for sure.

Moskos said...

No... but OR and WA probably would be if one were to take violent crime into account. Those are lower murder states. Hard to say.

campbell said...

NYPD's "secret" might be that they're really terrible at shooting people. In Aveni's paper NYPD's hit ratio (and there's some acknowledged weaknesses in the data) from 1990-2000 only once made it above 20 percent. Baltimore County, for roughly the same time period, reported 64 percent in daylight and 45 percent in low light. Offhand I couldn't find data for LAPD during that time period but their 2008-2010 numbers were 74, 75, and 65 percent.



Moskos said...

I don't know if one can accurately compare the stats between the NYPD and the LAPD. But the best I can try and do:

I believe The NYPD hit its target 30 out of 45 times (2012 firearms discharge report). 66% of the time.

The LAPD (2007-2010) hit 117 and missed 52. 74% hit hate. That's a difference, but not a huge difference. Not nearly as much as the kill numbers.

I don't know if it's related, but only 1 LAPD officer have been shot and killed since 2005. For the NYPD the number is 8. (From officer down)

CollegeCop said...

Judging law enforcment officers or whole agencies based on 'kill rates' is silly. You're talking about hundreds of thousands of individual human beings in thousands of agencies and somehow holding them accountable for actions the VAST majority of them probably won't ever take part in over the course of a 20-25 years career.

It has to be remembered that LE is +800k sworn members policing a country of 319 MILLION people that are virtually armed to the teeth with more guns than cars (no country has as many guns per capita as the U.S. does), and yet out of that potentiality volatile mix you get less than 1200 "killed by police" incidents per year. That's what, roughly 1 citizen out of every 266,000?

IMO professor, You're starting a bit to sound like the media types whose shows you appear on, losing sight of the grand scheme and larger context of things.

campbell said...

But we can still make fun of NYPD for the spray and pray shooting method, right?

"The Department does not consider average hit percentages for the Annual Firearms Discharge Report. Instead, the objective completion rate per incident is employed, as it is both more accurate and more instructive."

Haha, come on guys, as great fictional cop Oliver Queenan once said, we deeal in deception here. What we do not deal with is self deception. If you have two large departments and one is putting 18 percent of its rounds on target and the other is putting 70 percent of its rounds on then that is relevant and instructive as hell. I'm not buying this nonsense of "sure we're sending a lot of rounds downrange and fuck if I know where they all went but eventually one of those bullets hit a suspect and godamnit that's all you need to know".

campbell said...

Anyways, my joking around aside, as a guy who works a western state, here's my personal guess as to a big driving factor. That table you posted a couple posts down? The top and bottom of that chart seems to share a theme with some numbers on a different chart. It's not 100 percent correlation, but it's there. The suspects who dance with tina just might be a different animal then the guys who are on the horse.


Anonymous said...


Remember the NYPD has a 12 pound trigger pull.

Moskos said...

I don't think kill rate is a factor. But I wanted to check. You never know what you'll find.

I never would have suspected the disparity in being killed by police.

But I know quick access to top rate medical care is a factor in living or dying after being shot. In term of both crime and police use of lethal force, "shooting" data is better than "homicide" data. It's just a lot harder to get.

Moskos said...

Interesting about the meth connection. Could be. Though I'd guess that the states with a lot of killings were always that way. That the phenomenon precedes crystal meth.

But maybe meth is the factor. You know, meth is the only drug I have seen as a cop (Baltimore City, circa 2000) or a civilian.

That statement probably baffles the minds of cops in the poor rural west.

Moskos said...

I think it's safe to say that most NYPD have no experience with guns outside of the job. So it wouldn't surprise me that they are worse shots.

Similarly, I suspect (but have no proof) that the NYPD crashes cars more because a fair number of police in NYC only got the driver's license when they became cops.

Unknown said...

A Tale of Two Cities:

I might have shared this before, but it's relevant enough to this post to share it again


This is even more shocking when you compare rates of violent crimes, poverty rate, etc.

Anonymous said...


I'm a cop in Oregon. I've attended enough national training to see a wide difference in East and West coppers. I'm not one to say one is better than the other, just different based on society norms and demands. I think the California numbers are interesting. The pay in California and retirements were so lucrative that they have bankrupted some of the cities. Stockton and Fresno are great examples of that. They've cut their forces dramatically and now they are seeing more OIS's. The East coast pay is garbage but at least a town like NY has plenty of coppers. I think you have to find a balance with enough pay to recruit good cops, but not so much that a town (Fresno) goes bankrupt. Brett

Unknown said...

I've heard mention of the relationship between city size (presumably population size) and the rate of people killed by police. Can you fill me in any references or data sources? Or maybe you're using the Guardian or WashPo datasets and looking up population size? If there's a strong relationship that can be attributed to specific training deficiencies or use of force policies, regional training consortia might be an answer. Especially given the current lack of will at the federal level.